server: implement /json/list

Mimic Chrome's endpoints to enable Selenium CDP connection.

.github/actions/install/action.yml

@@ -0,0 +1,78 @@
+name: "Browsercore install"
+description: "Install deps for the project browsercore"
+
+inputs:
+  zig:
+    description: 'Zig version to install'
+    required: false
+    default: '0.13.0'
+  arch:
+    description: 'CPU arch used to select the v8 lib'
+    required: false
+    default: 'x86_64'
+  os:
+    description: 'OS used to select the v8 lib'
+    required: false
+    default: 'linux'
+  zig-v8:
+    description: 'zig v8 version to install'
+    required: false
+    default: 'v0.1.11'
+  v8:
+    description: 'v8 version to install'
+    required: false
+    default: '11.1.134'
+  cache-dir:
+    description: 'cache dir to use'
+    required: false
+    default: '~/.cache'
+
+runs:
+  using: "composite"
+
+  steps:
+    - name: Install apt deps
+      if: ${{ inputs.os == 'linux' }}
+      shell: bash
+      run: sudo apt-get install -y wget xz-utils python3 ca-certificates git pkg-config libglib2.0-dev gperf libexpat1-dev cmake clang
+
+    - uses: mlugg/setup-zig@v1
+      with:
+        version: ${{ inputs.zig }}
+
+    - name: Cache v8
+      id: cache-v8
+      uses: actions/cache@v4
+      env:
+        cache-name: cache-v8
+      with:
+        path: ${{ inputs.cache-dir }}/v8
+        key: libc_v8_${{ inputs.v8 }}_${{ inputs.os }}_${{ inputs.arch }}_${{ inputs.zig-v8 }}.a
+
+    - if: ${{ steps.cache-v8.outputs.cache-hit != 'true' }}
+      shell: bash
+      run: |
+        mkdir -p ${{ inputs.cache-dir }}/v8
+
+        wget -O ${{ inputs.cache-dir }}/v8/libc_v8.a https://github.com/lightpanda-io/zig-v8-fork/releases/download/${{ inputs.zig-v8 }}/libc_v8_${{ inputs.v8 }}_${{ inputs.os }}_${{ inputs.arch }}.a
+
+    - name: install v8
+      shell: bash
+      run: |
+        mkdir -p vendor/zig-js-runtime/vendor/v8/${{inputs.arch}}-${{inputs.os}}/debug
+        ln -s ${{ inputs.cache-dir }}/v8/libc_v8.a vendor/zig-js-runtime/vendor/v8/${{inputs.arch}}-${{inputs.os}}/debug/libc_v8.a
+
+        mkdir -p vendor/zig-js-runtime/vendor/v8/${{inputs.arch}}-${{inputs.os}}/release
+        ln -s ${{ inputs.cache-dir }}/v8/libc_v8.a vendor/zig-js-runtime/vendor/v8/${{inputs.arch}}-${{inputs.os}}/release/libc_v8.a
+
+    - name: libiconv
+      shell: bash
+      run: make install-libiconv
+
+    - name: build mimalloc
+      shell: bash
+      run: make install-mimalloc
+
+    - name: build netsurf
+      shell: bash
+      run: make install-netsurf

.github/workflows/build-deps.yml

@@ -1,60 +0,0 @@
-name: build-deps
-
-on:
-  push:
-    branches:
-      - "main"
-    paths:
-      - "vendor/lexbor-src"
-      - "vendor/netsurf/**"
-  # Allows you to run this workflow manually from the Actions tab
-  workflow_dispatch:
-
-env:
-  REGISTRY: ghcr.io
-  IMAGE_NAME: lightpanda-io/browsercore-deps
-  ZIG_DOCKER_VERSION: 0.12.0-dev.1773-8a8fd47d2
-
-jobs:
-  build-deps:
-    strategy:
-      matrix:
-        include:
-          - os: linux
-            build_arch: amd64
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      packages: write
-
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-          token: ${{ secrets.GH_CI_PAT }}
-          submodules: true
-
-      - name: Docker connect
-        uses: docker/login-action@65b78e6e13532edd9afa3aa52ac7964289d1a9c1
-        with:
-          registry: ${{ env.REGISTRY }}
-          username: ${{ github.actor }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Extract metadata (tags, labels) for Docker
-        id: meta
-        uses: docker/metadata-action@9ec57ed1fcdbf14dcef7dfbe97b2010124a938b7
-        with:
-          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
-
-      - name: Docker build
-        uses: docker/build-push-action@f2a1d5e99d037542a71f64918e516c093c6f3fc4
-        with:
-          context: .
-          file: Dockerfile.deps
-          push: true
-          tags: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}:${{env.ZIG_DOCKER_VERSION}}
-          labels: ${{ steps.meta.outputs.labels }}
-          build-args: |
-            ZIG_DOCKER_VERSION=${{ env.ZIG_DOCKER_VERSION }}
-          platforms: ${{matrix.os}}/${{matrix.build_arch}}

.github/workflows/build.yml

@@ -0,0 +1,73 @@
+name: nightly build
+
+on:
+  schedule:
+    - cron: "2 2 * * *"
+
+  # Allows you to run this workflow manually from the Actions tab
+  workflow_dispatch:
+
+permissions:
+  contents: write
+
+jobs:
+  build-linux-x86_64:
+    env:
+      ARCH: x86_64
+      OS: linux
+
+    runs-on: ubuntu-22.04
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
+          submodules: recursive
+
+      - uses: ./.github/actions/install
+
+      - name: zig build
+        run: zig build --release=safe -Doptimize=ReleaseSafe -Dengine=v8 -Dcpu=x86_64
+
+      - name: Rename binary
+        run: mv zig-out/bin/lightpanda lightpanda-${{ env.ARCH }}-${{ env.OS }}
+
+      - name: Upload the build
+        uses: ncipollo/release-action@v1
+        with:
+          allowUpdates: true
+          artifacts: lightpanda-${{ env.ARCH }}-${{ env.OS }}
+          tag: nightly
+
+  build-macos-aarch64:
+    env:
+      ARCH: aarch64
+      OS: macos
+
+    runs-on: macos-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
+          submodules: recursive
+
+      - uses: ./.github/actions/install
+        with:
+          os: ${{env.OS}}
+          arch: ${{env.ARCH}}
+
+      - name: zig build
+        run: zig build --release=safe -Doptimize=ReleaseSafe -Dengine=v8
+
+      - name: Rename binary
+        run: mv zig-out/bin/lightpanda lightpanda-${{ env.ARCH }}-${{ env.OS }}
+
+      - name: Upload the build
+        uses: ncipollo/release-action@v1
+        with:
+          allowUpdates: true
+          artifacts: lightpanda-${{ env.ARCH }}-${{ env.OS }}
+          tag: nightly

.github/workflows/cla.yml

@@ -0,0 +1,32 @@
+name: "CLA Assistant"
+on:
+  issue_comment:
+    types: [created]
+  pull_request_target:
+    types: [opened,closed,synchronize]
+
+permissions:
+  actions: write
+  contents: read
+  pull-requests: write
+  statuses: write
+
+jobs:
+  CLAAssistant:
+    runs-on: ubuntu-latest
+    steps:
+      - name: "CLA Assistant"
+        if: (github.event.comment.body == 'recheck' || github.event.comment.body == 'I have read the CLA Document and I hereby sign the CLA') || github.event_name == 'pull_request_target'
+        uses: contributor-assistant/github-action@v2.6.1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          PERSONAL_ACCESS_TOKEN: ${{ secrets.CLA_GH_PAT }}
+        with:
+          path-to-signatures: 'signatures/browser/version1/cla.json'
+          path-to-document: 'https://github.com/lightpanda-io/browser/blob/main/CLA.md'
+          # branch should not be protected
+          branch: 'main'
+          allowlist: krichprollsch,francisbouvier,katie-lpd
+
+          remote-organization-name: lightpanda-io
+          remote-repository-name: cla

.github/workflows/wpt.yml

@@ -1,7 +1,6 @@
 name: wpt
 
 env:
-  ARCH: x86_64-linux
   AWS_ACCESS_KEY_ID: ${{ vars.LPD_PERF_AWS_ACCESS_KEY_ID }}
   AWS_SECRET_ACCESS_KEY: ${{ secrets.LPD_PERF_AWS_SECRET_ACCESS_KEY }}
   AWS_BUCKET: ${{ vars.LPD_PERF_AWS_BUCKET }}
@@ -12,10 +11,12 @@ on:
     branches:
       - main
     paths:
+      - "build.zig"
       - "src/**/*.zig"
       - "src/*.zig"
       - "tests/wpt/**"
       - "vendor/**"
+      - ".github/**"
   pull_request:
 
     # By default GH trigger on types opened, synchronize and reopened.
@@ -26,10 +27,13 @@ on:
     types: [opened, synchronize, reopened, ready_for_review]
 
     paths:
+      - ".github/**"
+      - "build.zig"
       - "src/**/*.zig"
       - "src/*.zig"
       - "tests/wpt/**"
       - "vendor/**"
+      - ".github/**"
   # Allows you to run this workflow manually from the Actions tab
   workflow_dispatch:
 
@@ -41,37 +45,15 @@ jobs:
     if: github.event.pull_request.draft == false
 
     runs-on: ubuntu-latest
-    container:
-      image: ghcr.io/lightpanda-io/zig-browsercore:0.12.0-dev.1773-8a8fd47d2
-      credentials:
-       username: ${{ github.actor }}
-       password: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
         with:
           fetch-depth: 0
-          token: ${{ secrets.GH_CI_PAT }}
-          # fetch submodules recusively, to get jsruntime-lib submodules also.
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
           submodules: recursive
 
-      - name: install v8
-        run: |
-          mkdir -p vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/debug
-          ln -s /usr/local/lib/libc_v8.a vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/debug/libc_v8.a
-
-          mkdir -p vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/release
-          ln -s /usr/local/lib/libc_v8.a vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/release/libc_v8.a
-
-      - name: install deps
-        run: |
-          ln -s /usr/local/lib/lexbor vendor/lexbor
-
-          ln -s /usr/local/lib/libiconv vendor/libiconv
-
-          ln -s /usr/local/lib/netsurf/build vendor/netsurf/build
-          ln -s /usr/local/lib/netsurf/lib vendor/netsurf/lib
-          ln -s /usr/local/lib/netsurf/include vendor/netsurf/include
+      - uses: ./.github/actions/install
 
       - run: zig build wpt -Dengine=v8 -- --safe --summary
 
@@ -88,7 +70,7 @@ jobs:
           echo "${{github.sha}}" > commit.txt
 
       - name: upload artifact
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: wpt-results
           path: |
@@ -112,7 +94,7 @@ jobs:
 
     steps:
       - name: download artifact
-        uses: actions/download-artifact@v3
+        uses: actions/download-artifact@v4
         with:
           name: wpt-results
 

.github/workflows/zig-fmt.yml

@@ -1,5 +1,8 @@
 name: zig-fmt
 
+env:
+  ZIG_VERSION: 0.13.0
+
 on:
   pull_request:
 
@@ -11,6 +14,8 @@ on:
     types: [opened, synchronize, reopened, ready_for_review]
 
     paths:
+      - ".github/**"
+      - "build.zig"
       - "src/**/*.zig"
       - "src/*.zig"
   # Allows you to run this workflow manually from the Actions tab
@@ -24,16 +29,13 @@ jobs:
     if: github.event.pull_request.draft == false
 
     runs-on: ubuntu-latest
-    container:
-      image: ghcr.io/lightpanda-io/zig:0.12.0-dev.1773-8a8fd47d2
-      credentials:
-       username: ${{ github.actor }}
-       password: ${{ secrets.GITHUB_TOKEN }}
-    outputs:
-      zig_fmt_errs: ${{ steps.fmt.outputs.zig_fmt_errs }}
 
     steps:
-      - uses: actions/checkout@v3
+      - uses: mlugg/setup-zig@v1
+        with:
+          version: ${{ env.ZIG_VERSION }}
+
+      - uses: actions/checkout@v4
         with:
           fetch-depth: 0
 

.github/workflows/zig-test.yml

@@ -1,16 +1,22 @@
 name: zig-test
 
 env:
-  ARCH: x86_64-linux
+  AWS_ACCESS_KEY_ID: ${{ vars.LPD_PERF_AWS_ACCESS_KEY_ID }}
+  AWS_SECRET_ACCESS_KEY: ${{ secrets.LPD_PERF_AWS_SECRET_ACCESS_KEY }}
+  AWS_BUCKET: ${{ vars.LPD_PERF_AWS_BUCKET }}
+  AWS_REGION: ${{ vars.LPD_PERF_AWS_REGION }}
 
 on:
   push:
     branches:
       - main
     paths:
+      - "build.zig"
       - "src/**/*.zig"
       - "src/*.zig"
-      - "vendor/jsruntime-lib"
+      - "vendor/zig-js-runtime"
+      - ".github/**"
+      - "vendor/**"
   pull_request:
 
     # By default GH trigger on types opened, synchronize and reopened.
@@ -21,57 +27,144 @@ on:
     types: [opened, synchronize, reopened, ready_for_review]
 
     paths:
+      - ".github/**"
+      - "build.zig"
       - "src/**/*.zig"
       - "src/*.zig"
       - "vendor/**"
+      - ".github/**"
+      - "vendor/**"
   # Allows you to run this workflow manually from the Actions tab
   workflow_dispatch:
 
 jobs:
+  zig-build-dev:
+    name: zig build dev
+
+    # Don't run the CI with draft PR.
+    if: github.event.pull_request.draft == false
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
+          submodules: recursive
+
+      - uses: ./.github/actions/install
+
+      - name: zig build debug
+        run: zig build -Dengine=v8
+
+      - name: upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: lightpanda-build-dev
+          path: |
+            zig-out/bin/lightpanda
+          retention-days: 1
+
+  zig-build-release:
+    name: zig build release
+
+    # Don't run the CI on PR
+    if: github.event_name != 'pull_request'
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
+          submodules: recursive
+
+      - uses: ./.github/actions/install
+
+      - name: zig build release
+        run: zig build -Doptimize=ReleaseSafe -Dengine=v8
+
   zig-test:
     name: zig test
 
     # Don't run the CI with draft PR.
     if: github.event.pull_request.draft == false
 
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          # fetch submodules recusively, to get zig-js-runtime submodules also.
+          submodules: recursive
+
+      - uses: ./.github/actions/install
+
+      - name: zig build test
+        run: zig build test -Dengine=v8 -- --json > bench.json
+
+      - name: write commit
+        run: |
+          echo "${{github.sha}}" > commit.txt
+
+      - name: upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: bench-results
+          path: |
+            bench.json
+            commit.txt
+          retention-days: 10
+
+  bench-fmt:
+    name: perf-fmt
+    needs: zig-test
+
+    # Don't execute on PR
+    if: github.event_name != 'pull_request'
+
     runs-on: ubuntu-latest
     container:
-      image: ghcr.io/lightpanda-io/zig-browsercore:0.12.0-dev.1773-8a8fd47d2
+      image: ghcr.io/lightpanda-io/perf-fmt:latest
       credentials:
        username: ${{ github.actor }}
        password: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-      - uses: actions/checkout@v3
+      - name: download artifact
+        uses: actions/download-artifact@v4
         with:
+          name: bench-results
+
+      - name: format and send json result
+        run: /perf-fmt bench-browser ${{ github.sha }} bench.json
+
+  demo-puppeteer:
+    name: demo-puppeteer
+    needs: zig-build-dev
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          repository: 'lightpanda-io/demo'
           fetch-depth: 0
-          token: ${{ secrets.GH_CI_PAT }}
-          # fetch submodules recusively, to get jsruntime-lib submodules also.
-          submodules: recursive
 
-      - name: install v8
+      - run: npm install
+
+      - name: download artifact
+        uses: actions/download-artifact@v4
+        with:
+          name: lightpanda-build-dev
+
+      - run: chmod a+x ./lightpanda
+
+      - name: run puppeteer
         run: |
-          mkdir -p vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/debug
-          ln -s /usr/local/lib/libc_v8.a vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/debug/libc_v8.a
-
-          mkdir -p vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/release
-          ln -s /usr/local/lib/libc_v8.a vendor/jsruntime-lib/vendor/v8/${{env.ARCH}}/release/libc_v8.a
-
-      - name: install deps
-        run: |
-          ln -s /usr/local/lib/lexbor vendor/lexbor
-
-          ln -s /usr/local/lib/libiconv vendor/libiconv
-
-          ln -s /usr/local/lib/netsurf/build vendor/netsurf/build
-          ln -s /usr/local/lib/netsurf/lib vendor/netsurf/lib
-          ln -s /usr/local/lib/netsurf/include vendor/netsurf/include
-
-      - name: zig build debug
-        run: zig build -Dengine=v8
-
-      - name: zig build test
-        run: zig build test -Dengine=v8
-
-      - name: zig build release
-        run: zig build -Doptimize=ReleaseSafe -Dengine=v8
+          python3 -m http.server 1234 -d ./public &
+          ./lightpanda &
+          RUNS=2 npm run bench-puppeteer-cdp

.gitignore

@@ -1,7 +1,5 @@
 zig-cache
+/.zig-cache/
 zig-out
-/vendor/lexbor/
-/vendor/netsurf/build/
-/vendor/netsurf/lib/
-/vendor/netsurf/include/
+/vendor/netsurf/out
 /vendor/libiconv/

.gitmodules

@@ -1,24 +1,30 @@
-[submodule "vendor/jsruntime-lib"]
-	path = vendor/jsruntime-lib
-	url = git@github.com:lightpanda-io/jsruntime-lib.git
-[submodule "vendor/lexbor-src"]
-	path = vendor/lexbor-src
-	url = https://github.com/lexbor/lexbor
+[submodule "vendor/zig-js-runtime"]
+	path = vendor/zig-js-runtime
+	url = https://github.com/lightpanda-io/zig-js-runtime.git/
 [submodule "vendor/netsurf/libwapcaplet"]
 	path = vendor/netsurf/libwapcaplet
-	url = https://source.netsurf-browser.org/libwapcaplet.git
+	url = https://github.com/lightpanda-io/libwapcaplet.git/
 [submodule "vendor/netsurf/libparserutils"]
 	path = vendor/netsurf/libparserutils
-	url = https://source.netsurf-browser.org/libparserutils.git
+	url = https://github.com/lightpanda-io/libparserutils.git/
 [submodule "vendor/netsurf/libdom"]
 	path = vendor/netsurf/libdom
-	url = git@github.com:lightpanda-io/libdom.git
+	url = https://github.com/lightpanda-io/libdom.git/
 [submodule "vendor/netsurf/share/netsurf-buildsystem"]
 	path = vendor/netsurf/share/netsurf-buildsystem
 	url = https://source.netsurf-browser.org/buildsystem.git
 [submodule "vendor/netsurf/libhubbub"]
 	path = vendor/netsurf/libhubbub
-	url = https://source.netsurf-browser.org/libhubbub.git
+	url = https://github.com/lightpanda-io/libhubbub.git/
 [submodule "tests/wpt"]
 	path = tests/wpt
 	url = https://github.com/lightpanda-io/wpt
+[submodule "vendor/mimalloc"]
+	path = vendor/mimalloc
+	url = https://github.com/microsoft/mimalloc.git/
+[submodule "vendor/tls.zig"]
+	path = vendor/tls.zig
+	url = https://github.com/ianic/tls.zig.git/
+[submodule "vendor/zig-async-io"]
+	path = vendor/zig-async-io
+	url = https://github.com/lightpanda-io/zig-async-io.git/

CLA.md

@@ -0,0 +1,93 @@
+# Lightpanda (Selecy SAS) Grant and Contributor License Agreement (“Agreement”)
+
+This agreement is based on the Apache Software Foundation Contributor License
+Agreement. (v r190612)
+
+Thank you for your interest in software projects stewarded by Lightpanda
+(Selecy SAS) (“Lightpanda”). In order to clarify the intellectual property
+license granted with Contributions from any person or entity, Lightpanda must
+have a Contributor License Agreement (CLA) on file that has been agreed to by
+each Contributor, indicating agreement to the license terms below. This license
+is for your protection as a Contributor as well as the protection of Lightpanda
+and its users; it does not change your rights to use your own Contributions for
+any other purpose. This Agreement allows an individual to contribute to
+Lightpanda on that individual’s own behalf, or an entity (the “Corporation”) to
+submit Contributions to Lightpanda, to authorize Contributions submitted by its
+designated employees to Lightpanda, and to grant copyright and patent licenses
+thereto.
+
+You accept and agree to the following terms and conditions for Your present and
+future Contributions submitted to Lightpanda. Except for the license granted
+herein to Lightpanda and recipients of software distributed by Lightpanda, You
+reserve all right, title, and interest in and to Your Contributions.
+
+1. Definitions. “You” (or “Your”) shall mean the copyright owner or legal
+   entity authorized by the copyright owner that is making this Agreement with
+   Lightpanda. For legal entities, the entity making a Contribution and all
+   other entities that control, are controlled by, or are under common control
+   with that entity are considered to be a single Contributor. For the purposes
+   of this definition, “control” means (i) the power, direct or indirect, to
+   cause the direction or management of such entity, whether by contract or
+   otherwise, or (ii) ownership of fifty percent (50%) or more of the
+   outstanding shares, or (iii) beneficial ownership of such entity.
+   “Contribution” shall mean any work, as well as any modifications or
+   additions to an existing work, that is intentionally submitted by You to
+   Lightpanda for inclusion in, or documentation of, any of the products owned
+   or managed by Lightpanda (the “Work”). For the purposes of this definition,
+   “submitted” means any form of electronic, verbal, or written communication
+   sent to Lightpanda or its representatives, including but not limited to
+   communication on electronic mailing lists, source code control systems (such
+   as GitHub), and issue tracking systems that are managed by, or on behalf of,
+   Lightpanda for the purpose of discussing and improving the Work, but
+   excluding communication that is conspicuously marked or otherwise designated
+   in writing by You as “Not a Contribution.”
+
+2. Grant of Copyright License. Subject to the terms and conditions of this
+   Agreement, You hereby grant to Lightpanda and to recipients of software
+   distributed by Lightpanda a perpetual, worldwide, non-exclusive, no-charge,
+   royalty-free, irrevocable copyright license to reproduce, prepare derivative
+   works of, publicly display, publicly perform, sublicense, and distribute
+   Your Contributions and such derivative works.
+
+3. Grant of Patent License. Subject to the terms and conditions of this
+   Agreement, You hereby grant to Lightpanda and to recipients of software
+   distributed by Lightpanda a perpetual, worldwide, non-exclusive, no-charge,
+   royalty-free, irrevocable (except as stated in this section) patent license
+   to make, have made, use, offer to sell, sell, import, and otherwise transfer
+   the Work, where such license applies only to those patent claims licensable
+   by You that are necessarily infringed by Your Contribution(s) alone or by
+   combination of Your Contribution(s) with the Work to which such
+   Contribution(s) were submitted. If any entity institutes patent litigation
+   against You or any other entity (including a cross-claim or counterclaim in
+   a lawsuit) alleging that your Contribution, or the Work to which you have
+   contributed, constitutes direct or contributory patent infringement, then
+   any patent licenses granted to that entity under this Agreement for that
+   Contribution or Work shall terminate as of the date such litigation is
+   filed.
+
+4. You represent that You are legally entitled to grant the above license. If
+   You are an individual, and if Your employer(s) has rights to intellectual
+   property that you create that includes Your Contributions, you represent
+   that You have received permission to make Contributions on behalf of that
+   employer, or that Your employer has waived such rights for your
+   Contributions to Lightpanda. If You are a Corporation, any individual who
+   makes a contribution from an account associated with You will be considered
+   authorized to Contribute on Your behalf.
+
+5. You represent that each of Your Contributions is Your original creation (see
+   section 7 for submissions on behalf of others).
+
+6. You are not expected to provide support for Your Contributions,except to the
+   extent You desire to provide support. You may provide support for free, for
+   a fee, or not at all. Unless required by applicable law or agreed to in
+   writing, You provide Your Contributions on an “AS IS” BASIS, WITHOUT
+   WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including,
+   without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT,
+   MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE.
+
+7. Should You wish to submit work that is not Your original creation, You may
+   submit it to Lightpanda separately from any Contribution, identifying the
+   complete details of its source and of any license or other restriction
+   (including, but not limited to, related patents, trademarks, and license
+   agreements) of which you are personally aware, and conspicuously marking the
+   work as “Submitted on behalf of a third-party: [named here]”.

CONTRIBUTING.md

@@ -0,0 +1,10 @@
+# Contributing
+
+Lightpanda accepts pull requests through GitHub.
+
+You have to sign our [CLA](CLA.md) during your first pull request process
+otherwise we're not able to accept your contributions.
+
+The process signature uses the [CLA assistant
+lite](https://github.com/marketplace/actions/cla-assistant-lite). You can see
+an example of the process in [#303](https://github.com/lightpanda-io/browser/pull/303).

Dockerfile

@@ -0,0 +1,79 @@
+FROM ubuntu:22.04
+
+ARG ZIG=0.13.0
+ARG ZIG_MINISIG=RWSGOq2NVecA2UPNdBUZykf1CCb147pkmdtYxgb3Ti+JO/wCYvhbAb/U
+ARG OS=linux
+ARG ARCH=x86_64
+ARG V8=11.1.134
+ARG ZIG_V8=v0.1.11
+
+RUN apt-get update -yq && \
+    apt-get install -yq xz-utils \
+        python3 ca-certificates git \
+        pkg-config libglib2.0-dev \
+        gperf libexpat1-dev \
+        cmake clang \
+        curl git
+
+# install minisig
+RUN curl -L -O https://github.com/jedisct1/minisign/releases/download/0.11/minisign-0.11-linux.tar.gz && \
+    tar xvzf minisign-0.11-linux.tar.gz
+
+# install zig
+RUN curl -O https://ziglang.org/download/${ZIG}/zig-linux-x86_64-${ZIG}.tar.xz && \
+    curl -O https://ziglang.org/download/${ZIG}/zig-linux-x86_64-${ZIG}.tar.xz.minisig
+
+RUN minisign-linux/x86_64/minisign -Vm zig-linux-x86_64-${ZIG}.tar.xz -P ${ZIG_MINISIG}
+
+# clean minisg
+RUN rm -fr minisign-0.11-linux.tar.gz minisign-linux
+
+# install zig
+RUN tar xvf zig-linux-x86_64-${ZIG}.tar.xz && \
+    mv zig-linux-x86_64-${ZIG} /usr/local/lib && \
+    ln -s /usr/local/lib/zig-linux-x86_64-${ZIG}/zig /usr/local/bin/zig
+
+# clean up zig install
+RUN rm -fr zig-linux-x86_64-${ZIG}.tar.xz zig-linux-x86_64-${ZIG}.tar.xz.minisig
+
+# force use of http instead of ssh with github
+RUN cat <<EOF > /root/.gitconfig
+[url "https://github.com/"]
+    insteadOf="git@github.com:"
+EOF
+
+# clone lightpanda
+RUN git clone git@github.com:lightpanda-io/browser.git
+
+WORKDIR /browser
+
+# install deps
+RUN git submodule init && \
+    git submodule update --recursive
+
+RUN cd vendor/zig-js-runtime && \
+    git submodule init && \
+    git submodule update --recursive
+
+RUN make install-libiconv && \
+    make install-netsurf && \
+    make install-mimalloc
+
+# download and install v8
+RUN curl -L -o libc_v8.a https://github.com/lightpanda-io/zig-v8-fork/releases/download/${ZIG_V8}/libc_v8_${V8}_${OS}_${ARCH}.a && \
+    mkdir -p vendor/zig-js-runtime/vendor/v8/${ARCH}-${OS}/release && \
+    mv libc_v8.a vendor/zig-js-runtime/vendor/v8/${ARCH}-${OS}/release/libc_v8.a
+
+# build release
+RUN make build
+
+FROM ubuntu:22.04
+
+# copy ca certificates
+COPY --from=0 /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/ca-certificates.crt
+
+COPY --from=0 /browser/zig-out/bin/lightpanda /bin/lightpanda
+
+EXPOSE 9222/tcp
+
+CMD ["/bin/lightpanda", "--host", "0.0.0.0", "--port", "9222"]

Dockerfile.deps

@@ -1,34 +0,0 @@
-# This dockerfile is used to build browsercore vendor dependencies except
-# jsruntime-lib v8.
-# jsruntime-lib v8 is built via zig-v8-fork/Dockerfile.
-ARG ZIG_DOCKER_VERSION=0.11.0
-FROM ghcr.io/lightpanda-io/zig:${ZIG_DOCKER_VERSION} as build
-
-# Install required dependencies
-RUN apt update && \
-    apt install -y git curl bash xz-utils python3 ca-certificates pkg-config \
-    libglib2.0-dev gperf libexpat1-dev cmake build-essential
-
-COPY ./Makefile /src/
-WORKDIR /src
-
-# build lexbor
-ADD ./vendor/lexbor-src /src/vendor/lexbor-src
-RUN make install-lexbor
-
-# build libiconv
-RUN make install-libiconv
-
-# build netsurf
-ADD ./vendor/netsurf /src/vendor/netsurf
-RUN make install-netsurf
-
-FROM scratch as artifact
-
-COPY --from=build /src/vendor/libiconv /usr/local/lib/libiconv
-
-COPY --from=build /src/vendor/lexbor /usr/local/lib/lexbor
-
-COPY --from=build /src/vendor/netsurf/build /usr/local/lib/netsurf/build
-COPY --from=build /src/vendor/netsurf/lib /usr/local/lib/netsurf/lib
-COPY --from=build /src/vendor/netsurf/include /usr/local/lib/netsurf/include

LICENSE

@@ -0,0 +1,661 @@
+                    GNU AFFERO GENERAL PUBLIC LICENSE
+                       Version 3, 19 November 2007
+
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The GNU Affero General Public License is a free, copyleft license for
+software and other kinds of works, specifically designed to ensure
+cooperation with the community in the case of network server software.
+
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+our General Public Licenses are intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+  Developers that use our General Public Licenses protect your rights
+with two steps: (1) assert copyright on the software, and (2) offer
+you this License which gives you legal permission to copy, distribute
+and/or modify the software.
+
+  A secondary benefit of defending all users' freedom is that
+improvements made in alternate versions of the program, if they
+receive widespread use, become available for other developers to
+incorporate.  Many developers of free software are heartened and
+encouraged by the resulting cooperation.  However, in the case of
+software used on network servers, this result may fail to come about.
+The GNU General Public License permits making a modified version and
+letting the public access it on a server without ever releasing its
+source code to the public.
+
+  The GNU Affero General Public License is designed specifically to
+ensure that, in such cases, the modified source code becomes available
+to the community.  It requires the operator of a network server to
+provide the source code of the modified version running there to the
+users of that server.  Therefore, public use of a modified version, on
+a publicly accessible server, gives the public access to the source
+code of the modified version.
+
+  An older license, called the Affero General Public License and
+published by Affero, was designed to accomplish similar goals.  This is
+a different license, not a version of the Affero GPL, but Affero has
+released a new version of the Affero GPL which permits relicensing under
+this license.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                       TERMS AND CONDITIONS
+
+  0. Definitions.
+
+  "This License" refers to version 3 of the GNU Affero General Public License.
+
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
+
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+  1. Source Code.
+
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
+
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+  The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+  The Corresponding Source for a work in source code form is that
+same work.
+
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+  13. Remote Network Interaction; Use with the GNU General Public License.
+
+  Notwithstanding any other provision of this License, if you modify the
+Program, your modified version must prominently offer all users
+interacting with it remotely through a computer network (if your version
+supports such interaction) an opportunity to receive the Corresponding
+Source of your version by providing access to the Corresponding Source
+from a network server at no charge, through some standard or customary
+means of facilitating copying of software.  This Corresponding Source
+shall include the Corresponding Source for any work covered by version 3
+of the GNU General Public License that is incorporated pursuant to the
+following paragraph.
+
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the work with which it is combined will remain governed by version
+3 of the GNU General Public License.
+
+  14. Revised Versions of this License.
+
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU Affero General Public License from time to time.  Such new versions
+will be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU Affero General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU Affero General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+  If the Program specifies that a proxy can decide which future
+versions of the GNU Affero General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+  15. Disclaimer of Warranty.
+
+  THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW.  EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published
+    by the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+  If your software can interact with users remotely through a computer
+network, you should also make sure that it provides a way for users to
+get its source.  For example, if your program is a web application, its
+interface could display a "Source" link that leads users to an archive
+of the code.  There are many ways you could offer source, and different
+solutions will be better for different programs; see section 13 for the
+specific requirements.
+
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU AGPL, see
+<https://www.gnu.org/licenses/>.

LICENSING.md

@@ -0,0 +1,23 @@
+# Licensing
+
+License names used in this document are as per [SPDX License
+List](https://spdx.org/licenses/).
+
+The default license for this project is [AGPL-3.0-only](LICENSE).
+
+## MIT
+
+The following files are licensed under MIT:
+
+```
+src/http/Client.zig
+src/polyfill/fetch.js
+```
+
+The following directories and their subdirectories are licensed under their
+original upstream licenses:
+
+```
+vendor/
+tests/wpt/
+```

Makefile

@@ -3,6 +3,27 @@
 
 ZIG := zig
 BC := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))
+# option test filter make unittest F="server"
+F=
+
+# OS and ARCH
+kernel = $(shell uname -ms)
+ifeq ($(kernel), Darwin arm64)
+	OS := macos
+	ARCH := aarch64
+else ifeq ($(kernel), Linux aarch64)
+	OS := linux
+	ARCH := aarch64
+else ifeq ($(kernel), Linux arm64)
+	OS := linux
+	ARCH := aarch64
+else ifeq ($(kernel), Linux x86_64)
+	OS := linux
+	ARCH := x86_64
+else
+	$(error "Unhandled kernel: $(kernel)")
+endif
+
 
 # Infos
 # -----
@@ -23,54 +44,36 @@ help:
 
 # $(ZIG) commands
 # ------------
-.PHONY: build build-release run run-release shell test bench download-zig wpt
+.PHONY: build build-dev run run-release shell test bench download-zig wpt unittest
 
-zig_version = $(shell grep 'recommended_zig_version = "' "vendor/jsruntime-lib/build.zig" | cut -d'"' -f2)
-kernel = $(shell uname -ms)
+zig_version = $(shell grep 'recommended_zig_version = "' "vendor/zig-js-runtime/build.zig" | cut -d'"' -f2)
 
 ## Download the zig recommended version
 download-zig:
-ifeq ($(kernel), Darwin x86_64)
-	$(eval target="macos")
-	$(eval arch="x86_64")
-else ifeq ($(kernel), Darwin arm64)
-	$(eval target="macos")
-	$(eval arch="aarch64")
-else ifeq ($(kernel), Linux aarch64)
-	$(eval target="linux")
-	$(eval arch="aarch64")
-else ifeq ($(kernel), Linux arm64)
-	$(eval target="linux")
-	$(eval arch="aarch64")
-else ifeq ($(kernel), Linux x86_64)
-	$(eval target="linux")
-	$(eval arch="x86_64")
-else
-	$(error "Unhandled kernel: $(kernel)")
-endif
-	$(eval url = "https://ziglang.org/builds/zig-$(target)-$(arch)-$(zig_version).tar.xz")
-	$(eval dest = "/tmp/zig-$(target)-$(arch)-$(zig_version).tar.xz")
+	$(eval url = "https://ziglang.org/download/$(zig_version)/zig-$(OS)-$(ARCH)-$(zig_version).tar.xz")
+	$(eval dest = "/tmp/zig-$(OS)-$(ARCH)-$(zig_version).tar.xz")
 	@printf "\e[36mDownload zig version $(zig_version)...\e[0m\n"
 	@curl -o "$(dest)" -L "$(url)" || (printf "\e[33mBuild ERROR\e[0m\n"; exit 1;)
 	@printf "\e[33mDownloaded $(dest)\e[0m\n"
 
-## Build in debug mode
+## Build in release-safe mode
 build:
-	@printf "\e[36mBuilding (debug)...\e[0m\n"
-	@$(ZIG) build -Dengine=v8 || (printf "\e[33mBuild ERROR\e[0m\n"; exit 1;)
-	@printf "\e[33mBuild OK\e[0m\n"
-
-build-release:
 	@printf "\e[36mBuilding (release safe)...\e[0m\n"
 	@$(ZIG) build -Doptimize=ReleaseSafe -Dengine=v8 || (printf "\e[33mBuild ERROR\e[0m\n"; exit 1;)
 	@printf "\e[33mBuild OK\e[0m\n"
 
-## Run the server
+## Build in debug mode
+build-dev:
+	@printf "\e[36mBuilding (debug)...\e[0m\n"
+	@$(ZIG) build -Dengine=v8 || (printf "\e[33mBuild ERROR\e[0m\n"; exit 1;)
+	@printf "\e[33mBuild OK\e[0m\n"
+
+## Run the server in debug mode
 run: build
 	@printf "\e[36mRunning...\e[0m\n"
-	@./zig-out/bin/browsercore || (printf "\e[33mRun ERROR\e[0m\n"; exit 1;)
+	@./zig-out/bin/lightpanda || (printf "\e[33mRun ERROR\e[0m\n"; exit 1;)
 
-## Run a JS shell in release-safe mode
+## Run a JS shell in debug mode
 shell:
 	@printf "\e[36mBuilding shell...\e[0m\n"
 	@$(ZIG) build shell -Dengine=v8 || (printf "\e[33mBuild ERROR\e[0m\n"; exit 1;)
@@ -90,18 +93,22 @@ test:
 	@$(ZIG) build test -Dengine=v8 || (printf "\e[33mTest ERROR\e[0m\n"; exit 1;)
 	@printf "\e[33mTest OK\e[0m\n"
 
+unittest:
+	@TEST_FILTER='${F}' $(ZIG) build unittest -freference-trace --summary all
+
 # Install and build required dependencies commands
 # ------------
 .PHONY: install-submodule
-.PHONY: install-lexbor install-jsruntime install-jsruntime-dev install-libiconv
+.PHONY: install-zig-js-runtime install-zig-js-runtime-dev install-libiconv
 .PHONY: _install-netsurf install-netsurf clean-netsurf test-netsurf install-netsurf-dev
+.PHONY: install-mimalloc install-mimalloc-dev clean-mimalloc
 .PHONY: install-dev install
 
 ## Install and build dependencies for release
-install: install-submodule install-lexbor install-jsruntime install-netsurf
+install: install-submodule install-zig-js-runtime install-libiconv install-netsurf install-mimalloc
 
 ## Install and build dependencies for dev
-install-dev: install-submodule install-lexbor install-jsruntime-dev install-netsurf-dev
+install-dev: install-submodule install-zig-js-runtime-dev install-libiconv install-netsurf-dev install-mimalloc-dev
 
 install-netsurf-dev: _install-netsurf
 install-netsurf-dev: OPTCFLAGS := -O0 -g -DNDEBUG
@@ -109,14 +116,16 @@ install-netsurf-dev: OPTCFLAGS := -O0 -g -DNDEBUG
 install-netsurf: _install-netsurf
 install-netsurf: OPTCFLAGS := -DNDEBUG
 
-BC_NS := $(BC)vendor/netsurf
-ICONV := $(BC)vendor/libiconv
+BC_NS := $(BC)vendor/netsurf/out/$(OS)-$(ARCH)
+ICONV := $(BC)vendor/libiconv/out/$(OS)-$(ARCH)
 # TODO: add Linux iconv path (I guess it depends on the distro)
 # TODO: this way of linking libiconv is not ideal. We should have a more generic way
 # and stick to a specif version. Maybe build from source. Anyway not now.
-_install-netsurf: install-libiconv
+_install-netsurf: clean-netsurf
 	@printf "\e[36mInstalling NetSurf...\e[0m\n" && \
-	ls $(ICONV) 1> /dev/null || (printf "\e[33mERROR: you need to install libiconv in your system (on MacOS on with Homebrew)\e[0m\n"; exit 1;) && \
+	ls $(ICONV)/lib/libiconv.a 1> /dev/null || (printf "\e[33mERROR: you need to execute 'make install-libiconv'\e[0m\n"; exit 1;) && \
+	mkdir -p $(BC_NS) && \
+	cp -R vendor/netsurf/share $(BC_NS) && \
 	export PREFIX=$(BC_NS) && \
 	export OPTLDFLAGS="-L$(ICONV)/lib" && \
 	export OPTCFLAGS="$(OPTCFLAGS) -I$(ICONV)/include" && \
@@ -154,10 +163,7 @@ _install-netsurf: install-libiconv
 
 clean-netsurf:
 	@printf "\e[36mCleaning NetSurf build...\e[0m\n" && \
-	cd vendor/netsurf && \
-	rm -R build && \
-	rm -R lib && \
-	rm -R include
+	rm -Rf $(BC_NS)
 
 test-netsurf:
 	@printf "\e[36mTesting NetSurf...\e[0m\n" && \
@@ -167,30 +173,56 @@ test-netsurf:
 	cd vendor/netsurf/libdom && \
 	BUILDDIR=$(BC_NS)/build/libdom make test
 
-install-libiconv:
-ifeq ("$(wildcard vendor/libiconv/lib/libiconv.a)","")
+download-libiconv:
+ifeq ("$(wildcard vendor/libiconv/libiconv-1.17)","")
 	@mkdir -p vendor/libiconv
 	@cd vendor/libiconv && \
 	curl https://ftp.gnu.org/pub/gnu/libiconv/libiconv-1.17.tar.gz | tar -xvzf -
-	@cd vendor/libiconv/libiconv-1.17 && \
-	./configure --prefix=$(BC)vendor/libiconv --enable-static && \
-	make && make install
 endif
 
-install-lexbor:
-	@mkdir -p vendor/lexbor
-	@cd vendor/lexbor && \
-	cmake ../lexbor-src -DLEXBOR_BUILD_SHARED=OFF && \
-	make
+install-libiconv: download-libiconv clean-libiconv
+	@cd vendor/libiconv/libiconv-1.17 && \
+	./configure --prefix=$(ICONV) --enable-static && \
+	make && make install
 
-install-jsruntime-dev:
-	@cd vendor/jsruntime-lib && \
+clean-libiconv:
+ifneq ("$(wildcard vendor/libiconv/libiconv-1.17/Makefile)","")
+	@cd vendor/libiconv/libiconv-1.17 && \
+	make clean
+endif
+
+install-zig-js-runtime-dev:
+	@cd vendor/zig-js-runtime && \
 	make install-dev
 
-install-jsruntime:
-	@cd vendor/jsruntime-lib && \
+install-zig-js-runtime:
+	@cd vendor/zig-js-runtime && \
 	make install
 
+.PHONY: _build_mimalloc
+
+MIMALLOC := $(BC)vendor/mimalloc/out/$(OS)-$(ARCH)
+_build_mimalloc: clean-mimalloc
+	@mkdir -p $(MIMALLOC)/build && \
+	cd $(MIMALLOC)/build && \
+	cmake -DMI_BUILD_SHARED=OFF -DMI_BUILD_OBJECT=OFF -DMI_BUILD_TESTS=OFF -DMI_OVERRIDE=OFF $(OPTS) ../../.. && \
+	make && \
+	mkdir -p $(MIMALLOC)/lib
+
+install-mimalloc-dev: _build_mimalloc
+install-mimalloc-dev: OPTS=-DCMAKE_BUILD_TYPE=Debug
+install-mimalloc-dev:
+	@cd $(MIMALLOC) && \
+	mv build/libmimalloc-debug.a lib/libmimalloc.a
+
+install-mimalloc: _build_mimalloc
+install-mimalloc:
+	@cd $(MIMALLOC) && \
+	mv build/libmimalloc.a lib/libmimalloc.a
+
+clean-mimalloc:
+	@rm -Rf $(MIMALLOC)/build
+
 ## Init and update git submodule
 install-submodule:
 	@git submodule init && \

README.md

@@ -1,19 +1,125 @@
-# Browsercore
+<p align="center">
+  <a href="https://lightpanda.io"><img src="https://cdn.lightpanda.io/assets/images/logo/lpd-logo.png" alt="Logo" height=170></a>
+</p>
 
-## Build
+<h1 align="center">Lightpanda Browser</h1>
+
+<p align="center"><a href="https://lightpanda.io/">lightpanda.io</a></p>
+
+<div align="center">
+
+[![Commit Activity](https://img.shields.io/github/commit-activity/m/lightpanda-io/browser)](https://github.com/lightpanda-io/browser/commits/main)
+[![License](https://img.shields.io/github/license/lightpanda-io/browser)](https://github.com/lightpanda-io/browser/blob/main/LICENSE)
+[![Twitter Follow](https://img.shields.io/twitter/follow/lightpanda_io)](https://twitter.com/lightpanda_io)
+[![GitHub stars](https://img.shields.io/github/stars/lightpanda-io/browser)](https://github.com/lightpanda-io/browser)
+
+</div>
+
+<div align="center">
+  
+<a href="https://trendshift.io/repositories/12815" target="_blank"><img src="https://trendshift.io/api/badge/repositories/12815" alt="lightpanda-io%2Fbrowser | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+
+</div>
+
+Lightpanda is the open-source browser made for headless usage:
+
+- Javascript execution
+- Support of Web APIs (partial, WIP)
+- Compatible with Playwright, Puppeteer through CDP (WIP)
+
+Fast web automation for AI agents, LLM training, scraping and testing with minimal memory footprint:
+
+- Ultra-low memory footprint (9x less than Chrome)
+- Exceptionally fast execution (11x faster than Chrome) & instant startup
+
+<img width=500px src="https://cdn.lightpanda.io/assets/images/benchmark_2024-12-04.png">
+
+See [benchmark details](https://github.com/lightpanda-io/demo).
+
+## Quick start
+
+### Install from the nightly builds
+
+You can download the last binary from the [nightly
+builds](https://github.com/lightpanda-io/browser/releases/tag/nightly) for
+Linux x86_64 and MacOS aarch64.
+
+```console
+# Download the binary
+$ wget https://github.com/lightpanda-io/browser/releases/download/nightly/lightpanda-x86_64-linux
+$ chmod a+x ./lightpanda-x86_64-linux
+$ ./lightpanda-x86_64-linux -h
+usage: ./lightpanda-x86_64-linux [options] [URL]
+
+  start Lightpanda browser
+
+  * if an url is provided the browser will fetch the page and exit
+  * otherwhise the browser starts a CDP server
+
+  -h, --help      Print this help message and exit.
+  --host          Host of the CDP server (default "127.0.0.1")
+  --port          Port of the CDP server (default "9222")
+  --timeout       Timeout for incoming connections of the CDP server (in seconds, default "3")
+  --dump          Dump document in stdout (fetch mode only)
+```
+
+### Dump an URL
+
+```console
+$ ./lightpanda-x86_64-linux --dump https://lightpanda.io
+info(browser): GET https://lightpanda.io/ http.Status.ok
+info(browser): fetch script https://api.website.lightpanda.io/js/script.js: http.Status.ok
+info(browser): eval remote https://api.website.lightpanda.io/js/script.js: TypeError: Cannot read properties of undefined (reading 'pushState')
+<!DOCTYPE html>
+```
+
+### Start a CDP server
+
+```console
+$ ./lightpanda-x86_64-linux --host 127.0.0.1 --port 9222
+info(websocket): starting blocking worker to listen on 127.0.0.1:9222
+info(server): accepting new conn...
+```
+
+Once the CDP server started, you can run a Puppeteer script by configuring the
+`browserWSEndpoint`.
+
+```js
+'use strict'
+
+import puppeteer from 'puppeteer-core';
+
+// use browserWSEndpoint to pass the Lightpanda's CDP server address.
+const browser = await puppeteer.connect({
+  browserWSEndpoint: "ws://127.0.0.1:9222",
+});
+
+// The rest of your script remains the same.
+const context = await browser.createBrowserContext();
+const page = await context.newPage();
+
+await page.goto('https://wikipedia.com/');
+
+await page.close();
+await context.close();
+```
+
+## Build from sources
 
 ### Prerequisites
 
-Browsercore is written with [Zig](https://ziglang.org/) `0.11.0`. You have to
+Lightpanda is written with [Zig](https://ziglang.org/) `0.13.0`. You have to
 install it with the right version in order to build the project.
 
-Browsercore also depends on
-[js-runtimelib](https://github.com/francisbouvier/jsruntime-lib/) and
-[lexbor](https://github.com/lexbor/lexbor) libs.
+Lightpanda also depends on
+[zig-js-runtime](https://github.com/lightpanda-io/zig-js-runtime/) (with v8),
+[Netsurf libs](https://www.netsurf-browser.org/) and
+[Mimalloc](https://microsoft.github.io/mimalloc).
 
-To be able to build the v8 engine for js-runtimelib, you have to install some libs:
+To be able to build the v8 engine for zig-js-runtime, you have to install some libs:
 
 For Debian/Ubuntu based Linux:
+
 ```
 sudo apt install xz-utils \
     python3 ca-certificates git \
@@ -22,94 +128,157 @@ sudo apt install xz-utils \
     cmake clang
 ```
 
-For MacOS, you only need Python 3 and cmake.
+For MacOS, you only need cmake:
 
-To be able to build lexbor, you need to install also `cmake`.
+```
+brew install cmake
+```
 
 ### Install and build dependencies
 
-The project uses git submodule for dependencies.
-The `make install-submodule` will init and update the submodules in the `vendor/`
-directory.
+#### All in one build
+
+You can run `make install` to install deps all in one (or `make install-dev` if you need the development versions).
+
+Be aware that the build task is very long and cpu consuming, as you will build from sources all dependencies, including the v8 Javascript engine.
+
+#### Step by step build dependency
+
+The project uses git submodules for dependencies.
+
+To init or update the submodules in the `vendor/` directory:
 
 ```
 make install-submodule
 ```
 
-### Build netsurf
+**Netsurf libs**
+
+Netsurf libs are used for HTML parsing and DOM tree generation.
 
-The command `make install-netsurf` will build netsurf libs used by browsercore.
 ```
 make install-netsurf
 ```
 
-### Build lexbor
+For dev env, use `make install-netsurf-dev`.
 
-The command `make install-lexbor` will build lexbor lib used by browsercore.
-```
-make install-lexbor
-```
+**Mimalloc**
 
-### Build jsruntime-lib
-
-The command `make install-jsruntime-dev` uses jsruntime-lib's `zig-v8` dependency to build v8 engine lib.
-Be aware the build task is very long and cpu consuming.
-
-Build v8 engine for debug/dev version, it creates
-`vendor/jsruntime-lib/vendor/v8/$ARCH/debug/libc_v8.a` file.
+Mimalloc is used as a C memory allocator.
 
 ```
-make install-jsruntime-dev
+make install-mimalloc
 ```
 
-You should also build a release vesion of v8 with:
+For dev env, use `make install-mimalloc-dev`.
+
+Note: when Mimalloc is built in dev mode, you can dump memory stats with the
+env var `MIMALLOC_SHOW_STATS=1`. See
+[https://microsoft.github.io/mimalloc/environment.html](https://microsoft.github.io/mimalloc/environment.html).
+
+**zig-js-runtime**
+
+Our own Zig/Javascript runtime, which includes the v8 Javascript engine.
+
+This build task is very long and cpu consuming, as you will build v8 from sources.
 
 ```
-make install-jsruntime
+make install-zig-js-runtime
 ```
 
-### All in one build
-
-You can run `make intall` and `make install-dev` to install deps all in one.
+For dev env, use `make install-zig-js-runtime-dev`.
 
 ## Test
 
 ### Unit Tests
 
-You can test browsercore by running `make test`.
+You can test Lightpanda by running `make test`.
 
 ### Web Platform Tests
 
-Browsercore is tested against the standardized [Web Platform
+Lightpanda is tested against the standardized [Web Platform
 Tests](https://web-platform-tests.org/).
 
-The relevant tests cases for Browsercore are commit with the project.
-All the tests cases executed are located in `tests/wpt` dir and come from an
-external repository: https://github.com/lightpanda-io/wpt
+The relevant tests cases are committed in a [dedicated repository](https://github.com/lightpanda-io/wpt) which is fetched by the `make install-submodule` command.
+
+All the tests cases executed are located in the `tests/wpt` sub-directory.
 
 For reference, you can easily execute a WPT test case with your browser via
 [wpt.live](https://wpt.live).
 
-*Run WPT test suite*
+#### Run WPT test suite
+
+To run all the tests:
 
-You can run all the test.
-The runner execute all the tests ending with `.html`.
 ```
 make wpt
 ```
 
-Or one specific test by using a suffix.
+Or one specific test:
+
 ```
 make wpt Node-childNodes.html
 ```
 
-*Add a new WPT test case*
+#### Add a new WPT test case
 
-We add new tests cases files with implemented changes in Browsercore.
+We add new relevant tests cases files when we implemented changes in Lightpanda.
 
-Copy the test case you want to add from the [WPT
-repo](https://github.com/web-platform-tests/wpt) into `tests/wpt` dir, commit
-the files in the https://github.com/lightpanda-io/wpt repository and update the
-git submodule in browsercore.
+To add a new test, copy the file you want from the [WPT
+repo](https://github.com/web-platform-tests/wpt) into the `tests/wpt` directory.
 
-:warning: Please keep the original directory tree structure into `tests/wpt`.
+:warning: Please keep the original directory tree structure of `tests/wpt`.
+
+## Contributing
+
+Lightpanda accepts pull requests through GitHub.
+
+You have to sign our [CLA](CLA.md) during the pull request process otherwise
+we're not able to accept your contributions.
+
+## Why?
+
+### Javascript execution is mandatory for the modern web
+
+In the good old days, scraping a webpage was as easy as making an HTTP request, cURL-like. It’s not possible anymore, because Javascript is everywhere, like it or not:
+
+- Ajax, Single Page App, infinite loading, “click to display”, instant search, etc.
+- JS web frameworks: React, Vue, Angular & others
+
+### Chrome is not the right tool
+
+If we need Javascript, why not use a real web browser? Take a huge desktop application, hack it, and run it on the server. Hundreds or thousands of instances of Chrome if you use it at scale. Are you sure it’s such a good idea?
+
+- Heavy on RAM and CPU, expensive to run
+- Hard to package, deploy and maintain at scale
+- Bloated, lots of features are not useful in headless usage
+
+### Lightpanda is built for performance
+
+If we want both Javascript and performance in a true headless browser, we need to start from scratch. Not another iteration of Chromium, really from a blank page. Crazy right? But that’s what we did:
+
+- Not based on Chromium, Blink or WebKit
+- Low-level system programming language (Zig) with optimisations in mind
+- Opinionated: without graphical rendering
+
+## Status
+
+Lightpanda is still a work in progress and is currently at a Beta stage.
+
+:warning: You should expect most websites to fail or crash.
+
+Here are the key features we have implemented:
+
+- [x] HTTP loader
+- [x] HTML parser and DOM tree (based on Netsurf libs)
+- [x] Javascript support (v8)
+- [x] Basic DOM APIs
+- [x] Ajax
+  - [x] XHR API
+  - [x] Fetch API
+- [x] DOM dump
+- [x] Basic CDP/websockets server
+
+NOTE: There are hundreds of Web APIs. Developing a browser (even just for headless mode) is a huge task. Coverage will increase over time.
+
+You can also follow the progress of our Javascript support in our dedicated [zig-js-runtime](https://github.com/lightpanda-io/zig-js-runtime#development) project.

build.zig

@@ -1,16 +1,34 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const builtin = @import("builtin");
 
-const jsruntime_path = "vendor/jsruntime-lib/";
-const jsruntime = @import("vendor/jsruntime-lib/build.zig");
+const jsruntime_path = "vendor/zig-js-runtime/";
+const jsruntime = @import("vendor/zig-js-runtime/build.zig");
 const jsruntime_pkgs = jsruntime.packages(jsruntime_path);
 
 /// Do not rename this constant. It is scanned by some scripts to determine
 /// which zig version to install.
 const recommended_zig_version = jsruntime.recommended_zig_version;
 
-pub fn build(b: *std.build.Builder) !void {
+pub fn build(b: *std.Build) !void {
     switch (comptime builtin.zig_version.order(std.SemanticVersion.parse(recommended_zig_version) catch unreachable)) {
         .eq => {},
         .lt => {
@@ -27,24 +45,23 @@ pub fn build(b: *std.build.Builder) !void {
     const target = b.standardTargetOptions(.{});
     const mode = b.standardOptimizeOption(.{});
 
-    const options = try jsruntime.buildOptions(b);
+    const options = jsruntime.buildOptions(b);
 
     // browser
     // -------
 
     // compile and install
     const exe = b.addExecutable(.{
-        .name = "browsercore",
-        .root_source_file = .{ .path = "src/main.zig" },
+        .name = "lightpanda",
+        .root_source_file = b.path("src/main.zig"),
         .target = target,
         .optimize = mode,
     });
-    try common(exe, options);
+    try common(b, exe, options);
     b.installArtifact(exe);
 
     // run
     const run_cmd = b.addRunArtifact(exe);
-    run_cmd.step.dependOn(b.getInstallStep());
     if (b.args) |args| {
         run_cmd.addArgs(args);
     }
@@ -58,19 +75,16 @@ pub fn build(b: *std.build.Builder) !void {
 
     // compile and install
     const shell = b.addExecutable(.{
-        .name = "browsercore-shell",
-        .root_source_file = .{ .path = "src/main_shell.zig" },
+        .name = "lightpanda-shell",
+        .root_source_file = b.path("src/main_shell.zig"),
         .target = target,
         .optimize = mode,
     });
-    try common(shell, options);
+    try common(b, shell, options);
     try jsruntime_pkgs.add_shell(shell);
-    // do not install shell binary
-    b.installArtifact(shell);
 
     // run
     const shell_cmd = b.addRunArtifact(shell);
-    shell_cmd.step.dependOn(b.getInstallStep());
     if (b.args) |args| {
         shell_cmd.addArgs(args);
     }
@@ -83,87 +97,135 @@ pub fn build(b: *std.build.Builder) !void {
     // ----
 
     // compile
-    const tests = b.addTest(.{ .root_source_file = .{ .path = "src/run_tests.zig" } });
-    try common(tests, options);
-    tests.single_threaded = true;
-    tests.test_runner = "src/test_runner.zig";
+    const tests = b.addTest(.{
+        .root_source_file = b.path("src/main_tests.zig"),
+        .test_runner = b.path("src/main_tests.zig"),
+        .target = target,
+        .optimize = mode,
+    });
+    try common(b, tests, options);
+
+    // add jsruntime pretty deps
+    tests.root_module.addAnonymousImport("pretty", .{
+        .root_source_file = b.path("vendor/zig-js-runtime/src/pretty.zig"),
+    });
+
     const run_tests = b.addRunArtifact(tests);
+    if (b.args) |args| {
+        run_tests.addArgs(args);
+    }
 
     // step
     const test_step = b.step("test", "Run unit tests");
     test_step.dependOn(&run_tests.step);
 
+    // unittest
+    // ----
+
+    // compile
+    const unit_tests = b.addTest(.{
+        .root_source_file = b.path("src/unit_tests.zig"),
+        .test_runner = b.path("src/unit_tests.zig"),
+        .target = target,
+        .optimize = mode,
+    });
+    try common(b, unit_tests, options);
+
+    const run_unit_tests = b.addRunArtifact(unit_tests);
+    if (b.args) |args| {
+        run_unit_tests.addArgs(args);
+    }
+
+    // step
+    const unit_test_step = b.step("unittest", "Run unit tests");
+    unit_test_step.dependOn(&run_unit_tests.step);
+
     // wpt
     // -----
 
     // compile and install
     const wpt = b.addExecutable(.{
-        .name = "browsercore-wpt",
-        .root_source_file = .{ .path = "src/main_wpt.zig" },
+        .name = "lightpanda-wpt",
+        .root_source_file = b.path("src/main_wpt.zig"),
         .target = target,
         .optimize = mode,
     });
-    try common(wpt, options);
-    b.installArtifact(wpt);
+    try common(b, wpt, options);
 
     // run
     const wpt_cmd = b.addRunArtifact(wpt);
-    wpt_cmd.step.dependOn(b.getInstallStep());
     if (b.args) |args| {
         wpt_cmd.addArgs(args);
     }
     // step
     const wpt_step = b.step("wpt", "WPT tests");
     wpt_step.dependOn(&wpt_cmd.step);
-
-    // get
-    // -----
-
-    // compile and install
-    const get = b.addExecutable(.{
-        .name = "browsercore-get",
-        .root_source_file = .{ .path = "src/main_get.zig" },
-        .target = target,
-        .optimize = mode,
-    });
-    try common(get, options);
-    b.installArtifact(get);
-
-    // run
-    const get_cmd = b.addRunArtifact(get);
-    get_cmd.step.dependOn(b.getInstallStep());
-    if (b.args) |args| {
-        get_cmd.addArgs(args);
-    }
-    // step
-    const get_step = b.step("get", "request URL");
-    get_step.dependOn(&get_cmd.step);
 }
 
 fn common(
-    step: *std.Build.CompileStep,
+    b: *std.Build,
+    step: *std.Build.Step.Compile,
     options: jsruntime.Options,
 ) !void {
-    try jsruntime_pkgs.add(step, options);
-    linkLexbor(step);
-    linkNetSurf(step);
+    const target = step.root_module.resolved_target.?;
+    const jsruntimemod = try jsruntime_pkgs.module(
+        b,
+        options,
+        step.root_module.optimize.?,
+        target,
+    );
+    step.root_module.addImport("jsruntime", jsruntimemod);
+
+    const netsurf = try moduleNetSurf(b, target);
+    netsurf.addImport("jsruntime", jsruntimemod);
+    step.root_module.addImport("netsurf", netsurf);
+
+    const asyncio = b.addModule("asyncio", .{
+        .root_source_file = b.path("vendor/zig-async-io/src/lib.zig"),
+    });
+    step.root_module.addImport("asyncio", asyncio);
+
+    const tlsmod = b.addModule("tls", .{
+        .root_source_file = b.path("vendor/tls.zig/src/main.zig"),
+    });
+    step.root_module.addImport("tls", tlsmod);
 }
 
-fn linkLexbor(step: *std.build.LibExeObjStep) void {
-    // cmake . -DLEXBOR_BUILD_SHARED=OFF
-    const lib_path = "vendor/lexbor/liblexbor_static.a";
-    step.addObjectFile(.{ .path = lib_path });
-    step.addIncludePath(.{ .path = "vendor/lexbor-src/source" });
-}
+fn moduleNetSurf(b: *std.Build, target: std.Build.ResolvedTarget) !*std.Build.Module {
+    const mod = b.addModule("netsurf", .{
+        .root_source_file = b.path("src/netsurf/netsurf.zig"),
+        .target = target,
+    });
 
-fn linkNetSurf(step: *std.build.LibExeObjStep) void {
+    const os = target.result.os.tag;
+    const arch = target.result.cpu.arch;
 
     // iconv
-    step.addObjectFile(.{ .path = "vendor/libiconv/lib/libiconv.a" });
-    step.addIncludePath(.{ .path = "vendor/libiconv/include" });
+    const libiconv_lib_path = try std.fmt.allocPrint(
+        mod.owner.allocator,
+        "vendor/libiconv/out/{s}-{s}/lib/libiconv.a",
+        .{ @tagName(os), @tagName(arch) },
+    );
+    const libiconv_include_path = try std.fmt.allocPrint(
+        mod.owner.allocator,
+        "vendor/libiconv/out/{s}-{s}/lib/libiconv.a",
+        .{ @tagName(os), @tagName(arch) },
+    );
+    mod.addObjectFile(b.path(libiconv_lib_path));
+    mod.addIncludePath(b.path(libiconv_include_path));
+
+    // mimalloc
+    mod.addImport("mimalloc", (try moduleMimalloc(b, target)));
 
     // netsurf libs
-    const ns = "vendor/netsurf/";
+    const ns = "vendor/netsurf";
+    const ns_include_path = try std.fmt.allocPrint(
+        mod.owner.allocator,
+        ns ++ "/out/{s}-{s}/include",
+        .{ @tagName(os), @tagName(arch) },
+    );
+    mod.addIncludePath(b.path(ns_include_path));
+
     const libs: [4][]const u8 = .{
         "libdom",
         "libhubbub",
@@ -171,8 +233,35 @@ fn linkNetSurf(step: *std.build.LibExeObjStep) void {
         "libwapcaplet",
     };
     inline for (libs) |lib| {
-        step.addObjectFile(.{ .path = ns ++ "/lib/" ++ lib ++ ".a" });
-        step.addIncludePath(.{ .path = ns ++ lib ++ "/src" });
+        const ns_lib_path = try std.fmt.allocPrint(
+            mod.owner.allocator,
+            ns ++ "/out/{s}-{s}/lib/" ++ lib ++ ".a",
+            .{ @tagName(os), @tagName(arch) },
+        );
+        mod.addObjectFile(b.path(ns_lib_path));
+        mod.addIncludePath(b.path(ns ++ "/" ++ lib ++ "/src"));
     }
-    step.addIncludePath(.{ .path = ns ++ "/include" });
+
+    return mod;
+}
+
+fn moduleMimalloc(b: *std.Build, target: std.Build.ResolvedTarget) !*std.Build.Module {
+    const mod = b.addModule("mimalloc", .{
+        .root_source_file = b.path("src/mimalloc/mimalloc.zig"),
+        .target = target,
+    });
+
+    const os = target.result.os.tag;
+    const arch = target.result.cpu.arch;
+
+    const mimalloc = "vendor/mimalloc";
+    const lib_path = try std.fmt.allocPrint(
+        mod.owner.allocator,
+        mimalloc ++ "/out/{s}-{s}/lib/libmimalloc.a",
+        .{ @tagName(os), @tagName(arch) },
+    );
+    mod.addObjectFile(b.path(lib_path));
+    mod.addIncludePath(b.path(mimalloc ++ "/include"));
+
+    return mod;
 }

src/apiweb.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const generate = @import("generate.zig");
 
 const Console = @import("jsruntime").Console;
@@ -6,6 +24,10 @@ const DOM = @import("dom/dom.zig");
 const HTML = @import("html/html.zig");
 const Events = @import("events/event.zig");
 const XHR = @import("xhr/xhr.zig");
+const Storage = @import("storage/storage.zig");
+const URL = @import("url/url.zig");
+const Iterators = @import("iterator/iterator.zig");
+const XMLSerializer = @import("xmlserializer/xmlserializer.zig");
 
 pub const HTMLDocument = @import("html/document.zig").HTMLDocument;
 
@@ -16,4 +38,10 @@ pub const Interfaces = generate.Tuple(.{
     Events.Interfaces,
     HTML.Interfaces,
     XHR.Interfaces,
-});
+    Storage.Interfaces,
+    URL.Interfaces,
+    Iterators.Interfaces,
+    XMLSerializer.Interfaces,
+}){};
+
+pub const UserContext = @import("user_context.zig").UserContext;

src/async/stream.zig

@@ -1,115 +0,0 @@
-const std = @import("std");
-const builtin = @import("builtin");
-const os = std.os;
-const io = std.io;
-const assert = std.debug.assert;
-
-const tcp = @import("tcp.zig");
-
-pub const Stream = struct {
-    alloc: std.mem.Allocator,
-    conn: *tcp.Conn,
-
-    handle: std.os.socket_t,
-
-    pub fn close(self: Stream) void {
-        os.closeSocket(self.handle);
-        self.alloc.destroy(self.conn);
-    }
-
-    pub const ReadError = os.ReadError;
-    pub const WriteError = os.WriteError;
-
-    pub const Reader = io.Reader(Stream, ReadError, read);
-    pub const Writer = io.Writer(Stream, WriteError, write);
-
-    pub fn reader(self: Stream) Reader {
-        return .{ .context = self };
-    }
-
-    pub fn writer(self: Stream) Writer {
-        return .{ .context = self };
-    }
-
-    pub fn read(self: Stream, buffer: []u8) ReadError!usize {
-        return self.conn.receive(self.handle, buffer) catch |err| switch (err) {
-            else => return error.Unexpected,
-        };
-    }
-
-    pub fn readv(s: Stream, iovecs: []const os.iovec) ReadError!usize {
-        return os.readv(s.handle, iovecs);
-    }
-
-    /// Returns the number of bytes read. If the number read is smaller than
-    /// `buffer.len`, it means the stream reached the end. Reaching the end of
-    /// a stream is not an error condition.
-    pub fn readAll(s: Stream, buffer: []u8) ReadError!usize {
-        return readAtLeast(s, buffer, buffer.len);
-    }
-
-    /// Returns the number of bytes read, calling the underlying read function
-    /// the minimal number of times until the buffer has at least `len` bytes
-    /// filled. If the number read is less than `len` it means the stream
-    /// reached the end. Reaching the end of the stream is not an error
-    /// condition.
-    pub fn readAtLeast(s: Stream, buffer: []u8, len: usize) ReadError!usize {
-        assert(len <= buffer.len);
-        var index: usize = 0;
-        while (index < len) {
-            const amt = try s.read(buffer[index..]);
-            if (amt == 0) break;
-            index += amt;
-        }
-        return index;
-    }
-
-    /// TODO in evented I/O mode, this implementation incorrectly uses the event loop's
-    /// file system thread instead of non-blocking. It needs to be reworked to properly
-    /// use non-blocking I/O.
-    pub fn write(self: Stream, buffer: []const u8) WriteError!usize {
-        return self.conn.send(self.handle, buffer) catch |err| switch (err) {
-            error.AccessDenied => error.AccessDenied,
-            error.WouldBlock => error.WouldBlock,
-            error.ConnectionResetByPeer => error.ConnectionResetByPeer,
-            error.MessageTooBig => error.FileTooBig,
-            error.BrokenPipe => error.BrokenPipe,
-            else => return error.Unexpected,
-        };
-    }
-
-    pub fn writeAll(self: Stream, bytes: []const u8) WriteError!void {
-        var index: usize = 0;
-        while (index < bytes.len) {
-            index += try self.write(bytes[index..]);
-        }
-    }
-
-    /// See https://github.com/ziglang/zig/issues/7699
-    /// See equivalent function: `std.fs.File.writev`.
-    pub fn writev(self: Stream, iovecs: []const os.iovec_const) WriteError!usize {
-        if (iovecs.len == 0) return 0;
-        const first_buffer = iovecs[0].iov_base[0..iovecs[0].iov_len];
-        return try self.write(first_buffer);
-    }
-
-    /// The `iovecs` parameter is mutable because this function needs to mutate the fields in
-    /// order to handle partial writes from the underlying OS layer.
-    /// See https://github.com/ziglang/zig/issues/7699
-    /// See equivalent function: `std.fs.File.writevAll`.
-    pub fn writevAll(self: Stream, iovecs: []os.iovec_const) WriteError!void {
-        if (iovecs.len == 0) return;
-
-        var i: usize = 0;
-        while (true) {
-            var amt = try self.writev(iovecs[i..]);
-            while (amt >= iovecs[i].iov_len) {
-                amt -= iovecs[i].iov_len;
-                i += 1;
-                if (i >= iovecs.len) return;
-            }
-            iovecs[i].iov_base += amt;
-            iovecs[i].iov_len -= amt;
-        }
-    }
-};

src/async/tcp.zig

@@ -1,94 +0,0 @@
-const std = @import("std");
-const net = std.net;
-const Stream = @import("stream.zig").Stream;
-const Loop = @import("jsruntime").Loop;
-const NetworkImpl = Loop.Network(Conn.Command);
-
-// Conn is a TCP connection using jsruntime Loop async I/O.
-// connect, send and receive are blocking, but use async I/O in the background.
-// Client doesn't own the socket used for the connection, the caller is
-// responsible for closing it.
-pub const Conn = struct {
-    const Command = struct {
-        impl: NetworkImpl,
-
-        done: bool = false,
-        err: ?anyerror = null,
-        ln: usize = 0,
-
-        fn ok(self: *Command, err: ?anyerror, ln: usize) void {
-            self.err = err;
-            self.ln = ln;
-            self.done = true;
-        }
-
-        fn wait(self: *Command) !usize {
-            while (!self.done) try self.impl.tick();
-
-            if (self.err) |err| return err;
-            return self.ln;
-        }
-        pub fn onConnect(self: *Command, err: ?anyerror) void {
-            self.ok(err, 0);
-        }
-        pub fn onSend(self: *Command, ln: usize, err: ?anyerror) void {
-            self.ok(err, ln);
-        }
-        pub fn onReceive(self: *Command, ln: usize, err: ?anyerror) void {
-            self.ok(err, ln);
-        }
-    };
-
-    loop: *Loop,
-
-    pub fn connect(self: *Conn, socket: std.os.socket_t, address: std.net.Address) !void {
-        var cmd = Command{ .impl = NetworkImpl.init(self.loop) };
-        cmd.impl.connect(&cmd, socket, address);
-        _ = try cmd.wait();
-    }
-
-    pub fn send(self: *Conn, socket: std.os.socket_t, buffer: []const u8) !usize {
-        var cmd = Command{ .impl = NetworkImpl.init(self.loop) };
-        cmd.impl.send(&cmd, socket, buffer);
-        return try cmd.wait();
-    }
-
-    pub fn receive(self: *Conn, socket: std.os.socket_t, buffer: []u8) !usize {
-        var cmd = Command{ .impl = NetworkImpl.init(self.loop) };
-        cmd.impl.receive(&cmd, socket, buffer);
-        return try cmd.wait();
-    }
-};
-
-pub fn tcpConnectToHost(alloc: std.mem.Allocator, loop: *Loop, name: []const u8, port: u16) !Stream {
-    // TODO async resolve
-    const list = try net.getAddressList(alloc, name, port);
-    defer list.deinit();
-
-    if (list.addrs.len == 0) return error.UnknownHostName;
-
-    for (list.addrs) |addr| {
-        return tcpConnectToAddress(alloc, loop, addr) catch |err| switch (err) {
-            error.ConnectionRefused => {
-                continue;
-            },
-            else => return err,
-        };
-    }
-    return std.os.ConnectError.ConnectionRefused;
-}
-
-pub fn tcpConnectToAddress(alloc: std.mem.Allocator, loop: *Loop, addr: net.Address) !Stream {
-    const sockfd = try std.os.socket(addr.any.family, std.os.SOCK.STREAM, std.os.IPPROTO.TCP);
-    errdefer std.os.closeSocket(sockfd);
-
-    var conn = try alloc.create(Conn);
-    conn.* = Conn{ .loop = loop };
-    try conn.connect(sockfd, addr);
-
-    return Stream{
-        .alloc = alloc,
-        .conn = conn,
-        .handle = sockfd,
-    };
-}

src/async/test.zig

@@ -1,172 +0,0 @@
-const std = @import("std");
-const http = std.http;
-const Client = @import("Client.zig");
-const Request = @import("Client.zig").Request;
-
-pub const Loop = @import("jsruntime").Loop;
-
-const url = "https://w3.org";
-
-test "blocking mode fetch API" {
-    const alloc = std.testing.allocator;
-
-    var loop = try Loop.init(alloc);
-    defer loop.deinit();
-
-    var client: Client = .{
-        .allocator = alloc,
-        .loop = &loop,
-    };
-    defer client.deinit();
-
-    // force client's CA cert scan from system.
-    try client.ca_bundle.rescan(client.allocator);
-
-    var res = try client.fetch(alloc, .{
-        .location = .{ .uri = try std.Uri.parse(url) },
-        .payload = .none,
-    });
-    defer res.deinit();
-
-    try std.testing.expect(res.status == .ok);
-}
-
-test "blocking mode open/send/wait API" {
-    const alloc = std.testing.allocator;
-
-    var loop = try Loop.init(alloc);
-    defer loop.deinit();
-
-    var client: Client = .{
-        .allocator = alloc,
-        .loop = &loop,
-    };
-    defer client.deinit();
-
-    // force client's CA cert scan from system.
-    try client.ca_bundle.rescan(client.allocator);
-
-    var headers = try std.http.Headers.initList(alloc, &[_]std.http.Field{});
-    defer headers.deinit();
-
-    var req = try client.open(.GET, try std.Uri.parse(url), headers, .{});
-    defer req.deinit();
-
-    try req.send(.{});
-    try req.finish();
-    try req.wait();
-
-    try std.testing.expect(req.response.status == .ok);
-}
-
-// Example how to write an async http client using the modified standard client.
-const AsyncClient = struct {
-    cli: Client,
-
-    const YieldImpl = Loop.Yield(AsyncRequest);
-    const AsyncRequest = struct {
-        const State = enum { new, open, send, finish, wait, done };
-
-        cli: *Client,
-        uri: std.Uri,
-        headers: std.http.Headers,
-
-        req: ?Request = undefined,
-        state: State = .new,
-
-        impl: YieldImpl,
-        err: ?anyerror = null,
-
-        pub fn deinit(self: *AsyncRequest) void {
-            if (self.req) |*r| r.deinit();
-            self.headers.deinit();
-        }
-
-        pub fn fetch(self: *AsyncRequest) void {
-            self.state = .new;
-            return self.impl.yield(self);
-        }
-
-        fn onerr(self: *AsyncRequest, err: anyerror) void {
-            self.state = .done;
-            self.err = err;
-        }
-
-        pub fn onYield(self: *AsyncRequest, err: ?anyerror) void {
-            if (err) |e| return self.onerr(e);
-
-            switch (self.state) {
-                .new => {
-                    self.state = .open;
-                    self.req = self.cli.open(.GET, self.uri, self.headers, .{}) catch |e| return self.onerr(e);
-                },
-                .open => {
-                    self.state = .send;
-                    self.req.?.send(.{}) catch |e| return self.onerr(e);
-                },
-                .send => {
-                    self.state = .finish;
-                    self.req.?.finish() catch |e| return self.onerr(e);
-                },
-                .finish => {
-                    self.state = .wait;
-                    self.req.?.wait() catch |e| return self.onerr(e);
-                },
-                .wait => {
-                    self.state = .done;
-                    return;
-                },
-                .done => return,
-            }
-
-            return self.impl.yield(self);
-        }
-
-        pub fn wait(self: *AsyncRequest) !void {
-            while (self.state != .done) try self.impl.tick();
-            if (self.err) |err| return err;
-        }
-    };
-
-    pub fn init(alloc: std.mem.Allocator, loop: *Loop) AsyncClient {
-        return .{
-            .cli = .{
-                .allocator = alloc,
-                .loop = loop,
-            },
-        };
-    }
-
-    pub fn deinit(self: *AsyncClient) void {
-        self.cli.deinit();
-    }
-
-    pub fn createRequest(self: *AsyncClient, uri: std.Uri) !AsyncRequest {
-        return .{
-            .impl = YieldImpl.init(self.cli.loop),
-            .cli = &self.cli,
-            .uri = uri,
-            .headers = .{ .allocator = self.cli.allocator, .owned = false },
-        };
-    }
-};
-
-test "non blocking client" {
-    const alloc = std.testing.allocator;
-
-    var loop = try Loop.init(alloc);
-    defer loop.deinit();
-
-    var client = AsyncClient.init(alloc, &loop);
-    defer client.deinit();
-
-    var reqs: [3]AsyncClient.AsyncRequest = undefined;
-    for (0..reqs.len) |i| {
-        reqs[i] = try client.createRequest(try std.Uri.parse(url));
-        reqs[i].fetch();
-    }
-    for (0..reqs.len) |i| {
-        try reqs[i].wait();
-        reqs[i].deinit();
-    }
-}

src/browser/browser.zig

@@ -1,48 +1,92 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
+const builtin = @import("builtin");
 
 const Types = @import("root").Types;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 const Loader = @import("loader.zig").Loader;
 const Dump = @import("dump.zig");
-const Mime = @import("mime.zig");
+const Mime = @import("mime.zig").Mime;
 
 const jsruntime = @import("jsruntime");
 const Loop = jsruntime.Loop;
 const Env = jsruntime.Env;
+const Module = jsruntime.Module;
 
 const apiweb = @import("../apiweb.zig");
 
 const Window = @import("../html/window.zig").Window;
 const Walker = @import("../dom/walker.zig").WalkerDepthFirst;
 
-const FetchResult = std.http.Client.FetchResult;
+const URL = @import("../url/url.zig").URL;
+const Location = @import("../html/location.zig").Location;
+
+const storage = @import("../storage/storage.zig");
+
+const FetchResult = @import("../http/Client.zig").Client.FetchResult;
+
+const UserContext = @import("../user_context.zig").UserContext;
+const HttpClient = @import("asyncio").Client;
+
+const polyfill = @import("../polyfill/polyfill.zig");
 
 const log = std.log.scoped(.browser);
 
+pub const user_agent = "Lightpanda/1.0";
+
 // Browser is an instance of the browser.
 // You can create multiple browser instances.
 // A browser contains only one session.
 // TODO allow multiple sessions per browser.
 pub const Browser = struct {
-    session: *Session,
+    session: Session = undefined,
+    agent: []const u8 = user_agent,
 
-    pub fn init(alloc: std.mem.Allocator, vm: jsruntime.VM) !Browser {
+    const uri = "about:blank";
+
+    pub fn init(self: *Browser, alloc: std.mem.Allocator, loop: *Loop, vm: jsruntime.VM) !void {
         // We want to ensure the caller initialised a VM, but the browser
         // doesn't use it directly...
         _ = vm;
 
-        return Browser{
-            .session = try Session.init(alloc, "about:blank"),
-        };
+        try Session.init(&self.session, alloc, loop, uri);
     }
 
     pub fn deinit(self: *Browser) void {
         self.session.deinit();
     }
 
-    pub fn currentSession(self: *Browser) *Session {
-        return self.session;
+    pub fn newSession(
+        self: *Browser,
+        alloc: std.mem.Allocator,
+        loop: *jsruntime.Loop,
+    ) !void {
+        self.session.deinit();
+        try Session.init(&self.session, alloc, loop, uri);
+    }
+
+    pub fn currentPage(self: *Browser) ?*Page {
+        if (self.session.page == null) return null;
+
+        return &self.session.page.?;
     }
 };
 
@@ -66,39 +110,90 @@ pub const Session = struct {
     // TODO handle proxy
     loader: Loader,
     env: Env = undefined,
-    loop: Loop,
+    inspector: ?jsruntime.Inspector = null,
+
     window: Window,
 
+    // TODO move the shed to the browser?
+    storageShed: storage.Shed,
+    page: ?Page = null,
+    httpClient: HttpClient,
+
     jstypes: [Types.len]usize = undefined,
 
-    fn init(alloc: std.mem.Allocator, uri: []const u8) !*Session {
-        var self = try alloc.create(Session);
+    fn init(self: *Session, alloc: std.mem.Allocator, loop: *Loop, uri: []const u8) !void {
         self.* = Session{
             .uri = uri,
             .alloc = alloc,
             .arena = std.heap.ArenaAllocator.init(alloc),
-            .window = Window.create(null),
+            .window = Window.create(null, .{ .agent = user_agent }),
             .loader = Loader.init(alloc),
-            .loop = try Loop.init(alloc),
+            .storageShed = storage.Shed.init(alloc),
+            .httpClient = undefined,
         };
 
-        self.env = try Env.init(self.arena.allocator(), &self.loop);
+        Env.init(&self.env, self.arena.allocator(), loop, null);
+        self.httpClient = .{ .allocator = alloc };
         try self.env.load(&self.jstypes);
+    }
 
-        return self;
+    fn fetchModule(ctx: *anyopaque, referrer: ?jsruntime.Module, specifier: []const u8) !jsruntime.Module {
+        _ = referrer;
+
+        const self: *Session = @ptrCast(@alignCast(ctx));
+
+        if (self.page == null) return error.NoPage;
+
+        log.debug("fetch module: specifier: {s}", .{specifier});
+        const alloc = self.arena.allocator();
+        const body = try self.page.?.fetchData(alloc, specifier);
+        defer alloc.free(body);
+
+        return self.env.compileModule(body, specifier);
     }
 
     fn deinit(self: *Session) void {
+        if (self.page) |*p| p.deinit();
+
+        if (self.inspector) |inspector| {
+            inspector.deinit(self.alloc);
+        }
+
         self.env.deinit();
         self.arena.deinit();
 
+        self.httpClient.deinit();
         self.loader.deinit();
-        self.loop.deinit();
-        self.alloc.destroy(self);
+        self.storageShed.deinit();
     }
 
-    pub fn createPage(self: *Session) !Page {
-        return Page.init(self.alloc, self);
+    pub fn initInspector(
+        self: *Session,
+        ctx: anytype,
+        onResp: jsruntime.InspectorOnResponseFn,
+        onEvent: jsruntime.InspectorOnEventFn,
+    ) !void {
+        const ctx_opaque = @as(*anyopaque, @ptrCast(ctx));
+        self.inspector = try jsruntime.Inspector.init(self.alloc, self.env, ctx_opaque, onResp, onEvent);
+        self.env.setInspector(self.inspector.?);
+    }
+
+    pub fn callInspector(self: *Session, msg: []const u8) void {
+        if (self.inspector) |inspector| {
+            inspector.send(msg, self.env);
+        } else {
+            @panic("No Inspector");
+        }
+    }
+
+    // NOTE: the caller is not the owner of the returned value,
+    // the pointer on Page is just returned as a convenience
+    pub fn createPage(self: *Session) !*Page {
+        if (self.page != null) return error.SessionPageExists;
+        const p: Page = undefined;
+        self.page = p;
+        Page.init(&self.page.?, self.alloc, self);
+        return &self.page.?;
     }
 };
 
@@ -115,29 +210,73 @@ pub const Page = struct {
     // handle url
     rawuri: ?[]const u8 = null,
     uri: std.Uri = undefined,
+    origin: ?[]const u8 = null,
+
+    // html url and location
+    url: ?URL = null,
+    location: Location = .{},
 
     raw_data: ?[]const u8 = null,
 
     fn init(
+        self: *Page,
         alloc: std.mem.Allocator,
         session: *Session,
-    ) Page {
-        return Page{
+    ) void {
+        self.* = .{
             .arena = std.heap.ArenaAllocator.init(alloc),
             .session = session,
         };
     }
 
+    // start js env.
+    // - auxData: extra data forwarded to the Inspector
+    // see Inspector.contextCreated
+    pub fn start(self: *Page, auxData: ?[]const u8) !void {
+        // start JS env
+        log.debug("start js env", .{});
+        try self.session.env.start();
+
+        // register the module loader
+        try self.session.env.setModuleLoadFn(self.session, Session.fetchModule);
+
+        // add global objects
+        log.debug("setup global env", .{});
+        try self.session.env.bindGlobal(&self.session.window);
+
+        // load polyfills
+        try polyfill.load(self.arena.allocator(), self.session.env);
+
+        // inspector
+        if (self.session.inspector) |inspector| {
+            log.debug("inspector context created", .{});
+            inspector.contextCreated(self.session.env, "", self.origin orelse "://", auxData);
+        }
+    }
+
     // reset js env and mem arena.
     pub fn end(self: *Page) void {
         self.session.env.stop();
         // TODO unload document: https://html.spec.whatwg.org/#unloading-documents
 
+        if (self.url) |*u| u.deinit(self.arena.allocator());
+        self.url = null;
+        self.location.url = null;
+        self.session.window.replaceLocation(&self.location) catch |e| {
+            log.err("reset window location: {any}", .{e});
+        };
+        self.doc = null;
+
+        // clear netsurf memory arena.
+        parser.deinit();
+
         _ = self.arena.reset(.free_all);
     }
 
     pub fn deinit(self: *Page) void {
+        self.end();
         self.arena.deinit();
+        self.session.page = null;
     }
 
     // dump writes the page content into the given file.
@@ -151,19 +290,61 @@ pub const Page = struct {
         }
 
         // if the page has a pointer to a document, dumps the HTML.
-        try Dump.htmlFile(self.doc.?, out);
+        try Dump.writeHTML(self.doc.?, out);
+    }
+
+    pub fn wait(self: *Page) !void {
+
+        // try catch
+        var try_catch: jsruntime.TryCatch = undefined;
+        try_catch.init(self.session.env);
+        defer try_catch.deinit();
+
+        self.session.env.wait() catch |err| {
+            // the js env could not be started if the document wasn't an HTML.
+            if (err == error.EnvNotStarted) return;
+
+            const alloc = self.arena.allocator();
+            if (try try_catch.err(alloc, self.session.env)) |msg| {
+                defer alloc.free(msg);
+                log.info("wait error: {s}", .{msg});
+                return;
+            }
+        };
+        log.debug("wait: OK", .{});
     }
 
     // spec reference: https://html.spec.whatwg.org/#document-lifecycle
-    pub fn navigate(self: *Page, uri: []const u8) !void {
+    // - auxData: extra data forwarded to the Inspector
+    // see Inspector.contextCreated
+    pub fn navigate(self: *Page, uri: []const u8, auxData: ?[]const u8) !void {
         const alloc = self.arena.allocator();
 
         log.debug("starting GET {s}", .{uri});
 
+        // if the uri is about:blank, nothing to do.
+        if (std.mem.eql(u8, "about:blank", uri)) {
+            return;
+        }
+
         // own the url
         if (self.rawuri) |prev| alloc.free(prev);
         self.rawuri = try alloc.dupe(u8, uri);
-        self.uri = std.Uri.parse(self.rawuri.?) catch try std.Uri.parseWithoutScheme(self.rawuri.?);
+        self.uri = std.Uri.parse(self.rawuri.?) catch try std.Uri.parseAfterScheme("", self.rawuri.?);
+
+        if (self.url) |*prev| prev.deinit(alloc);
+        self.url = try URL.constructor(alloc, self.rawuri.?, null);
+        self.location.url = &self.url.?;
+        try self.session.window.replaceLocation(&self.location);
+
+        // prepare origin value.
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+        try self.uri.writeToStream(.{
+            .scheme = true,
+            .authority = true,
+        }, buf.writer());
+        self.origin = try buf.toOwnedSlice();
 
         // TODO handle fragment in url.
 
@@ -173,33 +354,42 @@ pub const Page = struct {
 
         const req = resp.req;
 
-        log.info("GET {any} {d}", .{ self.uri, req.response.status });
+        log.info("GET {any} {d}", .{ self.uri, @intFromEnum(req.response.status) });
 
         // TODO handle redirection
-        if (req.response.status != .ok) {
-            log.debug("{?} {d} {s}\n{any}", .{
+        log.debug("{?} {d} {s}", .{
             req.response.version,
-                req.response.status,
+            @intFromEnum(req.response.status),
             req.response.reason,
-                req.response.headers,
+            // TODO log headers
         });
-            return error.BadStatusCode;
-        }
 
         // TODO handle charset
         // https://html.spec.whatwg.org/#content-type
-        const ct = req.response.headers.getFirstValue("Content-Type") orelse {
+        var it = req.response.iterateHeaders();
+        var ct: ?[]const u8 = null;
+        while (true) {
+            const h = it.next() orelse break;
+            if (std.ascii.eqlIgnoreCase(h.name, "Content-Type")) {
+                ct = try alloc.dupe(u8, h.value);
+            }
+        }
+        if (ct == null) {
             // no content type in HTTP headers.
             // TODO try to sniff mime type from the body.
             log.info("no content-type HTTP header", .{});
             return;
-        };
-        log.debug("header content-type: {s}", .{ct});
-        const mime = try Mime.parse(ct);
-        if (mime.eql(Mime.HTML)) {
-            try self.loadHTMLDoc(req.reader(), mime.charset orelse "utf-8");
+        }
+        defer alloc.free(ct.?);
+
+        log.debug("header content-type: {s}", .{ct.?});
+        var mime = try Mime.parse(alloc, ct.?);
+        defer mime.deinit();
+
+        if (mime.isHTML()) {
+            try self.loadHTMLDoc(req.reader(), mime.charset orelse "utf-8", auxData);
         } else {
-            log.info("non-HTML document: {s}", .{ct});
+            log.info("non-HTML document: {s}", .{ct.?});
 
             // save the body into the page.
             self.raw_data = try req.reader().readAllAlloc(alloc, 16 * 1024 * 1024);
@@ -207,9 +397,12 @@ pub const Page = struct {
     }
 
     // https://html.spec.whatwg.org/#read-html
-    fn loadHTMLDoc(self: *Page, reader: anytype, charset: []const u8) !void {
+    fn loadHTMLDoc(self: *Page, reader: anytype, charset: []const u8, auxData: ?[]const u8) !void {
         const alloc = self.arena.allocator();
 
+        // start netsurf memory arena.
+        try parser.init();
+
         log.debug("parse html with charset {s}", .{charset});
 
         const ccharset = try alloc.dupeZ(u8, charset);
@@ -224,23 +417,28 @@ pub const Page = struct {
         // TODO set document.readyState to interactive
         // https://html.spec.whatwg.org/#reporting-document-loading-status
 
-        // TODO inject the URL to the document including the fragment.
+        // inject the URL to the document including the fragment.
+        try parser.documentSetDocumentURI(doc, self.rawuri orelse "about:blank");
+
         // TODO set the referrer to the document.
 
-        self.session.window.replaceDocument(doc);
+        try self.session.window.replaceDocument(html_doc);
+        self.session.window.setStorageShelf(
+            try self.session.storageShed.getOrPut(self.origin orelse "null"),
+        );
 
         // https://html.spec.whatwg.org/#read-html
 
-        // start JS env
-        // TODO load the js env concurrently with the HTML parsing.
-        log.debug("start js env", .{});
-        try self.session.env.start(alloc);
+        // inspector
+        if (self.session.inspector) |inspector| {
+            inspector.contextCreated(self.session.env, "", self.origin.?, auxData);
+        }
 
-        // add global objects
-        log.debug("setup global env", .{});
-        try self.session.env.addObject(self.session.window, "window");
-        try self.session.env.addObject(self.session.window, "self");
-        try self.session.env.addObject(html_doc, "document");
+        // replace the user context document with the new one.
+        try self.session.env.setUserContext(.{
+            .document = html_doc,
+            .httpClient = &self.session.httpClient,
+        });
 
         // browse the DOM tree to retrieve scripts
         // TODO execute the synchronous scripts during the HTL parsing.
@@ -250,7 +448,7 @@ pub const Page = struct {
         // sasync stores scripts which can be run asynchronously.
         // for now they are just run after the non-async one in order to
         // dispatch DOMContentLoaded the sooner as possible.
-        var sasync = std.ArrayList(*parser.Element).init(alloc);
+        var sasync = std.ArrayList(Script).init(alloc);
         defer sasync.deinit();
 
         const root = parser.documentToNode(doc);
@@ -265,21 +463,10 @@ pub const Page = struct {
             }
 
             const e = parser.nodeToElement(next.?);
-            const tag = try parser.elementHTMLGetTagType(@as(*parser.ElementHTML, @ptrCast(e)));
-
-            // ignore non-script tags
-            if (tag != .script) continue;
 
             // ignore non-js script.
-            // > type
-            // > Attribute is not set (default), an empty string, or a JavaScript MIME
-            // > type indicates that the script is a "classic script", containing
-            // > JavaScript code.
-            // https://developer.mozilla.org/en-US/docs/Web/HTML/Element/script#attribute_is_not_set_default_an_empty_string_or_a_javascript_mime_type
-            const stype = try parser.elementGetAttribute(e, "type");
-            if (!isJS(stype)) {
-                continue;
-            }
+            const script = try Script.init(e) orelse continue;
+            if (script.kind == .unknown) continue;
 
             // Ignore the defer attribute b/c we analyze all script
             // after the document has been parsed.
@@ -293,8 +480,8 @@ pub const Page = struct {
             // > then the classic script will be fetched in parallel to
             // > parsing and evaluated as soon as it is available.
             // https://developer.mozilla.org/en-US/docs/Web/HTML/Element/script#async
-            if (try parser.elementGetAttribute(e, "async") != null) {
-                try sasync.append(e);
+            if (script.isasync) {
+                try sasync.append(script);
                 continue;
             }
 
@@ -316,7 +503,9 @@ pub const Page = struct {
             // > immediately before the browser continues to parse the
             // > page.
             // https://developer.mozilla.org/en-US/docs/Web/HTML/Element/script#notes
-            self.evalScript(e) catch |err| log.warn("evaljs: {any}", .{err});
+            try parser.documentHTMLSetCurrentScript(html_doc, @ptrCast(e));
+            self.evalScript(script) catch |err| log.warn("evaljs: {any}", .{err});
+            try parser.documentHTMLSetCurrentScript(html_doc, null);
         }
 
         // TODO wait for deferred scripts
@@ -326,12 +515,16 @@ pub const Page = struct {
         // have loaded.
         // https://html.spec.whatwg.org/#reporting-document-loading-status
         const evt = try parser.eventCreate();
+        defer parser.eventDestroy(evt);
+
         try parser.eventInit(evt, "DOMContentLoaded", .{ .bubbles = true, .cancelable = true });
         _ = try parser.eventTargetDispatchEvent(parser.toEventTarget(parser.DocumentHTML, html_doc), evt);
 
         // eval async scripts.
-        for (sasync.items) |e| {
-            self.evalScript(e) catch |err| log.warn("evaljs: {any}", .{err});
+        for (sasync.items) |s| {
+            try parser.documentHTMLSetCurrentScript(html_doc, @ptrCast(s.element));
+            self.evalScript(s) catch |err| log.warn("evaljs: {any}", .{err});
+            try parser.documentHTMLSetCurrentScript(html_doc, null);
         }
 
         // TODO wait for async scripts
@@ -340,22 +533,27 @@ pub const Page = struct {
 
         // dispatch window.load event
         const loadevt = try parser.eventCreate();
+        defer parser.eventDestroy(loadevt);
+
         try parser.eventInit(loadevt, "load", .{});
-        _ = try parser.eventTargetDispatchEvent(parser.toEventTarget(Window, &self.session.window), loadevt);
+        _ = try parser.eventTargetDispatchEvent(
+            parser.toEventTarget(Window, &self.session.window),
+            loadevt,
+        );
     }
 
     // evalScript evaluates the src in priority.
     // if no src is present, we evaluate the text source.
     // https://html.spec.whatwg.org/multipage/scripting.html#script-processing-model
-    fn evalScript(self: *Page, e: *parser.Element) !void {
+    fn evalScript(self: *Page, s: Script) !void {
         const alloc = self.arena.allocator();
 
         // https://html.spec.whatwg.org/multipage/webappapis.html#fetch-a-classic-script
-        const opt_src = try parser.elementGetAttribute(e, "src");
+        const opt_src = try parser.elementGetAttribute(s.element, "src");
         if (opt_src) |src| {
             log.debug("starting GET {s}", .{src});
 
-            self.fetchScript(src) catch |err| {
+            self.fetchScript(s) catch |err| {
                 switch (err) {
                     FetchError.BadStatusCode => return err,
 
@@ -374,19 +572,10 @@ pub const Page = struct {
             return;
         }
 
-        const opt_text = try parser.nodeTextContent(parser.elementToNode(e));
-        if (opt_text) |text| {
         // TODO handle charset attribute
-            var res = jsruntime.JSResult{};
-            try self.session.env.run(alloc, text, "", &res, null);
-            defer res.deinit(alloc);
-
-            if (res.success) {
-                log.debug("eval inline: {s}", .{res.result});
-            } else {
-                log.info("eval inline: {s}", .{res.result});
-            }
-
+        const opt_text = try parser.nodeTextContent(parser.elementToNode(s.element));
+        if (opt_text) |text| {
+            try s.eval(alloc, self.session.env, text);
             return;
         }
 
@@ -401,38 +590,67 @@ pub const Page = struct {
         JsErr,
     };
 
-    // fetchScript senf a GET request to the src and execute the script
-    // received.
-    fn fetchScript(self: *Page, src: []const u8) !void {
-        const alloc = self.arena.allocator();
+    // the caller owns the returned string
+    fn fetchData(self: *Page, alloc: std.mem.Allocator, src: []const u8) ![]const u8 {
+        log.debug("starting fetch {s}", .{src});
 
-        log.debug("starting fetch script {s}", .{src});
+        var buffer: [1024]u8 = undefined;
+        var b: []u8 = buffer[0..];
+        const u = try std.Uri.resolve_inplace(self.uri, src, &b);
 
-        const u = std.Uri.parse(src) catch try std.Uri.parseWithoutScheme(src);
-        const ru = try std.Uri.resolve(self.uri, u, false, alloc);
-
-        var fetchres = try self.session.loader.fetch(alloc, ru);
+        var fetchres = try self.session.loader.get(alloc, u);
         defer fetchres.deinit();
 
-        log.info("fech script {any}: {d}", .{ ru, fetchres.status });
+        const resp = fetchres.req.response;
 
-        if (fetchres.status != .ok) return FetchError.BadStatusCode;
+        log.info("fetch {any}: {d}", .{ u, resp.status });
+
+        if (resp.status != .ok) return FetchError.BadStatusCode;
 
         // TODO check content-type
+        const body = try fetchres.req.reader().readAllAlloc(alloc, 16 * 1024 * 1024);
 
         // check no body
-        if (fetchres.body == null) return FetchError.NoBody;
+        if (body.len == 0) return FetchError.NoBody;
 
-        var res = jsruntime.JSResult{};
-        try self.session.env.run(alloc, fetchres.body.?, src, &res, null);
-        defer res.deinit(alloc);
-
-        if (res.success) {
-            log.debug("eval remote {s}: {s}", .{ src, res.result });
-        } else {
-            log.info("eval remote {s}: {s}", .{ src, res.result });
-            return FetchError.JsErr;
+        return body;
     }
+
+    // fetchScript senf a GET request to the src and execute the script
+    // received.
+    fn fetchScript(self: *Page, s: Script) !void {
+        const alloc = self.arena.allocator();
+        const body = try self.fetchData(alloc, s.src);
+        defer alloc.free(body);
+
+        try s.eval(alloc, self.session.env, body);
+    }
+
+    const Script = struct {
+        element: *parser.Element,
+        kind: Kind,
+        isasync: bool,
+
+        src: []const u8,
+
+        const Kind = enum {
+            unknown,
+            javascript,
+            module,
+        };
+
+        fn init(e: *parser.Element) !?Script {
+            // ignore non-script tags
+            const tag = try parser.elementHTMLGetTagType(@as(*parser.ElementHTML, @ptrCast(e)));
+            if (tag != .script) return null;
+
+            return .{
+                .element = e,
+                .kind = kind(try parser.elementGetAttribute(e, "type")),
+                .isasync = try parser.elementGetAttribute(e, "async") != null,
+
+                .src = try parser.elementGetAttribute(e, "src") orelse "inline",
+            };
         }
 
         // > type
@@ -440,11 +658,36 @@ pub const Page = struct {
         // > type indicates that the script is a "classic script", containing
         // > JavaScript code.
         // https://developer.mozilla.org/en-US/docs/Web/HTML/Element/script#attribute_is_not_set_default_an_empty_string_or_a_javascript_mime_type
-    fn isJS(stype: ?[]const u8) bool {
-        if (stype == null or stype.?.len == 0) return true;
-        if (std.mem.eql(u8, stype.?, "application/javascript")) return true;
-        if (!std.mem.eql(u8, stype.?, "module")) return true;
+        fn kind(stype: ?[]const u8) Kind {
+            if (stype == null or stype.?.len == 0) return .javascript;
+            if (std.mem.eql(u8, stype.?, "application/javascript")) return .javascript;
+            if (std.mem.eql(u8, stype.?, "module")) return .module;
 
-        return false;
+            return .unknown;
         }
+
+        fn eval(self: Script, alloc: std.mem.Allocator, env: Env, body: []const u8) !void {
+            var try_catch: jsruntime.TryCatch = undefined;
+            try_catch.init(env);
+            defer try_catch.deinit();
+
+            const res = switch (self.kind) {
+                .unknown => return error.UnknownScript,
+                .javascript => env.exec(body, self.src),
+                .module => env.module(body, self.src),
+            } catch {
+                if (try try_catch.err(alloc, env)) |msg| {
+                    defer alloc.free(msg);
+                    log.info("eval script {s}: {s}", .{ self.src, msg });
+                }
+                return FetchError.JsErr;
+            };
+
+            if (builtin.mode == .Debug) {
+                const msg = try res.toString(alloc, env);
+                defer alloc.free(msg);
+                log.debug("eval script {s}: {s}", .{ self.src, msg });
+            }
+        }
+    };
 };

src/browser/dump.zig

@@ -1,96 +1,206 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 const File = std.fs.File;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 const Walker = @import("../dom/walker.zig").WalkerChildren;
 
-pub fn htmlFile(doc: *parser.Document, out: File) !void {
-    try out.writeAll("<!DOCTYPE html>\n");
-    try nodeFile(parser.documentToNode(doc), out);
-    try out.writeAll("\n");
+// writer must be a std.io.Writer
+pub fn writeHTML(doc: *parser.Document, writer: anytype) !void {
+    try writer.writeAll("<!DOCTYPE html>\n");
+    try writeChildren(parser.documentToNode(doc), writer);
+    try writer.writeAll("\n");
 }
 
-fn nodeFile(root: *parser.Node, out: File) !void {
-    const walker = Walker{};
-    var next: ?*parser.Node = null;
-    while (true) {
-        next = try walker.get_next(root, next) orelse break;
-        switch (try parser.nodeType(next.?)) {
+pub fn writeNode(node: *parser.Node, writer: anytype) anyerror!void {
+    switch (try parser.nodeType(node)) {
         .element => {
             // open the tag
-                const tag = try parser.nodeLocalName(next.?);
-                try out.writeAll("<");
-                try out.writeAll(tag);
+            const tag = try parser.nodeLocalName(node);
+            try writer.writeAll("<");
+            try writer.writeAll(tag);
 
             // write the attributes
-                const map = try parser.nodeGetAttributes(next.?);
+            const map = try parser.nodeGetAttributes(node);
             const ln = try parser.namedNodeMapGetLength(map);
             var i: u32 = 0;
             while (i < ln) {
                 const attr = try parser.namedNodeMapItem(map, i) orelse break;
-                    try out.writeAll(" ");
-                    try out.writeAll(try parser.attributeGetName(attr));
-                    try out.writeAll("=\"");
-                    try out.writeAll(try parser.attributeGetValue(attr) orelse "");
-                    try out.writeAll("\"");
+                try writer.writeAll(" ");
+                try writer.writeAll(try parser.attributeGetName(attr));
+                try writer.writeAll("=\"");
+                const attribute_value = try parser.attributeGetValue(attr) orelse "";
+                try writeEscapedAttributeValue(writer, attribute_value);
+                try writer.writeAll("\"");
                 i += 1;
             }
 
-                try out.writeAll(">");
+            try writer.writeAll(">");
+
+            // void elements can't have any content.
+            if (try isVoid(parser.nodeToElement(node))) return;
 
             // write the children
             // TODO avoid recursion
-                try nodeFile(next.?, out);
+            try writeChildren(node, writer);
 
             // close the tag
-                try out.writeAll("</");
-                try out.writeAll(tag);
-                try out.writeAll(">");
+            try writer.writeAll("</");
+            try writer.writeAll(tag);
+            try writer.writeAll(">");
         },
         .text => {
-                const v = try parser.nodeValue(next.?) orelse continue;
-                try out.writeAll(v);
+            const v = try parser.nodeValue(node) orelse return;
+            try writeEscapedTextNode(writer, v);
         },
         .cdata_section => {
-                const v = try parser.nodeValue(next.?) orelse continue;
-                try out.writeAll("<![CDATA[");
-                try out.writeAll(v);
-                try out.writeAll("]]>");
+            const v = try parser.nodeValue(node) orelse return;
+            try writer.writeAll("<![CDATA[");
+            try writer.writeAll(v);
+            try writer.writeAll("]]>");
         },
         .comment => {
-                const v = try parser.nodeValue(next.?) orelse continue;
-                try out.writeAll("<!--");
-                try out.writeAll(v);
-                try out.writeAll("-->");
+            const v = try parser.nodeValue(node) orelse return;
+            try writer.writeAll("<!--");
+            try writer.writeAll(v);
+            try writer.writeAll("-->");
         },
         // TODO handle processing instruction dump
-            .processing_instruction => continue,
+        .processing_instruction => return,
         // document fragment is outside of the main document DOM, so we
         // don't output it.
-            .document_fragment => continue,
+        .document_fragment => return,
         // document will never be called, but required for completeness.
-            .document => continue,
+        .document => return,
         // done globally instead, but required for completeness.
-            .document_type => continue,
+        .document_type => return,
         // deprecated
-            .attribute => continue,
-            .entity_reference => continue,
-            .entity => continue,
-            .notation => continue,
-        }
+        .attribute => return,
+        .entity_reference => return,
+        .entity => return,
+        .notation => return,
     }
 }
 
-// HTMLFileTestFn is run by run_tests.zig
-pub fn HTMLFileTestFn(out: File) !void {
-    const file = try std.fs.cwd().openFile("test.html", .{});
-    defer file.close();
+// writer must be a std.io.Writer
+pub fn writeChildren(root: *parser.Node, writer: anytype) !void {
+    const walker = Walker{};
+    var next: ?*parser.Node = null;
+    while (true) {
+        next = try walker.get_next(root, next) orelse break;
+        try writeNode(next.?, writer);
+    }
+}
 
-    const doc_html = try parser.documentHTMLParse(file.reader(), "UTF-8");
-    // ignore close error
+// area, base, br, col, embed, hr, img, input, link, meta, source, track, wbr
+// https://html.spec.whatwg.org/#void-elements
+fn isVoid(elem: *parser.Element) !bool {
+    const tag = try parser.elementHTMLGetTagType(@as(*parser.ElementHTML, @ptrCast(elem)));
+    return switch (tag) {
+        .area, .base, .br, .col, .embed, .hr, .img, .input, .link => true,
+        .meta, .source, .track, .wbr => true,
+        else => false,
+    };
+}
+
+fn writeEscapedTextNode(writer: anytype, value: []const u8) !void {
+    var v = value;
+    while (v.len > 0) {
+        const index = std.mem.indexOfAnyPos(u8, v, 0, &.{ '&', '<', '>' }) orelse {
+            return writer.writeAll(v);
+        };
+        try writer.writeAll(v[0..index]);
+        switch (v[index]) {
+            '&' => try writer.writeAll("&amp;"),
+            '<' => try writer.writeAll("&lt;"),
+            '>' => try writer.writeAll("&gt;"),
+            else => unreachable,
+        }
+        v = v[index + 1 ..];
+    }
+}
+
+fn writeEscapedAttributeValue(writer: anytype, value: []const u8) !void {
+    var v = value;
+    while (v.len > 0) {
+        const index = std.mem.indexOfAnyPos(u8, v, 0, &.{ '&', '<', '>', '"' }) orelse {
+            return writer.writeAll(v);
+        };
+        try writer.writeAll(v[0..index]);
+        switch (v[index]) {
+            '&' => try writer.writeAll("&amp;"),
+            '<' => try writer.writeAll("&lt;"),
+            '>' => try writer.writeAll("&gt;"),
+            '"' => try writer.writeAll("&quot;"),
+            else => unreachable,
+        }
+        v = v[index + 1 ..];
+    }
+}
+
+const testing = std.testing;
+test "dump.writeHTML" {
+    try testWriteHTML(
+        "<div id=\"content\">Over 9000!</div>",
+        "<div id=\"content\">Over 9000!</div>",
+    );
+
+    try testWriteHTML(
+        "<root><!-- a comment --></root>",
+        "<root><!-- a comment --></root>",
+    );
+
+    try testWriteHTML(
+        "<p>&lt; &gt; &amp;</p>",
+        "<p>&lt; &gt; &amp;</p>",
+    );
+
+    try testWriteHTML(
+        "<p id=\"&quot;&gt;&lt;&amp;&quot;''\">wat?</p>",
+        "<p id='\">&lt;&amp;&quot;&#39;&apos;'>wat?</p>",
+    );
+
+    try testWriteFullHTML(
+        \\<!DOCTYPE html>
+        \\<html><head><title>It's over what?</title><meta name="a" value="b">
+        \\</head><body>9000</body></html>
+        \\
+    , "<html><title>It's over what?</title><meta name=a value=\"b\">\n<body>9000");
+}
+
+fn testWriteHTML(comptime expected_body: []const u8, src: []const u8) !void {
+    const expected =
+        "<!DOCTYPE html>\n<html><head></head><body>" ++
+        expected_body ++
+        "</body></html>\n";
+    return testWriteFullHTML(expected, src);
+}
+
+fn testWriteFullHTML(comptime expected: []const u8, src: []const u8) !void {
+    var buf = std.ArrayListUnmanaged(u8){};
+    defer buf.deinit(testing.allocator);
+
+    const doc_html = try parser.documentHTMLParseFromStr(src);
     defer parser.documentHTMLClose(doc_html) catch {};
 
     const doc = parser.documentHTMLToDocument(doc_html);
-
-    try htmlFile(doc, out);
+    try writeHTML(doc, buf.writer(testing.allocator));
+    try testing.expectEqualStrings(expected, buf.items);
 }

src/browser/loader.zig

@@ -1,13 +1,34 @@
-const std = @import("std");
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
-const user_agent = "Lightpanda.io/1.0";
+const std = @import("std");
+const Client = @import("../http/Client.zig");
+
+const user_agent = @import("browser.zig").user_agent;
 
 pub const Loader = struct {
-    client: std.http.Client,
+    client: Client,
+    // use 64KB for headers buffer size.
+    server_header_buffer: [1024 * 64]u8 = undefined,
 
     pub const Response = struct {
         alloc: std.mem.Allocator,
-        req: *std.http.Client.Request,
+        req: *Client.Request,
 
         pub fn deinit(self: *Response) void {
             self.req.deinit();
@@ -17,7 +38,7 @@ pub const Loader = struct {
 
     pub fn init(alloc: std.mem.Allocator) Loader {
         return Loader{
-            .client = std.http.Client{
+            .client = Client{
                 .allocator = alloc,
             },
         };
@@ -27,46 +48,30 @@ pub const Loader = struct {
         self.client.deinit();
     }
 
-    // the caller must deinit the FetchResult.
-    pub fn fetch(self: *Loader, alloc: std.mem.Allocator, uri: std.Uri) !std.http.Client.FetchResult {
-        var headers = try std.http.Headers.initList(alloc, &[_]std.http.Field{
-            .{ .name = "User-Agent", .value = user_agent },
-            .{ .name = "Accept", .value = "*/*" },
-            .{ .name = "Accept-Language", .value = "en-US,en;q=0.5" },
-        });
-        defer headers.deinit();
-
-        return try self.client.fetch(alloc, .{
-            .location = .{ .uri = uri },
-            .headers = headers,
-            .payload = .none,
-        });
-    }
-
     // see
     // https://ziglang.org/documentation/master/std/#A;std:http.Client.fetch
     // for reference.
     // The caller is responsible for calling `deinit()` on the `Response`.
     pub fn get(self: *Loader, alloc: std.mem.Allocator, uri: std.Uri) !Response {
-        var headers = try std.http.Headers.initList(alloc, &[_]std.http.Field{
-            .{ .name = "User-Agent", .value = user_agent },
-            .{ .name = "Accept", .value = "*/*" },
-            .{ .name = "Accept-Language", .value = "en-US,en;q=0.5" },
-        });
-        defer headers.deinit();
-
         var resp = Response{
             .alloc = alloc,
-            .req = try alloc.create(std.http.Client.Request),
+            .req = try alloc.create(Client.Request),
         };
         errdefer alloc.destroy(resp.req);
 
-        resp.req.* = try self.client.open(.GET, uri, headers, .{
-            .handle_redirects = true, // TODO handle redirects manually
+        resp.req.* = try self.client.open(.GET, uri, .{
+            .headers = .{
+                .user_agent = .{ .override = user_agent },
+            },
+            .extra_headers = &.{
+                .{ .name = "Accept", .value = "*/*" },
+                .{ .name = "Accept-Language", .value = "en-US,en;q=0.5" },
+            },
+            .server_header_buffer = &self.server_header_buffer,
         });
         errdefer resp.req.deinit();
 
-        try resp.req.send(.{});
+        try resp.req.send();
         try resp.req.finish();
         try resp.req.wait();
 
@@ -74,13 +79,19 @@ pub const Loader = struct {
     }
 };
 
-test "basic url fetch" {
+test "loader: get" {
     const alloc = std.testing.allocator;
     var loader = Loader.init(alloc);
     defer loader.deinit();
 
-    var result = try loader.fetch(alloc, "https://en.wikipedia.org/wiki/Main_Page");
+    const uri = try std.Uri.parse("http://localhost:9582/loader");
+    var result = try loader.get(alloc, uri);
     defer result.deinit();
 
-    try std.testing.expect(result.status == std.http.Status.ok);
+    try std.testing.expectEqual(.ok, result.req.response.status);
+
+    var res: [128]u8 = undefined;
+    const size = try result.req.readAll(&res);
+    try std.testing.expectEqual(6, size);
+    try std.testing.expectEqualStrings("Hello!", res[0..6]);
 }

src/browser/mime.zig

@@ -1,220 +1,391 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
-const testing = std.testing;
+const Allocator = std.mem.Allocator;
 
-const Self = @This();
+pub const Mime = struct {
+    content_type: ContentType,
+    params: []const u8 = "",
+    charset: ?[]const u8 = null,
+    arena: std.heap.ArenaAllocator,
 
-const MimeError = error{
-    Empty,
-    TooBig,
-    Invalid,
-    InvalidChar,
-};
-
-mtype: []const u8,
-msubtype: []const u8,
-params: []const u8 = "",
-
-charset: ?[]const u8 = null,
-boundary: ?[]const u8 = null,
-
-pub const Empty = Self{ .mtype = "", .msubtype = "" };
-pub const HTML = Self{ .mtype = "text", .msubtype = "html" };
-pub const Javascript = Self{ .mtype = "application", .msubtype = "javascript" };
-
-const reader = struct {
-    s: []const u8,
-    i: usize = 0,
-
-    fn until(self: *reader, c: u8) []const u8 {
-        const ln = self.s.len;
-        const start = self.i;
-        while (self.i < ln) {
-            if (c == self.s[self.i]) return self.s[start..self.i];
-            self.i += 1;
-        }
-
-        return self.s[start..self.i];
-    }
-
-    fn tail(self: *reader) []const u8 {
-        if (self.i > self.s.len) return "";
-        defer self.i = self.s.len;
-        return self.s[self.i..];
-    }
-
-    fn skip(self: *reader) bool {
-        if (self.i >= self.s.len) return false;
-        self.i += 1;
-        return true;
-    }
-};
-
-test "reader.skip" {
-    var r = reader{ .s = "foo" };
-    try testing.expect(r.skip());
-    try testing.expect(r.skip());
-    try testing.expect(r.skip());
-    try testing.expect(!r.skip());
-    try testing.expect(!r.skip());
-}
-
-test "reader.tail" {
-    var r = reader{ .s = "foo" };
-    try testing.expectEqualStrings("foo", r.tail());
-    try testing.expectEqualStrings("", r.tail());
-}
-
-test "reader.until" {
-    var r = reader{ .s = "foo.bar.baz" };
-    try testing.expectEqualStrings("foo", r.until('.'));
-    _ = r.skip();
-    try testing.expectEqualStrings("bar", r.until('.'));
-    _ = r.skip();
-    try testing.expectEqualStrings("baz", r.until('.'));
-
-    r = reader{ .s = "foo" };
-    try testing.expectEqualStrings("foo", r.until('.'));
-
-    r = reader{ .s = "" };
-    try testing.expectEqualStrings("", r.until('.'));
-}
-
-fn trim(s: []const u8) []const u8 {
-    const ln = s.len;
-    if (ln == 0) {
-        return "";
-    }
-    var start: usize = 0;
-    while (start < ln) {
-        if (!std.ascii.isWhitespace(s[start])) break;
-        start += 1;
-    }
-
-    var end: usize = ln;
-    while (end > 0) {
-        if (!std.ascii.isWhitespace(s[end - 1])) break;
-        end -= 1;
-    }
-
-    return s[start..end];
-}
-
-test "trim" {
-    try testing.expectEqualStrings("", trim(""));
-    try testing.expectEqualStrings("foo", trim("foo"));
-    try testing.expectEqualStrings("foo", trim(" \n\tfoo"));
-    try testing.expectEqualStrings("foo", trim("foo \n\t"));
-}
-
-// https://mimesniff.spec.whatwg.org/#http-token-code-point
-fn isHTTPCodePoint(c: u8) bool {
-    return switch (c) {
-        '!', '#', '$', '%', '&', '\'', '*', '+', '-', '.', '^' => return true,
-        '_', '`', '|', '~' => return true,
-        else => std.ascii.isAlphanumeric(c),
+    pub const ContentTypeEnum = enum {
+        text_xml,
+        text_html,
+        text_plain,
+        other,
     };
-}
 
-fn valid(s: []const u8) bool {
-    const ln = s.len;
-    var i: usize = 0;
-    while (i < ln) {
-        if (!isHTTPCodePoint(s[i])) return false;
-        i += 1;
+    pub const ContentType = union(ContentTypeEnum) {
+        text_xml: void,
+        text_html: void,
+        text_plain: void,
+        other: struct { type: []const u8, sub_type: []const u8 },
+    };
+
+    pub fn parse(allocator: Allocator, input: []const u8) !Mime {
+        if (input.len > 255) {
+            return error.TooBig;
+        }
+
+        var arena = std.heap.ArenaAllocator.init(allocator);
+        errdefer arena.deinit();
+
+        var trimmed = trim(input);
+
+        const content_type, const type_len = try parseContentType(trimmed);
+        if (type_len >= trimmed.len) {
+            return .{ .arena = arena, .content_type = content_type };
+        }
+
+        const params = trimLeft(trimmed[type_len..]);
+
+        var charset: ?[]const u8 = null;
+
+        var it = std.mem.splitScalar(u8, params, ';');
+        while (it.next()) |attr| {
+            const i = std.mem.indexOfScalarPos(u8, attr, 0, '=') orelse return error.Invalid;
+            const name = trimLeft(attr[0..i]);
+
+            const value = trimRight(attr[i + 1 ..]);
+            if (value.len == 0) {
+                return error.Invalid;
+            }
+
+            switch (name.len) {
+                7 => if (isCaseEqual("charset", name)) {
+                    charset = try parseValue(arena.allocator(), value);
+                },
+                else => {},
+            }
+        }
+
+        return .{
+            .arena = arena,
+            .params = params,
+            .charset = charset,
+            .content_type = content_type,
+        };
+    }
+
+    pub fn deinit(self: *Mime) void {
+        self.arena.deinit();
+    }
+
+    pub fn isHTML(self: *const Mime) bool {
+        return self.content_type == .text_html;
+    }
+
+    fn parseContentType(value: []const u8) !struct { ContentType, usize } {
+        const separator = std.mem.indexOfScalarPos(u8, value, 0, '/') orelse {
+            return error.Invalid;
+        };
+        const end = std.mem.indexOfScalarPos(u8, value, separator, ';') orelse blk: {
+            break :blk value.len;
+        };
+
+        const main_type = value[0..separator];
+        const sub_type = trimRight(value[separator + 1 .. end]);
+
+        if (parseCommonContentType(main_type, sub_type)) |content_type| {
+            return .{ content_type, end + 1 };
+        }
+
+        if (main_type.len == 0) {
+            return error.Invalid;
+        }
+        if (validType(main_type) == false) {
+            return error.Invalid;
+        }
+
+        if (sub_type.len == 0) {
+            return error.Invalid;
+        }
+        if (validType(sub_type) == false) {
+            return error.Invalid;
+        }
+
+        const content_type = ContentType{ .other = .{
+            .type = main_type,
+            .sub_type = sub_type,
+        } };
+
+        return .{ content_type, end + 1 };
+    }
+
+    fn parseCommonContentType(main_type: []const u8, sub_type: []const u8) ?ContentType {
+        switch (main_type.len) {
+            4 => if (isCaseEqual("text", main_type)) {
+                switch (sub_type.len) {
+                    3 => if (isCaseEqual("xml", sub_type)) {
+                        return .{ .text_xml = {} };
+                    },
+                    4 => if (isCaseEqual("html", sub_type)) {
+                        return .{ .text_html = {} };
+                    },
+                    5 => if (isCaseEqual("plain", sub_type)) {
+                        return .{ .text_plain = {} };
+                    },
+                    else => {},
+                }
+            },
+            else => {},
+        }
+        return null;
+    }
+
+    const T_SPECIAL = blk: {
+        var v = [_]bool{false} ** 256;
+        for ("()<>@,;:\\\"/[]?=") |b| {
+            v[b] = true;
+        }
+        break :blk v;
+    };
+
+    fn parseValue(allocator: Allocator, value: []const u8) ![]const u8 {
+        if (value[0] != '"') {
+            return value;
+        }
+
+        // 1 to skip the opening quote
+        var value_pos: usize = 1;
+        var unescaped_len: usize = 0;
+        const last = value.len - 1;
+
+        while (value_pos < value.len) {
+            switch (value[value_pos]) {
+                '"' => break,
+                '\\' => {
+                    if (value_pos == last) {
+                        return error.Invalid;
+                    }
+                    const next = value[value_pos + 1];
+                    if (T_SPECIAL[next] == false) {
+                        return error.Invalid;
+                    }
+                    value_pos += 2;
+                },
+                else => value_pos += 1,
+            }
+            unescaped_len += 1;
+        }
+
+        if (unescaped_len == 0) {
+            return error.Invalid;
+        }
+
+        value_pos = 1;
+        const owned = try allocator.alloc(u8, unescaped_len);
+        for (0..unescaped_len) |i| {
+            switch (value[value_pos]) {
+                '"' => break,
+                '\\' => {
+                    owned[i] = value[value_pos + 1];
+                    value_pos += 2;
+                },
+                else => |c| {
+                    owned[i] = c;
+                    value_pos += 1;
+                },
+            }
+        }
+        return owned;
+    }
+
+    const VALID_CODEPOINTS = blk: {
+        var v: [256]bool = undefined;
+        for (0..256) |i| {
+            v[i] = std.ascii.isAlphanumeric(i);
+        }
+        for ("!#$%&\\*+-.^'_`|~") |b| {
+            v[b] = true;
+        }
+        break :blk v;
+    };
+
+    fn validType(value: []const u8) bool {
+        for (value) |b| {
+            if (VALID_CODEPOINTS[b] == false) {
+                return false;
+            }
         }
         return true;
-}
-
-// https://mimesniff.spec.whatwg.org/#parsing-a-mime-type
-pub fn parse(s: []const u8) Self.MimeError!Self {
-    const ln = s.len;
-    if (ln == 0) return MimeError.Empty;
-    // limit input size
-    if (ln > 255) return MimeError.TooBig;
-
-    var res = Self{ .mtype = "", .msubtype = "" };
-    var r = reader{ .s = s };
-
-    res.mtype = trim(r.until('/'));
-    if (res.mtype.len == 0) return MimeError.Invalid;
-    if (!valid(res.mtype)) return MimeError.InvalidChar;
-
-    if (!r.skip()) return MimeError.Invalid;
-    res.msubtype = trim(r.until(';'));
-    if (res.msubtype.len == 0) return MimeError.Invalid;
-    if (!valid(res.msubtype)) return MimeError.InvalidChar;
-
-    if (!r.skip()) return res;
-    res.params = trim(r.tail());
-    if (res.params.len == 0) return MimeError.Invalid;
-
-    // parse well known parameters.
-    // don't check invalid parameter format.
-    var rp = reader{ .s = res.params };
-    while (true) {
-        const name = trim(rp.until('='));
-        if (!rp.skip()) return res;
-        const value = trim(rp.until(';'));
-
-        if (std.ascii.eqlIgnoreCase(name, "charset")) {
-            res.charset = value;
-        }
-        if (std.ascii.eqlIgnoreCase(name, "boundary")) {
-            res.boundary = value;
     }
 
-        if (!rp.skip()) return res;
+    fn trim(s: []const u8) []const u8 {
+        return std.mem.trim(u8, s, &std.ascii.whitespace);
     }
 
-    return res;
-}
-
-test "parse valid" {
-    for ([_][]const u8{
-        "text/html",
-        " \ttext/html",
-        "text \t/html",
-        "text/ \thtml",
-        "text/html \t",
-    }) |tc| {
-        const m = try Self.parse(tc);
-        try testing.expectEqualStrings("text", m.mtype);
-        try testing.expectEqualStrings("html", m.msubtype);
+    fn trimLeft(s: []const u8) []const u8 {
+        return std.mem.trimLeft(u8, s, &std.ascii.whitespace);
     }
-    const m2 = try Self.parse("text/javascript1.5");
-    try testing.expectEqualStrings("text", m2.mtype);
-    try testing.expectEqualStrings("javascript1.5", m2.msubtype);
 
-    const m3 = try Self.parse("text/html; charset=utf-8");
-    try testing.expectEqualStrings("text", m3.mtype);
-    try testing.expectEqualStrings("html", m3.msubtype);
-    try testing.expectEqualStrings("charset=utf-8", m3.params);
-    try testing.expectEqualStrings("utf-8", m3.charset.?);
+    fn trimRight(s: []const u8) []const u8 {
+        return std.mem.trimRight(u8, s, &std.ascii.whitespace);
+    }
 
-    const m4 = try Self.parse("text/html; boundary=----");
-    try testing.expectEqualStrings("text", m4.mtype);
-    try testing.expectEqualStrings("html", m4.msubtype);
-    try testing.expectEqualStrings("boundary=----", m4.params);
-    try testing.expectEqualStrings("----", m4.boundary.?);
-}
+    fn isCaseEqual(comptime target: anytype, value: []const u8) bool {
+        // - 8 beause we don't care about the sentinel
+        const bit_len = @bitSizeOf(@TypeOf(target.*)) - 8;
+        const byte_len = bit_len / 8;
 
-test "parse invalid" {
-    for ([_][]const u8{
+        const T = @Type(.{ .Int = .{
+            .bits = bit_len,
+            .signedness = .unsigned,
+        } });
+
+        const bit_target: T = @bitCast(@as(*const [byte_len]u8, target).*);
+
+        if (@as(T, @bitCast(value[0..byte_len].*)) == bit_target) {
+            return true;
+        }
+        return std.ascii.eqlIgnoreCase(value, target);
+    }
+};
+
+const testing = std.testing;
+test "Mime: invalid " {
+    const invalids = [_][]const u8{
         "",
-        "te xt/html;",
-        "te@xt/html;",
-        "text/ht@ml;",
-        "text/html;",
-        "/text/html",
-        "/html",
-    }) |tc| {
-        _ = Self.parse(tc) catch continue;
-        try testing.expect(false);
+        "text",
+        "text /html",
+        "text/ html",
+        "text / html",
+        "text/html other",
+        "text/html; x",
+        "text/html; x=",
+        "text/html; x=  ",
+        "text/html; = ",
+        "text/html;=",
+        "text/html; charset=\"\"",
+        "text/html; charset=\"",
+        "text/html; charset=\"\\",
+        "text/html; charset=\"\\a\"", // invalid to escape non special characters
+    };
+
+    for (invalids) |invalid| {
+        try testing.expectError(error.Invalid, Mime.parse(undefined, invalid));
     }
 }
 
-// Compare type and subtype.
-pub fn eql(self: Self, b: Self) bool {
-    if (!std.mem.eql(u8, self.mtype, b.mtype)) return false;
-    return std.mem.eql(u8, self.msubtype, b.msubtype);
+test "Mime: parse common" {
+    try expect(.{ .content_type = .{ .text_xml = {} } }, "text/xml");
+    try expect(.{ .content_type = .{ .text_html = {} } }, "text/html");
+    try expect(.{ .content_type = .{ .text_plain = {} } }, "text/plain");
+
+    try expect(.{ .content_type = .{ .text_xml = {} } }, "text/xml;");
+    try expect(.{ .content_type = .{ .text_html = {} } }, "text/html;");
+    try expect(.{ .content_type = .{ .text_plain = {} } }, "text/plain;");
+
+    try expect(.{ .content_type = .{ .text_xml = {} } }, "  \ttext/xml");
+    try expect(.{ .content_type = .{ .text_html = {} } }, "text/html   ");
+    try expect(.{ .content_type = .{ .text_plain = {} } }, "text/plain \t\t");
+
+    try expect(.{ .content_type = .{ .text_xml = {} } }, "TEXT/xml");
+    try expect(.{ .content_type = .{ .text_html = {} } }, "text/Html");
+    try expect(.{ .content_type = .{ .text_plain = {} } }, "TEXT/PLAIN");
+
+    try expect(.{ .content_type = .{ .text_xml = {} } }, " TeXT/xml");
+    try expect(.{ .content_type = .{ .text_html = {} } }, "teXt/HtML  ;");
+    try expect(.{ .content_type = .{ .text_plain = {} } }, "tExT/PlAiN;");
+}
+
+test "Mime: parse uncommon" {
+    const text_javascript = Expectation{
+        .content_type = .{ .other = .{ .type = "text", .sub_type = "javascript" } },
+    };
+    try expect(text_javascript, "text/javascript");
+    try expect(text_javascript, "text/javascript;");
+    try expect(text_javascript, "  text/javascript\t  ");
+    try expect(text_javascript, "  text/javascript\t  ;");
+
+    try expect(
+        .{ .content_type = .{ .other = .{ .type = "Text", .sub_type = "Javascript" } } },
+        "Text/Javascript",
+    );
+}
+
+test "Mime: parse charset" {
+    try expect(.{
+        .content_type = .{ .text_xml = {} },
+        .charset = "utf-8",
+        .params = "charset=utf-8",
+    }, "text/xml; charset=utf-8");
+
+    try expect(.{
+        .content_type = .{ .text_xml = {} },
+        .charset = "utf-8",
+        .params = "charset=\"utf-8\"",
+    }, "text/xml;charset=\"utf-8\"");
+
+    try expect(.{
+        .content_type = .{ .text_xml = {} },
+        .charset = "\\ \" ",
+        .params = "charset=\"\\\\ \\\" \"",
+    }, "text/xml;charset=\"\\\\ \\\" \"   ");
+}
+
+test "Mime: isHTML" {
+    const isHTML = struct {
+        fn isHTML(expected: bool, input: []const u8) !void {
+            var mime = try Mime.parse(testing.allocator, input);
+            defer mime.deinit();
+            try testing.expectEqual(expected, mime.isHTML());
+        }
+    }.isHTML;
+    try isHTML(true, "text/html");
+    try isHTML(true, "text/html;");
+    try isHTML(true, "text/html; charset=utf-8");
+    try isHTML(false, "text/htm"); // htm not html
+    try isHTML(false, "text/plain");
+    try isHTML(false, "over/9000");
+}
+
+const Expectation = struct {
+    content_type: Mime.ContentType,
+    params: []const u8 = "",
+    charset: ?[]const u8 = null,
+};
+
+fn expect(expected: Expectation, input: []const u8) !void {
+    var actual = try Mime.parse(testing.allocator, input);
+    defer actual.deinit();
+
+    try testing.expectEqual(
+        std.meta.activeTag(expected.content_type),
+        std.meta.activeTag(actual.content_type),
+    );
+
+    switch (expected.content_type) {
+        .other => |e| {
+            const a = actual.content_type.other;
+            try testing.expectEqualStrings(e.type, a.type);
+            try testing.expectEqualStrings(e.sub_type, a.sub_type);
+        },
+        else => {}, // already asserted above
+    }
+
+    try testing.expectEqualStrings(expected.params, actual.params);
+
+    if (expected.charset) |ec| {
+        try testing.expectEqualStrings(ec, actual.charset.?);
+    } else {
+        try testing.expectEqual(null, actual.charset);
+    }
 }

src/cdp/browser.zig

@@ -0,0 +1,148 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    getVersion,
+    setDownloadBehavior,
+    getWindowForTarget,
+    setWindowBounds,
+};
+
+pub fn browser(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+    return switch (method) {
+        .getVersion => getVersion(alloc, msg, ctx),
+        .setDownloadBehavior => setDownloadBehavior(alloc, msg, ctx),
+        .getWindowForTarget => getWindowForTarget(alloc, msg, ctx),
+        .setWindowBounds => setWindowBounds(alloc, msg, ctx),
+    };
+}
+
+// TODO: hard coded data
+const ProtocolVersion = "1.3";
+const Product = "Chrome/124.0.6367.29";
+const Revision = "@9e6ded5ac1ff5e38d930ae52bd9aec09bd1a68e4";
+const UserAgent = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36";
+const JsVersion = "12.4.254.8";
+
+fn getVersion(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "browser.getVersion" });
+
+    // ouput
+    const Res = struct {
+        protocolVersion: []const u8 = ProtocolVersion,
+        product: []const u8 = Product,
+        revision: []const u8 = Revision,
+        userAgent: []const u8 = UserAgent,
+        jsVersion: []const u8 = JsVersion,
+    };
+    return result(alloc, input.id, Res, .{}, null);
+}
+
+// TODO: noop method
+fn setDownloadBehavior(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        behavior: []const u8,
+        browserContextId: ?[]const u8 = null,
+        downloadPath: ?[]const u8 = null,
+        eventsEnabled: ?bool = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("REQ > id {d}, method {s}", .{ input.id, "browser.setDownloadBehavior" });
+
+    // output
+    return result(alloc, input.id, null, null, null);
+}
+
+// TODO: hard coded ID
+const DevToolsWindowID = 1923710101;
+
+fn getWindowForTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+
+    // input
+    const Params = struct {
+        targetId: ?[]const u8 = null,
+    };
+    const input = try Input(?Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "browser.getWindowForTarget" });
+
+    // output
+    const Resp = struct {
+        windowId: u64 = DevToolsWindowID,
+        bounds: struct {
+            left: ?u64 = null,
+            top: ?u64 = null,
+            width: ?u64 = null,
+            height: ?u64 = null,
+            windowState: []const u8 = "normal",
+        } = .{},
+    };
+    return result(alloc, input.id, Resp, Resp{}, input.sessionId);
+}
+
+// TODO: noop method
+fn setWindowBounds(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "browser.setWindowBounds" });
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/cdp.zig

@@ -0,0 +1,277 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+
+const browser = @import("browser.zig").browser;
+const target = @import("target.zig").target;
+const page = @import("page.zig").page;
+const log = @import("log.zig").log;
+const runtime = @import("runtime.zig").runtime;
+const network = @import("network.zig").network;
+const emulation = @import("emulation.zig").emulation;
+const fetch = @import("fetch.zig").fetch;
+const performance = @import("performance.zig").performance;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+const inspector = @import("inspector.zig").inspector;
+const dom = @import("dom.zig").dom;
+const cdpdom = @import("dom.zig");
+const css = @import("css.zig").css;
+const security = @import("security.zig").security;
+
+const log_cdp = std.log.scoped(.cdp);
+
+pub const Error = error{
+    UnknonwDomain,
+    UnknownMethod,
+    NoResponse,
+    RequestWithoutID,
+};
+
+pub fn isCdpError(err: anyerror) ?Error {
+    // see https://github.com/ziglang/zig/issues/2473
+    const errors = @typeInfo(Error).ErrorSet.?;
+    inline for (errors) |e| {
+        if (std.mem.eql(u8, e.name, @errorName(err))) {
+            return @errorCast(err);
+        }
+    }
+    return null;
+}
+
+const Domains = enum {
+    Browser,
+    Target,
+    Page,
+    Log,
+    Runtime,
+    Network,
+    DOM,
+    CSS,
+    Inspector,
+    Emulation,
+    Fetch,
+    Performance,
+    Security,
+};
+
+// The caller is responsible for calling `free` on the returned slice.
+pub fn do(
+    alloc: std.mem.Allocator,
+    s: []const u8,
+    ctx: *Ctx,
+) anyerror![]const u8 {
+
+    // incoming message parser
+    var msg = IncomingMessage.init(alloc, s);
+    defer msg.deinit();
+
+    return dispatch(alloc, &msg, ctx);
+}
+
+pub fn dispatch(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) anyerror![]const u8 {
+    const method = try msg.getMethod();
+
+    // retrieve domain from method
+    var iter = std.mem.splitScalar(u8, method, '.');
+    const domain = std.meta.stringToEnum(Domains, iter.first()) orelse
+        return error.UnknonwDomain;
+
+    // select corresponding domain
+    const action = iter.next() orelse return error.BadMethod;
+    return switch (domain) {
+        .Browser => browser(alloc, msg, action, ctx),
+        .Target => target(alloc, msg, action, ctx),
+        .Page => page(alloc, msg, action, ctx),
+        .Log => log(alloc, msg, action, ctx),
+        .Runtime => runtime(alloc, msg, action, ctx),
+        .Network => network(alloc, msg, action, ctx),
+        .DOM => dom(alloc, msg, action, ctx),
+        .CSS => css(alloc, msg, action, ctx),
+        .Inspector => inspector(alloc, msg, action, ctx),
+        .Emulation => emulation(alloc, msg, action, ctx),
+        .Fetch => fetch(alloc, msg, action, ctx),
+        .Performance => performance(alloc, msg, action, ctx),
+        .Security => security(alloc, msg, action, ctx),
+    };
+}
+
+pub const State = struct {
+    executionContextId: u32 = 0,
+    contextID: ?[]const u8 = null,
+    sessionID: SessionID = .CONTEXTSESSIONID0497A05C95417CF4,
+    frameID: []const u8 = FrameID,
+    url: []const u8 = URLBase,
+    securityOrigin: []const u8 = URLBase,
+    secureContextType: []const u8 = "Secure", // TODO: enum
+    loaderID: []const u8 = LoaderID,
+
+    page_life_cycle_events: bool = false, // TODO; Target based value
+
+    // DOM
+    nodelist: cdpdom.NodeList,
+    nodesearchlist: cdpdom.NodeSearchList,
+
+    pub fn init(alloc: std.mem.Allocator) State {
+        return .{
+            .nodelist = cdpdom.NodeList.init(alloc),
+            .nodesearchlist = cdpdom.NodeSearchList.init(alloc),
+        };
+    }
+
+    pub fn deinit(self: *State) void {
+        self.nodelist.deinit();
+
+        // deinit all node searches.
+        for (self.nodesearchlist.items) |*s| s.deinit();
+        self.nodesearchlist.deinit();
+    }
+
+    pub fn reset(self: *State) void {
+        self.nodelist.reset();
+
+        // deinit all node searches.
+        for (self.nodesearchlist.items) |*s| s.deinit();
+        self.nodesearchlist.clearAndFree();
+    }
+};
+
+// Utils
+// -----
+
+pub fn dumpFile(
+    alloc: std.mem.Allocator,
+    id: u16,
+    script: []const u8,
+) !void {
+    const name = try std.fmt.allocPrint(alloc, "id_{d}.js", .{id});
+    defer alloc.free(name);
+    var dir = try std.fs.cwd().makeOpenPath("zig-cache/tmp", .{});
+    defer dir.close();
+    const f = try dir.createFile(name, .{});
+    defer f.close();
+    const nb = try f.write(script);
+    std.debug.assert(nb == script.len);
+    const p = try dir.realpathAlloc(alloc, name);
+    defer alloc.free(p);
+}
+
+// caller owns the slice returned
+pub fn stringify(alloc: std.mem.Allocator, res: anytype) ![]const u8 {
+    var out = std.ArrayList(u8).init(alloc);
+    defer out.deinit();
+
+    // Do not emit optional null fields
+    const options: std.json.StringifyOptions = .{ .emit_null_optional_fields = false };
+
+    try std.json.stringify(res, options, out.writer());
+    const ret = try alloc.alloc(u8, out.items.len);
+    @memcpy(ret, out.items);
+    return ret;
+}
+
+const resultNull = "{{\"id\": {d}, \"result\": {{}}}}";
+const resultNullSession = "{{\"id\": {d}, \"result\": {{}}, \"sessionId\": \"{s}\"}}";
+
+// caller owns the slice returned
+pub fn result(
+    alloc: std.mem.Allocator,
+    id: u16,
+    comptime T: ?type,
+    res: anytype,
+    sessionID: ?[]const u8,
+) ![]const u8 {
+    log_cdp.debug(
+        "Res > id {d}, sessionID {?s}, result {any}",
+        .{ id, sessionID, res },
+    );
+    if (T == null) {
+        // No need to stringify a custom JSON msg, just use string templates
+        if (sessionID) |sID| {
+            return try std.fmt.allocPrint(alloc, resultNullSession, .{ id, sID });
+        }
+        return try std.fmt.allocPrint(alloc, resultNull, .{id});
+    }
+
+    const Resp = struct {
+        id: u16,
+        result: T.?,
+        sessionId: ?[]const u8,
+    };
+    const resp = Resp{ .id = id, .result = res, .sessionId = sessionID };
+
+    return stringify(alloc, resp);
+}
+
+pub fn sendEvent(
+    alloc: std.mem.Allocator,
+    ctx: *Ctx,
+    name: []const u8,
+    comptime T: type,
+    params: T,
+    sessionID: ?[]const u8,
+) !void {
+    // some clients like chromedp expects empty parameters structs.
+    if (T == void) @compileError("sendEvent: use struct{} instead of void for empty parameters");
+
+    log_cdp.debug("Event > method {s}, sessionID {?s}", .{ name, sessionID });
+    const Resp = struct {
+        method: []const u8,
+        params: T,
+        sessionId: ?[]const u8,
+    };
+    const resp = Resp{ .method = name, .params = params, .sessionId = sessionID };
+
+    const event_msg = try stringify(alloc, resp);
+    try ctx.send(event_msg);
+}
+
+// Common
+// ------
+
+// TODO: hard coded IDs
+pub const SessionID = enum {
+    BROWSERSESSIONID597D9875C664CAC0,
+    CONTEXTSESSIONID0497A05C95417CF4,
+
+    pub fn parse(str: []const u8) !SessionID {
+        inline for (@typeInfo(SessionID).Enum.fields) |enumField| {
+            if (std.mem.eql(u8, str, enumField.name)) {
+                return @field(SessionID, enumField.name);
+            }
+        }
+        return error.InvalidSessionID;
+    }
+};
+pub const BrowserSessionID = @tagName(SessionID.BROWSERSESSIONID597D9875C664CAC0);
+pub const ContextSessionID = @tagName(SessionID.CONTEXTSESSIONID0497A05C95417CF4);
+pub const URLBase = "chrome://newtab/";
+pub const LoaderID = "LOADERID24DD2FD56CF1EF33C965C79C";
+pub const FrameID = "FRAMEIDD8AED408A0467AC93100BCDBE";
+
+pub const TimestampEvent = struct {
+    timestamp: f64,
+};

src/cdp/css.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+};
+
+pub fn css(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "inspector.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/dom.zig

@@ -0,0 +1,342 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+const css = @import("../dom/css.zig");
+
+const parser = @import("netsurf");
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+    getDocument,
+    performSearch,
+    getSearchResults,
+    discardSearchResults,
+};
+
+pub fn dom(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+        .getDocument => getDocument(alloc, msg, ctx),
+        .performSearch => performSearch(alloc, msg, ctx),
+        .getSearchResults => getSearchResults(alloc, msg, ctx),
+        .discardSearchResults => discardSearchResults(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "inspector.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// NodeList references tree nodes with an array id.
+pub const NodeList = struct {
+    coll: List,
+
+    const List = std.ArrayList(*parser.Node);
+
+    pub fn init(alloc: std.mem.Allocator) NodeList {
+        return .{
+            .coll = List.init(alloc),
+        };
+    }
+
+    pub fn deinit(self: *NodeList) void {
+        self.coll.deinit();
+    }
+
+    pub fn reset(self: *NodeList) void {
+        self.coll.clearAndFree();
+    }
+
+    pub fn set(self: *NodeList, node: *parser.Node) !NodeId {
+        for (self.coll.items, 0..) |n, i| {
+            if (n == node) return @intCast(i);
+        }
+
+        try self.coll.append(node);
+        return @intCast(self.coll.items.len);
+    }
+};
+
+const NodeId = u32;
+
+const Node = struct {
+    nodeId: NodeId,
+    parentId: ?NodeId = null,
+    backendNodeId: NodeId,
+    nodeType: u32,
+    nodeName: []const u8 = "",
+    localName: []const u8 = "",
+    nodeValue: []const u8 = "",
+    childNodeCount: ?u32 = null,
+    children: ?[]const Node = null,
+    documentURL: ?[]const u8 = null,
+    baseURL: ?[]const u8 = null,
+    xmlVersion: []const u8 = "",
+    compatibilityMode: []const u8 = "NoQuirksMode",
+    isScrollable: bool = false,
+
+    fn init(n: *parser.Node, nlist: *NodeList) !Node {
+        const id = try nlist.set(n);
+        return .{
+            .nodeId = id,
+            .backendNodeId = id,
+            .nodeType = @intFromEnum(try parser.nodeType(n)),
+            .nodeName = try parser.nodeName(n),
+            .localName = try parser.nodeLocalName(n),
+            .nodeValue = try parser.nodeValue(n) orelse "",
+        };
+    }
+
+    fn initChildren(
+        self: *Node,
+        alloc: std.mem.Allocator,
+        n: *parser.Node,
+        nlist: *NodeList,
+    ) !std.ArrayList(Node) {
+        const children = try parser.nodeGetChildNodes(n);
+        const ln = try parser.nodeListLength(children);
+        self.childNodeCount = ln;
+
+        var list = try std.ArrayList(Node).initCapacity(alloc, ln);
+
+        var i: u32 = 0;
+        while (i < ln) {
+            defer i += 1;
+            const child = try parser.nodeListItem(children, i) orelse continue;
+            try list.append(try Node.init(child, nlist));
+        }
+
+        self.children = list.items;
+
+        return list;
+    }
+};
+
+// https://chromedevtools.github.io/devtools-protocol/tot/DOM/#method-getDocument
+fn getDocument(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        depth: ?u32 = null,
+        pierce: ?bool = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "DOM.getDocument" });
+
+    // retrieve the root node
+    const page = ctx.browser.currentPage() orelse return error.NoPage;
+
+    if (page.doc == null) return error.NoDocument;
+
+    const node = parser.documentToNode(page.doc.?);
+    var n = try Node.init(node, &ctx.state.nodelist);
+    var list = try n.initChildren(alloc, node, &ctx.state.nodelist);
+    defer list.deinit();
+
+    // output
+    const Resp = struct {
+        root: Node,
+    };
+    const resp: Resp = .{
+        .root = n,
+    };
+
+    const res = try result(alloc, input.id, Resp, resp, input.sessionId);
+    try ctx.send(res);
+
+    return "";
+}
+
+pub const NodeSearch = struct {
+    coll: List,
+    name: []u8,
+    alloc: std.mem.Allocator,
+
+    var count: u8 = 0;
+
+    const List = std.ArrayListUnmanaged(NodeId);
+
+    pub fn initCapacity(alloc: std.mem.Allocator, ln: usize) !NodeSearch {
+        count += 1;
+
+        return .{
+            .alloc = alloc,
+            .coll = try List.initCapacity(alloc, ln),
+            .name = try std.fmt.allocPrint(alloc, "{d}", .{count}),
+        };
+    }
+
+    pub fn deinit(self: *NodeSearch) void {
+        self.coll.deinit(self.alloc);
+        self.alloc.free(self.name);
+    }
+
+    pub fn append(self: *NodeSearch, id: NodeId) !void {
+        try self.coll.append(self.alloc, id);
+    }
+};
+pub const NodeSearchList = std.ArrayList(NodeSearch);
+
+// https://chromedevtools.github.io/devtools-protocol/tot/DOM/#method-performSearch
+fn performSearch(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        query: []const u8,
+        includeUserAgentShadowDOM: ?bool = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "DOM.performSearch" });
+
+    // retrieve the root node
+    const page = ctx.browser.currentPage() orelse return error.NoPage;
+
+    if (page.doc == null) return error.NoDocument;
+
+    const list = try css.querySelectorAll(alloc, parser.documentToNode(page.doc.?), input.params.query);
+    const ln = list.nodes.items.len;
+    var ns = try NodeSearch.initCapacity(alloc, ln);
+
+    for (list.nodes.items) |n| {
+        const id = try ctx.state.nodelist.set(n);
+        try ns.append(id);
+    }
+
+    try ctx.state.nodesearchlist.append(ns);
+
+    // output
+    const Resp = struct {
+        searchId: []const u8,
+        resultCount: u32,
+    };
+    const resp: Resp = .{
+        .searchId = ns.name,
+        .resultCount = @intCast(ln),
+    };
+
+    return result(alloc, input.id, Resp, resp, input.sessionId);
+}
+
+// https://chromedevtools.github.io/devtools-protocol/tot/DOM/#method-discardSearchResults
+fn discardSearchResults(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        searchId: []const u8,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "DOM.discardSearchResults" });
+
+    // retrieve the search from context
+    for (ctx.state.nodesearchlist.items, 0..) |*s, i| {
+        if (!std.mem.eql(u8, s.name, input.params.searchId)) continue;
+
+        s.deinit();
+        _ = ctx.state.nodesearchlist.swapRemove(i);
+        break;
+    }
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// https://chromedevtools.github.io/devtools-protocol/tot/DOM/#method-getSearchResults
+fn getSearchResults(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        searchId: []const u8,
+        fromIndex: u32,
+        toIndex: u32,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "DOM.getSearchResults" });
+
+    if (input.params.fromIndex >= input.params.toIndex) return error.BadIndices;
+
+    // retrieve the search from context
+    var ns: ?*const NodeSearch = undefined;
+    for (ctx.state.nodesearchlist.items) |s| {
+        if (!std.mem.eql(u8, s.name, input.params.searchId)) continue;
+
+        ns = &s;
+        break;
+    }
+
+    if (ns == null) return error.searchResultNotFound;
+    const items = ns.?.coll.items;
+
+    if (input.params.fromIndex >= items.len) return error.BadFromIndex;
+    if (input.params.toIndex > items.len) return error.BadToIndex;
+
+    // output
+    const Resp = struct {
+        nodeIds: []NodeId,
+    };
+    const resp: Resp = .{
+        .nodeIds = ns.?.coll.items[input.params.fromIndex..input.params.toIndex],
+    };
+
+    return result(alloc, input.id, Resp, resp, input.sessionId);
+}

src/cdp/emulation.zig

@@ -0,0 +1,123 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const stringify = cdp.stringify;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    setEmulatedMedia,
+    setFocusEmulationEnabled,
+    setDeviceMetricsOverride,
+    setTouchEmulationEnabled,
+};
+
+pub fn emulation(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+    return switch (method) {
+        .setEmulatedMedia => setEmulatedMedia(alloc, msg, ctx),
+        .setFocusEmulationEnabled => setFocusEmulationEnabled(alloc, msg, ctx),
+        .setDeviceMetricsOverride => setDeviceMetricsOverride(alloc, msg, ctx),
+        .setTouchEmulationEnabled => setTouchEmulationEnabled(alloc, msg, ctx),
+    };
+}
+
+const MediaFeature = struct {
+    name: []const u8,
+    value: []const u8,
+};
+
+// TODO: noop method
+fn setEmulatedMedia(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+
+    // input
+    const Params = struct {
+        media: ?[]const u8 = null,
+        features: ?[]MediaFeature = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "emulation.setEmulatedMedia" });
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// TODO: noop method
+fn setFocusEmulationEnabled(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        enabled: bool,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "emulation.setFocusEmulationEnabled" });
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// TODO: noop method
+fn setDeviceMetricsOverride(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "emulation.setDeviceMetricsOverride" });
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// TODO: noop method
+fn setTouchEmulationEnabled(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "emulation.setTouchEmulationEnabled" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/fetch.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    disable,
+};
+
+pub fn fetch(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .disable => disable(alloc, msg, ctx),
+    };
+}
+
+// TODO: noop method
+fn disable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "fetch.disable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/inspector.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+};
+
+pub fn inspector(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "inspector.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/log.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+const stringify = cdp.stringify;
+
+const log_cdp = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+};
+
+pub fn log(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log_cdp.debug("Req > id {d}, method {s}", .{ input.id, "log.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/msg.zig

@@ -0,0 +1,293 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+// Parse incoming protocol message in json format.
+pub const IncomingMessage = struct {
+    scanner: std.json.Scanner,
+    json: []const u8,
+
+    obj_begin: bool = false,
+    obj_end: bool = false,
+
+    id: ?u16 = null,
+    scan_sessionId: bool = false,
+    sessionId: ?[]const u8 = null,
+    method: ?[]const u8 = null,
+    params_skip: bool = false,
+
+    pub fn init(alloc: std.mem.Allocator, json: []const u8) IncomingMessage {
+        return .{
+            .json = json,
+            .scanner = std.json.Scanner.initCompleteInput(alloc, json),
+        };
+    }
+
+    pub fn deinit(self: *IncomingMessage) void {
+        self.scanner.deinit();
+    }
+
+    fn scanUntil(self: *IncomingMessage, key: []const u8) !void {
+        while (true) {
+            switch (try self.scanner.next()) {
+                .end_of_document => return error.EndOfDocument,
+                .object_begin => {
+                    if (self.obj_begin) return error.InvalidObjectBegin;
+                    self.obj_begin = true;
+                },
+                .object_end => {
+                    if (!self.obj_begin) return error.InvalidObjectEnd;
+                    if (self.obj_end) return error.InvalidObjectEnd;
+                    self.obj_end = true;
+                },
+                .string => |s| {
+                    // is the key what we expects?
+                    if (std.mem.eql(u8, s, key)) return;
+
+                    // save other known keys
+                    if (std.mem.eql(u8, s, "id")) try self.scanId();
+                    if (std.mem.eql(u8, s, "sessionId")) try self.scanSessionId();
+                    if (std.mem.eql(u8, s, "method")) try self.scanMethod();
+                    if (std.mem.eql(u8, s, "params")) try self.scanParams();
+
+                    // TODO should we skip unknown key?
+                },
+                else => return error.InvalidToken,
+            }
+        }
+    }
+
+    fn scanId(self: *IncomingMessage) !void {
+        const t = try self.scanner.next();
+        if (t != .number) return error.InvalidId;
+        self.id = try std.fmt.parseUnsigned(u16, t.number, 10);
+    }
+
+    fn getId(self: *IncomingMessage) !u16 {
+        if (self.id != null) return self.id.?;
+
+        try self.scanUntil("id");
+        try self.scanId();
+        return self.id.?;
+    }
+
+    fn scanSessionId(self: *IncomingMessage) !void {
+        switch (try self.scanner.next()) {
+            // session id can be null.
+            .null => return,
+            .string => |s| self.sessionId = s,
+            else => return error.InvalidSessionId,
+        }
+
+        self.scan_sessionId = true;
+    }
+
+    fn getSessionId(self: *IncomingMessage) !?[]const u8 {
+        if (self.scan_sessionId) return self.sessionId;
+
+        self.scanUntil("sessionId") catch |err| {
+            if (err != error.EndOfDocument) return err;
+            // if the document doesn't contains any session id key, we must
+            // return null value.
+            self.scan_sessionId = true;
+            return null;
+        };
+        try self.scanSessionId();
+        return self.sessionId;
+    }
+
+    fn scanMethod(self: *IncomingMessage) !void {
+        const t = try self.scanner.next();
+        if (t != .string) return error.InvalidMethod;
+        self.method = t.string;
+    }
+
+    pub fn getMethod(self: *IncomingMessage) ![]const u8 {
+        if (self.method != null) return self.method.?;
+
+        try self.scanUntil("method");
+        try self.scanMethod();
+        return self.method.?;
+    }
+
+    // scanParams skip found parameters b/c if we encounter params *before*
+    // asking for getParams, we don't know how to parse them.
+    fn scanParams(self: *IncomingMessage) !void {
+        const tt = try self.scanner.peekNextTokenType();
+        // accept object begin or null JSON value.
+        if (tt != .object_begin and tt != .null) return error.InvalidParams;
+        try self.scanner.skipValue();
+        self.params_skip = true;
+    }
+
+    // getParams restart the JSON parsing
+    fn getParams(self: *IncomingMessage, alloc: ?std.mem.Allocator, T: type) !T {
+        if (T == void) return void{};
+        std.debug.assert(alloc != null); // if T is not void, alloc should not be null
+
+        if (self.params_skip) {
+            // TODO if the params have been skipped, we have to retart the
+            // parsing from start.
+            return error.SkippedParams;
+        }
+
+        self.scanUntil("params") catch |err| {
+            // handle nullable type
+            if (@typeInfo(T) == .Optional) {
+                if (err == error.InvalidToken or err == error.EndOfDocument) {
+                    return null;
+                }
+            }
+            return err;
+        };
+
+        // parse "params"
+        const options = std.json.ParseOptions{
+            .ignore_unknown_fields = true,
+            .max_value_len = self.scanner.input.len,
+            .allocate = .alloc_always,
+        };
+        return try std.json.innerParse(T, alloc.?, &self.scanner, options);
+    }
+};
+
+pub fn Input(T: type) type {
+    return struct {
+        arena: ?*std.heap.ArenaAllocator = null,
+        id: u16,
+        params: T,
+        sessionId: ?[]const u8,
+
+        const Self = @This();
+
+        pub fn get(alloc: std.mem.Allocator, msg: *IncomingMessage) !Self {
+            var arena: ?*std.heap.ArenaAllocator = null;
+            var allocator: ?std.mem.Allocator = null;
+
+            if (T != void) {
+                arena = try alloc.create(std.heap.ArenaAllocator);
+                arena.?.* = std.heap.ArenaAllocator.init(alloc);
+                allocator = arena.?.allocator();
+            }
+
+            errdefer {
+                if (arena) |_arena| {
+                    _arena.deinit();
+                    alloc.destroy(_arena);
+                }
+            }
+
+            return .{
+                .arena = arena,
+                .params = try msg.getParams(allocator, T),
+                .id = try msg.getId(),
+                .sessionId = try msg.getSessionId(),
+            };
+        }
+
+        pub fn deinit(self: Self) void {
+            if (self.arena) |arena| {
+                const allocator = arena.child_allocator;
+                arena.deinit();
+                allocator.destroy(arena);
+            }
+        }
+    };
+}
+
+test "read incoming message" {
+    const inputs = [_][]const u8{
+        \\{"id":1,"method":"foo","sessionId":"bar","params":{"bar":"baz"}}
+        ,
+        \\{"params":{"bar":"baz"},"id":1,"method":"foo","sessionId":"bar"}
+        ,
+        \\{"sessionId":"bar","params":{"bar":"baz"},"id":1,"method":"foo"}
+        ,
+        \\{"method":"foo","sessionId":"bar","params":{"bar":"baz"},"id":1}
+        ,
+    };
+
+    for (inputs) |input| {
+        var msg = IncomingMessage.init(std.testing.allocator, input);
+        defer msg.deinit();
+
+        try std.testing.expectEqual(1, try msg.getId());
+        try std.testing.expectEqualSlices(u8, "foo", try msg.getMethod());
+        try std.testing.expectEqualSlices(u8, "bar", (try msg.getSessionId()).?);
+
+        const T = struct { bar: []const u8 };
+        const in = Input(T).get(std.testing.allocator, &msg) catch |err| {
+            if (err != error.SkippedParams) return err;
+            // TODO remove this check when params in the beginning is handled.
+            continue;
+        };
+        defer in.deinit();
+        try std.testing.expectEqualSlices(u8, "baz", in.params.bar);
+    }
+}
+
+test "read incoming message with null session id" {
+    const inputs = [_][]const u8{
+        \\{"id":1}
+        ,
+        \\{"params":{"bar":"baz"},"id":1,"method":"foo"}
+        ,
+        \\{"sessionId":null,"params":{"bar":"baz"},"id":1,"method":"foo"}
+        ,
+    };
+
+    for (inputs) |input| {
+        var msg = IncomingMessage.init(std.testing.allocator, input);
+        defer msg.deinit();
+
+        try std.testing.expect(try msg.getSessionId() == null);
+        try std.testing.expectEqual(1, try msg.getId());
+    }
+}
+
+test "message with nullable params" {
+    const T = struct {
+        bar: []const u8,
+    };
+
+    // nullable type, params is present => value
+    const not_null =
+        \\{"id": 1,"method":"foo","params":{"bar":"baz"}}
+    ;
+    var msg = IncomingMessage.init(std.testing.allocator, not_null);
+    defer msg.deinit();
+    const input = try Input(?T).get(std.testing.allocator, &msg);
+    defer input.deinit();
+    try std.testing.expectEqualStrings(input.params.?.bar, "baz");
+
+    // nullable type, params is not present => null
+    const is_null =
+        \\{"id": 1,"method":"foo","sessionId":"AAA"}
+    ;
+    var msg_null = IncomingMessage.init(std.testing.allocator, is_null);
+    defer msg_null.deinit();
+    const input_null = try Input(?T).get(std.testing.allocator, &msg_null);
+    defer input_null.deinit();
+    try std.testing.expectEqual(null, input_null.params);
+    try std.testing.expectEqualStrings("AAA", input_null.sessionId.?);
+
+    // not nullable type, params is not present => error
+    const params_or_error = msg_null.getParams(std.testing.allocator, T);
+    try std.testing.expectError(error.EndOfDocument, params_or_error);
+}

src/cdp/network.zig

@@ -0,0 +1,75 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+    setCacheDisabled,
+};
+
+pub fn network(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+        .setCacheDisabled => setCacheDisabled(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "network.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// TODO: noop method
+fn setCacheDisabled(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "network.setCacheDisabled" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/page.zig

@@ -0,0 +1,464 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const stringify = cdp.stringify;
+const sendEvent = cdp.sendEvent;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Runtime = @import("runtime.zig");
+
+const Methods = enum {
+    enable,
+    getFrameTree,
+    setLifecycleEventsEnabled,
+    addScriptToEvaluateOnNewDocument,
+    createIsolatedWorld,
+    navigate,
+};
+
+pub fn page(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+        .getFrameTree => getFrameTree(alloc, msg, ctx),
+        .setLifecycleEventsEnabled => setLifecycleEventsEnabled(alloc, msg, ctx),
+        .addScriptToEvaluateOnNewDocument => addScriptToEvaluateOnNewDocument(alloc, msg, ctx),
+        .createIsolatedWorld => createIsolatedWorld(alloc, msg, ctx),
+        .navigate => navigate(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+const Frame = struct {
+    id: []const u8,
+    loaderId: []const u8,
+    url: []const u8,
+    domainAndRegistry: []const u8 = "",
+    securityOrigin: []const u8,
+    mimeType: []const u8 = "text/html",
+    adFrameStatus: struct {
+        adFrameType: []const u8 = "none",
+    } = .{},
+    secureContextType: []const u8,
+    crossOriginIsolatedContextType: []const u8 = "NotIsolated",
+    gatedAPIFeatures: [][]const u8 = &[0][]const u8{},
+};
+
+fn getFrameTree(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.getFrameTree" });
+
+    // output
+    const FrameTree = struct {
+        frameTree: struct {
+            frame: Frame,
+        },
+        childFrames: ?[]@This() = null,
+
+        pub fn format(
+            self: @This(),
+            comptime _: []const u8,
+            options: std.fmt.FormatOptions,
+            writer: anytype,
+        ) !void {
+            try writer.writeAll("cdp.page.getFrameTree { ");
+            try writer.writeAll(".frameTree = { ");
+            try writer.writeAll(".frame = { ");
+            const frame = self.frameTree.frame;
+            try writer.writeAll(".id = ");
+            try std.fmt.formatText(frame.id, "s", options, writer);
+            try writer.writeAll(", .loaderId = ");
+            try std.fmt.formatText(frame.loaderId, "s", options, writer);
+            try writer.writeAll(", .url = ");
+            try std.fmt.formatText(frame.url, "s", options, writer);
+            try writer.writeAll(" } } }");
+        }
+    };
+    const frameTree = FrameTree{
+        .frameTree = .{
+            .frame = .{
+                .id = ctx.state.frameID,
+                .url = ctx.state.url,
+                .securityOrigin = ctx.state.securityOrigin,
+                .secureContextType = ctx.state.secureContextType,
+                .loaderId = ctx.state.loaderID,
+            },
+        },
+    };
+    return result(alloc, input.id, FrameTree, frameTree, input.sessionId);
+}
+
+fn setLifecycleEventsEnabled(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        enabled: bool,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.setLifecycleEventsEnabled" });
+
+    ctx.state.page_life_cycle_events = true;
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+const LifecycleEvent = struct {
+    frameId: []const u8,
+    loaderId: ?[]const u8,
+    name: []const u8 = undefined,
+    timestamp: f32 = undefined,
+};
+
+// TODO: hard coded method
+fn addScriptToEvaluateOnNewDocument(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        source: []const u8,
+        worldName: ?[]const u8 = null,
+        includeCommandLineAPI: bool = false,
+        runImmediately: bool = false,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.addScriptToEvaluateOnNewDocument" });
+
+    // output
+    const Res = struct {
+        identifier: []const u8 = "1",
+
+        pub fn format(
+            self: @This(),
+            comptime _: []const u8,
+            options: std.fmt.FormatOptions,
+            writer: anytype,
+        ) !void {
+            try writer.writeAll("cdp.page.addScriptToEvaluateOnNewDocument { ");
+            try writer.writeAll(".identifier = ");
+            try std.fmt.formatText(self.identifier, "s", options, writer);
+            try writer.writeAll(" }");
+        }
+    };
+    return result(alloc, input.id, Res, Res{}, input.sessionId);
+}
+
+// TODO: hard coded method
+fn createIsolatedWorld(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        frameId: []const u8,
+        worldName: []const u8,
+        grantUniveralAccess: bool,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.createIsolatedWorld" });
+
+    // noop executionContextCreated event
+    try Runtime.executionContextCreated(
+        alloc,
+        ctx,
+        0,
+        "",
+        input.params.worldName,
+        // TODO: hard coded ID
+        "7102379147004877974.3265385113993241162",
+        .{
+            .isDefault = false,
+            .type = "isolated",
+            .frameId = input.params.frameId,
+        },
+        input.sessionId,
+    );
+
+    // output
+    const Resp = struct {
+        executionContextId: u8 = 0,
+    };
+
+    return result(alloc, input.id, Resp, .{}, input.sessionId);
+}
+
+fn navigate(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        url: []const u8,
+        referrer: ?[]const u8 = null,
+        transitionType: ?[]const u8 = null, // TODO: enum
+        frameId: ?[]const u8 = null,
+        referrerPolicy: ?[]const u8 = null, // TODO: enum
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    std.debug.assert(input.sessionId != null);
+    log.debug("Req > id {d}, method {s}", .{ input.id, "page.navigate" });
+
+    // change state
+    ctx.state.reset();
+    ctx.state.url = input.params.url;
+    // TODO: hard coded ID
+    ctx.state.loaderID = "AF8667A203C5392DBE9AC290044AA4C2";
+
+    var life_event = LifecycleEvent{
+        .frameId = ctx.state.frameID,
+        .loaderId = ctx.state.loaderID,
+    };
+    var ts_event: cdp.TimestampEvent = undefined;
+
+    // frameStartedLoading event
+    // TODO: event partially hard coded
+    const FrameStartedLoading = struct {
+        frameId: []const u8,
+    };
+    const frame_started_loading = FrameStartedLoading{ .frameId = ctx.state.frameID };
+    try sendEvent(
+        alloc,
+        ctx,
+        "Page.frameStartedLoading",
+        FrameStartedLoading,
+        frame_started_loading,
+        input.sessionId,
+    );
+    if (ctx.state.page_life_cycle_events) {
+        life_event.name = "init";
+        life_event.timestamp = 343721.796037;
+        try sendEvent(
+            alloc,
+            ctx,
+            "Page.lifecycleEvent",
+            LifecycleEvent,
+            life_event,
+            input.sessionId,
+        );
+    }
+
+    // output
+    const Resp = struct {
+        frameId: []const u8,
+        loaderId: ?[]const u8,
+        errorText: ?[]const u8 = null,
+
+        pub fn format(
+            self: @This(),
+            comptime _: []const u8,
+            options: std.fmt.FormatOptions,
+            writer: anytype,
+        ) !void {
+            try writer.writeAll("cdp.page.navigate.Resp { ");
+            try writer.writeAll(".frameId = ");
+            try std.fmt.formatText(self.frameId, "s", options, writer);
+            if (self.loaderId) |loaderId| {
+                try writer.writeAll(", .loaderId = '");
+                try std.fmt.formatText(loaderId, "s", options, writer);
+            }
+            try writer.writeAll(" }");
+        }
+    };
+    const resp = Resp{
+        .frameId = ctx.state.frameID,
+        .loaderId = ctx.state.loaderID,
+    };
+    const res = try result(alloc, input.id, Resp, resp, input.sessionId);
+    try ctx.send(res);
+
+    // TODO: at this point do we need async the following actions to be async?
+
+    // Send Runtime.executionContextsCleared event
+    // TODO: noop event, we have no env context at this point, is it necesarry?
+    try sendEvent(alloc, ctx, "Runtime.executionContextsCleared", struct {}, .{}, input.sessionId);
+
+    // Launch navigate, the page must have been created by a
+    // target.createTarget.
+    var p = ctx.browser.currentPage() orelse return error.NoPage;
+    ctx.state.executionContextId += 1;
+    const auxData = try std.fmt.allocPrint(
+        alloc,
+        // NOTE: we assume this is the default web page
+        "{{\"isDefault\":true,\"type\":\"default\",\"frameId\":\"{s}\"}}",
+        .{ctx.state.frameID},
+    );
+    defer alloc.free(auxData);
+    try p.navigate(input.params.url, auxData);
+
+    // Events
+
+    // lifecycle init event
+    // TODO: partially hard coded
+    if (ctx.state.page_life_cycle_events) {
+        life_event.name = "init";
+        life_event.timestamp = 343721.796037;
+        try sendEvent(
+            alloc,
+            ctx,
+            "Page.lifecycleEvent",
+            LifecycleEvent,
+            life_event,
+            input.sessionId,
+        );
+    }
+
+    // DOM.documentUpdated
+    try sendEvent(
+        alloc,
+        ctx,
+        "DOM.documentUpdated",
+        struct {},
+        .{},
+        input.sessionId,
+    );
+
+    // frameNavigated event
+    const FrameNavigated = struct {
+        frame: Frame,
+        type: []const u8 = "Navigation",
+    };
+    const frame_navigated = FrameNavigated{
+        .frame = .{
+            .id = ctx.state.frameID,
+            .url = ctx.state.url,
+            .securityOrigin = ctx.state.securityOrigin,
+            .secureContextType = ctx.state.secureContextType,
+            .loaderId = ctx.state.loaderID,
+        },
+    };
+    try sendEvent(
+        alloc,
+        ctx,
+        "Page.frameNavigated",
+        FrameNavigated,
+        frame_navigated,
+        input.sessionId,
+    );
+
+    // domContentEventFired event
+    // TODO: partially hard coded
+    ts_event = .{ .timestamp = 343721.803338 };
+    try sendEvent(
+        alloc,
+        ctx,
+        "Page.domContentEventFired",
+        cdp.TimestampEvent,
+        ts_event,
+        input.sessionId,
+    );
+
+    // lifecycle DOMContentLoaded event
+    // TODO: partially hard coded
+    if (ctx.state.page_life_cycle_events) {
+        life_event.name = "DOMContentLoaded";
+        life_event.timestamp = 343721.803338;
+        try sendEvent(
+            alloc,
+            ctx,
+            "Page.lifecycleEvent",
+            LifecycleEvent,
+            life_event,
+            input.sessionId,
+        );
+    }
+
+    // loadEventFired event
+    // TODO: partially hard coded
+    ts_event = .{ .timestamp = 343721.824655 };
+    try sendEvent(
+        alloc,
+        ctx,
+        "Page.loadEventFired",
+        cdp.TimestampEvent,
+        ts_event,
+        input.sessionId,
+    );
+
+    // lifecycle DOMContentLoaded event
+    // TODO: partially hard coded
+    if (ctx.state.page_life_cycle_events) {
+        life_event.name = "load";
+        life_event.timestamp = 343721.824655;
+        try sendEvent(
+            alloc,
+            ctx,
+            "Page.lifecycleEvent",
+            LifecycleEvent,
+            life_event,
+            input.sessionId,
+        );
+    }
+
+    // frameStoppedLoading
+    const FrameStoppedLoading = struct { frameId: []const u8 };
+    try sendEvent(
+        alloc,
+        ctx,
+        "Page.frameStoppedLoading",
+        FrameStoppedLoading,
+        .{ .frameId = ctx.state.frameID },
+        input.sessionId,
+    );
+
+    return "";
+}

src/cdp/performance.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+};
+
+pub fn performance(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "performance.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/runtime.zig

@@ -0,0 +1,195 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const builtin = @import("builtin");
+
+const jsruntime = @import("jsruntime");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+const stringify = cdp.stringify;
+const target = @import("target.zig");
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+    runIfWaitingForDebugger,
+    evaluate,
+    addBinding,
+    callFunctionOn,
+    releaseObject,
+};
+
+pub fn runtime(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        // NOTE: we could send it anyway to the JS runtime but it's good to check it
+        return error.UnknownMethod;
+    return switch (method) {
+        .runIfWaitingForDebugger => runIfWaitingForDebugger(alloc, msg, ctx),
+        else => sendInspector(alloc, method, msg, ctx),
+    };
+}
+
+fn sendInspector(
+    alloc: std.mem.Allocator,
+    method: Methods,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+
+    // save script in file at debug mode
+    if (std.log.defaultLogEnabled(.debug)) {
+
+        // input
+        var id: u16 = undefined;
+        var script: ?[]const u8 = null;
+
+        if (method == .evaluate) {
+            const Params = struct {
+                expression: []const u8,
+                contextId: ?u8 = null,
+                returnByValue: ?bool = null,
+                awaitPromise: ?bool = null,
+                userGesture: ?bool = null,
+            };
+
+            const input = try Input(Params).get(alloc, msg);
+            defer input.deinit();
+            log.debug("Req > id {d}, method {s} (script saved on cache)", .{ input.id, "runtime.evaluate" });
+            const params = input.params;
+            const func = try alloc.alloc(u8, params.expression.len);
+            @memcpy(func, params.expression);
+            script = func;
+            id = input.id;
+        } else if (method == .callFunctionOn) {
+            const Params = struct {
+                functionDeclaration: []const u8,
+                objectId: ?[]const u8 = null,
+                executionContextId: ?u8 = null,
+                arguments: ?[]struct {
+                    value: ?[]const u8 = null,
+                    objectId: ?[]const u8 = null,
+                } = null,
+                returnByValue: ?bool = null,
+                awaitPromise: ?bool = null,
+                userGesture: ?bool = null,
+            };
+
+            const input = try Input(Params).get(alloc, msg);
+            defer input.deinit();
+            log.debug("Req > id {d}, method {s} (script saved on cache)", .{ input.id, "runtime.callFunctionOn" });
+            const params = input.params;
+            const func = try alloc.alloc(u8, params.functionDeclaration.len);
+            @memcpy(func, params.functionDeclaration);
+            script = func;
+            id = input.id;
+        }
+
+        if (script) |src| {
+            try cdp.dumpFile(alloc, id, src);
+            alloc.free(src);
+        }
+    }
+
+    if (msg.sessionId) |s| {
+        ctx.state.sessionID = cdp.SessionID.parse(s) catch |err| {
+            log.err("parse sessionID: {s} {any}", .{ s, err });
+            return err;
+        };
+    }
+
+    // remove awaitPromise true params
+    // TODO: delete when Promise are correctly handled by zig-js-runtime
+    if (method == .callFunctionOn or method == .evaluate) {
+        if (std.mem.indexOf(u8, msg.json, "\"awaitPromise\":true")) |_| {
+            const buf = try alloc.alloc(u8, msg.json.len + 1);
+            defer alloc.free(buf);
+            _ = std.mem.replace(u8, msg.json, "\"awaitPromise\":true", "\"awaitPromise\":false", buf);
+            try ctx.sendInspector(buf);
+            return "";
+        }
+    }
+
+    try ctx.sendInspector(msg.json);
+
+    if (msg.id == null) return "";
+
+    return result(alloc, msg.id.?, null, null, msg.sessionId);
+}
+
+pub const AuxData = struct {
+    isDefault: bool = true,
+    type: []const u8 = "default",
+    frameId: []const u8 = cdp.FrameID,
+};
+
+pub fn executionContextCreated(
+    alloc: std.mem.Allocator,
+    ctx: *Ctx,
+    id: u16,
+    origin: []const u8,
+    name: []const u8,
+    uniqueID: []const u8,
+    auxData: ?AuxData,
+    sessionID: ?[]const u8,
+) !void {
+    const Params = struct {
+        context: struct {
+            id: u64,
+            origin: []const u8,
+            name: []const u8,
+            uniqueId: []const u8,
+            auxData: ?AuxData = null,
+        },
+    };
+    const params = Params{
+        .context = .{
+            .id = id,
+            .origin = origin,
+            .name = name,
+            .uniqueId = uniqueID,
+            .auxData = auxData,
+        },
+    };
+    try cdp.sendEvent(alloc, ctx, "Runtime.executionContextCreated", Params, params, sessionID);
+}
+
+// TODO: noop method
+// should we be passing this also to the JS Inspector?
+fn runIfWaitingForDebugger(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "runtime.runIfWaitingForDebugger" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/security.zig

@@ -0,0 +1,59 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    enable,
+};
+
+pub fn security(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+
+    return switch (method) {
+        .enable => enable(alloc, msg, ctx),
+    };
+}
+
+fn enable(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "security.enable" });
+
+    return result(alloc, input.id, null, null, input.sessionId);
+}

src/cdp/target.zig

@@ -0,0 +1,529 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const server = @import("../server.zig");
+const Ctx = server.Ctx;
+const cdp = @import("cdp.zig");
+const result = cdp.result;
+const stringify = cdp.stringify;
+const IncomingMessage = @import("msg.zig").IncomingMessage;
+const Input = @import("msg.zig").Input;
+
+const log = std.log.scoped(.cdp);
+
+const Methods = enum {
+    setDiscoverTargets,
+    setAutoAttach,
+    attachToTarget,
+    getTargetInfo,
+    getBrowserContexts,
+    createBrowserContext,
+    disposeBrowserContext,
+    createTarget,
+    closeTarget,
+    sendMessageToTarget,
+    detachFromTarget,
+};
+
+pub fn target(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    action: []const u8,
+    ctx: *Ctx,
+) ![]const u8 {
+    const method = std.meta.stringToEnum(Methods, action) orelse
+        return error.UnknownMethod;
+    return switch (method) {
+        .setDiscoverTargets => setDiscoverTargets(alloc, msg, ctx),
+        .setAutoAttach => setAutoAttach(alloc, msg, ctx),
+        .attachToTarget => attachToTarget(alloc, msg, ctx),
+        .getTargetInfo => getTargetInfo(alloc, msg, ctx),
+        .getBrowserContexts => getBrowserContexts(alloc, msg, ctx),
+        .createBrowserContext => createBrowserContext(alloc, msg, ctx),
+        .disposeBrowserContext => disposeBrowserContext(alloc, msg, ctx),
+        .createTarget => createTarget(alloc, msg, ctx),
+        .closeTarget => closeTarget(alloc, msg, ctx),
+        .sendMessageToTarget => sendMessageToTarget(alloc, msg, ctx),
+        .detachFromTarget => detachFromTarget(alloc, msg, ctx),
+    };
+}
+
+// TODO: hard coded IDs
+pub const PageTargetID = "PAGETARGETIDB638E9DC0F52DDC";
+pub const BrowserTargetID = "browser9-targ-et6f-id0e-83f3ab73a30c";
+pub const BrowserContextID = "BROWSERCONTEXTIDA95049E9DFE95EA9";
+
+// TODO: noop method
+fn setDiscoverTargets(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.setDiscoverTargets" });
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+const AttachToTarget = struct {
+    sessionId: []const u8,
+    targetInfo: struct {
+        targetId: []const u8,
+        type: []const u8 = "page",
+        title: []const u8,
+        url: []const u8,
+        attached: bool = true,
+        canAccessOpener: bool = false,
+        browserContextId: []const u8,
+    },
+    waitingForDebugger: bool = false,
+};
+
+const TargetCreated = struct {
+    sessionId: []const u8,
+    targetInfo: struct {
+        targetId: []const u8,
+        type: []const u8 = "page",
+        title: []const u8,
+        url: []const u8,
+        attached: bool = true,
+        canAccessOpener: bool = false,
+        browserContextId: []const u8,
+    },
+};
+
+const TargetFilter = struct {
+    type: ?[]const u8 = null,
+    exclude: ?bool = null,
+};
+
+// TODO: noop method
+fn setAutoAttach(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        autoAttach: bool,
+        waitForDebuggerOnStart: bool,
+        flatten: bool = true,
+        filter: ?[]TargetFilter = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.setAutoAttach" });
+
+    // attachedToTarget event
+    if (input.sessionId == null) {
+        const attached = AttachToTarget{
+            .sessionId = cdp.BrowserSessionID,
+            .targetInfo = .{
+                .targetId = PageTargetID,
+                .title = "about:blank",
+                .url = cdp.URLBase,
+                .browserContextId = BrowserContextID,
+            },
+        };
+        try cdp.sendEvent(alloc, ctx, "Target.attachedToTarget", AttachToTarget, attached, null);
+    }
+
+    // output
+    return result(alloc, input.id, null, null, input.sessionId);
+}
+
+// TODO: noop method
+fn attachToTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+
+    // input
+    const Params = struct {
+        targetId: []const u8,
+        flatten: bool = true,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.attachToTarget" });
+
+    // attachedToTarget event
+    if (input.sessionId == null) {
+        const attached = AttachToTarget{
+            .sessionId = cdp.BrowserSessionID,
+            .targetInfo = .{
+                .targetId = input.params.targetId,
+                .title = "about:blank",
+                .url = cdp.URLBase,
+                .browserContextId = BrowserContextID,
+            },
+        };
+        try cdp.sendEvent(alloc, ctx, "Target.attachedToTarget", AttachToTarget, attached, null);
+    }
+
+    // output
+    const SessionId = struct {
+        sessionId: []const u8,
+    };
+    const output = SessionId{
+        .sessionId = input.sessionId orelse cdp.BrowserSessionID,
+    };
+    return result(alloc, input.id, SessionId, output, null);
+}
+
+fn getTargetInfo(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        targetId: ?[]const u8 = null,
+    };
+    const input = try Input(?Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.getTargetInfo" });
+
+    // output
+    const TargetInfo = struct {
+        targetId: []const u8,
+        type: []const u8,
+        title: []const u8 = "",
+        url: []const u8 = "",
+        attached: bool = true,
+        openerId: ?[]const u8 = null,
+        canAccessOpener: bool = false,
+        openerFrameId: ?[]const u8 = null,
+        browserContextId: ?[]const u8 = null,
+        subtype: ?[]const u8 = null,
+    };
+    const targetInfo = TargetInfo{
+        .targetId = BrowserTargetID,
+        .type = "browser",
+    };
+    return result(alloc, input.id, TargetInfo, targetInfo, null);
+}
+
+// Browser context are not handled and not in the roadmap for now
+// The following methods are "fake"
+
+// TODO: noop method
+fn getBrowserContexts(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.getBrowserContexts" });
+
+    // ouptut
+    const Resp = struct {
+        browserContextIds: [][]const u8,
+    };
+    var resp: Resp = undefined;
+    if (ctx.state.contextID) |contextID| {
+        var contextIDs = [1][]const u8{contextID};
+        resp = .{ .browserContextIds = &contextIDs };
+    } else {
+        const contextIDs = [0][]const u8{};
+        resp = .{ .browserContextIds = &contextIDs };
+    }
+    return result(alloc, input.id, Resp, resp, null);
+}
+
+const ContextID = "CONTEXTIDDCCDD11109E2D4FEFBE4F89";
+
+// TODO: noop method
+fn createBrowserContext(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        disposeOnDetach: bool = false,
+        proxyServer: ?[]const u8 = null,
+        proxyBypassList: ?[]const u8 = null,
+        originsWithUniversalNetworkAccess: ?[][]const u8 = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.createBrowserContext" });
+
+    ctx.state.contextID = ContextID;
+
+    // output
+    const Resp = struct {
+        browserContextId: []const u8 = ContextID,
+
+        pub fn format(
+            self: @This(),
+            comptime _: []const u8,
+            options: std.fmt.FormatOptions,
+            writer: anytype,
+        ) !void {
+            try writer.writeAll("cdp.target.createBrowserContext { ");
+            try writer.writeAll(".browserContextId = ");
+            try std.fmt.formatText(self.browserContextId, "s", options, writer);
+            try writer.writeAll(" }");
+        }
+    };
+    return result(alloc, input.id, Resp, Resp{}, input.sessionId);
+}
+
+fn disposeBrowserContext(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        browserContextId: []const u8,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.disposeBrowserContext" });
+
+    // output
+    const res = try result(alloc, input.id, null, .{}, null);
+    try ctx.send(res);
+
+    return error.DisposeBrowserContext;
+}
+
+// TODO: hard coded IDs
+const TargetID = "TARGETID460A8F29706A2ADF14316298";
+const LoaderID = "LOADERID42AA389647D702B4D805F49A";
+
+fn createTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        url: []const u8,
+        width: ?u64 = null,
+        height: ?u64 = null,
+        browserContextId: ?[]const u8 = null,
+        enableBeginFrameControl: bool = false,
+        newWindow: bool = false,
+        background: bool = false,
+        forTab: ?bool = null,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.createTarget" });
+
+    // change CDP state
+    ctx.state.frameID = TargetID;
+    ctx.state.url = "about:blank";
+    ctx.state.securityOrigin = "://";
+    ctx.state.secureContextType = "InsecureScheme";
+    ctx.state.loaderID = LoaderID;
+
+    if (msg.sessionId) |s| {
+        ctx.state.sessionID = cdp.SessionID.parse(s) catch |err| {
+            log.err("parse sessionID: {s} {any}", .{ s, err });
+            return err;
+        };
+    }
+
+    // TODO stop the previous page instead?
+    if (ctx.browser.currentPage() != null) return error.pageAlreadyExists;
+
+    // create the page
+    const p = try ctx.browser.session.createPage();
+    ctx.state.executionContextId += 1;
+    // start the js env
+    const auxData = try std.fmt.allocPrint(
+        alloc,
+        // NOTE: we assume this is the default web page
+        "{{\"isDefault\":true,\"type\":\"default\",\"frameId\":\"{s}\"}}",
+        .{ctx.state.frameID},
+    );
+    defer alloc.free(auxData);
+    try p.start(auxData);
+
+    // send targetCreated event
+    const created = TargetCreated{
+        .sessionId = cdp.ContextSessionID,
+        .targetInfo = .{
+            .targetId = ctx.state.frameID,
+            .title = "about:blank",
+            .url = ctx.state.url,
+            .browserContextId = input.params.browserContextId orelse ContextID,
+            .attached = true,
+        },
+    };
+    try cdp.sendEvent(alloc, ctx, "Target.targetCreated", TargetCreated, created, input.sessionId);
+
+    // send attachToTarget event
+    const attached = AttachToTarget{
+        .sessionId = cdp.ContextSessionID,
+        .targetInfo = .{
+            .targetId = ctx.state.frameID,
+            .title = "about:blank",
+            .url = ctx.state.url,
+            .browserContextId = input.params.browserContextId orelse ContextID,
+            .attached = true,
+        },
+        .waitingForDebugger = true,
+    };
+    try cdp.sendEvent(alloc, ctx, "Target.attachedToTarget", AttachToTarget, attached, input.sessionId);
+
+    // output
+    const Resp = struct {
+        targetId: []const u8 = TargetID,
+
+        pub fn format(
+            self: @This(),
+            comptime _: []const u8,
+            options: std.fmt.FormatOptions,
+            writer: anytype,
+        ) !void {
+            try writer.writeAll("cdp.target.createTarget { ");
+            try writer.writeAll(".targetId = ");
+            try std.fmt.formatText(self.targetId, "s", options, writer);
+            try writer.writeAll(" }");
+        }
+    };
+    return result(alloc, input.id, Resp, Resp{}, input.sessionId);
+}
+
+fn closeTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        targetId: []const u8,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.closeTarget" });
+
+    // output
+    const Resp = struct {
+        success: bool = true,
+    };
+    const res = try result(alloc, input.id, Resp, Resp{}, null);
+    try ctx.send(res);
+
+    // Inspector.detached event
+    const InspectorDetached = struct {
+        reason: []const u8 = "Render process gone.",
+    };
+    try cdp.sendEvent(
+        alloc,
+        ctx,
+        "Inspector.detached",
+        InspectorDetached,
+        .{},
+        input.sessionId orelse cdp.ContextSessionID,
+    );
+
+    // detachedFromTarget event
+    const TargetDetached = struct {
+        sessionId: []const u8,
+        targetId: []const u8,
+    };
+    try cdp.sendEvent(
+        alloc,
+        ctx,
+        "Target.detachedFromTarget",
+        TargetDetached,
+        .{
+            .sessionId = input.sessionId orelse cdp.ContextSessionID,
+            .targetId = input.params.targetId,
+        },
+        null,
+    );
+
+    if (ctx.browser.currentPage()) |page| page.end();
+
+    return "";
+}
+
+fn sendMessageToTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    ctx: *Ctx,
+) ![]const u8 {
+    // input
+    const Params = struct {
+        message: []const u8,
+        sessionId: []const u8,
+    };
+    const input = try Input(Params).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s} ({s})", .{ input.id, "target.sendMessageToTarget", input.params.message });
+
+    // get the wrapped message.
+    var wmsg = IncomingMessage.init(alloc, input.params.message);
+    defer wmsg.deinit();
+
+    const res = cdp.dispatch(alloc, &wmsg, ctx) catch |e| {
+        log.err("send message {d} ({s}): {any}", .{ input.id, input.params.message, e });
+        // TODO dispatch error correctly.
+        return e;
+    };
+
+    // receivedMessageFromTarget event
+    const ReceivedMessageFromTarget = struct {
+        message: []const u8,
+        sessionId: []const u8,
+    };
+    try cdp.sendEvent(
+        alloc,
+        ctx,
+        "Target.receivedMessageFromTarget",
+        ReceivedMessageFromTarget,
+        .{
+            .message = res,
+            .sessionId = input.params.sessionId,
+        },
+        null,
+    );
+
+    return "";
+}
+
+// noop
+fn detachFromTarget(
+    alloc: std.mem.Allocator,
+    msg: *IncomingMessage,
+    _: *Ctx,
+) ![]const u8 {
+    // input
+    const input = try Input(void).get(alloc, msg);
+    defer input.deinit();
+    log.debug("Req > id {d}, method {s}", .{ input.id, "target.detachFromTarget" });
+
+    // output
+    return result(alloc, input.id, bool, true, input.sessionId);
+}

src/css/README.md

@@ -0,0 +1,218 @@
+# css
+
+Lightpanda css implements CSS selectors parsing and matching in Zig.
+This package is a port of the Go lib [andybalholm/cascadia](https://github.com/andybalholm/cascadia).
+
+## Usage
+
+### Query parser
+
+```zig
+const css = @import("css.zig");
+
+const selector = try css.parse(alloc, "h1", .{});
+defer selector.deinit(alloc);
+```
+
+### DOM tree match
+
+The lib expects a `Node` interface implementation to match your DOM tree.
+
+```zig
+pub const Node = struct {
+    pub fn firstChild(_: Node) !?Node {
+        return error.TODO;
+    }
+
+    pub fn lastChild(_: Node) !?Node {
+        return error.TODO;
+    }
+
+    pub fn nextSibling(_: Node) !?Node {
+        return error.TODO;
+    }
+
+    pub fn prevSibling(_: Node) !?Node {
+        return error.TODO;
+    }
+
+    pub fn parent(_: Node) !?Node {
+        return error.TODO;
+    }
+
+    pub fn isElement(_: Node) bool {
+        return false;
+    }
+
+    pub fn isDocument(_: Node) bool {
+        return false;
+    }
+
+    pub fn isComment(_: Node) bool {
+        return false;
+    }
+
+    pub fn isText(_: Node) bool {
+        return false;
+    }
+
+    pub fn isEmptyText(_: Node) !bool {
+        return error.TODO;
+    }
+
+    pub fn tag(_: Node) ![]const u8 {
+        return error.TODO;
+    }
+
+    pub fn attr(_: Node, _: []const u8) !?[]const u8 {
+        return error.TODO;
+    }
+
+    pub fn eql(_: Node, _: Node) bool {
+        return false;
+    }
+};
+```
+
+You also need do define a `Matcher` implementing a `match` function to
+accumulate the results.
+
+```zig
+const Matcher = struct {
+    const Nodes = std.ArrayList(Node);
+
+    nodes: Nodes,
+
+    fn init(alloc: std.mem.Allocator) Matcher {
+        return .{ .nodes = Nodes.init(alloc) };
+    }
+
+    fn deinit(m: *Matcher) void {
+        m.nodes.deinit();
+    }
+
+    pub fn match(m: *Matcher, n: Node) !void {
+        try m.nodes.append(n);
+    }
+};
+```
+
+Then you can use the lib itself.
+
+```zig
+var matcher = Matcher.init(alloc);
+defer matcher.deinit();
+
+try css.matchAll(selector, node, &matcher);
+_ = try css.matchFirst(selector, node, &matcher); // returns true if a node matched.
+```
+
+## Features
+
+* [x] parse query selector
+* [x] `matchAll`
+* [x] `matchFirst`
+* [ ] specificity
+
+### Selectors implemented
+
+#### Selectors
+
+* [x] Class selectors
+* [x] Id selectors
+* [x] Type selectors
+* [x] Universal selectors
+* [ ] Nesting selectors
+
+#### Combinators
+
+* [x] Child combinator
+* [ ] Column combinator
+* [x] Descendant combinator
+* [ ] Namespace combinator
+* [x] Next-sibling combinator
+* [x] Selector list combinator
+* [x] Subsequent-sibling combinator
+
+#### Attribute
+
+* [x] `[attr]`
+* [x] `[attr=value]`
+* [x] `[attr|=value]`
+* [x] `[attr^=value]`
+* [x] `[attr$=value]`
+* [ ] `[attr*=value]`
+* [x] `[attr operator value i]`
+* [ ] `[attr operator value s]`
+
+#### Pseudo classes
+
+* [ ] `:active`
+* [ ] `:any-link`
+* [ ] `:autofill`
+* [ ] `:blank Experimental`
+* [x] `:checked`
+* [ ] `:current Experimental`
+* [ ] `:default`
+* [ ] `:defined`
+* [ ] `:dir() Experimental`
+* [x] `:disabled`
+* [x] `:empty`
+* [x] `:enabled`
+* [ ] `:first`
+* [x] `:first-child`
+* [x] `:first-of-type`
+* [ ] `:focus`
+* [ ] `:focus-visible`
+* [ ] `:focus-within`
+* [ ] `:fullscreen`
+* [ ] `:future Experimental`
+* [x] `:has() Experimental`
+* [ ] `:host`
+* [ ] `:host()`
+* [ ] `:host-context() Experimental`
+* [ ] `:hover`
+* [ ] `:indeterminate`
+* [ ] `:in-range`
+* [ ] `:invalid`
+* [ ] `:is()`
+* [x] `:lang()`
+* [x] `:last-child`
+* [x] `:last-of-type`
+* [ ] `:left`
+* [x] `:link`
+* [ ] `:local-link Experimental`
+* [ ] `:modal`
+* [x] `:not()`
+* [x] `:nth-child()`
+* [x] `:nth-last-child()`
+* [x] `:nth-last-of-type()`
+* [x] `:nth-of-type()`
+* [x] `:only-child`
+* [x] `:only-of-type`
+* [ ] `:optional`
+* [ ] `:out-of-range`
+* [ ] `:past Experimental`
+* [ ] `:paused`
+* [ ] `:picture-in-picture`
+* [ ] `:placeholder-shown`
+* [ ] `:playing`
+* [ ] `:read-only`
+* [ ] `:read-write`
+* [ ] `:required`
+* [ ] `:right`
+* [x] `:root`
+* [ ] `:scope`
+* [ ] `:state() Experimental`
+* [ ] `:target`
+* [ ] `:target-within Experimental`
+* [ ] `:user-invalid Experimental`
+* [ ] `:valid`
+* [ ] `:visited`
+* [ ] `:where()`
+* [ ] `:contains()`
+* [ ] `:containsown()`
+* [ ] `:matched()`
+* [ ] `:matchesown()`
+* [x] `:root`
+

src/css/css.zig

@@ -0,0 +1,176 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+// CSS Selector parser and query
+// This package is a rewrite in Zig of Cascadia CSS Selector parser.
+// see https://github.com/andybalholm/cascadia
+const std = @import("std");
+const Selector = @import("selector.zig").Selector;
+const parser = @import("parser.zig");
+
+// parse parse a selector string and returns the parsed result or an error.
+pub fn parse(alloc: std.mem.Allocator, s: []const u8, opts: parser.ParseOptions) parser.ParseError!Selector {
+    var p = parser.Parser{ .s = s, .i = 0, .opts = opts };
+    return p.parse(alloc);
+}
+
+// matchFirst call m.match with the first node that matches the selector s, from the
+// descendants of n and returns true. If none matches, it returns false.
+pub fn matchFirst(s: Selector, node: anytype, m: anytype) !bool {
+    var c = try node.firstChild();
+    while (true) {
+        if (c == null) break;
+
+        if (try s.match(c.?)) {
+            try m.match(c.?);
+            return true;
+        }
+
+        if (try matchFirst(s, c.?, m)) return true;
+        c = try c.?.nextSibling();
+    }
+    return false;
+}
+
+// matchAll call m.match with the all the nodes that matches the selector s, from the
+// descendants of n.
+pub fn matchAll(s: Selector, node: anytype, m: anytype) !void {
+    var c = try node.firstChild();
+    while (true) {
+        if (c == null) break;
+
+        if (try s.match(c.?)) try m.match(c.?);
+        try matchAll(s, c.?, m);
+        c = try c.?.nextSibling();
+    }
+}
+
+test "parse" {
+    const alloc = std.testing.allocator;
+
+    const testcases = [_][]const u8{
+        "address",
+        "*",
+        "#foo",
+        "li#t1",
+        "*#t4",
+        ".t1",
+        "p.t1",
+        "div.teST",
+        ".t1.fail",
+        "p.t1.t2",
+        "p.--t1",
+        "p.--t1.--t2",
+        "p[title]",
+        "div[class=\"red\" i]",
+        "address[title=\"foo\"]",
+        "address[title=\"FoOIgnoRECaSe\" i]",
+        "address[title!=\"foo\"]",
+        "address[title!=\"foo\" i]",
+        "p[title!=\"FooBarUFoo\" i]",
+        "[  \t title        ~=       foo    ]",
+        "p[title~=\"FOO\" i]",
+        "p[title~=toofoo i]",
+        "[title~=\"hello world\"]",
+        "[title~=\"hello\" i]",
+        "[title~=\"hello\"          I]",
+        "[lang|=\"en\"]",
+        "[lang|=\"EN\" i]",
+        "[lang|=\"EN\"     i]",
+        "[title^=\"foo\"]",
+        "[title^=\"foo\" i]",
+        "[title$=\"bar\"]",
+        "[title$=\"BAR\" i]",
+        "[title*=\"bar\"]",
+        "[title*=\"BaRu\" i]",
+        "[title*=\"BaRu\" I]",
+        "p[class$=\" \"]",
+        "p[class$=\"\"]",
+        "p[class^=\" \"]",
+        "p[class^=\"\"]",
+        "p[class*=\" \"]",
+        "p[class*=\"\"]",
+        "input[name=Sex][value=F]",
+        "table[border=\"0\"][cellpadding=\"0\"][cellspacing=\"0\"]",
+        ".t1:not(.t2)",
+        "div:not(.t1)",
+        "div:not([class=\"t2\"])",
+        "li:nth-child(odd)",
+        "li:nth-child(even)",
+        "li:nth-child(-n+2)",
+        "li:nth-child(3n+1)",
+        "li:nth-last-child(odd)",
+        "li:nth-last-child(even)",
+        "li:nth-last-child(-n+2)",
+        "li:nth-last-child(3n+1)",
+        "span:first-child",
+        "span:last-child",
+        "p:nth-of-type(2)",
+        "p:nth-last-of-type(2)",
+        "p:last-of-type",
+        "p:first-of-type",
+        "p:only-child",
+        "p:only-of-type",
+        ":empty",
+        "div p",
+        "div table p",
+        "div > p",
+        "p ~ p",
+        "p + p",
+        "li, p",
+        "p +/*This is a comment*/ p",
+        "p:contains(\"that wraps\")",
+        "p:containsOwn(\"that wraps\")",
+        ":containsOwn(\"inner\")",
+        "p:containsOwn(\"block\")",
+        "div:has(#p1)",
+        "div:has(:containsOwn(\"2\"))",
+        "body :has(:containsOwn(\"2\"))",
+        "body :haschild(:containsOwn(\"2\"))",
+        "p:matches([\\d])",
+        "p:matches([a-z])",
+        "p:matches([a-zA-Z])",
+        "p:matches([^\\d])",
+        "p:matches(^(0|a))",
+        "p:matches(^\\d+$)",
+        "p:not(:matches(^\\d+$))",
+        "div :matchesOwn(^\\d+$)",
+        "[href#=(fina)]:not([href#=(\\/\\/[^\\/]+untrusted)])",
+        "[href#=(^https:\\/\\/[^\\/]*\\/?news)]",
+        ":input",
+        ":root",
+        "*:root",
+        "html:nth-child(1)",
+        "*:root:first-child",
+        "*:root:nth-child(1)",
+        "a:not(:root)",
+        "body > *:nth-child(3n+2)",
+        "input:disabled",
+        ":disabled",
+        ":enabled",
+        "div.class1, div.class2",
+    };
+
+    for (testcases) |tc| {
+        const s = parse(alloc, tc, .{}) catch |e| {
+            std.debug.print("query {s}", .{tc});
+            return e;
+        };
+        defer s.deinit(alloc);
+    }
+}

src/css/libdom.zig

@@ -0,0 +1,102 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const parser = @import("netsurf");
+
+// Node implementation with Netsurf Libdom C lib.
+pub const Node = struct {
+    node: *parser.Node,
+
+    pub fn firstChild(n: Node) !?Node {
+        const c = try parser.nodeFirstChild(n.node);
+        if (c) |cc| return .{ .node = cc };
+
+        return null;
+    }
+
+    pub fn lastChild(n: Node) !?Node {
+        const c = try parser.nodeLastChild(n.node);
+        if (c) |cc| return .{ .node = cc };
+
+        return null;
+    }
+
+    pub fn nextSibling(n: Node) !?Node {
+        const c = try parser.nodeNextSibling(n.node);
+        if (c) |cc| return .{ .node = cc };
+
+        return null;
+    }
+
+    pub fn prevSibling(n: Node) !?Node {
+        const c = try parser.nodePreviousSibling(n.node);
+        if (c) |cc| return .{ .node = cc };
+
+        return null;
+    }
+
+    pub fn parent(n: Node) !?Node {
+        const c = try parser.nodeParentNode(n.node);
+        if (c) |cc| return .{ .node = cc };
+
+        return null;
+    }
+
+    pub fn isElement(n: Node) bool {
+        const t = parser.nodeType(n.node) catch return false;
+        return t == .element;
+    }
+
+    pub fn isDocument(n: Node) bool {
+        const t = parser.nodeType(n.node) catch return false;
+        return t == .document;
+    }
+
+    pub fn isComment(n: Node) bool {
+        const t = parser.nodeType(n.node) catch return false;
+        return t == .comment;
+    }
+
+    pub fn isText(n: Node) bool {
+        const t = parser.nodeType(n.node) catch return false;
+        return t == .text;
+    }
+
+    pub fn isEmptyText(n: Node) !bool {
+        const data = try parser.nodeTextContent(n.node);
+        if (data == null) return true;
+        if (data.?.len == 0) return true;
+
+        return std.mem.trim(u8, data.?, &std.ascii.whitespace).len == 0;
+    }
+
+    pub fn tag(n: Node) ![]const u8 {
+        return try parser.nodeName(n.node);
+    }
+
+    pub fn attr(n: Node, key: []const u8) !?[]const u8 {
+        if (!n.isElement()) return null;
+        return try parser.elementGetAttribute(parser.nodeToElement(n.node), key);
+    }
+
+    pub fn eql(a: Node, b: Node) bool {
+        return a.node == b.node;
+    }
+};

src/css/libdom_test.zig

@@ -0,0 +1,325 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const css = @import("css.zig");
+const Node = @import("libdom.zig").Node;
+const parser = @import("netsurf");
+
+const Matcher = struct {
+    const Nodes = std.ArrayList(Node);
+
+    nodes: Nodes,
+
+    fn init(alloc: std.mem.Allocator) Matcher {
+        return .{ .nodes = Nodes.init(alloc) };
+    }
+
+    fn deinit(m: *Matcher) void {
+        m.nodes.deinit();
+    }
+
+    fn reset(m: *Matcher) void {
+        m.nodes.clearRetainingCapacity();
+    }
+
+    pub fn match(m: *Matcher, n: Node) !void {
+        try m.nodes.append(n);
+    }
+};
+
+test "matchFirst" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    const testcases = [_]struct {
+        q: []const u8,
+        html: []const u8,
+        exp: usize,
+    }{
+        .{ .q = "address", .html = "<body><address>This address...</address></body>", .exp = 1 },
+        .{ .q = "*", .html = "<!-- comment --><html><head></head><body>text</body></html>", .exp = 1 },
+        .{ .q = "*", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "#foo", .html = "<p id=\"foo\"><p id=\"bar\">", .exp = 1 },
+        .{ .q = "li#t1", .html = "<ul><li id=\"t1\"><p id=\"t1\">", .exp = 1 },
+        .{ .q = ".t3", .html = "<ul><li class=\"t1\"><li class=\"t2 t3\">", .exp = 1 },
+        .{ .q = "*#t4", .html = "<ol><li id=\"t4\"><li id=\"t44\">", .exp = 1 },
+        .{ .q = ".t1", .html = "<ul><li class=\"t1\"><li class=\"t2\">", .exp = 1 },
+        .{ .q = "p.t1", .html = "<p class=\"t1 t2\">", .exp = 1 },
+        .{ .q = "div.teST", .html = "<div class=\"test\">", .exp = 0 },
+        .{ .q = ".t1.fail", .html = "<p class=\"t1 t2\">", .exp = 0 },
+        .{ .q = "p.t1.t2", .html = "<p class=\"t1 t2\">", .exp = 1 },
+        .{ .q = "p.--t1", .html = "<p class=\"--t1 --t2\">", .exp = 1 },
+        .{ .q = "p.--t1.--t2", .html = "<p class=\"--t1 --t2\">", .exp = 1 },
+        .{ .q = "p[title]", .html = "<p><p title=\"title\">", .exp = 1 },
+        .{ .q = "div[class=\"red\" i]", .html = "<div><div class=\"Red\">", .exp = 1 },
+        .{ .q = "address[title=\"foo\"]", .html = "<address><address title=\"foo\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "address[title=\"FoOIgnoRECaSe\" i]", .html = "<address><address title=\"fooIgnoreCase\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "address[title!=\"foo\"]", .html = "<address><address title=\"foo\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "address[title!=\"foo\" i]", .html = "<address><address title=\"FOO\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "p[title!=\"FooBarUFoo\" i]", .html = "<p title=\"fooBARuFOO\"><p title=\"varfoo\">", .exp = 1 },
+        .{ .q = "[   title        ~=       foo    ]", .html = "<p title=\"tot foo bar\">", .exp = 1 },
+        .{ .q = "p[title~=\"FOO\" i]", .html = "<p title=\"tot foo bar\">", .exp = 1 },
+        .{ .q = "p[title~=toofoo i]", .html = "<p title=\"tot foo bar\">", .exp = 0 },
+        .{ .q = "[title~=\"hello world\"]", .html = "<p title=\"hello world\">", .exp = 0 },
+        .{ .q = "[title~=\"hello\" i]", .html = "<p title=\"HELLO world\">", .exp = 1 },
+        .{ .q = "[title~=\"hello\"          I]", .html = "<p title=\"HELLO world\">", .exp = 1 },
+        .{ .q = "[lang|=\"en\"]", .html = "<p lang=\"en\"><p lang=\"en-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 1 },
+        .{ .q = "[lang|=\"EN\" i]", .html = "<p lang=\"en\"><p lang=\"En-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 1 },
+        .{ .q = "[lang|=\"EN\"     i]", .html = "<p lang=\"en\"><p lang=\"En-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 1 },
+        .{ .q = "[title^=\"foo\"]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title^=\"foo\" i]", .html = "<p title=\"FooBAR\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title$=\"bar\"]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title$=\"BAR\" i]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title*=\"bar\"]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "[title*=\"BaRu\" i]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "[title*=\"BaRu\" I]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "p[class$=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class$=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class^=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class^=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class*=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class*=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "input[name=Sex][value=F]", .html = "<input type=\"radio\" name=\"Sex\" value=\"F\"/>", .exp = 1 },
+        .{ .q = "table[border=\"0\"][cellpadding=\"0\"][cellspacing=\"0\"]", .html = "<table border=\"0\" cellpadding=\"0\" cellspacing=\"0\" style=\"table-layout: fixed; width: 100%; border: 0 dashed; border-color: #FFFFFF\"><tr style=\"height:64px\">aaa</tr></table>", .exp = 1 },
+        .{ .q = ".t1:not(.t2)", .html = "<p class=\"t1 t2\">", .exp = 0 },
+        .{ .q = "div:not(.t1)", .html = "<div class=\"t3\">", .exp = 1 },
+        .{ .q = "div:not([class=\"t2\"])", .html = "<div><div class=\"t2\"><div class=\"t3\">", .exp = 1 },
+        .{ .q = "li:nth-child(odd)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-child(even)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-child(-n+2)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-child(3n+1)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-last-child(odd)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 1 },
+        .{ .q = "li:nth-last-child(even)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 1 },
+        .{ .q = "li:nth-last-child(-n+2)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 1 },
+        .{ .q = "li:nth-last-child(3n+1)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 1 },
+        .{ .q = "span:first-child", .html = "<p>some text <span id=\"1\">and a span</span><span id=\"2\"> and another</span></p>", .exp = 1 },
+        .{ .q = "span:last-child", .html = "<span>a span</span> and some text", .exp = 1 },
+        .{ .q = "p:nth-of-type(2)", .html = "<address></address><p id=1><p id=2>", .exp = 1 },
+        .{ .q = "p:nth-last-of-type(2)", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:last-of-type", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:first-of-type", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:only-child", .html = "<div><p id=\"1\"></p><a></a></div><div><p id=\"2\"></p></div>", .exp = 1 },
+        .{ .q = "p:only-of-type", .html = "<div><p id=\"1\"></p><a></a></div><div><p id=\"2\"></p><p id=\"3\"></p></div>", .exp = 1 },
+        .{ .q = ":empty", .html = "<p id=\"1\"><!-- --><p id=\"2\">Hello<p id=\"3\"><span>", .exp = 1 },
+        .{ .q = "div p", .html = "<div><p id=\"1\"><table><tr><td><p id=\"2\"></table></div><p id=\"3\">", .exp = 1 },
+        .{ .q = "div table p", .html = "<div><p id=\"1\"><table><tr><td><p id=\"2\"></table></div><p id=\"3\">", .exp = 1 },
+        .{ .q = "div > p", .html = "<div><p id=\"1\"><div><p id=\"2\"></div><table><tr><td><p id=\"3\"></table></div>", .exp = 1 },
+        .{ .q = "p ~ p", .html = "<p id=\"1\"><p id=\"2\"></p><address></address><p id=\"3\">", .exp = 1 },
+        .{ .q = "p + p", .html = "<p id=\"1\"></p> <!--comment--> <p id=\"2\"></p><address></address><p id=\"3\">", .exp = 1 },
+        .{ .q = "li, p", .html = "<ul><li></li><li></li></ul><p>", .exp = 1 },
+        .{ .q = "p +/*This is a comment*/ p", .html = "<p id=\"1\"><p id=\"2\"></p><address></address><p id=\"3\">", .exp = 1 },
+        // .{ .q = "p:contains(\"that wraps\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        // .{ .q = "p:containsOwn(\"that wraps\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 0 },
+        // .{ .q = ":containsOwn(\"inner\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        // .{ .q = "p:containsOwn(\"block\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        // .{ .q = "div:has(#p1)", .html = "<div id=\"d1\"><p id=\"p1\"><span>text content</span></p></div><div id=\"d2\"/>", .exp = 1 },
+        // .{ .q = "div:has(:containsOwn(\"2\"))", .html = "<div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p>contents <em>2</em></p></div>", .exp = 1 },
+        // .{ .q = "body :has(:containsOwn(\"2\"))", .html = "<body><div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p id=\"p2\">contents <em>2</em></p></div></body>", .exp = 1 },
+        // .{ .q = "body :haschild(:containsOwn(\"2\"))", .html = "<body><div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p id=\"p2\">contents <em>2</em></p></div></body>", .exp = 1 },
+        // .{ .q = "p:matches([\\d])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches([a-z])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches([a-zA-Z])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches([^\\d])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches(^(0|a))", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches(^\\d+$)", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:not(:matches(^\\d+$))", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "div :matchesOwn(^\\d+$)", .html = "<div><p id=\"p1\">01234<em>567</em>89</p><div>", .exp = 1 },
+        // .{ .q = "[href#=(fina)]:not([href#=(\\/\\/[^\\/]+untrusted)])", .html = "<ul> <li><a id=\"a1\" href=\"http://www.google.com/finance\"></a> <li><a id=\"a2\" href=\"http://finance.yahoo.com/\"></a> <li><a id=\"a2\" href=\"http://finance.untrusted.com/\"/> <li><a id=\"a3\" href=\"https://www.google.com/news\"/> <li><a id=\"a4\" href=\"http://news.yahoo.com\"/> </ul>", .exp = 1 },
+        // .{ .q = "[href#=(^https:\\/\\/[^\\/]*\\/?news)]", .html = "<ul> <li><a id=\"a1\" href=\"http://www.google.com/finance\"/> <li><a id=\"a2\" href=\"http://finance.yahoo.com/\"/> <li><a id=\"a3\" href=\"https://www.google.com/news\"></a> <li><a id=\"a4\" href=\"http://news.yahoo.com\"/> </ul>", .exp = 1 },
+        .{ .q = ":input", .html = "<form> <label>Username <input type=\"text\" name=\"username\" /></label> <label>Password <input type=\"password\" name=\"password\" /></label> <label>Country <select name=\"country\"> <option value=\"ca\">Canada</option> <option value=\"us\">United States</option> </select> </label> <label>Bio <textarea name=\"bio\"></textarea></label> <button>Sign up</button> </form>", .exp = 1 },
+        .{ .q = ":root", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "html:nth-child(1)", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root:first-child", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root:nth-child(1)", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "a:not(:root)", .html = "<html><head></head><body><a href=\"http://www.foo.com\"></a></body></html>", .exp = 1 },
+        .{ .q = "body > *:nth-child(3n+2)", .html = "<html><head></head><body><p></p><div></div><span></span><a></a><form></form></body></html>", .exp = 1 },
+        .{ .q = "input:disabled", .html = "<html><head></head><body><fieldset disabled><legend id=\"1\"><input id=\"i1\"/></legend><legend id=\"2\"><input id=\"i2\"/></legend></fieldset></body></html>", .exp = 1 },
+        .{ .q = ":disabled", .html = "<html><head></head><body><fieldset disabled></fieldset></body></html>", .exp = 1 },
+        .{ .q = ":enabled", .html = "<html><head></head><body><fieldset></fieldset></body></html>", .exp = 1 },
+        .{ .q = "div.class1, div.class2", .html = "<div class=class1></div><div class=class2></div><div class=class3></div>", .exp = 1 },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const doc = try parser.documentHTMLParseFromStr(tc.html);
+        defer parser.documentHTMLClose(doc) catch {};
+
+        const s = css.parse(alloc, tc.q, .{}) catch |e| {
+            std.debug.print("parse, query: {s}\n", .{tc.q});
+            return e;
+        };
+
+        defer s.deinit(alloc);
+
+        const node = Node{ .node = parser.documentHTMLToNode(doc) };
+
+        _ = css.matchFirst(s, node, &matcher) catch |e| {
+            std.debug.print("match, query: {s}\n", .{tc.q});
+            return e;
+        };
+        std.testing.expectEqual(tc.exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("expectation, query: {s}\n", .{tc.q});
+            return e;
+        };
+    }
+}
+
+test "matchAll" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    const testcases = [_]struct {
+        q: []const u8,
+        html: []const u8,
+        exp: usize,
+    }{
+        .{ .q = "address", .html = "<body><address>This address...</address></body>", .exp = 1 },
+        .{ .q = "*", .html = "<!-- comment --><html><head></head><body>text</body></html>", .exp = 3 },
+        .{ .q = "*", .html = "<html><head></head><body></body></html>", .exp = 3 },
+        .{ .q = "#foo", .html = "<p id=\"foo\"><p id=\"bar\">", .exp = 1 },
+        .{ .q = "li#t1", .html = "<ul><li id=\"t1\"><p id=\"t1\">", .exp = 1 },
+        .{ .q = ".t3", .html = "<ul><li class=\"t1\"><li class=\"t2 t3\">", .exp = 1 },
+        .{ .q = "*#t4", .html = "<ol><li id=\"t4\"><li id=\"t44\">", .exp = 1 },
+        .{ .q = ".t1", .html = "<ul><li class=\"t1\"><li class=\"t2\">", .exp = 1 },
+        .{ .q = "p.t1", .html = "<p class=\"t1 t2\">", .exp = 1 },
+        .{ .q = "div.teST", .html = "<div class=\"test\">", .exp = 0 },
+        .{ .q = ".t1.fail", .html = "<p class=\"t1 t2\">", .exp = 0 },
+        .{ .q = "p.t1.t2", .html = "<p class=\"t1 t2\">", .exp = 1 },
+        .{ .q = "p.--t1", .html = "<p class=\"--t1 --t2\">", .exp = 1 },
+        .{ .q = "p.--t1.--t2", .html = "<p class=\"--t1 --t2\">", .exp = 1 },
+        .{ .q = "p[title]", .html = "<p><p title=\"title\">", .exp = 1 },
+        .{ .q = "div[class=\"red\" i]", .html = "<div><div class=\"Red\">", .exp = 1 },
+        .{ .q = "address[title=\"foo\"]", .html = "<address><address title=\"foo\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "address[title=\"FoOIgnoRECaSe\" i]", .html = "<address><address title=\"fooIgnoreCase\"><address title=\"bar\">", .exp = 1 },
+        .{ .q = "address[title!=\"foo\"]", .html = "<address><address title=\"foo\"><address title=\"bar\">", .exp = 2 },
+        .{ .q = "address[title!=\"foo\" i]", .html = "<address><address title=\"FOO\"><address title=\"bar\">", .exp = 2 },
+        .{ .q = "p[title!=\"FooBarUFoo\" i]", .html = "<p title=\"fooBARuFOO\"><p title=\"varfoo\">", .exp = 1 },
+        .{ .q = "[   title        ~=       foo    ]", .html = "<p title=\"tot foo bar\">", .exp = 1 },
+        .{ .q = "p[title~=\"FOO\" i]", .html = "<p title=\"tot foo bar\">", .exp = 1 },
+        .{ .q = "p[title~=toofoo i]", .html = "<p title=\"tot foo bar\">", .exp = 0 },
+        .{ .q = "[title~=\"hello world\"]", .html = "<p title=\"hello world\">", .exp = 0 },
+        .{ .q = "[title~=\"hello\" i]", .html = "<p title=\"HELLO world\">", .exp = 1 },
+        .{ .q = "[title~=\"hello\"          I]", .html = "<p title=\"HELLO world\">", .exp = 1 },
+        .{ .q = "[lang|=\"en\"]", .html = "<p lang=\"en\"><p lang=\"en-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 2 },
+        .{ .q = "[lang|=\"EN\" i]", .html = "<p lang=\"en\"><p lang=\"En-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 2 },
+        .{ .q = "[lang|=\"EN\"     i]", .html = "<p lang=\"en\"><p lang=\"En-gb\"><p lang=\"enough\"><p lang=\"fr-en\">", .exp = 2 },
+        .{ .q = "[title^=\"foo\"]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title^=\"foo\" i]", .html = "<p title=\"FooBAR\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title$=\"bar\"]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title$=\"BAR\" i]", .html = "<p title=\"foobar\"><p title=\"barfoo\">", .exp = 1 },
+        .{ .q = "[title*=\"bar\"]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "[title*=\"BaRu\" i]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "[title*=\"BaRu\" I]", .html = "<p title=\"foobarufoo\">", .exp = 1 },
+        .{ .q = "p[class$=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class$=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class^=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class^=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class*=\" \"]", .html = "<p class=\" \">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "p[class*=\"\"]", .html = "<p class=\"\">This text should be green.</p><p>This text should be green.</p>", .exp = 0 },
+        .{ .q = "input[name=Sex][value=F]", .html = "<input type=\"radio\" name=\"Sex\" value=\"F\"/>", .exp = 1 },
+        .{ .q = "table[border=\"0\"][cellpadding=\"0\"][cellspacing=\"0\"]", .html = "<table border=\"0\" cellpadding=\"0\" cellspacing=\"0\" style=\"table-layout: fixed; width: 100%; border: 0 dashed; border-color: #FFFFFF\"><tr style=\"height:64px\">aaa</tr></table>", .exp = 1 },
+        .{ .q = ".t1:not(.t2)", .html = "<p class=\"t1 t2\">", .exp = 0 },
+        .{ .q = "div:not(.t1)", .html = "<div class=\"t3\">", .exp = 1 },
+        .{ .q = "div:not([class=\"t2\"])", .html = "<div><div class=\"t2\"><div class=\"t3\">", .exp = 2 },
+        .{ .q = "li:nth-child(odd)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 2 },
+        .{ .q = "li:nth-child(even)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-child(-n+2)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 2 },
+        .{ .q = "li:nth-child(3n+1)", .html = "<ol><li id=1><li id=2><li id=3></ol>", .exp = 1 },
+        .{ .q = "li:nth-last-child(odd)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 2 },
+        .{ .q = "li:nth-last-child(even)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 2 },
+        .{ .q = "li:nth-last-child(-n+2)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 2 },
+        .{ .q = "li:nth-last-child(3n+1)", .html = "<ol><li id=1><li id=2><li id=3><li id=4></ol>", .exp = 2 },
+        .{ .q = "span:first-child", .html = "<p>some text <span id=\"1\">and a span</span><span id=\"2\"> and another</span></p>", .exp = 1 },
+        .{ .q = "span:last-child", .html = "<span>a span</span> and some text", .exp = 1 },
+        .{ .q = "p:nth-of-type(2)", .html = "<address></address><p id=1><p id=2>", .exp = 1 },
+        .{ .q = "p:nth-last-of-type(2)", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:last-of-type", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:first-of-type", .html = "<address></address><p id=1><p id=2></p><a>", .exp = 1 },
+        .{ .q = "p:only-child", .html = "<div><p id=\"1\"></p><a></a></div><div><p id=\"2\"></p></div>", .exp = 1 },
+        .{ .q = "p:only-of-type", .html = "<div><p id=\"1\"></p><a></a></div><div><p id=\"2\"></p><p id=\"3\"></p></div>", .exp = 1 },
+        .{ .q = ":empty", .html = "<p id=\"1\"><!-- --><p id=\"2\">Hello<p id=\"3\"><span>", .exp = 3 },
+        .{ .q = "div p", .html = "<div><p id=\"1\"><table><tr><td><p id=\"2\"></table></div><p id=\"3\">", .exp = 2 },
+        .{ .q = "div table p", .html = "<div><p id=\"1\"><table><tr><td><p id=\"2\"></table></div><p id=\"3\">", .exp = 1 },
+        .{ .q = "div > p", .html = "<div><p id=\"1\"><div><p id=\"2\"></div><table><tr><td><p id=\"3\"></table></div>", .exp = 2 },
+        .{ .q = "p ~ p", .html = "<p id=\"1\"><p id=\"2\"></p><address></address><p id=\"3\">", .exp = 2 },
+        .{ .q = "p + p", .html = "<p id=\"1\"></p> <!--comment--> <p id=\"2\"></p><address></address><p id=\"3\">", .exp = 1 },
+        .{ .q = "li, p", .html = "<ul><li></li><li></li></ul><p>", .exp = 3 },
+        .{ .q = "p +/*This is a comment*/ p", .html = "<p id=\"1\"><p id=\"2\"></p><address></address><p id=\"3\">", .exp = 1 },
+        // .{ .q = "p:contains(\"that wraps\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        // .{ .q = "p:containsOwn(\"that wraps\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 0 },
+        // .{ .q = ":containsOwn(\"inner\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        // .{ .q = "p:containsOwn(\"block\")", .html = "<p>Text block that <span>wraps inner text</span> and continues</p>", .exp = 1 },
+        .{ .q = "div:has(#p1)", .html = "<div id=\"d1\"><p id=\"p1\"><span>text content</span></p></div><div id=\"d2\"/>", .exp = 1 },
+        // .{ .q = "div:has(:containsOwn(\"2\"))", .html = "<div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p>contents <em>2</em></p></div>", .exp = 1 },
+        // .{ .q = "body :has(:containsOwn(\"2\"))", .html = "<body><div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p id=\"p2\">contents <em>2</em></p></div></body>", .exp = 2 },
+        // .{ .q = "body :haschild(:containsOwn(\"2\"))", .html = "<body><div id=\"d1\"><p id=\"p1\"><span>contents 1</span></p></div> <div id=\"d2\"><p id=\"p2\">contents <em>2</em></p></div></body>", .exp = 1 },
+        // .{ .q = "p:matches([\\d])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 2 },
+        // .{ .q = "p:matches([a-z])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:matches([a-zA-Z])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 2 },
+        // .{ .q = "p:matches([^\\d])", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 2 },
+        // .{ .q = "p:matches(^(0|a))", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 3 },
+        // .{ .q = "p:matches(^\\d+$)", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 1 },
+        // .{ .q = "p:not(:matches(^\\d+$))", .html = "<p id=\"p1\">0123456789</p><p id=\"p2\">abcdef</p><p id=\"p3\">0123ABCD</p>", .exp = 2 },
+        // .{ .q = "div :matchesOwn(^\\d+$)", .html = "<div><p id=\"p1\">01234<em>567</em>89</p><div>", .exp = 2 },
+        // .{ .q = "[href#=(fina)]:not([href#=(\\/\\/[^\\/]+untrusted)])", .html = "<ul> <li><a id=\"a1\" href=\"http://www.google.com/finance\"></a> <li><a id=\"a2\" href=\"http://finance.yahoo.com/\"></a> <li><a id=\"a2\" href=\"http://finance.untrusted.com/\"/> <li><a id=\"a3\" href=\"https://www.google.com/news\"/> <li><a id=\"a4\" href=\"http://news.yahoo.com\"/> </ul>", .exp = 2 },
+        // .{ .q = "[href#=(^https:\\/\\/[^\\/]*\\/?news)]", .html = "<ul> <li><a id=\"a1\" href=\"http://www.google.com/finance\"/> <li><a id=\"a2\" href=\"http://finance.yahoo.com/\"/> <li><a id=\"a3\" href=\"https://www.google.com/news\"></a> <li><a id=\"a4\" href=\"http://news.yahoo.com\"/> </ul>", .exp = 1 },
+        .{ .q = ":input", .html = "<form> <label>Username <input type=\"text\" name=\"username\" /></label> <label>Password <input type=\"password\" name=\"password\" /></label> <label>Country <select name=\"country\"> <option value=\"ca\">Canada</option> <option value=\"us\">United States</option> </select> </label> <label>Bio <textarea name=\"bio\"></textarea></label> <button>Sign up</button> </form>", .exp = 5 },
+        .{ .q = ":root", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "html:nth-child(1)", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root:first-child", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "*:root:nth-child(1)", .html = "<html><head></head><body></body></html>", .exp = 1 },
+        .{ .q = "a:not(:root)", .html = "<html><head></head><body><a href=\"http://www.foo.com\"></a></body></html>", .exp = 1 },
+        .{ .q = "body > *:nth-child(3n+2)", .html = "<html><head></head><body><p></p><div></div><span></span><a></a><form></form></body></html>", .exp = 2 },
+        .{ .q = "input:disabled", .html = "<html><head></head><body><fieldset disabled><legend id=\"1\"><input id=\"i1\"/></legend><legend id=\"2\"><input id=\"i2\"/></legend></fieldset></body></html>", .exp = 1 },
+        .{ .q = ":disabled", .html = "<html><head></head><body><fieldset disabled></fieldset></body></html>", .exp = 1 },
+        .{ .q = ":enabled", .html = "<html><head></head><body><fieldset></fieldset></body></html>", .exp = 1 },
+        .{ .q = "div.class1, div.class2", .html = "<div class=class1></div><div class=class2></div><div class=class3></div>", .exp = 2 },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const doc = try parser.documentHTMLParseFromStr(tc.html);
+        defer parser.documentHTMLClose(doc) catch {};
+
+        const s = css.parse(alloc, tc.q, .{}) catch |e| {
+            std.debug.print("parse, query: {s}\n", .{tc.q});
+            return e;
+        };
+        defer s.deinit(alloc);
+
+        const node = Node{ .node = parser.documentHTMLToNode(doc) };
+
+        _ = css.matchAll(s, node, &matcher) catch |e| {
+            std.debug.print("match, query: {s}\n", .{tc.q});
+            return e;
+        };
+        std.testing.expectEqual(tc.exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("expectation, query: {s}\n", .{tc.q});
+            return e;
+        };
+    }
+}

src/css/match_test.zig

@@ -0,0 +1,587 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const css = @import("css.zig");
+
+// Node mock implementation for test only.
+pub const Node = struct {
+    child: ?*const Node = null,
+    last: ?*const Node = null,
+    sibling: ?*const Node = null,
+    prev: ?*const Node = null,
+    par: ?*const Node = null,
+
+    name: []const u8 = "",
+    att: ?[]const u8 = null,
+
+    pub fn firstChild(n: *const Node) !?*const Node {
+        return n.child;
+    }
+
+    pub fn lastChild(n: *const Node) !?*const Node {
+        return n.last;
+    }
+
+    pub fn nextSibling(n: *const Node) !?*const Node {
+        return n.sibling;
+    }
+
+    pub fn prevSibling(n: *const Node) !?*const Node {
+        return n.prev;
+    }
+
+    pub fn parent(n: *const Node) !?*const Node {
+        return n.par;
+    }
+
+    pub fn isElement(_: *const Node) bool {
+        return true;
+    }
+
+    pub fn isDocument(_: *const Node) bool {
+        return false;
+    }
+
+    pub fn isComment(_: *const Node) bool {
+        return false;
+    }
+
+    pub fn isText(_: *const Node) bool {
+        return false;
+    }
+
+    pub fn isEmptyText(_: *const Node) !bool {
+        return false;
+    }
+
+    pub fn tag(n: *const Node) ![]const u8 {
+        return n.name;
+    }
+
+    pub fn attr(n: *const Node, _: []const u8) !?[]const u8 {
+        return n.att;
+    }
+
+    pub fn eql(a: *const Node, b: *const Node) bool {
+        return a == b;
+    }
+};
+
+const Matcher = struct {
+    const Nodes = std.ArrayList(*const Node);
+
+    nodes: Nodes,
+
+    fn init(alloc: std.mem.Allocator) Matcher {
+        return .{ .nodes = Nodes.init(alloc) };
+    }
+
+    fn deinit(m: *Matcher) void {
+        m.nodes.deinit();
+    }
+
+    fn reset(m: *Matcher) void {
+        m.nodes.clearRetainingCapacity();
+    }
+
+    pub fn match(m: *Matcher, n: *const Node) !void {
+        try m.nodes.append(n);
+    }
+};
+
+test "matchFirst" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    const testcases = [_]struct {
+        q: []const u8,
+        n: Node,
+        exp: usize,
+    }{
+        .{
+            .q = "address",
+            .n = .{ .child = &.{ .name = "body", .child = &.{ .name = "address" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "#foo",
+            .n = .{ .child = &.{ .name = "p", .att = "foo", .child = &.{ .name = "p" } } },
+            .exp = 1,
+        },
+        .{
+            .q = ".t1",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "t1" } } },
+            .exp = 1,
+        },
+        .{
+            .q = ".t1",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "foo t1" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo!=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo!=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo~=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "baz bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo~=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo^=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo$=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo*=rb]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar-baz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "ba" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "strong, a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a", .par = &.{ .name = "p" } }, .sibling = &.{ .name = "a" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "span", .child = &.{
+                .name = "a",
+                .par = &.{ .name = "span", .par = &.{ .name = "p" } },
+            } } } },
+            .exp = 1,
+        },
+        .{
+            .q = ":not(p)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p:has(a)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p:has(strong)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "p:haschild(a)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p:haschild(strong)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "p:lang(en)",
+            .n = .{ .child = &.{ .name = "p", .att = "en-US", .child = &.{ .name = "a" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "a:lang(en)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a", .par = &.{ .att = "en-US" } } } },
+            .exp = 1,
+        },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const s = try css.parse(alloc, tc.q, .{});
+        defer s.deinit(alloc);
+
+        _ = css.matchFirst(s, &tc.n, &matcher) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+
+        std.testing.expectEqual(tc.exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+    }
+}
+
+test "matchAll" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    const testcases = [_]struct {
+        q: []const u8,
+        n: Node,
+        exp: usize,
+    }{
+        .{
+            .q = "address",
+            .n = .{ .child = &.{ .name = "body", .child = &.{ .name = "address" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "#foo",
+            .n = .{ .child = &.{ .name = "p", .att = "foo", .child = &.{ .name = "p" } } },
+            .exp = 1,
+        },
+        .{
+            .q = ".t1",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "t1" } } },
+            .exp = 1,
+        },
+        .{
+            .q = ".t1",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "foo t1" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo!=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo!=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 2,
+        },
+        .{
+            .q = "[foo~=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "baz bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo~=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "[foo^=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo$=baz]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo*=rb]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "barbaz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "bar-baz" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "[foo|=bar]",
+            .n = .{ .child = &.{ .name = "p", .sibling = &.{ .name = "p", .att = "ba" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "strong, a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 2,
+        },
+        .{
+            .q = "p a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a", .par = &.{ .name = "p" } }, .sibling = &.{ .name = "a" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p a",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "span", .child = &.{
+                .name = "a",
+                .par = &.{ .name = "span", .par = &.{ .name = "p" } },
+            } } } },
+            .exp = 1,
+        },
+        .{
+            .q = ":not(p)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 2,
+        },
+        .{
+            .q = "p:has(a)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p:has(strong)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "p:haschild(a)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "p:haschild(strong)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a" }, .sibling = &.{ .name = "strong" } } },
+            .exp = 0,
+        },
+        .{
+            .q = "p:lang(en)",
+            .n = .{ .child = &.{ .name = "p", .att = "en-US", .child = &.{ .name = "a" } } },
+            .exp = 1,
+        },
+        .{
+            .q = "a:lang(en)",
+            .n = .{ .child = &.{ .name = "p", .child = &.{ .name = "a", .par = &.{ .att = "en-US" } } } },
+            .exp = 1,
+        },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const s = try css.parse(alloc, tc.q, .{});
+        defer s.deinit(alloc);
+
+        css.matchAll(s, &tc.n, &matcher) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+
+        std.testing.expectEqual(tc.exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+    }
+}
+
+test "pseudo class" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    var p1: Node = .{ .name = "p" };
+    var p2: Node = .{ .name = "p" };
+    var a1: Node = .{ .name = "a" };
+
+    p1.sibling = &p2;
+    p2.prev = &p1;
+
+    p2.sibling = &a1;
+    a1.prev = &p2;
+
+    var root: Node = .{ .child = &p1, .last = &a1 };
+    p1.par = &root;
+    p2.par = &root;
+    a1.par = &root;
+
+    const testcases = [_]struct {
+        q: []const u8,
+        n: Node,
+        exp: ?*const Node,
+    }{
+        .{ .q = "p:only-child", .n = root, .exp = null },
+        .{ .q = "a:only-of-type", .n = root, .exp = &a1 },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const s = try css.parse(alloc, tc.q, .{});
+        defer s.deinit(alloc);
+
+        css.matchAll(s, &tc.n, &matcher) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+
+        if (tc.exp) |exp_n| {
+            const exp: usize = 1;
+            std.testing.expectEqual(exp, matcher.nodes.items.len) catch |e| {
+                std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+                return e;
+            };
+
+            std.testing.expectEqual(exp_n, matcher.nodes.items[0]) catch |e| {
+                std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+                return e;
+            };
+
+            continue;
+        }
+
+        const exp: usize = 0;
+        std.testing.expectEqual(exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+    }
+}
+
+test "nth pseudo class" {
+    const alloc = std.testing.allocator;
+
+    var matcher = Matcher.init(alloc);
+    defer matcher.deinit();
+
+    var p1: Node = .{ .name = "p" };
+    var p2: Node = .{ .name = "p" };
+
+    p1.sibling = &p2;
+    p2.prev = &p1;
+
+    var root: Node = .{ .child = &p1, .last = &p2 };
+    p1.par = &root;
+    p2.par = &root;
+
+    const testcases = [_]struct {
+        q: []const u8,
+        n: Node,
+        exp: ?*const Node,
+    }{
+        .{ .q = "a:nth-of-type(1)", .n = root, .exp = null },
+        .{ .q = "p:nth-of-type(1)", .n = root, .exp = &p1 },
+        .{ .q = "p:nth-of-type(2)", .n = root, .exp = &p2 },
+        .{ .q = "p:nth-of-type(0)", .n = root, .exp = null },
+        .{ .q = "p:nth-of-type(2n)", .n = root, .exp = &p2 },
+        .{ .q = "p:nth-last-child(1)", .n = root, .exp = &p2 },
+        .{ .q = "p:nth-last-child(2)", .n = root, .exp = &p1 },
+        .{ .q = "p:nth-child(1)", .n = root, .exp = &p1 },
+        .{ .q = "p:nth-child(2)", .n = root, .exp = &p2 },
+        .{ .q = "p:nth-child(odd)", .n = root, .exp = &p1 },
+        .{ .q = "p:nth-child(even)", .n = root, .exp = &p2 },
+        .{ .q = "p:nth-child(n+2)", .n = root, .exp = &p2 },
+    };
+
+    for (testcases) |tc| {
+        matcher.reset();
+
+        const s = try css.parse(alloc, tc.q, .{});
+        defer s.deinit(alloc);
+
+        css.matchAll(s, &tc.n, &matcher) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+
+        if (tc.exp) |exp_n| {
+            const exp: usize = 1;
+            std.testing.expectEqual(exp, matcher.nodes.items.len) catch |e| {
+                std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+                return e;
+            };
+
+            std.testing.expectEqual(exp_n, matcher.nodes.items[0]) catch |e| {
+                std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+                return e;
+            };
+
+            continue;
+        }
+
+        const exp: usize = 0;
+        std.testing.expectEqual(exp, matcher.nodes.items.len) catch |e| {
+            std.debug.print("query: {s}, parsed selector: {any}\n", .{ tc.q, s });
+            return e;
+        };
+    }
+}

src/css/parser.zig

@@ -0,0 +1,917 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+// CSS Selector parser
+// This file is a rewrite in Zig of Cascadia CSS Selector parser.
+// see https://github.com/andybalholm/cascadia
+// see https://github.com/andybalholm/cascadia/blob/master/parser.go
+const std = @import("std");
+const ascii = std.ascii;
+
+const selector = @import("selector.zig");
+const Selector = selector.Selector;
+const PseudoClass = selector.PseudoClass;
+const AttributeOP = selector.AttributeOP;
+const Combinator = selector.Combinator;
+
+pub const ParseError = error{
+    ExpectedSelector,
+    ExpectedIdentifier,
+    ExpectedName,
+    ExpectedIDSelector,
+    ExpectedClassSelector,
+    ExpectedAttributeSelector,
+    ExpectedString,
+    ExpectedRegexp,
+    ExpectedPseudoClassSelector,
+    ExpectedParenthesis,
+    ExpectedParenthesisClose,
+    ExpectedNthExpression,
+    ExpectedInteger,
+    InvalidEscape,
+    EscapeLineEndingOutsideString,
+    InvalidUnicode,
+    UnicodeIsNotHandled,
+    WriteError,
+    PseudoElementNotAtSelectorEnd,
+    PseudoElementNotUnique,
+    PseudoElementDisabled,
+    InvalidAttributeOperator,
+    InvalidAttributeSelector,
+    InvalidString,
+    InvalidRegexp,
+    InvalidPseudoClassSelector,
+    EmptyPseudoClassSelector,
+    InvalidPseudoClass,
+    InvalidPseudoElement,
+    UnmatchParenthesis,
+    NotHandled,
+    UnknownPseudoSelector,
+    InvalidNthExpression,
+} || PseudoClass.Error || Combinator.Error || std.mem.Allocator.Error;
+
+pub const ParseOptions = struct {
+    accept_pseudo_elts: bool = true,
+};
+
+pub const Parser = struct {
+    s: []const u8, // string to parse
+    i: usize = 0, // current position
+
+    opts: ParseOptions,
+
+    pub fn parse(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        return p.parseSelectorGroup(alloc);
+    }
+
+    // skipWhitespace consumes whitespace characters and comments.
+    // It returns true if there was actually anything to skip.
+    fn skipWhitespace(p: *Parser) bool {
+        var i = p.i;
+        while (i < p.s.len) {
+            const c = p.s[i];
+            // Whitespaces.
+            if (ascii.isWhitespace(c)) {
+                i += 1;
+                continue;
+            }
+
+            // Comments.
+            if (c == '/') {
+                if (std.mem.startsWith(u8, p.s[i..], "/*")) {
+                    if (std.mem.indexOf(u8, p.s[i..], "*/")) |end| {
+                        i += end + "*/".len;
+                        continue;
+                    }
+                }
+            }
+            break;
+        }
+
+        if (i > p.i) {
+            p.i = i;
+            return true;
+        }
+
+        return false;
+    }
+
+    // parseSimpleSelectorSequence parses a selector sequence that applies to
+    // a single element.
+    fn parseSimpleSelectorSequence(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        if (p.i >= p.s.len) {
+            return ParseError.ExpectedSelector;
+        }
+
+        var buf = std.ArrayList(Selector).init(alloc);
+        defer buf.deinit();
+
+        switch (p.s[p.i]) {
+            '*' => {
+                // It's the universal selector. Just skip over it, since it
+                // doesn't affect the meaning.
+                p.i += 1;
+
+                // other version of universal selector
+                if (p.i + 2 < p.s.len and std.mem.eql(u8, "|*", p.s[p.i .. p.i + 2])) {
+                    p.i += 2;
+                }
+            },
+            '#', '.', '[', ':' => {
+                // There's no type selector. Wait to process the other till the
+                // main loop.
+            },
+            else => try buf.append(try p.parseTypeSelector(alloc)),
+        }
+
+        var pseudo_elt: ?PseudoClass = null;
+
+        loop: while (p.i < p.s.len) {
+            var ns: Selector = switch (p.s[p.i]) {
+                '#' => try p.parseIDSelector(alloc),
+                '.' => try p.parseClassSelector(alloc),
+                '[' => try p.parseAttributeSelector(alloc),
+                ':' => try p.parsePseudoclassSelector(alloc),
+                else => break :loop,
+            };
+            errdefer ns.deinit(alloc);
+
+            // From https://drafts.csswg.org/selectors-3/#pseudo-elements :
+            // "Only one pseudo-element may appear per selector, and if present
+            // it must appear after the sequence of simple selectors that
+            // represents the subjects of the selector.""
+            switch (ns) {
+                .pseudo_element => |e| {
+                    //  We found a pseudo-element.
+                    //  Only one pseudo-element is accepted per selector.
+                    if (pseudo_elt != null) return ParseError.PseudoElementNotUnique;
+                    if (!p.opts.accept_pseudo_elts) return ParseError.PseudoElementDisabled;
+
+                    pseudo_elt = e;
+                    ns.deinit(alloc);
+                },
+                else => {
+                    if (pseudo_elt != null) return ParseError.PseudoElementNotAtSelectorEnd;
+                    try buf.append(ns);
+                },
+            }
+        }
+
+        // no need wrap the selectors in compoundSelector
+        if (buf.items.len == 1 and pseudo_elt == null) return buf.items[0];
+
+        return .{ .compound = .{ .selectors = try buf.toOwnedSlice(), .pseudo_elt = pseudo_elt } };
+    }
+
+    // parseTypeSelector parses a type selector (one that matches by tag name).
+    fn parseTypeSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+        try p.parseIdentifier(buf.writer());
+
+        return .{ .tag = try buf.toOwnedSlice() };
+    }
+
+    // parseIdentifier parses an identifier.
+    fn parseIdentifier(p: *Parser, w: anytype) ParseError!void {
+        const prefix = '-';
+        var numPrefix: usize = 0;
+
+        while (p.s.len > p.i and p.s[p.i] == prefix) {
+            p.i += 1;
+            numPrefix += 1;
+        }
+
+        if (p.s.len <= p.i) {
+            return ParseError.ExpectedSelector;
+        }
+
+        const c = p.s[p.i];
+        if (!nameStart(c) or c == '\\') {
+            return ParseError.ExpectedSelector;
+        }
+
+        var ii: usize = 0;
+        while (ii < numPrefix) {
+            w.writeByte(prefix) catch return ParseError.WriteError;
+            ii += 1;
+        }
+        try parseName(p, w);
+    }
+
+    // parseName parses a name (which is like an identifier, but doesn't have
+    // extra restrictions on the first character).
+    fn parseName(p: *Parser, w: anytype) ParseError!void {
+        var i = p.i;
+        var ok = false;
+
+        while (i < p.s.len) {
+            const c = p.s[i];
+
+            if (nameChar(c)) {
+                const start = i;
+                while (i < p.s.len and nameChar(p.s[i])) i += 1;
+                w.writeAll(p.s[start..i]) catch return ParseError.WriteError;
+                ok = true;
+            } else if (c == '\\') {
+                p.i = i;
+                try p.parseEscape(w);
+                i = p.i;
+                ok = true;
+            } else {
+                // default:
+                break;
+            }
+        }
+
+        if (!ok) return ParseError.ExpectedName;
+        p.i = i;
+    }
+
+    // parseEscape parses a backslash escape.
+    // The returned string is owned by the caller.
+    fn parseEscape(p: *Parser, w: anytype) ParseError!void {
+        if (p.s.len < p.i + 2 or p.s[p.i] != '\\') {
+            return ParseError.InvalidEscape;
+        }
+
+        const start = p.i + 1;
+        const c = p.s[start];
+        if (ascii.isWhitespace(c)) return ParseError.EscapeLineEndingOutsideString;
+
+        // unicode escape (hex)
+        if (ascii.isHex(c)) {
+            var i: usize = start;
+            while (i < start + 6 and i < p.s.len and ascii.isHex(p.s[i])) {
+                i += 1;
+            }
+            const v = std.fmt.parseUnsigned(u21, p.s[start..i], 16) catch return ParseError.InvalidUnicode;
+            if (p.s.len > i) {
+                switch (p.s[i]) {
+                    '\r' => {
+                        i += 1;
+                        if (p.s.len > i and p.s[i] == '\n') i += 1;
+                    },
+                    ' ', '\t', '\n', std.ascii.control_code.ff => i += 1,
+                    else => {},
+                }
+                p.i = i;
+                var buf: [4]u8 = undefined;
+                const ln = std.unicode.utf8Encode(v, &buf) catch return ParseError.InvalidUnicode;
+                w.writeAll(buf[0..ln]) catch return ParseError.WriteError;
+                return;
+            }
+        }
+
+        // Return the literal character after the backslash.
+        p.i += 2;
+        w.writeAll(p.s[start .. start + 1]) catch return ParseError.WriteError;
+    }
+
+    // parseIDSelector parses a selector that matches by id attribute.
+    fn parseIDSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        if (p.i >= p.s.len) return ParseError.ExpectedIDSelector;
+        if (p.s[p.i] != '#') return ParseError.ExpectedIDSelector;
+
+        p.i += 1;
+
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try p.parseName(buf.writer());
+        return .{ .id = try buf.toOwnedSlice() };
+    }
+
+    // parseClassSelector parses a selector that matches by class attribute.
+    fn parseClassSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        if (p.i >= p.s.len) return ParseError.ExpectedClassSelector;
+        if (p.s[p.i] != '.') return ParseError.ExpectedClassSelector;
+
+        p.i += 1;
+
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try p.parseIdentifier(buf.writer());
+        return .{ .class = try buf.toOwnedSlice() };
+    }
+
+    // parseAttributeSelector parses a selector that matches by attribute value.
+    fn parseAttributeSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        if (p.i >= p.s.len) return ParseError.ExpectedAttributeSelector;
+        if (p.s[p.i] != '[') return ParseError.ExpectedAttributeSelector;
+
+        p.i += 1;
+        _ = p.skipWhitespace();
+
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try p.parseIdentifier(buf.writer());
+        const key = try buf.toOwnedSlice();
+        errdefer alloc.free(key);
+
+        lowerstr(key);
+
+        _ = p.skipWhitespace();
+        if (p.i >= p.s.len) return ParseError.ExpectedAttributeSelector;
+        if (p.s[p.i] == ']') {
+            p.i += 1;
+            return .{ .attribute = .{ .key = key } };
+        }
+
+        if (p.i + 2 >= p.s.len) return ParseError.ExpectedAttributeSelector;
+
+        const op = try parseAttributeOP(p.s[p.i .. p.i + 2]);
+        p.i += op.len();
+
+        _ = p.skipWhitespace();
+        if (p.i >= p.s.len) return ParseError.ExpectedAttributeSelector;
+
+        buf.clearRetainingCapacity();
+        var is_val: bool = undefined;
+        if (op == .regexp) {
+            is_val = false;
+            try p.parseRegex(buf.writer());
+        } else {
+            is_val = true;
+            switch (p.s[p.i]) {
+                '\'', '"' => try p.parseString(buf.writer()),
+                else => try p.parseIdentifier(buf.writer()),
+            }
+        }
+
+        _ = p.skipWhitespace();
+        if (p.i >= p.s.len) return ParseError.ExpectedAttributeSelector;
+
+        // check if the attribute contains an ignore case flag
+        var ci = false;
+        if (p.s[p.i] == 'i' or p.s[p.i] == 'I') {
+            ci = true;
+            p.i += 1;
+        }
+
+        _ = p.skipWhitespace();
+        if (p.i >= p.s.len) return ParseError.ExpectedAttributeSelector;
+
+        if (p.s[p.i] != ']') return ParseError.InvalidAttributeSelector;
+        p.i += 1;
+
+        return .{ .attribute = .{
+            .key = key,
+            .val = if (is_val) try buf.toOwnedSlice() else null,
+            .regexp = if (!is_val) try buf.toOwnedSlice() else null,
+            .op = op,
+            .ci = ci,
+        } };
+    }
+
+    // parseString parses a single- or double-quoted string.
+    fn parseString(p: *Parser, writer: anytype) ParseError!void {
+        var i = p.i;
+        if (p.s.len < i + 2) return ParseError.ExpectedString;
+
+        const quote = p.s[i];
+        i += 1;
+
+        loop: while (i < p.s.len) {
+            switch (p.s[i]) {
+                '\\' => {
+                    if (p.s.len > i + 1) {
+                        const c = p.s[i + 1];
+                        switch (c) {
+                            '\r' => {
+                                if (p.s.len > i + 2 and p.s[i + 2] == '\n') {
+                                    i += 3;
+                                    continue :loop;
+                                }
+                                i += 2;
+                                continue :loop;
+                            },
+                            '\n', std.ascii.control_code.ff => {
+                                i += 2;
+                                continue :loop;
+                            },
+                            else => {},
+                        }
+                    }
+                    p.i = i;
+                    try p.parseEscape(writer);
+                    i = p.i;
+                },
+                '\r', '\n', std.ascii.control_code.ff => return ParseError.InvalidString,
+                else => |c| {
+                    if (c == quote) break :loop;
+                    const start = i;
+                    while (i < p.s.len) {
+                        const cc = p.s[i];
+                        if (cc == quote or cc == '\\' or c == '\r' or c == '\n' or c == std.ascii.control_code.ff) break;
+                        i += 1;
+                    }
+                    writer.writeAll(p.s[start..i]) catch return ParseError.WriteError;
+                },
+            }
+        }
+
+        if (i >= p.s.len) return ParseError.InvalidString;
+
+        // Consume the final quote.
+        i += 1;
+        p.i = i;
+    }
+
+    // parseRegex parses a regular expression; the end is defined by encountering an
+    // unmatched closing ')' or ']' which is not consumed
+    fn parseRegex(p: *Parser, writer: anytype) ParseError!void {
+        var i = p.i;
+        if (p.s.len < i + 2) return ParseError.ExpectedRegexp;
+
+        // number of open parens or brackets;
+        // when it becomes negative, finished parsing regex
+        var open: isize = 0;
+
+        loop: while (i < p.s.len) {
+            switch (p.s[i]) {
+                '(', '[' => open += 1,
+                ')', ']' => {
+                    open -= 1;
+                    if (open < 0) break :loop;
+                },
+                else => {},
+            }
+            i += 1;
+        }
+
+        if (i >= p.s.len) return ParseError.InvalidRegexp;
+        writer.writeAll(p.s[p.i..i]) catch return ParseError.WriteError;
+        p.i = i;
+    }
+
+    // parsePseudoclassSelector parses a pseudoclass selector like :not(p) or a pseudo-element
+    // For backwards compatibility, both ':' and '::' prefix are allowed for pseudo-elements.
+    // https://drafts.csswg.org/selectors-3/#pseudo-elements
+    fn parsePseudoclassSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        if (p.i >= p.s.len) return ParseError.ExpectedPseudoClassSelector;
+        if (p.s[p.i] != ':') return ParseError.ExpectedPseudoClassSelector;
+
+        p.i += 1;
+
+        var must_pseudo_elt: bool = false;
+        if (p.i >= p.s.len) return ParseError.EmptyPseudoClassSelector;
+        if (p.s[p.i] == ':') { // we found a pseudo-element
+            must_pseudo_elt = true;
+            p.i += 1;
+        }
+
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try p.parseIdentifier(buf.writer());
+
+        const pseudo_class = try PseudoClass.parse(buf.items);
+
+        // reset the buffer to reuse it.
+        buf.clearRetainingCapacity();
+
+        if (must_pseudo_elt and !pseudo_class.isPseudoElement()) return ParseError.InvalidPseudoElement;
+
+        switch (pseudo_class) {
+            .not, .has, .haschild => {
+                if (!p.consumeParenthesis()) return ParseError.ExpectedParenthesis;
+
+                const sel = try p.parseSelectorGroup(alloc);
+                if (!p.consumeClosingParenthesis()) return ParseError.ExpectedParenthesisClose;
+
+                const s = try alloc.create(Selector);
+                errdefer alloc.destroy(s);
+                s.* = sel;
+
+                return .{ .pseudo_class_relative = .{ .pseudo_class = pseudo_class, .match = s } };
+            },
+            .contains, .containsown => {
+                if (!p.consumeParenthesis()) return ParseError.ExpectedParenthesis;
+                if (p.i == p.s.len) return ParseError.UnmatchParenthesis;
+
+                switch (p.s[p.i]) {
+                    '\'', '"' => try p.parseString(buf.writer()),
+                    else => try p.parseString(buf.writer()),
+                }
+
+                _ = p.skipWhitespace();
+                if (p.i >= p.s.len) return ParseError.InvalidPseudoClass;
+                if (!p.consumeClosingParenthesis()) return ParseError.ExpectedParenthesisClose;
+
+                const val = try buf.toOwnedSlice();
+                errdefer alloc.free(val);
+
+                lowerstr(val);
+
+                return .{ .pseudo_class_contains = .{ .own = pseudo_class == .containsown, .val = val } };
+            },
+            .matches, .matchesown => {
+                if (!p.consumeParenthesis()) return ParseError.ExpectedParenthesis;
+
+                try p.parseRegex(buf.writer());
+                if (p.i >= p.s.len) return ParseError.InvalidPseudoClassSelector;
+                if (!p.consumeClosingParenthesis()) return ParseError.ExpectedParenthesisClose;
+
+                return .{ .pseudo_class_regexp = .{ .own = pseudo_class == .matchesown, .regexp = try buf.toOwnedSlice() } };
+            },
+            .nth_child, .nth_last_child, .nth_of_type, .nth_last_of_type => {
+                if (!p.consumeParenthesis()) return ParseError.ExpectedParenthesis;
+                const nth = try p.parseNth(alloc);
+                if (!p.consumeClosingParenthesis()) return ParseError.ExpectedParenthesisClose;
+
+                const last = pseudo_class == .nth_last_child or pseudo_class == .nth_last_of_type;
+                const of_type = pseudo_class == .nth_of_type or pseudo_class == .nth_last_of_type;
+                return .{ .pseudo_class_nth = .{ .a = nth[0], .b = nth[1], .of_type = of_type, .last = last } };
+            },
+            .first_child => return .{ .pseudo_class_nth = .{ .a = 0, .b = 1, .of_type = false, .last = false } },
+            .last_child => return .{ .pseudo_class_nth = .{ .a = 0, .b = 1, .of_type = false, .last = true } },
+            .first_of_type => return .{ .pseudo_class_nth = .{ .a = 0, .b = 1, .of_type = true, .last = false } },
+            .last_of_type => return .{ .pseudo_class_nth = .{ .a = 0, .b = 1, .of_type = true, .last = true } },
+            .only_child => return .{ .pseudo_class_only_child = false },
+            .only_of_type => return .{ .pseudo_class_only_child = true },
+            .input, .empty, .root, .link => return .{ .pseudo_class = pseudo_class },
+            .enabled, .disabled, .checked => return .{ .pseudo_class = pseudo_class },
+            .lang => {
+                if (!p.consumeParenthesis()) return ParseError.ExpectedParenthesis;
+                if (p.i == p.s.len) return ParseError.UnmatchParenthesis;
+
+                try p.parseIdentifier(buf.writer());
+
+                _ = p.skipWhitespace();
+                if (p.i >= p.s.len) return ParseError.InvalidPseudoClass;
+                if (!p.consumeClosingParenthesis()) return ParseError.ExpectedParenthesisClose;
+
+                const val = try buf.toOwnedSlice();
+                errdefer alloc.free(val);
+                lowerstr(val);
+
+                return .{ .pseudo_class_lang = val };
+            },
+            .visited, .hover, .active, .focus, .target => {
+                // Not applicable in a static context: never match.
+                return .{ .never_match = pseudo_class };
+            },
+            .after, .backdrop, .before, .cue, .first_letter => return .{ .pseudo_element = pseudo_class },
+            .first_line, .grammar_error, .marker, .placeholder => return .{ .pseudo_element = pseudo_class },
+            .selection, .spelling_error => return .{ .pseudo_element = pseudo_class },
+        }
+    }
+
+    // consumeParenthesis consumes an opening parenthesis and any following
+    // whitespace. It returns true if there was actually a parenthesis to skip.
+    fn consumeParenthesis(p: *Parser) bool {
+        if (p.i < p.s.len and p.s[p.i] == '(') {
+            p.i += 1;
+            _ = p.skipWhitespace();
+            return true;
+        }
+        return false;
+    }
+
+    // parseSelectorGroup parses a group of selectors, separated by commas.
+    fn parseSelectorGroup(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        const s = try p.parseSelector(alloc);
+
+        var buf = std.ArrayList(Selector).init(alloc);
+        defer buf.deinit();
+
+        try buf.append(s);
+
+        while (p.i < p.s.len) {
+            if (p.s[p.i] != ',') break;
+            p.i += 1;
+            const ss = try p.parseSelector(alloc);
+            try buf.append(ss);
+        }
+
+        if (buf.items.len == 1) return buf.items[0];
+
+        return .{ .group = try buf.toOwnedSlice() };
+    }
+
+    // parseSelector parses a selector that may include combinators.
+    fn parseSelector(p: *Parser, alloc: std.mem.Allocator) ParseError!Selector {
+        _ = p.skipWhitespace();
+        var s = try p.parseSimpleSelectorSequence(alloc);
+
+        while (true) {
+            var combinator: Combinator = .empty;
+            if (p.skipWhitespace()) {
+                combinator = .descendant;
+            }
+            if (p.i >= p.s.len) {
+                return s;
+            }
+
+            switch (p.s[p.i]) {
+                '+', '>', '~' => {
+                    combinator = try Combinator.parse(p.s[p.i]);
+                    p.i += 1;
+                    _ = p.skipWhitespace();
+                },
+                // These characters can't begin a selector, but they can legally occur after one.
+                ',', ')' => {
+                    return s;
+                },
+                else => {},
+            }
+
+            if (combinator == .empty) {
+                return s;
+            }
+
+            const c = try p.parseSimpleSelectorSequence(alloc);
+
+            const first = try alloc.create(Selector);
+            errdefer alloc.destroy(first);
+            first.* = s;
+
+            const second = try alloc.create(Selector);
+            errdefer alloc.destroy(second);
+            second.* = c;
+
+            s = Selector{ .combined = .{ .first = first, .second = second, .combinator = combinator } };
+        }
+
+        return s;
+    }
+
+    // consumeClosingParenthesis consumes a closing parenthesis and any preceding
+    // whitespace. It returns true if there was actually a parenthesis to skip.
+    fn consumeClosingParenthesis(p: *Parser) bool {
+        const i = p.i;
+        _ = p.skipWhitespace();
+        if (p.i < p.s.len and p.s[p.i] == ')') {
+            p.i += 1;
+            return true;
+        }
+        p.i = i;
+        return false;
+    }
+
+    // parseInteger parses a  decimal integer.
+    fn parseInteger(p: *Parser) ParseError!isize {
+        var i = p.i;
+        const start = i;
+        while (i < p.s.len and '0' <= p.s[i] and p.s[i] <= '9') i += 1;
+        if (i == start) return ParseError.ExpectedInteger;
+        p.i = i;
+
+        return std.fmt.parseUnsigned(isize, p.s[start..i], 10) catch ParseError.ExpectedInteger;
+    }
+
+    fn parseNthReadN(p: *Parser, a: isize) ParseError![2]isize {
+        _ = p.skipWhitespace();
+        if (p.i >= p.s.len) return ParseError.ExpectedNthExpression;
+
+        return switch (p.s[p.i]) {
+            '+' => {
+                p.i += 1;
+                _ = p.skipWhitespace();
+                const b = try p.parseInteger();
+                return .{ a, b };
+            },
+            '-' => {
+                p.i += 1;
+                _ = p.skipWhitespace();
+                const b = try p.parseInteger();
+                return .{ a, -b };
+            },
+            else => .{ a, 0 },
+        };
+    }
+
+    fn parseNthReadA(p: *Parser, a: isize) ParseError![2]isize {
+        if (p.i >= p.s.len) return ParseError.ExpectedNthExpression;
+        return switch (p.s[p.i]) {
+            'n', 'N' => {
+                p.i += 1;
+                return p.parseNthReadN(a);
+            },
+            else => .{ 0, a },
+        };
+    }
+
+    fn parseNthNegativeA(p: *Parser) ParseError![2]isize {
+        if (p.i >= p.s.len) return ParseError.ExpectedNthExpression;
+        const c = p.s[p.i];
+        if (std.ascii.isDigit(c)) {
+            const a = try p.parseInteger() * -1;
+            return p.parseNthReadA(a);
+        }
+        if (c == 'n' or c == 'N') {
+            p.i += 1;
+            return p.parseNthReadN(-1);
+        }
+
+        return ParseError.InvalidNthExpression;
+    }
+
+    fn parseNthPositiveA(p: *Parser) ParseError![2]isize {
+        if (p.i >= p.s.len) return ParseError.ExpectedNthExpression;
+        const c = p.s[p.i];
+        if (std.ascii.isDigit(c)) {
+            const a = try p.parseInteger();
+            return p.parseNthReadA(a);
+        }
+        if (c == 'n' or c == 'N') {
+            p.i += 1;
+            return p.parseNthReadN(1);
+        }
+
+        return ParseError.InvalidNthExpression;
+    }
+
+    // parseNth parses the argument for :nth-child (normally of the form an+b).
+    fn parseNth(p: *Parser, alloc: std.mem.Allocator) ParseError![2]isize {
+        // initial state
+        if (p.i >= p.s.len) return ParseError.ExpectedNthExpression;
+        return switch (p.s[p.i]) {
+            '-' => {
+                p.i += 1;
+                return p.parseNthNegativeA();
+            },
+            '+' => {
+                p.i += 1;
+                return p.parseNthPositiveA();
+            },
+            '0', '1', '2', '3', '4', '5', '6', '7', '8', '9' => p.parseNthPositiveA(),
+            'n', 'N' => {
+                p.i += 1;
+                return p.parseNthReadN(1);
+            },
+            'o', 'O', 'e', 'E' => {
+                var buf = std.ArrayList(u8).init(alloc);
+                defer buf.deinit();
+
+                try p.parseName(buf.writer());
+
+                if (std.ascii.eqlIgnoreCase("odd", buf.items)) return .{ 2, 1 };
+                if (std.ascii.eqlIgnoreCase("even", buf.items)) return .{ 2, 0 };
+
+                return ParseError.InvalidNthExpression;
+            },
+            else => ParseError.InvalidNthExpression,
+        };
+    }
+};
+
+// nameStart returns whether c can be the first character of an identifier
+// (not counting an initial hyphen, or an escape sequence).
+fn nameStart(c: u8) bool {
+    return 'a' <= c and c <= 'z' or 'A' <= c and c <= 'Z' or c == '_' or c > 127;
+}
+
+// nameChar returns whether c can be a character within an identifier
+// (not counting an escape sequence).
+fn nameChar(c: u8) bool {
+    return 'a' <= c and c <= 'z' or 'A' <= c and c <= 'Z' or c == '_' or c > 127 or
+        c == '-' or '0' <= c and c <= '9';
+}
+
+fn lowerstr(str: []u8) void {
+    for (str, 0..) |c, i| {
+        str[i] = std.ascii.toLower(c);
+    }
+}
+
+// parseAttributeOP parses an AttributeOP from a string of 1 or 2 bytes.
+fn parseAttributeOP(s: []const u8) ParseError!AttributeOP {
+    if (s.len < 1 or s.len > 2) return ParseError.InvalidAttributeOperator;
+
+    // if the first sign is equal, we don't check anything else.
+    if (s[0] == '=') return .eql;
+
+    if (s.len != 2 or s[1] != '=') return ParseError.InvalidAttributeOperator;
+
+    return switch (s[0]) {
+        '=' => .eql,
+        '!' => .not_eql,
+        '~' => .one_of,
+        '|' => .prefix_hyphen,
+        '^' => .prefix,
+        '$' => .suffix,
+        '*' => .contains,
+        '#' => .regexp,
+        else => ParseError.InvalidAttributeOperator,
+    };
+}
+
+test "parser.skipWhitespace" {
+    const testcases = [_]struct {
+        s: []const u8,
+        i: usize,
+        r: bool,
+    }{
+        .{ .s = "", .i = 0, .r = false },
+        .{ .s = "foo", .i = 0, .r = false },
+        .{ .s = " ", .i = 1, .r = true },
+        .{ .s = " foo", .i = 1, .r = true },
+        .{ .s = "/* foo */ bar", .i = 10, .r = true },
+        .{ .s = "/* foo", .i = 0, .r = false },
+    };
+
+    for (testcases) |tc| {
+        var p = Parser{ .s = tc.s, .opts = .{} };
+        const res = p.skipWhitespace();
+        try std.testing.expectEqual(tc.r, res);
+        try std.testing.expectEqual(tc.i, p.i);
+    }
+}
+
+test "parser.parseIdentifier" {
+    const alloc = std.testing.allocator;
+
+    const testcases = [_]struct {
+        s: []const u8, // given value
+        exp: []const u8, // expected value
+        err: bool = false,
+    }{
+        .{ .s = "x", .exp = "x" },
+        .{ .s = "96", .exp = "", .err = true },
+        .{ .s = "-x", .exp = "-x" },
+        .{ .s = "r\\e9 sumé", .exp = "résumé" },
+        .{ .s = "r\\0000e9 sumé", .exp = "résumé" },
+        .{ .s = "r\\0000e9sumé", .exp = "résumé" },
+        .{ .s = "a\\\"b", .exp = "a\"b" },
+    };
+
+    var buf = std.ArrayList(u8).init(alloc);
+    defer buf.deinit();
+
+    for (testcases) |tc| {
+        buf.clearRetainingCapacity();
+
+        var p = Parser{ .s = tc.s, .opts = .{} };
+        p.parseIdentifier(buf.writer()) catch |e| {
+            // if error was expected, continue.
+            if (tc.err) continue;
+
+            std.debug.print("test case {s}\n", .{tc.s});
+            return e;
+        };
+        std.testing.expectEqualDeep(tc.exp, buf.items) catch |e| {
+            std.debug.print("test case {s} : {s}\n", .{ tc.s, buf.items });
+            return e;
+        };
+    }
+}
+
+test "parser.parseString" {
+    const alloc = std.testing.allocator;
+
+    const testcases = [_]struct {
+        s: []const u8, // given value
+        exp: []const u8, // expected value
+        err: bool = false,
+    }{
+        .{ .s = "\"x\"", .exp = "x" },
+        .{ .s = "'x'", .exp = "x" },
+        .{ .s = "'x", .exp = "", .err = true },
+        .{ .s = "'x\\\r\nx'", .exp = "xx" },
+        .{ .s = "\"r\\e9 sumé\"", .exp = "résumé" },
+        .{ .s = "\"r\\0000e9 sumé\"", .exp = "résumé" },
+        .{ .s = "\"r\\0000e9sumé\"", .exp = "résumé" },
+        .{ .s = "\"a\\\"b\"", .exp = "a\"b" },
+        .{ .s = "\"\\\n\"", .exp = "" },
+        .{ .s = "\"hello world\"", .exp = "hello world" },
+    };
+
+    var buf = std.ArrayList(u8).init(alloc);
+    defer buf.deinit();
+
+    for (testcases) |tc| {
+        buf.clearRetainingCapacity();
+
+        var p = Parser{ .s = tc.s, .opts = .{} };
+        p.parseString(buf.writer()) catch |e| {
+            // if error was expected, continue.
+            if (tc.err) continue;
+
+            std.debug.print("test case {s}\n", .{tc.s});
+            return e;
+        };
+        std.testing.expectEqualDeep(tc.exp, buf.items) catch |e| {
+            std.debug.print("test case {s} : {s}\n", .{ tc.s, buf.items });
+            return e;
+        };
+    }
+}

src/css/selector.zig

@@ -0,0 +1,767 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+pub const AttributeOP = enum {
+    eql, // =
+    not_eql, // !=
+    one_of, // ~=
+    prefix_hyphen, // |=
+    prefix, // ^=
+    suffix, // $=
+    contains, // *=
+    regexp, // #=
+
+    pub fn len(op: AttributeOP) u2 {
+        if (op == .eql) return 1;
+        return 2;
+    }
+};
+
+pub const Combinator = enum {
+    empty,
+    descendant, // space
+    child, // >
+    next_sibling, // +
+    subsequent_sibling, // ~
+
+    pub const Error = error{
+        InvalidCombinator,
+    };
+
+    pub fn parse(c: u8) Error!Combinator {
+        return switch (c) {
+            ' ' => .descendant,
+            '>' => .child,
+            '+' => .next_sibling,
+            '~' => .subsequent_sibling,
+            else => Error.InvalidCombinator,
+        };
+    }
+};
+
+pub const PseudoClass = enum {
+    not,
+    has,
+    haschild,
+    contains,
+    containsown,
+    matches,
+    matchesown,
+    nth_child,
+    nth_last_child,
+    nth_of_type,
+    nth_last_of_type,
+    first_child,
+    last_child,
+    first_of_type,
+    last_of_type,
+    only_child,
+    only_of_type,
+    input,
+    empty,
+    root,
+    link,
+    lang,
+    enabled,
+    disabled,
+    checked,
+    visited,
+    hover,
+    active,
+    focus,
+    target,
+    after,
+    backdrop,
+    before,
+    cue,
+    first_letter,
+    first_line,
+    grammar_error,
+    marker,
+    placeholder,
+    selection,
+    spelling_error,
+
+    pub const Error = error{
+        InvalidPseudoClass,
+    };
+
+    pub fn isPseudoElement(pc: PseudoClass) bool {
+        return switch (pc) {
+            .after, .backdrop, .before, .cue, .first_letter => true,
+            .first_line, .grammar_error, .marker, .placeholder => true,
+            .selection, .spelling_error => true,
+            else => false,
+        };
+    }
+
+    pub fn parse(s: []const u8) Error!PseudoClass {
+        if (std.ascii.eqlIgnoreCase(s, "not")) return .not;
+        if (std.ascii.eqlIgnoreCase(s, "has")) return .has;
+        if (std.ascii.eqlIgnoreCase(s, "haschild")) return .haschild;
+        if (std.ascii.eqlIgnoreCase(s, "contains")) return .contains;
+        if (std.ascii.eqlIgnoreCase(s, "containsown")) return .containsown;
+        if (std.ascii.eqlIgnoreCase(s, "matches")) return .matches;
+        if (std.ascii.eqlIgnoreCase(s, "matchesown")) return .matchesown;
+        if (std.ascii.eqlIgnoreCase(s, "nth-child")) return .nth_child;
+        if (std.ascii.eqlIgnoreCase(s, "nth-last-child")) return .nth_last_child;
+        if (std.ascii.eqlIgnoreCase(s, "nth-of-type")) return .nth_of_type;
+        if (std.ascii.eqlIgnoreCase(s, "nth-last-of-type")) return .nth_last_of_type;
+        if (std.ascii.eqlIgnoreCase(s, "first-child")) return .first_child;
+        if (std.ascii.eqlIgnoreCase(s, "last-child")) return .last_child;
+        if (std.ascii.eqlIgnoreCase(s, "first-of-type")) return .first_of_type;
+        if (std.ascii.eqlIgnoreCase(s, "last-of-type")) return .last_of_type;
+        if (std.ascii.eqlIgnoreCase(s, "only-child")) return .only_child;
+        if (std.ascii.eqlIgnoreCase(s, "only-of-type")) return .only_of_type;
+        if (std.ascii.eqlIgnoreCase(s, "input")) return .input;
+        if (std.ascii.eqlIgnoreCase(s, "empty")) return .empty;
+        if (std.ascii.eqlIgnoreCase(s, "root")) return .root;
+        if (std.ascii.eqlIgnoreCase(s, "link")) return .link;
+        if (std.ascii.eqlIgnoreCase(s, "lang")) return .lang;
+        if (std.ascii.eqlIgnoreCase(s, "enabled")) return .enabled;
+        if (std.ascii.eqlIgnoreCase(s, "disabled")) return .disabled;
+        if (std.ascii.eqlIgnoreCase(s, "checked")) return .checked;
+        if (std.ascii.eqlIgnoreCase(s, "visited")) return .visited;
+        if (std.ascii.eqlIgnoreCase(s, "hover")) return .hover;
+        if (std.ascii.eqlIgnoreCase(s, "active")) return .active;
+        if (std.ascii.eqlIgnoreCase(s, "focus")) return .focus;
+        if (std.ascii.eqlIgnoreCase(s, "target")) return .target;
+        if (std.ascii.eqlIgnoreCase(s, "after")) return .after;
+        if (std.ascii.eqlIgnoreCase(s, "backdrop")) return .backdrop;
+        if (std.ascii.eqlIgnoreCase(s, "before")) return .before;
+        if (std.ascii.eqlIgnoreCase(s, "cue")) return .cue;
+        if (std.ascii.eqlIgnoreCase(s, "first-letter")) return .first_letter;
+        if (std.ascii.eqlIgnoreCase(s, "first-line")) return .first_line;
+        if (std.ascii.eqlIgnoreCase(s, "grammar-error")) return .grammar_error;
+        if (std.ascii.eqlIgnoreCase(s, "marker")) return .marker;
+        if (std.ascii.eqlIgnoreCase(s, "placeholder")) return .placeholder;
+        if (std.ascii.eqlIgnoreCase(s, "selection")) return .selection;
+        if (std.ascii.eqlIgnoreCase(s, "spelling-error")) return .spelling_error;
+        return Error.InvalidPseudoClass;
+    }
+};
+
+pub const Selector = union(enum) {
+    pub const Error = error{
+        UnknownCombinedCombinator,
+        UnsupportedRelativePseudoClass,
+        UnsupportedContainsPseudoClass,
+        UnsupportedPseudoClass,
+        UnsupportedPseudoElement,
+        UnsupportedRegexpPseudoClass,
+        UnsupportedAttrRegexpOperator,
+    };
+
+    compound: struct {
+        selectors: []Selector,
+        pseudo_elt: ?PseudoClass,
+    },
+    group: []Selector,
+    tag: []const u8,
+    id: []const u8,
+    class: []const u8,
+    attribute: struct {
+        key: []const u8,
+        val: ?[]const u8 = null,
+        op: ?AttributeOP = null,
+        regexp: ?[]const u8 = null,
+        ci: bool = false,
+    },
+    combined: struct {
+        first: *Selector,
+        second: *Selector,
+        combinator: Combinator,
+    },
+
+    never_match: PseudoClass,
+
+    pseudo_class: PseudoClass,
+    pseudo_class_only_child: bool,
+    pseudo_class_lang: []const u8,
+    pseudo_class_relative: struct {
+        pseudo_class: PseudoClass,
+        match: *Selector,
+    },
+    pseudo_class_contains: struct {
+        own: bool,
+        val: []const u8,
+    },
+    pseudo_class_regexp: struct {
+        own: bool,
+        regexp: []const u8,
+    },
+    pseudo_class_nth: struct {
+        a: isize,
+        b: isize,
+        of_type: bool,
+        last: bool,
+    },
+    pseudo_element: PseudoClass,
+
+    // returns true if s is a whitespace-separated list that includes val.
+    fn word(haystack: []const u8, needle: []const u8, ci: bool) bool {
+        if (haystack.len == 0) return false;
+        var it = std.mem.splitAny(u8, haystack, " \t\r\n"); // TODO add \f
+        while (it.next()) |part| {
+            if (eql(part, needle, ci)) return true;
+        }
+        return false;
+    }
+
+    fn eql(a: []const u8, b: []const u8, ci: bool) bool {
+        if (ci) return std.ascii.eqlIgnoreCase(a, b);
+        return std.mem.eql(u8, a, b);
+    }
+
+    fn starts(haystack: []const u8, needle: []const u8, ci: bool) bool {
+        if (ci) return std.ascii.startsWithIgnoreCase(haystack, needle);
+        return std.mem.startsWith(u8, haystack, needle);
+    }
+
+    fn ends(haystack: []const u8, needle: []const u8, ci: bool) bool {
+        if (ci) return std.ascii.endsWithIgnoreCase(haystack, needle);
+        return std.mem.endsWith(u8, haystack, needle);
+    }
+
+    fn contains(haystack: []const u8, needle: []const u8, ci: bool) bool {
+        if (ci) return std.ascii.indexOfIgnoreCase(haystack, needle) != null;
+        return std.mem.indexOf(u8, haystack, needle) != null;
+    }
+
+    // match returns true if the node matches the selector query.
+    pub fn match(s: Selector, n: anytype) !bool {
+        return switch (s) {
+            .tag => |v| n.isElement() and std.ascii.eqlIgnoreCase(v, try n.tag()),
+            .id => |v| return n.isElement() and std.mem.eql(u8, v, try n.attr("id") orelse return false),
+            .class => |v| return n.isElement() and word(try n.attr("class") orelse return false, v, false),
+            .group => |v| {
+                for (v) |sel| {
+                    if (try sel.match(n)) return true;
+                }
+                return false;
+            },
+            .compound => |v| {
+                if (v.selectors.len == 0) return n.isElement();
+
+                for (v.selectors) |sel| {
+                    if (!try sel.match(n)) return false;
+                }
+                return true;
+            },
+            .combined => |v| {
+                return switch (v.combinator) {
+                    .empty => try v.first.match(n),
+                    .descendant => {
+                        if (!try v.second.match(n)) return false;
+
+                        // The first must match a ascendent.
+                        var p = try n.parent();
+                        while (p != null) {
+                            if (try v.first.match(p.?)) {
+                                return true;
+                            }
+                            p = try p.?.parent();
+                        }
+
+                        return false;
+                    },
+                    .child => {
+                        const p = try n.parent();
+                        if (p == null) return false;
+
+                        return try v.second.match(n) and try v.first.match(p.?);
+                    },
+                    .next_sibling => {
+                        if (!try v.second.match(n)) return false;
+                        var c = try n.prevSibling();
+                        while (c != null) {
+                            if (c.?.isText() or c.?.isComment()) {
+                                c = try c.?.prevSibling();
+                                continue;
+                            }
+                            return try v.first.match(c.?);
+                        }
+                        return false;
+                    },
+                    .subsequent_sibling => {
+                        if (!try v.second.match(n)) return false;
+
+                        var c = try n.prevSibling();
+                        while (c != null) {
+                            if (try v.first.match(c.?)) return true;
+                            c = try c.?.prevSibling();
+                        }
+                        return false;
+                    },
+                };
+            },
+            .attribute => |v| {
+                var attr = try n.attr(v.key);
+
+                if (v.op == null) return attr != null;
+                if (v.val == null or v.val.?.len == 0) return false;
+
+                const val = v.val.?;
+
+                return switch (v.op.?) {
+                    .eql => attr != null and eql(attr.?, val, v.ci),
+                    .not_eql => attr == null or !eql(attr.?, val, v.ci),
+                    .one_of => attr != null and word(attr.?, val, v.ci),
+                    .prefix => {
+                        if (attr == null) return false;
+                        attr.? = std.mem.trim(u8, attr.?, &std.ascii.whitespace);
+
+                        if (attr.?.len == 0) return false;
+
+                        return starts(attr.?, val, v.ci);
+                    },
+                    .suffix => {
+                        if (attr == null) return false;
+                        attr.? = std.mem.trim(u8, attr.?, &std.ascii.whitespace);
+
+                        if (attr.?.len == 0) return false;
+
+                        return ends(attr.?, val, v.ci);
+                    },
+                    .contains => {
+                        if (attr == null) return false;
+                        attr.? = std.mem.trim(u8, attr.?, &std.ascii.whitespace);
+
+                        if (attr.?.len == 0) return false;
+
+                        return contains(attr.?, val, v.ci);
+                    },
+                    .prefix_hyphen => {
+                        if (attr == null) return false;
+                        if (eql(attr.?, val, v.ci)) return true;
+
+                        if (attr.?.len <= val.len) return false;
+
+                        if (!starts(attr.?, val, v.ci)) return false;
+
+                        return attr.?[val.len] == '-';
+                    },
+                    .regexp => return Error.UnsupportedAttrRegexpOperator, // TODO handle regexp attribute operator.
+                };
+            },
+            .never_match => return false,
+            .pseudo_class_relative => |v| {
+                if (!n.isElement()) return false;
+
+                return switch (v.pseudo_class) {
+                    .not => !try v.match.match(n),
+                    .has => try hasDescendantMatch(v.match, n),
+                    .haschild => try hasChildMatch(v.match, n),
+                    else => Error.UnsupportedRelativePseudoClass,
+                };
+            },
+            .pseudo_class_contains => return Error.UnsupportedContainsPseudoClass, // TODO, need mem allocation.
+            .pseudo_class_regexp => return Error.UnsupportedRegexpPseudoClass, // TODO need mem allocation.
+            .pseudo_class_nth => |v| {
+                if (v.a == 0) {
+                    if (v.last) {
+                        return simpleNthLastChildMatch(v.b, v.of_type, n);
+                    }
+                    return simpleNthChildMatch(v.b, v.of_type, n);
+                }
+                return nthChildMatch(v.a, v.b, v.last, v.of_type, n);
+            },
+            .pseudo_class => |v| {
+                return switch (v) {
+                    .input => {
+                        if (!n.isElement()) return false;
+                        const ntag = try n.tag();
+
+                        return std.ascii.eqlIgnoreCase("input", ntag) or
+                            std.ascii.eqlIgnoreCase("select", ntag) or
+                            std.ascii.eqlIgnoreCase("button", ntag) or
+                            std.ascii.eqlIgnoreCase("textarea", ntag);
+                    },
+                    .empty => {
+                        if (!n.isElement()) return false;
+
+                        var c = try n.firstChild();
+                        while (c != null) {
+                            if (c.?.isElement()) return false;
+
+                            if (c.?.isText()) {
+                                if (try c.?.isEmptyText()) continue;
+                                return false;
+                            }
+
+                            c = try c.?.nextSibling();
+                        }
+
+                        return true;
+                    },
+                    .root => {
+                        if (!n.isElement()) return false;
+
+                        const p = try n.parent();
+                        return (p != null and p.?.isDocument());
+                    },
+                    .link => {
+                        const ntag = try n.tag();
+
+                        return std.ascii.eqlIgnoreCase("a", ntag) or
+                            std.ascii.eqlIgnoreCase("area", ntag) or
+                            std.ascii.eqlIgnoreCase("link", ntag);
+                    },
+                    .enabled => {
+                        if (!n.isElement()) return false;
+
+                        const ntag = try n.tag();
+
+                        if (std.ascii.eqlIgnoreCase("a", ntag) or
+                            std.ascii.eqlIgnoreCase("area", ntag) or
+                            std.ascii.eqlIgnoreCase("link", ntag))
+                        {
+                            return try n.attr("href") != null;
+                        }
+
+                        if (std.ascii.eqlIgnoreCase("optgroup", ntag) or
+                            std.ascii.eqlIgnoreCase("menuitem", ntag) or
+                            std.ascii.eqlIgnoreCase("fieldset", ntag))
+                        {
+                            return try n.attr("disabled") == null;
+                        }
+
+                        if (std.ascii.eqlIgnoreCase("input", ntag) or
+                            std.ascii.eqlIgnoreCase("button", ntag) or
+                            std.ascii.eqlIgnoreCase("select", ntag) or
+                            std.ascii.eqlIgnoreCase("textarea", ntag) or
+                            std.ascii.eqlIgnoreCase("option", ntag))
+                        {
+                            return try n.attr("disabled") == null and
+                                !try inDisabledFieldset(n);
+                        }
+
+                        return false;
+                    },
+                    .disabled => {
+                        if (!n.isElement()) return false;
+
+                        const ntag = try n.tag();
+
+                        if (std.ascii.eqlIgnoreCase("optgroup", ntag) or
+                            std.ascii.eqlIgnoreCase("menuitem", ntag) or
+                            std.ascii.eqlIgnoreCase("fieldset", ntag))
+                        {
+                            return try n.attr("disabled") != null;
+                        }
+
+                        if (std.ascii.eqlIgnoreCase("input", ntag) or
+                            std.ascii.eqlIgnoreCase("button", ntag) or
+                            std.ascii.eqlIgnoreCase("select", ntag) or
+                            std.ascii.eqlIgnoreCase("textarea", ntag) or
+                            std.ascii.eqlIgnoreCase("option", ntag))
+                        {
+                            return try n.attr("disabled") != null or
+                                try inDisabledFieldset(n);
+                        }
+
+                        return false;
+                    },
+                    .checked => {
+                        if (!n.isElement()) return false;
+
+                        const ntag = try n.tag();
+
+                        if (std.ascii.eqlIgnoreCase("intput", ntag)) {
+                            const ntype = try n.attr("type");
+                            if (ntype == null) return false;
+
+                            if (std.mem.eql(u8, ntype.?, "checkbox") or
+                                std.mem.eql(u8, ntype.?, "radio"))
+                            {
+                                return try n.attr("checked") != null;
+                            }
+
+                            return false;
+                        }
+                        if (std.ascii.eqlIgnoreCase("option", ntag)) {
+                            return try n.attr("selected") != null;
+                        }
+
+                        return false;
+                    },
+                    .visited => return false,
+                    .hover => return false,
+                    .active => return false,
+                    .focus => return false,
+                    // TODO implement using the url fragment.
+                    // see https://developer.mozilla.org/en-US/docs/Web/CSS/:target
+                    .target => return false,
+
+                    // all others pseudo class are handled by specialized
+                    // pseudo_class_X selectors.
+                    else => return Error.UnsupportedPseudoClass,
+                };
+            },
+            .pseudo_class_only_child => |v| onlyChildMatch(v, n),
+            .pseudo_class_lang => |v| langMatch(v, n),
+
+            // pseudo elements doesn't make sense in the matching process.
+            // > A CSS pseudo-element is a keyword added to a selector that
+            // > lets you style a specific part of the selected element(s).
+            // https://developer.mozilla.org/en-US/docs/Web/CSS/Pseudo-elements
+            .pseudo_element => return Error.UnsupportedPseudoElement,
+        };
+    }
+
+    fn hasLegendInPreviousSiblings(n: anytype) anyerror!bool {
+        var c = try n.prevSibling();
+        while (c != null) {
+            const ctag = try c.?.tag();
+            if (std.ascii.eqlIgnoreCase("legend", ctag)) return true;
+            c = try c.?.prevSibling();
+        }
+        return false;
+    }
+
+    fn inDisabledFieldset(n: anytype) anyerror!bool {
+        const p = try n.parent();
+        if (p == null) return false;
+
+        const ntag = try n.tag();
+        const ptag = try p.?.tag();
+
+        if (std.ascii.eqlIgnoreCase("fieldset", ptag) and
+            try p.?.attr("disabled") != null and
+            (!std.ascii.eqlIgnoreCase("legend", ntag) or try hasLegendInPreviousSiblings(n)))
+        {
+            return true;
+        }
+
+        // TODO should we handle legend like cascadia does?
+        // The implemention below looks suspicious, I didn't find a test case
+        // in cascadia and I didn't find the reference about legend in the
+        // specs. For now I do prefer ignoring this part.
+        //
+        // ```
+        // (n.DataAtom != atom.Legend || hasLegendInPreviousSiblings(n)) {
+        // ```
+        // https://github.com/andybalholm/cascadia/blob/master/pseudo_classes.go#L434
+
+        return try inDisabledFieldset(p.?);
+    }
+
+    fn langMatch(lang: []const u8, n: anytype) anyerror!bool {
+        if (try n.attr("lang")) |own| {
+            if (std.mem.eql(u8, own, lang)) return true;
+
+            // check if the lang attr starts with lang+'-'
+            if (std.mem.startsWith(u8, own, lang)) {
+                if (own.len > lang.len and own[lang.len] == '-') return true;
+            }
+        }
+
+        // if the tag doesn't match, try the parent.
+        const p = try n.parent();
+        if (p == null) return false;
+
+        return langMatch(lang, p.?);
+    }
+
+    // onlyChildMatch implements :only-child
+    //  If `ofType` is true, it implements :only-of-type instead.
+    fn onlyChildMatch(of_type: bool, n: anytype) anyerror!bool {
+        if (!n.isElement()) return false;
+
+        const p = try n.parent();
+        if (p == null) return false;
+
+        const ntag = try n.tag();
+
+        var count: usize = 0;
+        var c = try p.?.firstChild();
+        // loop hover all n siblings.
+        while (c != null) {
+            // ignore non elements or others tags if of-type is true.
+            if (!c.?.isElement() or (of_type and !std.mem.eql(u8, ntag, try c.?.tag()))) {
+                c = try c.?.nextSibling();
+                continue;
+            }
+
+            count += 1;
+            if (count > 1) return false;
+
+            c = try c.?.nextSibling();
+        }
+
+        return count == 1;
+    }
+
+    // simpleNthLastChildMatch implements :nth-last-child(b).
+    // If ofType is true, implements :nth-last-of-type instead.
+    fn simpleNthLastChildMatch(b: isize, of_type: bool, n: anytype) anyerror!bool {
+        if (!n.isElement()) return false;
+
+        const p = try n.parent();
+        if (p == null) return false;
+
+        const ntag = try n.tag();
+
+        var count: isize = 0;
+        var c = try p.?.lastChild();
+        // loop hover all n siblings.
+        while (c != null) {
+            // ignore non elements or others tags if of-type is true.
+            if (!c.?.isElement() or (of_type and !std.mem.eql(u8, ntag, try c.?.tag()))) {
+                c = try c.?.prevSibling();
+                continue;
+            }
+
+            count += 1;
+
+            if (n.eql(c.?)) return count == b;
+            if (count >= b) return false;
+
+            c = try c.?.prevSibling();
+        }
+
+        return false;
+    }
+
+    // simpleNthChildMatch implements :nth-child(b).
+    // If ofType is true, implements :nth-of-type instead.
+    fn simpleNthChildMatch(b: isize, of_type: bool, n: anytype) anyerror!bool {
+        if (!n.isElement()) return false;
+
+        const p = try n.parent();
+        if (p == null) return false;
+
+        const ntag = try n.tag();
+
+        var count: isize = 0;
+        var c = try p.?.firstChild();
+        // loop hover all n siblings.
+        while (c != null) {
+            // ignore non elements or others tags if of-type is true.
+            if (!c.?.isElement() or (of_type and !std.mem.eql(u8, ntag, try c.?.tag()))) {
+                c = try c.?.nextSibling();
+                continue;
+            }
+
+            count += 1;
+
+            if (n.eql(c.?)) return count == b;
+            if (count >= b) return false;
+
+            c = try c.?.nextSibling();
+        }
+
+        return false;
+    }
+
+    // nthChildMatch implements :nth-child(an+b).
+    // If last is true, implements :nth-last-child instead.
+    // If ofType is true, implements :nth-of-type instead.
+    fn nthChildMatch(a: isize, b: isize, last: bool, of_type: bool, n: anytype) anyerror!bool {
+        if (!n.isElement()) return false;
+
+        const p = try n.parent();
+        if (p == null) return false;
+
+        const ntag = try n.tag();
+
+        var i: isize = -1;
+        var count: isize = 0;
+        var c = try p.?.firstChild();
+        // loop hover all n siblings.
+        while (c != null) {
+            // ignore non elements or others tags if of-type is true.
+            if (!c.?.isElement() or (of_type and !std.mem.eql(u8, ntag, try c.?.tag()))) {
+                c = try c.?.nextSibling();
+                continue;
+            }
+            count += 1;
+
+            if (n.eql(c.?)) {
+                i = count;
+                if (!last) break;
+            }
+
+            c = try c.?.nextSibling();
+        }
+
+        if (i == -1) return false;
+
+        if (last) i = count - i + 1;
+
+        i -= b;
+        if (a == 0) return i == 0;
+        return @mod(i, a) == 0 and @divTrunc(i, a) >= 0;
+    }
+
+    fn hasDescendantMatch(s: *const Selector, n: anytype) anyerror!bool {
+        var c = try n.firstChild();
+        while (c != null) {
+            if (try s.match(c.?)) return true;
+            if (c.?.isElement() and try hasDescendantMatch(s, c.?)) return true;
+            c = try c.?.nextSibling();
+        }
+
+        return false;
+    }
+
+    fn hasChildMatch(s: *const Selector, n: anytype) anyerror!bool {
+        var c = try n.firstChild();
+        while (c != null) {
+            if (try s.match(c.?)) return true;
+            c = try c.?.nextSibling();
+        }
+
+        return false;
+    }
+
+    pub fn deinit(sel: Selector, alloc: std.mem.Allocator) void {
+        switch (sel) {
+            .group => |v| {
+                for (v) |vv| vv.deinit(alloc);
+                alloc.free(v);
+            },
+            .compound => |v| {
+                for (v.selectors) |vv| vv.deinit(alloc);
+                alloc.free(v.selectors);
+            },
+            .tag, .id, .class, .pseudo_class_lang => |v| alloc.free(v),
+            .attribute => |att| {
+                alloc.free(att.key);
+                if (att.val) |v| alloc.free(v);
+                if (att.regexp) |v| alloc.free(v);
+            },
+            .combined => |c| {
+                c.first.deinit(alloc);
+                alloc.destroy(c.first);
+                c.second.deinit(alloc);
+                alloc.destroy(c.second);
+            },
+            .pseudo_class_relative => |v| {
+                v.match.deinit(alloc);
+                alloc.destroy(v.match);
+            },
+            .pseudo_class_contains => |v| alloc.free(v.val),
+            .pseudo_class_regexp => |v| alloc.free(v.regexp),
+            .pseudo_class, .pseudo_element, .never_match => {},
+            .pseudo_class_nth, .pseudo_class_only_child => {},
+        }
+    }
+};

src/dom/attribute.zig

@@ -1,10 +1,28 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const Node = @import("node.zig").Node;
 const DOMException = @import("exceptions.zig").DOMException;

src/dom/cdata_section.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const Text = @import("text.zig").Text;
 

src/dom/character_data.zig

@@ -1,11 +1,28 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
-const generate = @import("../generate.zig");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const Node = @import("node.zig").Node;
 const Comment = @import("comment.zig").Comment;
@@ -14,12 +31,12 @@ const ProcessingInstruction = @import("processing_instruction.zig").ProcessingIn
 const HTMLElem = @import("../html/elements.zig");
 
 // CharacterData interfaces
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     Comment,
     Text.Text,
     Text.Interfaces,
     ProcessingInstruction,
-});
+};
 
 // CharacterData implementation
 pub const CharacterData = struct {

src/dom/comment.zig

@@ -1,9 +1,59 @@
-const parser = @import("../netsurf.zig");
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+const std = @import("std");
+
+const parser = @import("netsurf");
+
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
 
 const CharacterData = @import("character_data.zig").CharacterData;
 
+const UserContext = @import("../user_context.zig").UserContext;
+
+// https://dom.spec.whatwg.org/#interface-comment
 pub const Comment = struct {
     pub const Self = parser.Comment;
     pub const prototype = *CharacterData;
     pub const mem_guarantied = true;
+
+    pub fn constructor(userctx: UserContext, data: ?[]const u8) !*parser.Comment {
+        return parser.documentCreateComment(
+            parser.documentHTMLToDocument(userctx.document),
+            data orelse "",
+        );
+    }
 };
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var constructor = [_]Case{
+        .{ .src = "let comment = new Comment('foo')", .ex = "undefined" },
+        .{ .src = "comment.data", .ex = "foo" },
+
+        .{ .src = "let emptycomment = new Comment()", .ex = "undefined" },
+        .{ .src = "emptycomment.data", .ex = "" },
+    };
+    try checkCases(js_env, &constructor);
+}

src/dom/css.zig

@@ -0,0 +1,79 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const parser = @import("netsurf");
+
+const css = @import("../css/css.zig");
+const Node = @import("../css/libdom.zig").Node;
+const NodeList = @import("nodelist.zig").NodeList;
+
+const MatchFirst = struct {
+    n: ?*parser.Node = null,
+
+    pub fn match(m: *MatchFirst, n: Node) !void {
+        m.n = n.node;
+    }
+};
+
+pub fn querySelector(alloc: std.mem.Allocator, n: *parser.Node, selector: []const u8) !?*parser.Node {
+    const ps = try css.parse(alloc, selector, .{ .accept_pseudo_elts = true });
+    defer ps.deinit(alloc);
+
+    var m = MatchFirst{};
+
+    _ = try css.matchFirst(ps, Node{ .node = n }, &m);
+    return m.n;
+}
+
+const MatchAll = struct {
+    alloc: std.mem.Allocator,
+    nl: NodeList,
+
+    fn init(alloc: std.mem.Allocator) MatchAll {
+        return .{
+            .alloc = alloc,
+            .nl = NodeList.init(),
+        };
+    }
+
+    fn deinit(m: *MatchAll) void {
+        m.nl.deinit(m.alloc);
+    }
+
+    pub fn match(m: *MatchAll, n: Node) !void {
+        try m.nl.append(m.alloc, n.node);
+    }
+
+    fn toOwnedList(m: *MatchAll) NodeList {
+        defer m.nl = NodeList.init();
+        return m.nl;
+    }
+};
+
+pub fn querySelectorAll(alloc: std.mem.Allocator, n: *parser.Node, selector: []const u8) !NodeList {
+    const ps = try css.parse(alloc, selector, .{ .accept_pseudo_elts = true });
+    defer ps.deinit(alloc);
+
+    var m = MatchAll.init(alloc);
+    defer m.deinit();
+
+    try css.matchAll(ps, Node{ .node = n }, &m);
+    return m.toOwnedList();
+}

src/dom/document.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
@@ -11,8 +29,8 @@ const Node = @import("node.zig").Node;
 const NodeList = @import("nodelist.zig").NodeList;
 const NodeUnion = @import("node.zig").Union;
 
-const Walker = @import("walker.zig").WalkerDepthFirst;
 const collection = @import("html_collection.zig");
+const css = @import("css.zig");
 
 const Element = @import("element.zig").Element;
 const ElementUnion = @import("element.zig").Union;
@@ -21,14 +39,26 @@ const DocumentType = @import("document_type.zig").DocumentType;
 const DocumentFragment = @import("document_fragment.zig").DocumentFragment;
 const DOMImplementation = @import("implementation.zig").DOMImplementation;
 
+const UserContext = @import("../user_context.zig").UserContext;
+
 // WEB IDL https://dom.spec.whatwg.org/#document
 pub const Document = struct {
     pub const Self = parser.Document;
     pub const prototype = *Node;
     pub const mem_guarantied = true;
 
-    pub fn constructor() !*parser.Document {
-        return try parser.domImplementationCreateHTMLDocument(null);
+    pub fn constructor(userctx: UserContext) !*parser.DocumentHTML {
+        const doc = try parser.documentCreateDocument(
+            try parser.documentHTMLGetTitle(userctx.document),
+        );
+
+        // we have to work w/ document instead of html document.
+        const ddoc = parser.documentHTMLToDocument(doc);
+        const ccur = parser.documentHTMLToDocument(userctx.document);
+        try parser.documentSetDocumentURI(ddoc, try parser.documentGetDocumentURI(ccur));
+        try parser.documentSetInputEncoding(ddoc, try parser.documentGetInputEncoding(ccur));
+
+        return doc;
     }
 
     // JS funcs
@@ -188,54 +218,18 @@ pub const Document = struct {
         return 1;
     }
 
-    // TODO netsurf doesn't handle query selectors. We have to implement a
-    // solution by ourselves.
-    // For now we handle only * and single id selector like `#foo`.
-    pub fn _querySelector(self: *parser.Document, selectors: []const u8) !?ElementUnion {
-        if (selectors.len == 0) return null;
+    pub fn _querySelector(self: *parser.Document, alloc: std.mem.Allocator, selector: []const u8) !?ElementUnion {
+        if (selector.len == 0) return null;
 
-        // catch-all, return the firstElementChild
-        if (selectors[0] == '*') return try get_firstElementChild(self);
+        const n = try css.querySelector(alloc, parser.documentToNode(self), selector);
 
-        // support only simple id selector.
-        if (selectors[0] != '#' or std.mem.indexOf(u8, selectors, " ") != null) return null;
+        if (n == null) return null;
 
-        return try _getElementById(self, selectors[1..]);
+        return try Element.toInterface(parser.nodeToElement(n.?));
     }
 
-    // TODO netsurf doesn't handle query selectors. We have to implement a
-    // solution by ourselves.
-    // We handle only * and single id selector like `#foo`.
-    pub fn _querySelectorAll(self: *parser.Document, alloc: std.mem.Allocator, selectors: []const u8) !NodeList {
-        var list = try NodeList.init();
-        errdefer list.deinit(alloc);
-
-        if (selectors.len == 0) return list;
-
-        // catch-all, return all elements
-        if (selectors[0] == '*') {
-            // walk over the node tree fo find the node by id.
-            const root = parser.documentToNode(self);
-            const walker = Walker{};
-            var next: ?*parser.Node = null;
-            while (true) {
-                next = try walker.get_next(root, next) orelse return list;
-                // ignore non-element nodes.
-                if (try parser.nodeType(next.?) != .element) {
-                    continue;
-                }
-                try list.append(alloc, next.?);
-            }
-        }
-
-        // support only simple id selector.
-        if (selectors[0] != '#' or std.mem.indexOf(u8, selectors, " ") != null) return list;
-
-        // walk over the node tree fo find the node by id.
-        const e = try parser.documentGetElementById(self, selectors[1..]) orelse return list;
-        try list.append(alloc, parser.elementToNode(e));
-
-        return list;
+    pub fn _querySelectorAll(self: *parser.Document, alloc: std.mem.Allocator, selector: []const u8) !NodeList {
+        return css.querySelectorAll(alloc, parser.documentToNode(self), selector);
     }
 
     // TODO according with https://dom.spec.whatwg.org/#parentnode, the
@@ -279,6 +273,13 @@ pub fn testExecFn(
         .{ .src = "newdoc.children.length", .ex = "0" },
         .{ .src = "newdoc.getElementsByTagName('*').length", .ex = "0" },
         .{ .src = "newdoc.getElementsByTagName('*').item(0)", .ex = "null" },
+        .{ .src = "newdoc.inputEncoding === document.inputEncoding", .ex = "true" },
+        .{ .src = "newdoc.documentURI === document.documentURI", .ex = "true" },
+        .{ .src = "newdoc.URL === document.URL", .ex = "true" },
+        .{ .src = "newdoc.compatMode === document.compatMode", .ex = "true" },
+        .{ .src = "newdoc.characterSet === document.characterSet", .ex = "true" },
+        .{ .src = "newdoc.charset === document.charset", .ex = "true" },
+        .{ .src = "newdoc.contentType === document.contentType", .ex = "true" },
     };
     try checkCases(js_env, &constructor);
 
@@ -377,12 +378,14 @@ pub fn testExecFn(
     var createComment = [_]Case{
         .{ .src = "var v = document.createComment('foo')", .ex = "undefined" },
         .{ .src = "v.nodeName", .ex = "#comment" },
+        .{ .src = "let v2 = v.cloneNode()", .ex = "undefined" },
     };
     try checkCases(js_env, &createComment);
 
     var createProcessingInstruction = [_]Case{
         .{ .src = "let pi = document.createProcessingInstruction('foo', 'bar')", .ex = "undefined" },
         .{ .src = "pi.target", .ex = "foo" },
+        .{ .src = "let pi2 = pi.cloneNode()", .ex = "undefined" },
     };
     try checkCases(js_env, &createProcessingInstruction);
 
@@ -426,6 +429,16 @@ pub fn testExecFn(
         .{ .src = "document.querySelector('*').nodeName", .ex = "HTML" },
         .{ .src = "document.querySelector('#content').id", .ex = "content" },
         .{ .src = "document.querySelector('#para').id", .ex = "para" },
+        .{ .src = "document.querySelector('.ok').id", .ex = "link" },
+        .{ .src = "document.querySelector('a ~ p').id", .ex = "para-empty" },
+        .{ .src = "document.querySelector(':root').nodeName", .ex = "HTML" },
+
+        .{ .src = "document.querySelectorAll('p').length", .ex = "2" },
+        .{ .src = 
+        \\Array.from(document.querySelectorAll('#content > p#para-empty'))
+        \\.map(row => row.querySelector('span').textContent)
+        \\.length;
+        , .ex = "1" },
     };
     try checkCases(js_env, &querySelector);
 
@@ -439,7 +452,7 @@ pub fn testExecFn(
     try checkCases(js_env, &adoptNode);
 
     const tags = comptime parser.Tag.all();
-    comptime var createElements: [(tags.len) * 2]Case = undefined;
+    var createElements: [(tags.len) * 2]Case = undefined;
     inline for (tags, 0..) |tag, i| {
         const tag_name = @tagName(tag);
         createElements[i * 2] = Case{

src/dom/document_fragment.zig

@@ -1,21 +1,56 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
+
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
 
 const Node = @import("node.zig").Node;
 
+const UserContext = @import("../user_context.zig").UserContext;
+
 // WEB IDL https://dom.spec.whatwg.org/#documentfragment
 pub const DocumentFragment = struct {
     pub const Self = parser.DocumentFragment;
     pub const prototype = *Node;
     pub const mem_guarantied = true;
 
-    // TODO add constructor, but I need to associate the new DocumentFragment
-    // with the current document global object...
-    // > The new DocumentFragment() constructor steps are to set this’s node
-    // > document to current global object’s associated Document.
-    // https://dom.spec.whatwg.org/#dom-documentfragment-documentfragment
-    pub fn constructor() !*parser.DocumentFragment {
-        return error.NotImplemented;
+    pub fn constructor(userctx: UserContext) !*parser.DocumentFragment {
+        return parser.documentCreateDocumentFragment(
+            parser.documentHTMLToDocument(userctx.document),
+        );
     }
 };
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var constructor = [_]Case{
+        .{ .src = "const dc = new DocumentFragment()", .ex = "undefined" },
+        .{ .src = "dc.constructor.name", .ex = "DocumentFragment" },
+    };
+    try checkCases(js_env, &constructor);
+}

src/dom/document_type.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const Node = @import("node.zig").Node;
 

src/dom/dom.zig

@@ -1,20 +1,38 @@
-const generate = @import("../generate.zig");
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
 
 const DOMException = @import("exceptions.zig").DOMException;
 const EventTarget = @import("event_target.zig").EventTarget;
 const DOMImplementation = @import("implementation.zig").DOMImplementation;
 const NamedNodeMap = @import("namednodemap.zig").NamedNodeMap;
 const DOMTokenList = @import("token_list.zig").DOMTokenList;
-const NodeList = @import("nodelist.zig").NodeList;
+const NodeList = @import("nodelist.zig");
 const Nod = @import("node.zig");
+const MutationObserver = @import("mutation_observer.zig");
 
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     DOMException,
     EventTarget,
     DOMImplementation,
     NamedNodeMap,
     DOMTokenList,
-    NodeList,
+    NodeList.Interfaces,
     Nod.Node,
     Nod.Interfaces,
-});
+    MutationObserver.Interfaces,
+};

src/dom/element.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
@@ -8,6 +26,8 @@ const checkCases = jsruntime.test_utils.checkCases;
 const Variadic = jsruntime.Variadic;
 
 const collection = @import("html_collection.zig");
+const dump = @import("../browser/dump.zig");
+const css = @import("css.zig");
 
 const Node = @import("node.zig").Node;
 const Walker = @import("walker.zig").WalkerDepthFirst;
@@ -78,6 +98,48 @@ pub const Element = struct {
         return try parser.nodeGetAttributes(parser.elementToNode(self));
     }
 
+    pub fn get_innerHTML(self: *parser.Element, alloc: std.mem.Allocator) ![]const u8 {
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try dump.writeChildren(parser.elementToNode(self), buf.writer());
+        // TODO express the caller owned the slice.
+        // https://github.com/lightpanda-io/jsruntime-lib/issues/195
+        return buf.toOwnedSlice();
+    }
+
+    pub fn get_outerHTML(self: *parser.Element, alloc: std.mem.Allocator) ![]const u8 {
+        var buf = std.ArrayList(u8).init(alloc);
+        defer buf.deinit();
+
+        try dump.writeNode(parser.elementToNode(self), buf.writer());
+        // TODO express the caller owned the slice.
+        // https://github.com/lightpanda-io/jsruntime-lib/issues/195
+        return buf.toOwnedSlice();
+    }
+
+    pub fn set_innerHTML(self: *parser.Element, str: []const u8) !void {
+        const node = parser.elementToNode(self);
+        const doc = try parser.nodeOwnerDocument(node) orelse return parser.DOMError.WrongDocument;
+        // parse the fragment
+        const fragment = try parser.documentParseFragmentFromStr(doc, str);
+
+        // remove existing children
+        try Node.removeChildren(node);
+
+        // get fragment body children
+        const children = try parser.documentFragmentBodyChildren(fragment) orelse return;
+
+        // append children to the node
+        const ln = try parser.nodeListLength(children);
+        var i: u32 = 0;
+        while (i < ln) {
+            defer i += 1;
+            const child = try parser.nodeListItem(children, i) orelse continue;
+            _ = try parser.nodeAppendChild(node, child);
+        }
+    }
+
     pub fn _hasAttributes(self: *parser.Element) !bool {
         return try parser.nodeHasAttributes(parser.elementToNode(self));
     }
@@ -86,14 +148,26 @@ pub const Element = struct {
         return try parser.elementGetAttribute(self, qname);
     }
 
+    pub fn _getAttributeNS(self: *parser.Element, ns: []const u8, qname: []const u8) !?[]const u8 {
+        return try parser.elementGetAttributeNS(self, ns, qname);
+    }
+
     pub fn _setAttribute(self: *parser.Element, qname: []const u8, value: []const u8) !void {
         return try parser.elementSetAttribute(self, qname, value);
     }
 
+    pub fn _setAttributeNS(self: *parser.Element, ns: []const u8, qname: []const u8, value: []const u8) !void {
+        return try parser.elementSetAttributeNS(self, ns, qname, value);
+    }
+
     pub fn _removeAttribute(self: *parser.Element, qname: []const u8) !void {
         return try parser.elementRemoveAttribute(self, qname);
     }
 
+    pub fn _removeAttributeNS(self: *parser.Element, ns: []const u8, qname: []const u8) !void {
+        return try parser.elementRemoveAttributeNS(self, ns, qname);
+    }
+
     pub fn _hasAttribute(self: *parser.Element, qname: []const u8) !bool {
         return try parser.elementHasAttribute(self, qname);
     }
@@ -230,56 +304,18 @@ pub const Element = struct {
         }
     }
 
-    // TODO netsurf doesn't handle query selectors. We have to implement a
-    // solution by ourselves.
-    // We handle only * and single id selector like `#foo`.
-    pub fn _querySelector(self: *parser.Element, selectors: []const u8) !?Union {
-        if (selectors.len == 0) return null;
+    pub fn _querySelector(self: *parser.Element, alloc: std.mem.Allocator, selector: []const u8) !?Union {
+        if (selector.len == 0) return null;
 
-        // catch-all, return the firstElementChild
-        if (selectors[0] == '*') return try get_firstElementChild(self);
+        const n = try css.querySelector(alloc, parser.elementToNode(self), selector);
 
-        // support only simple id selector.
-        if (selectors[0] != '#' or std.mem.indexOf(u8, selectors, " ") != null) return null;
+        if (n == null) return null;
 
-        // walk over the node tree fo find the node by id.
-        const n = try getElementById(self, selectors[1..]) orelse return null;
-        return try toInterface(parser.nodeToElement(n));
+        return try toInterface(parser.nodeToElement(n.?));
     }
 
-    // TODO netsurf doesn't handle query selectors. We have to implement a
-    // solution by ourselves.
-    // We handle only * and single id selector like `#foo`.
-    pub fn _querySelectorAll(self: *parser.Element, alloc: std.mem.Allocator, selectors: []const u8) !NodeList {
-        var list = try NodeList.init();
-        errdefer list.deinit(alloc);
-
-        if (selectors.len == 0) return list;
-
-        // catch-all, return all elements
-        if (selectors[0] == '*') {
-            // walk over the node tree fo find the node by id.
-            const root = parser.elementToNode(self);
-            const walker = Walker{};
-            var next: ?*parser.Node = null;
-            while (true) {
-                next = try walker.get_next(root, next) orelse return list;
-                // ignore non-element nodes.
-                if (try parser.nodeType(next.?) != .element) {
-                    continue;
-                }
-                try list.append(alloc, next.?);
-            }
-        }
-
-        // support only simple id selector.
-        if (selectors[0] != '#' or std.mem.indexOf(u8, selectors, " ") != null) return list;
-
-        // walk over the node tree fo find the node by id.
-        const n = try getElementById(self, selectors[1..]) orelse return list;
-        try list.append(alloc, n);
-
-        return list;
+    pub fn _querySelectorAll(self: *parser.Element, alloc: std.mem.Allocator, selector: []const u8) !NodeList {
+        return css.querySelectorAll(alloc, parser.elementToNode(self), selector);
     }
 
     // TODO according with https://dom.spec.whatwg.org/#parentnode, the
@@ -400,6 +436,12 @@ pub fn testExecFn(
         .{ .src = "e.querySelector('#link').id", .ex = "link" },
         .{ .src = "e.querySelector('#para').id", .ex = "para" },
         .{ .src = "e.querySelector('*').id", .ex = "link" },
+        .{ .src = "e.querySelector('')", .ex = "null" },
+        .{ .src = "e.querySelector('*').id", .ex = "link" },
+        .{ .src = "e.querySelector('#content')", .ex = "null" },
+        .{ .src = "e.querySelector('#para').id", .ex = "para" },
+        .{ .src = "e.querySelector('.ok').id", .ex = "link" },
+        .{ .src = "e.querySelector('a ~ p').id", .ex = "para-empty" },
 
         .{ .src = "e.querySelectorAll('foo').length", .ex = "0" },
         .{ .src = "e.querySelectorAll('#foo').length", .ex = "0" },
@@ -408,6 +450,8 @@ pub fn testExecFn(
         .{ .src = "e.querySelectorAll('#para').length", .ex = "1" },
         .{ .src = "e.querySelectorAll('#para').item(0).id", .ex = "para" },
         .{ .src = "e.querySelectorAll('*').length", .ex = "4" },
+        .{ .src = "e.querySelectorAll('p').length", .ex = "2" },
+        .{ .src = "e.querySelectorAll('.ok').item(0).id", .ex = "link" },
     };
     try checkCases(js_env, &querySelector);
 
@@ -420,4 +464,25 @@ pub fn testExecFn(
         .{ .src = "f.getAttributeNode('bar')", .ex = "null" },
     };
     try checkCases(js_env, &attrNode);
+
+    var innerHTML = [_]Case{
+        .{ .src = "document.getElementById('para').innerHTML", .ex = " And" },
+        .{ .src = "document.getElementById('para-empty').innerHTML.trim()", .ex = "<span id=\"para-empty-child\"></span>" },
+
+        .{ .src = "let h = document.getElementById('para-empty')", .ex = "undefined" },
+        .{ .src = "const prev = h.innerHTML", .ex = "undefined" },
+        .{ .src = "h.innerHTML = '<p id=\"hello\">hello world</p>'", .ex = "<p id=\"hello\">hello world</p>" },
+        .{ .src = "h.innerHTML", .ex = "<p id=\"hello\">hello world</p>" },
+        .{ .src = "h.firstChild.nodeName", .ex = "P" },
+        .{ .src = "h.firstChild.id", .ex = "hello" },
+        .{ .src = "h.firstChild.textContent", .ex = "hello world" },
+        .{ .src = "h.innerHTML = prev; true", .ex = "true" },
+        .{ .src = "document.getElementById('para-empty').innerHTML.trim()", .ex = "<span id=\"para-empty-child\"></span>" },
+    };
+    try checkCases(js_env, &innerHTML);
+
+    var outerHTML = [_]Case{
+        .{ .src = "document.getElementById('para').outerHTML", .ex = "<p id=\"para\"> And</p>" },
+    };
+    try checkCases(js_env, &outerHTML);
 }

src/dom/event_target.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
@@ -6,7 +24,8 @@ const JSObjectID = jsruntime.JSObjectID;
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
+const EventHandler = @import("../events/event.zig").EventHandler;
 
 const DOMException = @import("exceptions.zig").DOMException;
 const Nod = @import("node.zig");
@@ -54,7 +73,8 @@ pub const EventTarget = struct {
             self,
             alloc,
             eventType,
-            cbk,
+            EventHandler,
+            .{ .cbk = cbk },
             capture orelse false,
         );
     }

src/dom/exceptions.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 const allocPrint = std.fmt.allocPrint;
 
@@ -5,7 +23,7 @@ const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 // https://webidl.spec.whatwg.org/#idl-DOMException
 pub const DOMException = struct {
@@ -91,6 +109,12 @@ pub const DOMException = struct {
             error.InvalidNodeType => "InvalidNodeTypeError",
             error.DataClone => "DataCloneError",
             error.NoError => unreachable,
+
+            // custom netsurf error
+            error.UnspecifiedEventType => "UnspecifiedEventTypeError",
+            error.DispatchRequest => "DispatchRequestError",
+            error.NoMemory => "NoMemoryError",
+            error.AttributeWrongType => "AttributeWrongTypeError",
         };
     }
 
@@ -124,6 +148,12 @@ pub const DOMException = struct {
             error.InvalidNodeType => 24,
             error.DataClone => 25,
             error.NoError => unreachable,
+
+            // custom netsurf error
+            error.UnspecifiedEventType => 128,
+            error.DispatchRequest => 129,
+            error.NoMemory => 130,
+            error.AttributeWrongType => 131,
         };
     }
 

src/dom/html_collection.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;

src/dom/implementation.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
@@ -57,7 +75,7 @@ pub const DOMImplementation = struct {
         return try parser.domImplementationCreateDocument(cnamespace, cqname, doctype);
     }
 
-    pub fn _createHTMLDocument(_: *DOMImplementation, title: ?[]const u8) !*parser.Document {
+    pub fn _createHTMLDocument(_: *DOMImplementation, title: ?[]const u8) !*parser.DocumentHTML {
         return try parser.domImplementationCreateHTMLDocument(title);
     }
 
@@ -77,7 +95,8 @@ pub fn testExecFn(
 ) anyerror!void {
     var getImplementation = [_]Case{
         .{ .src = "let impl = document.implementation", .ex = "undefined" },
-        .{ .src = "impl.createHTMLDocument();", .ex = "[object Document]" },
+        .{ .src = "impl.createHTMLDocument();", .ex = "[object HTMLDocument]" },
+        .{ .src = "impl.createHTMLDocument('foo');", .ex = "[object HTMLDocument]" },
         .{ .src = "impl.createDocument(null, 'foo');", .ex = "[object Document]" },
         .{ .src = "impl.createDocumentType('foo', 'bar', 'baz')", .ex = "[object DocumentType]" },
         .{ .src = "impl.hasFeature()", .ex = "true" },

src/dom/mutation_observer.zig

@@ -0,0 +1,405 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const parser = @import("netsurf");
+
+const jsruntime = @import("jsruntime");
+const Callback = jsruntime.Callback;
+const CallbackResult = jsruntime.CallbackResult;
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+const NodeList = @import("nodelist.zig").NodeList;
+
+pub const Interfaces = .{
+    MutationObserver,
+    MutationRecord,
+    MutationRecords,
+};
+
+const Walker = @import("../dom/walker.zig").WalkerChildren;
+
+const log = std.log.scoped(.events);
+
+// WEB IDL https://dom.spec.whatwg.org/#interface-mutationobserver
+pub const MutationObserver = struct {
+    cbk: Callback,
+    observers: Observers,
+
+    pub const mem_guarantied = true;
+
+    const Observer = struct {
+        node: *parser.Node,
+        options: MutationObserverInit,
+    };
+
+    const deinitFunc = struct {
+        fn deinit(ctx: ?*anyopaque, alloc: std.mem.Allocator) void {
+            const o: *Observer = @ptrCast(@alignCast(ctx));
+            alloc.destroy(o);
+        }
+    }.deinit;
+
+    const Observers = std.ArrayListUnmanaged(*Observer);
+
+    pub const MutationObserverInit = struct {
+        childList: bool = false,
+        attributes: bool = false,
+        characterData: bool = false,
+        subtree: bool = false,
+        attributeOldValue: bool = false,
+        characterDataOldValue: bool = false,
+        // TODO
+        // attributeFilter: [][]const u8,
+
+        fn attr(self: MutationObserverInit) bool {
+            return self.attributes or self.attributeOldValue;
+        }
+
+        fn cdata(self: MutationObserverInit) bool {
+            return self.characterData or self.characterDataOldValue;
+        }
+    };
+
+    pub fn constructor(cbk: Callback) !MutationObserver {
+        return MutationObserver{
+            .cbk = cbk,
+            .observers = .{},
+        };
+    }
+
+    // TODO
+    fn resolveOptions(opt: ?MutationObserverInit) MutationObserverInit {
+        return opt orelse .{};
+    }
+
+    pub fn _observe(self: *MutationObserver, alloc: std.mem.Allocator, node: *parser.Node, options: ?MutationObserverInit) !void {
+        const o = try alloc.create(Observer);
+        o.* = .{
+            .node = node,
+            .options = resolveOptions(options),
+        };
+        errdefer alloc.destroy(o);
+
+        // register the new observer.
+        try self.observers.append(alloc, o);
+
+        // register node's events.
+        if (o.options.childList or o.options.subtree) {
+            try parser.eventTargetAddEventListener(
+                parser.toEventTarget(parser.Node, node),
+                alloc,
+                "DOMNodeInserted",
+                EventHandler,
+                .{ .cbk = self.cbk, .data = o, .deinitFunc = deinitFunc },
+                false,
+            );
+            try parser.eventTargetAddEventListener(
+                parser.toEventTarget(parser.Node, node),
+                alloc,
+                "DOMNodeRemoved",
+                EventHandler,
+                .{ .cbk = self.cbk, .data = o, .deinitFunc = deinitFunc },
+                false,
+            );
+        }
+        if (o.options.attr()) {
+            try parser.eventTargetAddEventListener(
+                parser.toEventTarget(parser.Node, node),
+                alloc,
+                "DOMAttrModified",
+                EventHandler,
+                .{ .cbk = self.cbk, .data = o, .deinitFunc = deinitFunc },
+                false,
+            );
+        }
+        if (o.options.cdata()) {
+            try parser.eventTargetAddEventListener(
+                parser.toEventTarget(parser.Node, node),
+                alloc,
+                "DOMCharacterDataModified",
+                EventHandler,
+                .{ .cbk = self.cbk, .data = o, .deinitFunc = deinitFunc },
+                false,
+            );
+        }
+        if (o.options.subtree) {
+            try parser.eventTargetAddEventListener(
+                parser.toEventTarget(parser.Node, node),
+                alloc,
+                "DOMSubtreeModified",
+                EventHandler,
+                .{ .cbk = self.cbk, .data = o, .deinitFunc = deinitFunc },
+                false,
+            );
+        }
+    }
+
+    // TODO
+    pub fn _disconnect(_: *MutationObserver) !void {
+        // TODO unregister listeners.
+    }
+
+    pub fn deinit(self: *MutationObserver, alloc: std.mem.Allocator) void {
+        // TODO unregister listeners.
+        for (self.observers.items) |o| alloc.destroy(o);
+        self.observers.deinit(alloc);
+    }
+
+    // TODO
+    pub fn _takeRecords(_: MutationObserver) ?[]const u8 {
+        return &[_]u8{};
+    }
+};
+
+// Handle multiple record?
+pub const MutationRecords = struct {
+    first: ?MutationRecord = null,
+
+    pub const mem_guarantied = true;
+
+    pub fn get_length(self: *MutationRecords) u32 {
+        if (self.first == null) return 0;
+
+        return 1;
+    }
+
+    pub fn postAttach(self: *MutationRecords, js_obj: jsruntime.JSObject) !void {
+        if (self.first) |mr| {
+            try js_obj.set("0", mr);
+        }
+    }
+};
+
+pub const MutationRecord = struct {
+    type: []const u8,
+    target: *parser.Node,
+    addedNodes: NodeList = NodeList.init(),
+    removedNodes: NodeList = NodeList.init(),
+    previousSibling: ?*parser.Node = null,
+    nextSibling: ?*parser.Node = null,
+    attributeName: ?[]const u8 = null,
+    attributeNamespace: ?[]const u8 = null,
+    oldValue: ?[]const u8 = null,
+
+    pub const mem_guarantied = true;
+
+    pub fn get_type(self: MutationRecord) []const u8 {
+        return self.type;
+    }
+
+    pub fn get_addedNodes(self: MutationRecord) NodeList {
+        return self.addedNodes;
+    }
+
+    pub fn get_removedNodes(self: MutationRecord) NodeList {
+        return self.addedNodes;
+    }
+
+    pub fn get_target(self: MutationRecord) *parser.Node {
+        return self.target;
+    }
+
+    pub fn get_attributeName(self: MutationRecord) ?[]const u8 {
+        return self.attributeName;
+    }
+
+    pub fn get_attributeNamespace(self: MutationRecord) ?[]const u8 {
+        return self.attributeNamespace;
+    }
+
+    pub fn get_previousSibling(self: MutationRecord) ?*parser.Node {
+        return self.previousSibling;
+    }
+
+    pub fn get_nextSibling(self: MutationRecord) ?*parser.Node {
+        return self.nextSibling;
+    }
+
+    pub fn get_oldValue(self: MutationRecord) ?[]const u8 {
+        return self.oldValue;
+    }
+};
+
+// EventHandler dedicated to mutation events.
+const EventHandler = struct {
+    fn apply(o: *MutationObserver.Observer, target: *parser.Node) bool {
+        // mutation on any target is always ok.
+        if (o.options.subtree) return true;
+        // if target equals node, alway ok.
+        if (target == o.node) return true;
+
+        // no subtree, no same target and no childlist, always noky.
+        if (!o.options.childList) return false;
+
+        // target must be a child of o.node
+        const walker = Walker{};
+        var next: ?*parser.Node = null;
+        while (true) {
+            next = walker.get_next(o.node, next) catch break orelse break;
+            if (next.? == target) return true;
+        }
+
+        return false;
+    }
+
+    fn handle(evt: ?*parser.Event, data: parser.EventHandlerData) void {
+        if (evt == null) return;
+
+        var mrs: MutationRecords = .{};
+
+        const t = parser.eventType(evt.?) catch |e| {
+            log.err("mutation observer event type: {any}", .{e});
+            return;
+        };
+        const et = parser.eventTarget(evt.?) catch |e| {
+            log.err("mutation observer event target: {any}", .{e});
+            return;
+        } orelse return;
+        const node = parser.eventTargetToNode(et);
+
+        // retrieve the observer from the data.
+        const o: *MutationObserver.Observer = @ptrCast(@alignCast(data.data));
+
+        if (!apply(o, node)) return;
+
+        const muevt = parser.eventToMutationEvent(evt.?);
+
+        // TODO get the allocator by another way?
+        const alloc = data.cbk.nat_ctx.alloc;
+
+        if (std.mem.eql(u8, t, "DOMAttrModified")) {
+            mrs.first = .{
+                .type = "attributes",
+                .target = o.node,
+                .attributeName = parser.mutationEventAttributeName(muevt) catch null,
+            };
+
+            // record old value if required.
+            if (o.options.attributeOldValue) {
+                mrs.first.?.oldValue = parser.mutationEventPrevValue(muevt) catch null;
+            }
+        } else if (std.mem.eql(u8, t, "DOMCharacterDataModified")) {
+            mrs.first = .{
+                .type = "characterData",
+                .target = o.node,
+            };
+
+            // record old value if required.
+            if (o.options.characterDataOldValue) {
+                mrs.first.?.oldValue = parser.mutationEventPrevValue(muevt) catch null;
+            }
+        } else if (std.mem.eql(u8, t, "DOMNodeInserted")) {
+            mrs.first = .{
+                .type = "childList",
+                .target = o.node,
+                .addedNodes = NodeList.init(),
+                .removedNodes = NodeList.init(),
+            };
+
+            const rn = parser.mutationEventRelatedNode(muevt) catch null;
+            if (rn) |n| {
+                mrs.first.?.addedNodes.append(alloc, n) catch |e| {
+                    log.err("mutation event handler error: {any}", .{e});
+                    return;
+                };
+            }
+        } else if (std.mem.eql(u8, t, "DOMNodeRemoved")) {
+            mrs.first = .{
+                .type = "childList",
+                .target = o.node,
+                .addedNodes = NodeList.init(),
+                .removedNodes = NodeList.init(),
+            };
+
+            const rn = parser.mutationEventRelatedNode(muevt) catch null;
+            if (rn) |n| {
+                mrs.first.?.removedNodes.append(alloc, n) catch |e| {
+                    log.err("mutation event handler error: {any}", .{e});
+                    return;
+                };
+            }
+        } else {
+            return;
+        }
+
+        var res = CallbackResult.init(alloc);
+        defer res.deinit();
+
+        // TODO pass MutationRecords and MutationObserver
+        data.cbk.trycall(.{mrs}, &res) catch |e| log.err("mutation event handler error: {any}", .{e});
+
+        // in case of function error, we log the result and the trace.
+        if (!res.success) {
+            log.info("mutation observer event handler error: {s}", .{res.result orelse "unknown"});
+            log.debug("{s}", .{res.stack orelse "no stack trace"});
+        }
+    }
+}.handle;
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var constructor = [_]Case{
+        .{ .src = "new MutationObserver(() => {}).observe(document, { childList: true });", .ex = "undefined" },
+    };
+    try checkCases(js_env, &constructor);
+
+    var attr = [_]Case{
+        .{ .src = 
+        \\var nb = 0;
+        \\var mrs;
+        \\new MutationObserver((mu) => {
+        \\    mrs = mu;
+        \\    nb++;
+        \\}).observe(document.firstElementChild, { attributes: true, attributeOldValue: true });
+        \\document.firstElementChild.setAttribute("foo", "bar");
+        \\// ignored b/c it's about another target.
+        \\document.firstElementChild.firstChild.setAttribute("foo", "bar");
+        \\nb;
+        , .ex = "1" },
+        .{ .src = "mrs[0].type", .ex = "attributes" },
+        .{ .src = "mrs[0].target == document.firstElementChild", .ex = "true" },
+        .{ .src = "mrs[0].target.getAttribute('foo')", .ex = "bar" },
+        .{ .src = "mrs[0].attributeName", .ex = "foo" },
+        .{ .src = "mrs[0].oldValue", .ex = "null" },
+    };
+    try checkCases(js_env, &attr);
+
+    var cdata = [_]Case{
+        .{ .src = 
+        \\var node = document.getElementById("para").firstChild;
+        \\var nb2 = 0;
+        \\var mrs2;
+        \\new MutationObserver((mu) => {
+        \\    mrs2 = mu;
+        \\    nb2++;
+        \\}).observe(node, { characterData: true, characterDataOldValue: true });
+        \\node.data = "foo";
+        \\nb2;
+        , .ex = "1" },
+        .{ .src = "mrs2[0].type", .ex = "characterData" },
+        .{ .src = "mrs2[0].target == node", .ex = "true" },
+        .{ .src = "mrs2[0].target.data", .ex = "foo" },
+        .{ .src = "mrs2[0].oldValue", .ex = " And" },
+    };
+    try checkCases(js_env, &cdata);
+}

src/dom/namednodemap.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;

src/dom/node.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
@@ -8,7 +26,7 @@ const Variadic = jsruntime.Variadic;
 
 const generate = @import("../generate.zig");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const EventTarget = @import("event_target.zig").EventTarget;
 
@@ -22,13 +40,14 @@ const DocumentType = @import("document_type.zig").DocumentType;
 const DocumentFragment = @import("document_fragment.zig").DocumentFragment;
 const HTMLCollection = @import("html_collection.zig").HTMLCollection;
 const HTMLCollectionIterator = @import("html_collection.zig").HTMLCollectionIterator;
+const Walker = @import("walker.zig").WalkerDepthFirst;
 
 // HTML
 const HTML = @import("../html/html.zig");
 const HTMLElem = @import("../html/elements.zig");
 
 // Node interfaces
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     Attr,
     CData.CharacterData,
     CData.Interfaces,
@@ -38,12 +57,10 @@ pub const Interfaces = generate.Tuple(.{
     DocumentFragment,
     HTMLCollection,
     HTMLCollectionIterator,
-
     HTML.Interfaces,
-});
-const Generated = generate.Union.compile(Interfaces);
-pub const Union = Generated._union;
-pub const Tags = Generated._enum;
+};
+
+pub const Union = generate.Union(Interfaces);
 
 // Node implementation
 pub const Node = struct {
@@ -177,21 +194,68 @@ pub const Node = struct {
         return try Node.toInterface(clone);
     }
 
-    pub fn _compareDocumentPosition(self: *parser.Node, other: *parser.Node) void {
-        // TODO
-        _ = other;
-        _ = self;
-        std.log.err("Not implemented {s}", .{"node.compareDocumentPosition()"});
+    pub fn _compareDocumentPosition(self: *parser.Node, other: *parser.Node) !u32 {
+        if (self == other) return 0;
+
+        const docself = try parser.nodeOwnerDocument(self);
+        const docother = try parser.nodeOwnerDocument(other);
+
+        // Both are in different document.
+        if (docself == null or docother == null or docother.? != docself.?) {
+            return @intFromEnum(parser.DocumentPosition.disconnected);
+        }
+
+        // TODO Both are in a different trees in the same document.
+
+        const w = Walker{};
+        var next: ?*parser.Node = null;
+
+        // Is other a descendant of self?
+        while (true) {
+            next = try w.get_next(self, next) orelse break;
+            if (other == next) {
+                return @intFromEnum(parser.DocumentPosition.following) +
+                    @intFromEnum(parser.DocumentPosition.contained_by);
+            }
+        }
+
+        // Is self a descendant of other?
+        next = null;
+        while (true) {
+            next = try w.get_next(other, next) orelse break;
+            if (self == next) {
+                return @intFromEnum(parser.DocumentPosition.contains) +
+                    @intFromEnum(parser.DocumentPosition.preceding);
+            }
+        }
+
+        next = null;
+        while (true) {
+            next = try w.get_next(parser.documentToNode(docself.?), next) orelse break;
+            if (other == next) {
+                // other precedes self.
+                return @intFromEnum(parser.DocumentPosition.preceding);
+            }
+            if (self == next) {
+                // other follows self.
+                return @intFromEnum(parser.DocumentPosition.following);
+            }
+        }
+
+        return 0;
     }
 
     pub fn _contains(self: *parser.Node, other: *parser.Node) !bool {
         return try parser.nodeContains(self, other);
     }
 
-    pub fn _getRootNode(self: *parser.Node) void {
-        // TODO
-        _ = self;
-        std.log.err("Not implemented {s}", .{"node.getRootNode()"});
+    pub fn _getRootNode(self: *parser.Node) !?HTMLElem.Union {
+        // TODO return this’s shadow-including root if options["composed"] is true
+        const res = try parser.nodeOwnerDocument(self);
+        if (res == null) {
+            return null;
+        }
+        return try HTMLElem.toInterface(HTMLElem.Union, @as(*parser.Element, @ptrCast(res.?)));
     }
 
     pub fn _hasChildNodes(self: *parser.Node) !bool {
@@ -199,7 +263,7 @@ pub const Node = struct {
     }
 
     pub fn get_childNodes(self: *parser.Node, alloc: std.mem.Allocator) !NodeList {
-        var list = try NodeList.init();
+        var list = NodeList.init();
         errdefer list.deinit(alloc);
 
         var n = try parser.nodeFirstChild(self) orelse return list;
@@ -259,14 +323,30 @@ pub const Node = struct {
         return try Node.toInterface(res);
     }
 
+    // Check if the hierarchy node tree constraints are respected.
+    // For now, it checks only if new nodes are not self.
+    // TODO implements the others contraints.
+    // see https://dom.spec.whatwg.org/#concept-node-tree
+    pub fn hierarchy(self: *parser.Node, nodes: ?Variadic(*parser.Node)) !bool {
+        if (nodes == null) return true;
+        if (nodes.?.slice.len == 0) return true;
+
+        for (nodes.?.slice) |node| if (self == node) return false;
+
+        return true;
+    }
+
     // TODO according with https://dom.spec.whatwg.org/#parentnode, the
     // function must accept either node or string.
     // blocked by https://github.com/lightpanda-io/jsruntime-lib/issues/114
     pub fn prepend(self: *parser.Node, nodes: ?Variadic(*parser.Node)) !void {
         if (nodes == null) return;
         if (nodes.?.slice.len == 0) return;
-        const first = try parser.nodeFirstChild(self);
 
+        // check hierarchy
+        if (!try hierarchy(self, nodes)) return parser.DOMError.HierarchyRequest;
+
+        const first = try parser.nodeFirstChild(self);
         if (first == null) {
             for (nodes.?.slice) |node| {
                 _ = try parser.nodeAppendChild(self, node);
@@ -285,6 +365,10 @@ pub const Node = struct {
     pub fn append(self: *parser.Node, nodes: ?Variadic(*parser.Node)) !void {
         if (nodes == null) return;
         if (nodes.?.slice.len == 0) return;
+
+        // check hierarchy
+        if (!try hierarchy(self, nodes)) return parser.DOMError.HierarchyRequest;
+
         for (nodes.?.slice) |node| {
             _ = try parser.nodeAppendChild(self, node);
         }
@@ -297,24 +381,33 @@ pub const Node = struct {
         if (nodes == null) return;
         if (nodes.?.slice.len == 0) return;
 
+        // check hierarchy
+        if (!try hierarchy(self, nodes)) return parser.DOMError.HierarchyRequest;
+
         // remove existing children
-        if (try parser.nodeHasChildNodes(self)) {
+        try removeChildren(self);
+
+        // add new children
+        for (nodes.?.slice) |node| {
+            _ = try parser.nodeAppendChild(self, node);
+        }
+    }
+
+    pub fn removeChildren(self: *parser.Node) !void {
+        if (!try parser.nodeHasChildNodes(self)) return;
+
         const children = try parser.nodeGetChildNodes(self);
         const ln = try parser.nodeListLength(children);
         var i: u32 = 0;
         while (i < ln) {
             defer i += 1;
-                const child = try parser.nodeListItem(children, i) orelse continue;
+            // we always retrieve the 0 index child on purpose: libdom nodelist
+            // are dynamic. So the next child to remove is always as pos 0.
+            const child = try parser.nodeListItem(children, 0) orelse continue;
             _ = try parser.nodeRemoveChild(self, child);
         }
     }
 
-        // add new children
-        for (nodes.?.slice) |node| {
-            _ = try parser.nodeAppendChild(self, node);
-        }
-    }
-
     pub fn deinit(_: *parser.Node, _: std.mem.Allocator) void {}
 };
 
@@ -337,6 +430,21 @@ pub fn testExecFn(
     ;
     try runScript(js_env, alloc, trim_and_replace, "proto_test");
 
+    var node_compare_document_position = [_]Case{
+        .{ .src = "document.body.compareDocumentPosition(document.firstChild); ", .ex = "10" },
+        .{ .src = "document.getElementById(\"para-empty\").compareDocumentPosition(document.getElementById(\"content\"));", .ex = "10" },
+        .{ .src = "document.getElementById(\"content\").compareDocumentPosition(document.getElementById(\"para-empty\"));", .ex = "20" },
+        .{ .src = "document.getElementById(\"link\").compareDocumentPosition(document.getElementById(\"link\"));", .ex = "0" },
+        .{ .src = "document.getElementById(\"para-empty\").compareDocumentPosition(document.getElementById(\"link\"));", .ex = "2" },
+        .{ .src = "document.getElementById(\"link\").compareDocumentPosition(document.getElementById(\"para-empty\"));", .ex = "4" },
+    };
+    try checkCases(js_env, &node_compare_document_position);
+
+    var get_root_node = [_]Case{
+        .{ .src = "document.getElementById('content').getRootNode().__proto__.constructor.name", .ex = "HTMLDocument" },
+    };
+    try checkCases(js_env, &get_root_node);
+
     var first_child = [_]Case{
         // for next test cases
         .{ .src = "let content = document.getElementById('content')", .ex = "undefined" },

src/dom/nodelist.zig

@@ -1,16 +1,101 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
+const Callback = jsruntime.Callback;
+const CallbackResult = jsruntime.CallbackResult;
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
 
 const NodeUnion = @import("node.zig").Union;
 const Node = @import("node.zig").Node;
 
+const U32Iterator = @import("../iterator/iterator.zig").U32Iterator;
+
+const log = std.log.scoped(.nodelist);
+
 const DOMException = @import("exceptions.zig").DOMException;
 
+pub const Interfaces = .{
+    NodeListIterator,
+    NodeList,
+};
+
+pub const NodeListIterator = struct {
+    pub const mem_guarantied = true;
+
+    coll: *NodeList,
+    index: u32 = 0,
+
+    pub const Return = struct {
+        value: ?NodeUnion,
+        done: bool,
+    };
+
+    pub fn _next(self: *NodeListIterator) !Return {
+        const e = try self.coll._item(self.index);
+        if (e == null) {
+            return Return{
+                .value = null,
+                .done = true,
+            };
+        }
+
+        self.index += 1;
+        return Return{
+            .value = e,
+            .done = false,
+        };
+    }
+};
+
+pub const NodeListEntriesIterator = struct {
+    pub const mem_guarantied = true;
+
+    coll: *NodeList,
+    index: u32 = 0,
+
+    pub const Return = struct {
+        value: ?NodeUnion,
+        done: bool,
+    };
+
+    pub fn _next(self: *NodeListEntriesIterator) !Return {
+        const e = try self.coll._item(self.index);
+        if (e == null) {
+            return Return{
+                .value = null,
+                .done = true,
+            };
+        }
+
+        self.index += 1;
+        return Return{
+            .value = e,
+            .done = false,
+        };
+    }
+};
+
 // Nodelist is implemented in pure Zig b/c libdom's NodeList doesn't allow to
 // append nodes.
 // WEB IDL https://dom.spec.whatwg.org/#nodelist
@@ -26,7 +111,7 @@ pub const NodeList = struct {
 
     nodes: NodesArrayList,
 
-    pub fn init() !NodeList {
+    pub fn init() NodeList {
         return NodeList{
             .nodes = NodesArrayList{},
         };
@@ -54,9 +139,50 @@ pub const NodeList = struct {
         return try Node.toInterface(n);
     }
 
-    // TODO _symbol_iterator
+    pub fn _forEach(self: *NodeList, alloc: std.mem.Allocator, cbk: Callback) !void { // TODO handle thisArg
+        var res = CallbackResult.init(alloc);
+        defer res.deinit();
 
-    // TODO implement postAttach
+        for (self.nodes.items, 0..) |n, i| {
+            const ii: u32 = @intCast(i);
+            cbk.trycall(.{ n, ii, self }, &res) catch |e| {
+                log.err("callback error: {s}", .{res.result orelse "unknown"});
+                log.debug("{s}", .{res.stack orelse "no stack trace"});
+
+                return e;
+            };
+        }
+    }
+
+    pub fn _keys(self: *NodeList) U32Iterator {
+        return .{
+            .length = self.get_length(),
+        };
+    }
+
+    pub fn _values(self: *NodeList) NodeListIterator {
+        return .{
+            .coll = self,
+        };
+    }
+
+    pub fn _symbol_iterator(self: *NodeList) NodeListIterator {
+        return self._values();
+    }
+
+    // TODO entries() https://developer.mozilla.org/en-US/docs/Web/API/NodeList/entries
+
+    pub fn postAttach(self: *NodeList, alloc: std.mem.Allocator, js_obj: jsruntime.JSObject) !void {
+        const ln = self.get_length();
+        var i: u32 = 0;
+        while (i < ln) {
+            defer i += 1;
+            const k = try std.fmt.allocPrint(alloc, "{d}", .{i});
+
+            const node = try self._item(i) orelse unreachable;
+            try js_obj.set(k, node);
+        }
+    }
 };
 
 // Tests
@@ -69,6 +195,14 @@ pub fn testExecFn(
     var childnodes = [_]Case{
         .{ .src = "let list = document.getElementById('content').childNodes", .ex = "undefined" },
         .{ .src = "list.length", .ex = "9" },
+        .{ .src = "list[0].__proto__.constructor.name", .ex = "Text" },
+        .{ .src = 
+        \\let i = 0;
+        \\list.forEach(function (n, idx) {
+        \\  i += idx;
+        \\});
+        \\i;
+        , .ex = "36" },
     };
     try checkCases(js_env, &childnodes);
 }

src/dom/processing_instruction.zig

@@ -1,6 +1,29 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+const parser = @import("netsurf");
+const Node = @import("node.zig").Node;
 
 // https://dom.spec.whatwg.org/#processinginstruction
 pub const ProcessingInstruction = struct {
@@ -8,18 +31,39 @@ pub const ProcessingInstruction = struct {
 
     // TODO for libdom processing instruction inherit from node.
     // But the spec says it must inherit from CDATA.
-    // Moreover, inherit from Node causes also a crash with cloneNode.
-    // https://github.com/lightpanda-io/browsercore/issues/123
-    //
-    // In consequence, for now, we don't implement all these func for
-    // ProcessingInstruction.
-    //
-    //pub const prototype = *CharacterData;
-
+    pub const prototype = *Node;
     pub const mem_guarantied = true;
 
     pub fn get_target(self: *parser.ProcessingInstruction) ![]const u8 {
         // libdom stores the ProcessingInstruction target in the node's name.
-        return try parser.nodeName(@as(*parser.Node, @ptrCast(self)));
+        return try parser.nodeName(parser.processingInstructionToNode(self));
+    }
+
+    pub fn _cloneNode(self: *parser.ProcessingInstruction, _: ?bool) !*parser.ProcessingInstruction {
+        return try parser.processInstructionCopy(self);
+    }
+
+    pub fn get_data(self: *parser.ProcessingInstruction) !?[]const u8 {
+        return try parser.nodeValue(parser.processingInstructionToNode(self));
+    }
+
+    pub fn set_data(self: *parser.ProcessingInstruction, data: []u8) !void {
+        try parser.nodeSetValue(parser.processingInstructionToNode(self), data);
     }
 };
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var createProcessingInstruction = [_]Case{
+        .{ .src = "let pi = document.createProcessingInstruction('foo', 'bar')", .ex = "undefined" },
+        .{ .src = "pi.target", .ex = "foo" },
+        .{ .src = "pi.data", .ex = "bar" },
+        .{ .src = "pi.data = 'foo'", .ex = "foo" },
+        .{ .src = "pi.data", .ex = "foo" },
+
+        .{ .src = "let pi2 = pi.cloneNode()", .ex = "undefined" },
+    };
+    try checkCases(js_env, &createProcessingInstruction);
+}

src/dom/text.zig

@@ -1,25 +1,51 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
-const generate = @import("../generate.zig");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const CharacterData = @import("character_data.zig").CharacterData;
 const CDATASection = @import("cdata_section.zig").CDATASection;
 
+const UserContext = @import("../user_context.zig").UserContext;
+
 // Text interfaces
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     CDATASection,
-});
+};
 
 pub const Text = struct {
     pub const Self = parser.Text;
     pub const prototype = *CharacterData;
     pub const mem_guarantied = true;
 
+    pub fn constructor(userctx: UserContext, data: ?[]const u8) !*parser.Text {
+        return parser.documentCreateTextNode(
+            parser.documentHTMLToDocument(userctx.document),
+            data orelse "",
+        );
+    }
+
     // JS funcs
     // --------
 
@@ -44,6 +70,15 @@ pub fn testExecFn(
     _: std.mem.Allocator,
     js_env: *jsruntime.Env,
 ) anyerror!void {
+    var constructor = [_]Case{
+        .{ .src = "let t = new Text('foo')", .ex = "undefined" },
+        .{ .src = "t.data", .ex = "foo" },
+
+        .{ .src = "let emptyt = new Text()", .ex = "undefined" },
+        .{ .src = "emptyt.data", .ex = "" },
+    };
+    try checkCases(js_env, &constructor);
+
     var get_whole_text = [_]Case{
         .{ .src = "let text = document.getElementById('link').firstChild", .ex = "undefined" },
         .{ .src = "text.wholeText === 'OK'", .ex = "true" },

src/dom/token_list.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;

src/dom/walker.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 pub const Walker = union(enum) {
     walkerDepthFirst: WalkerDepthFirst,

src/events/event.zig

@@ -1,13 +1,32 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const generate = @import("../generate.zig");
 
 const jsruntime = @import("jsruntime");
 const Callback = jsruntime.Callback;
+const CallbackResult = jsruntime.CallbackResult;
 const Case = jsruntime.test_utils.Case;
 const checkCases = jsruntime.test_utils.checkCases;
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const DOMException = @import("../dom/exceptions.zig").DOMException;
 const EventTarget = @import("../dom/event_target.zig").EventTarget;
@@ -15,13 +34,15 @@ const EventTargetUnion = @import("../dom/event_target.zig").Union;
 
 const ProgressEvent = @import("../xhr/progress_event.zig").ProgressEvent;
 
+const log = std.log.scoped(.events);
+
 // Event interfaces
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     Event,
     ProgressEvent,
-});
-const Generated = generate.Union.compile(Interfaces);
-pub const Union = Generated._union;
+};
+
+pub const Union = generate.Union(Interfaces);
 
 // https://dom.spec.whatwg.org/#event
 pub const Event = struct {
@@ -218,3 +239,25 @@ pub fn testExecFn(
     };
     try checkCases(js_env, &remove);
 }
+
+pub const EventHandler = struct {
+    fn handle(event: ?*parser.Event, data: parser.EventHandlerData) void {
+        // TODO get the allocator by another way?
+        var res = CallbackResult.init(data.cbk.nat_ctx.alloc);
+        defer res.deinit();
+
+        if (event) |evt| {
+            data.cbk.trycall(.{
+                Event.toInterface(evt) catch unreachable,
+            }, &res) catch |e| log.err("event handler error: {any}", .{e});
+        } else {
+            data.cbk.trycall(.{event}, &res) catch |e| log.err("event handler error (null event): {any}", .{e});
+        }
+
+        // in case of function error, we log the result and the trace.
+        if (!res.success) {
+            log.info("event handler error try catch: {s}", .{res.result orelse "unknown"});
+            log.debug("{s}", .{res.stack orelse "no stack trace"});
+        }
+    }
+}.handle;

src/generate.zig

@@ -1,418 +1,229 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
-const builtin = @import("builtin");
 
-// Utils
-// -----
-
-fn itoa(comptime i: u8) ![]const u8 {
-    var len: usize = undefined;
-    if (i < 10) {
-        len = 1;
-    } else if (i < 100) {
-        len = 2;
-    } else {
-        return error.GenerateTooMuchMembers;
-    }
-    var buf: [len]u8 = undefined;
-    return try std.fmt.bufPrint(buf[0..], "{d}", .{i});
-}
-
-fn fmtName(comptime T: type) []const u8 {
-    var it = std.mem.splitBackwards(u8, @typeName(T), ".");
-    return it.first();
-}
+// ----
+const Type = std.builtin.Type;
 
 // Union
 // -----
 
-// Generate a flatten tagged Union from various structs and union of structs
-// TODO: make this function more generic
-// TODO: dedup
-pub const Union = struct {
-    _enum: type,
-    _union: type,
+// Generate a flatten tagged Union from a Tuple
+pub fn Union(interfaces: anytype) type {
+    // @setEvalBranchQuota(10000);
+    const tuple = Tuple(interfaces){};
+    const fields = std.meta.fields(@TypeOf(tuple));
 
-    pub fn compile(comptime tuple: anytype) Union {
-        return private_compile(tuple) catch |err| @compileError(@errorName(err));
-    }
-
-    fn private_compile(comptime tuple: anytype) !Union {
-        @setEvalBranchQuota(10000);
-
-        // check types provided
-        const tuple_T = @TypeOf(tuple);
-        const tuple_info = @typeInfo(tuple_T);
-        if (tuple_info != .Struct or !tuple_info.Struct.is_tuple) {
-            return error.GenerateArgNotTuple;
-        }
-
-        const tuple_members = tuple_info.Struct.fields;
-
-        // first iteration to get the total number of members
-        var members_nb = 0;
-        for (tuple_members) |member| {
-            const member_T = @field(tuple, member.name);
-            const member_info = @typeInfo(member_T);
-            if (member_info == .Union) {
-                const member_union = member_info.Union;
-                members_nb += member_union.fields.len;
-            } else if (member_info == .Struct) {
-                members_nb += 1;
-            } else {
-                return error.GenerateMemberNotUnionOrStruct;
-            }
-        }
-
-        // define the tag type regarding the members nb
-        var tag_type: type = undefined;
-        if (members_nb < 3) {
-            tag_type = u1;
-        } else if (members_nb < 4) {
-            tag_type = u2;
-        } else if (members_nb < 8) {
-            tag_type = u3;
-        } else if (members_nb < 16) {
-            tag_type = u4;
-        } else if (members_nb < 32) {
-            tag_type = u5;
-        } else if (members_nb < 64) {
-            tag_type = u6;
-        } else if (members_nb < 128) {
-            tag_type = u7;
-        } else if (members_nb < 256) {
-            tag_type = u8;
-        } else if (members_nb < 65536) {
-            tag_type = u16;
-        } else {
-            return error.GenerateTooMuchMembers;
-        }
+    const tag_type = switch (fields.len) {
+        0 => unreachable,
+        1 => u0,
+        2 => u1,
+        3...4 => u2,
+        5...8 => u3,
+        9...16 => u4,
+        17...32 => u5,
+        33...64 => u6,
+        65...128 => u7,
+        129...256 => u8,
+        else => @compileError("Too many interfaces to generate union"),
+    };
 
     // second iteration to generate tags
-        var enum_fields: [members_nb]std.builtin.Type.EnumField = undefined;
-        var done = 0;
-        for (tuple_members) |member| {
-            const member_T = @field(tuple, member.name);
-            const member_info = @typeInfo(member_T);
-            if (member_info == .Union) {
-                const member_union = member_info.Union;
-                for (member_union.fields) |field| {
-                    enum_fields[done] = .{
-                        .name = fmtName(field.type),
-                        .value = done,
+    var enum_fields: [fields.len]Type.EnumField = undefined;
+    for (fields, 0..) |field, index| {
+        const member = @field(tuple, field.name);
+        const full_name = @typeName(member);
+        const separator = std.mem.lastIndexOfScalar(u8, full_name, '.') orelse unreachable;
+        const name = full_name[separator + 1 ..];
+        enum_fields[index] = .{
+            .name = name ++ "",
+            .value = index,
         };
-                    done += 1;
     }
-            } else if (member_info == .Struct) {
-                enum_fields[done] = .{
-                    .name = fmtName(member_T),
-                    .value = done,
-                };
-                done += 1;
-            }
-        }
-        const decls: [0]std.builtin.Type.Declaration = undefined;
-        const enum_info = std.builtin.Type.Enum{
+
+    const enum_info = Type.Enum{
         .tag_type = tag_type,
         .fields = &enum_fields,
-            .decls = &decls,
+        .decls = &.{},
         .is_exhaustive = true,
     };
-        const enum_T = @Type(std.builtin.Type{ .Enum = enum_info });
+    const enum_T = @Type(.{ .Enum = enum_info });
 
     // third iteration to generate union type
-        var union_fields: [members_nb]std.builtin.Type.UnionField = undefined;
-        done = 0;
-        for (tuple_members, 0..) |member, i| {
-            const member_T = @field(tuple, member.name);
-            const member_info = @typeInfo(member_T);
-            if (member_info == .Union) {
-                const member_union = member_info.Union;
-                for (member_union.fields) |field| {
-                    var T: type = undefined;
-                    if (@hasDecl(field.type, "Self")) {
-                        T = @field(field.type, "Self");
-                        T = *T;
-                    } else {
-                        T = field.type;
+    var union_fields: [fields.len]Type.UnionField = undefined;
+    for (fields, enum_fields, 0..) |field, e, index| {
+        var FT = @field(tuple, field.name);
+        if (@hasDecl(FT, "Self")) {
+            FT = *(@field(FT, "Self"));
         }
-                    union_fields[done] = .{
-                        .name = fmtName(field.type),
-                        .type = T,
-                        .alignment = @alignOf(T),
+        union_fields[index] = .{
+            .type = FT,
+            .name = e.name,
+            .alignment = @alignOf(FT),
         };
-                    done += 1;
     }
-            } else if (member_info == .Struct) {
-                const member_name = try itoa(i);
-                var T = @field(tuple, member_name);
-                if (@hasDecl(T, "Self")) {
-                    T = @field(T, "Self");
-                    T = *T;
-                }
-                union_fields[done] = .{
-                    .name = fmtName(member_T),
-                    .type = T,
-                    .alignment = @alignOf(T),
-                };
-                done += 1;
-            }
-        }
-        const union_info = std.builtin.Type.Union{
-            .layout = .Auto,
+
+    return @Type(.{ .Union = .{
+        .layout = .auto,
         .tag_type = enum_T,
         .fields = &union_fields,
-            .decls = &decls,
-        };
-        const union_T = @Type(std.builtin.Type{ .Union = union_info });
-
-        return .{
-            ._enum = enum_T,
-            ._union = union_T,
-        };
-    }
-};
+        .decls = &.{},
+    } });
+}
 
 // Tuple
 // -----
 
-fn tupleNb(comptime tuple: anytype) usize {
-    var nb = 0;
-    for (@typeInfo(@TypeOf(tuple)).Struct.fields) |member| {
-        const member_T = @field(tuple, member.name);
-        if (@TypeOf(member_T) == type) {
-            nb += 1;
-        } else {
-            const member_info = @typeInfo(@TypeOf(member_T));
-            if (member_info != .Struct and !member_info.Struct.is_tuple) {
-                @compileError("GenerateMemberNotTypeOrTuple");
-            }
-            for (member_info.Struct.fields) |field| {
-                if (@TypeOf(@field(member_T, field.name)) != type) {
-                    @compileError("GenerateMemberTupleChildNotType");
-                }
-            }
-            nb += member_info.Struct.fields.len;
-        }
-    }
-    return nb;
-}
-
-fn tupleTypes(comptime nb: usize, comptime tuple: anytype) [nb]type {
-    var types: [nb]type = undefined;
-    var done = 0;
-    for (@typeInfo(@TypeOf(tuple)).Struct.fields) |member| {
-        const T = @field(tuple, member.name);
-        if (@TypeOf(T) == type) {
-            types[done] = T;
-            done += 1;
-        } else {
-            const info = @typeInfo(@TypeOf(T));
-            for (info.Struct.fields) |field| {
-                types[done] = @field(T, field.name);
-                done += 1;
-            }
-        }
-    }
-    return types;
-}
-
-fn isDup(comptime nb: usize, comptime list: [nb]type, comptime T: type, comptime i: usize) bool {
-    for (list, 0..) |item, index| {
-        if (i >= index) {
-            // check sequentially
-            continue;
-        }
-        if (T == item) {
-            return true;
-        }
-    }
-    return false;
-}
-
-fn dedupIndexes(comptime nb: usize, comptime types: [nb]type) [nb]i32 {
-    var dedup_indexes: [nb]i32 = undefined;
-    for (types, 0..) |T, i| {
-        if (isDup(nb, types, T, i)) {
-            dedup_indexes[i] = -1;
-        } else {
-            dedup_indexes[i] = i;
-        }
-    }
-    return dedup_indexes;
-}
-
-fn dedupNb(comptime nb: usize, comptime dedup_indexes: [nb]i32) usize {
-    var dedup_nb = 0;
-    for (dedup_indexes) |index| {
-        if (index != -1) {
-            dedup_nb += 1;
-        }
-    }
-    return dedup_nb;
-}
-
-fn TupleT(comptime tuple: anytype) type {
+// Flattens and depuplicates a list of nested tuples. For example
+// input: {A, B, {C, B, D}, {A, E}}
+// output {A, B, C, D, E}
+pub fn Tuple(args: anytype) type {
     @setEvalBranchQuota(100000);
 
-    // logic
-    const nb = tupleNb(tuple);
-    const types = tupleTypes(nb, tuple);
-    const dedup_indexes = dedupIndexes(nb, types);
-    const dedup_nb = dedupNb(nb, dedup_indexes);
+    const count = countInterfaces(args, 0);
+    var interfaces: [count]type = undefined;
+    _ = flattenInterfaces(args, &interfaces, 0);
 
-    // generate the tuple type
-    var fields: [dedup_nb]std.builtin.Type.StructField = undefined;
-    var done = 0;
-    for (dedup_indexes) |index| {
-        if (index == -1) {
+    const unfiltered_count, const filter_set = filterMap(count, interfaces);
+
+    var field_index: usize = 0;
+    var fields: [unfiltered_count]Type.StructField = undefined;
+
+    for (filter_set, 0..) |filter, i| {
+        if (filter) {
             continue;
         }
-        fields[done] = .{
-            .name = try itoa(done),
+        fields[field_index] = .{
+            .name = std.fmt.comptimePrint("{d}", .{field_index}),
             .type = type,
-            .default_value = null,
-            .is_comptime = false,
+            // has to be true in order to properly capture the default value
+            .is_comptime = true,
             .alignment = @alignOf(type),
+            .default_value = @ptrCast(&interfaces[i]),
         };
-        done += 1;
+        field_index += 1;
     }
-    const decls: [0]std.builtin.Type.Declaration = undefined;
-    const info = std.builtin.Type.Struct{
-        .layout = .Auto,
+
+    return @Type(.{ .Struct = .{
+        .layout = .auto,
         .fields = &fields,
-        .decls = &decls,
+        .decls = &.{},
         .is_tuple = true,
+    } });
+}
+
+fn countInterfaces(args: anytype, count: usize) usize {
+    var new_count = count;
+    for (@typeInfo(@TypeOf(args)).Struct.fields) |f| {
+        const member = @field(args, f.name);
+        if (@TypeOf(member) == type) {
+            new_count += 1;
+        } else {
+            new_count = countInterfaces(member, new_count);
+        }
+    }
+    return new_count;
+}
+
+fn flattenInterfaces(args: anytype, interfaces: []type, index: usize) usize {
+    var new_index = index;
+    for (@typeInfo(@TypeOf(args)).Struct.fields) |f| {
+        const member = @field(args, f.name);
+        if (@TypeOf(member) == type) {
+            interfaces[new_index] = member;
+            new_index += 1;
+        } else {
+            new_index = flattenInterfaces(member, interfaces, new_index);
+        }
+    }
+    return new_index;
+}
+
+fn filterMap(comptime count: usize, interfaces: [count]type) struct { usize, [count]bool } {
+    var map: [count]bool = undefined;
+    var unfiltered_count: usize = 0;
+    outer: for (interfaces, 0..) |iface, i| {
+        for (interfaces[i + 1 ..]) |check| {
+            if (iface == check) {
+                map[i] = true;
+                continue :outer;
+            }
+        }
+        map[i] = false;
+        unfiltered_count += 1;
+    }
+    return .{ unfiltered_count, map };
+}
+
+test "generate.Union" {
+    const Astruct = struct {
+        pub const Self = Other;
+        const Other = struct {};
     };
-    return @Type(std.builtin.Type{ .Struct = info });
+
+    const Bstruct = struct {
+        value: u8 = 0,
+    };
+
+    const Cstruct = struct {
+        value: u8 = 0,
+    };
+
+    const value = Union(.{ Astruct, Bstruct, .{Cstruct} });
+    const ti = @typeInfo(value).Union;
+    try std.testing.expectEqual(3, ti.fields.len);
+    try std.testing.expectEqualStrings("*generate.test.generate.Union.Astruct.Other", @typeName(ti.fields[0].type));
+    try std.testing.expectEqualStrings(ti.fields[0].name, "Astruct");
+    try std.testing.expectEqual(Bstruct, ti.fields[1].type);
+    try std.testing.expectEqualStrings(ti.fields[1].name, "Bstruct");
+    try std.testing.expectEqual(Cstruct, ti.fields[2].type);
+    try std.testing.expectEqualStrings(ti.fields[2].name, "Cstruct");
 }
 
-// Create a flatten tuple from various structs and tuple of structs
-// Duplicates will be removed.
-// TODO: make this function more generic
-pub fn Tuple(comptime tuple: anytype) TupleT(tuple) {
+test "generate.Tuple" {
+    const Astruct = struct {};
 
-    // check types provided
-    const tuple_T = @TypeOf(tuple);
-    const tuple_info = @typeInfo(tuple_T);
-    if (tuple_info != .Struct or !tuple_info.Struct.is_tuple) {
-        @compileError("GenerateArgNotTuple");
+    const Bstruct = struct {
+        value: u8 = 0,
+    };
+
+    const Cstruct = struct {
+        value: u8 = 0,
+    };
+
+    {
+        const tuple = Tuple(.{ Astruct, Bstruct }){};
+        const ti = @typeInfo(@TypeOf(tuple)).Struct;
+        try std.testing.expectEqual(true, ti.is_tuple);
+        try std.testing.expectEqual(2, ti.fields.len);
+        try std.testing.expectEqual(Astruct, tuple.@"0");
+        try std.testing.expectEqual(Bstruct, tuple.@"1");
     }
 
-    // generate the type
-    const T = TupleT(tuple);
-
-    // logic
-    const nb = tupleNb(tuple);
-    const types = tupleTypes(nb, tuple);
-    const dedup_indexes = dedupIndexes(nb, types);
-
-    // instantiate the tuple
-    var t: T = undefined;
-    var done = 0;
-    for (dedup_indexes) |index| {
-        if (index == -1) {
-            continue;
+    {
+        // dedupe
+        const tuple = Tuple(.{ Cstruct, Astruct, .{Astruct}, Bstruct, .{ Astruct, .{ Astruct, Bstruct } } }){};
+        const ti = @typeInfo(@TypeOf(tuple)).Struct;
+        try std.testing.expectEqual(true, ti.is_tuple);
+        try std.testing.expectEqual(3, ti.fields.len);
+        try std.testing.expectEqual(Cstruct, tuple.@"0");
+        try std.testing.expectEqual(Astruct, tuple.@"1");
+        try std.testing.expectEqual(Bstruct, tuple.@"2");
     }
-        const name = try itoa(done);
-        @field(t, name) = types[index];
-        done += 1;
-    }
-    return t;
-}
-
-// Tests
-// -----
-
-const Error = error{
-    GenerateArgNotTuple,
-    GenerateMemberNotUnionOrStruct,
-    GenerateMemberNotTupleOrStruct,
-    GenerateMemberTupleNotStruct,
-    GenerateTooMuchMembers,
-};
-
-const Astruct = struct {
-    value: u8 = 0,
-};
-const Bstruct = struct {
-    value: u8 = 0,
-};
-const Cstruct = struct {
-    value: u8 = 0,
-};
-const Dstruct = struct {
-    value: u8 = 0,
-};
-
-pub fn tests() !void {
-
-    // Union from structs
-    const FromStructs = try Union.private_compile(.{ Astruct, Bstruct, Cstruct });
-
-    const from_structs_enum = @typeInfo(FromStructs._enum);
-    try std.testing.expect(from_structs_enum == .Enum);
-    try std.testing.expect(from_structs_enum.Enum.fields.len == 3);
-    try std.testing.expect(from_structs_enum.Enum.tag_type == u2);
-    try std.testing.expect(from_structs_enum.Enum.fields[0].value == 0);
-    try std.testing.expectEqualStrings(from_structs_enum.Enum.fields[0].name, "Astruct");
-
-    const from_structs_union = @typeInfo(FromStructs._union);
-    try std.testing.expect(from_structs_union == .Union);
-    try std.testing.expect(from_structs_union.Union.tag_type == FromStructs._enum);
-    try std.testing.expect(from_structs_union.Union.fields.len == 3);
-    try std.testing.expect(from_structs_union.Union.fields[0].type == Astruct);
-    try std.testing.expectEqualStrings(from_structs_union.Union.fields[0].name, "Astruct");
-
-    // Union from union and structs
-    const FromMix = try Union.private_compile(.{ FromStructs._union, Dstruct });
-
-    const from_mix_enum = @typeInfo(FromMix._enum);
-    try std.testing.expect(from_mix_enum == .Enum);
-    try std.testing.expect(from_mix_enum.Enum.fields.len == 4);
-    try std.testing.expect(from_mix_enum.Enum.tag_type == u3);
-    try std.testing.expect(from_mix_enum.Enum.fields[0].value == 0);
-    try std.testing.expectEqualStrings(from_mix_enum.Enum.fields[3].name, "Dstruct");
-
-    const from_mix_union = @typeInfo(FromMix._union);
-    try std.testing.expect(from_mix_union == .Union);
-    try std.testing.expect(from_mix_union.Union.tag_type == FromMix._enum);
-    try std.testing.expect(from_mix_union.Union.fields.len == 4);
-    try std.testing.expect(from_mix_union.Union.fields[3].type == Dstruct);
-    try std.testing.expectEqualStrings(from_mix_union.Union.fields[3].name, "Dstruct");
-
-    std.debug.print("Generate Union: OK\n", .{});
-
-    // Tuple from structs
-    const tuple_structs = .{ Astruct, Bstruct };
-    const tFromStructs = Tuple(tuple_structs);
-    const t_from_structs = @typeInfo(@TypeOf(tFromStructs));
-    try std.testing.expect(t_from_structs == .Struct);
-    try std.testing.expect(t_from_structs.Struct.is_tuple);
-    try std.testing.expect(t_from_structs.Struct.fields.len == 2);
-    try std.testing.expect(@field(tFromStructs, "0") == Astruct);
-    try std.testing.expect(@field(tFromStructs, "1") == Bstruct);
-
-    // Tuple from tuple and structs
-    const tuple_mix = .{ tFromStructs, Cstruct };
-    const tFromMix = Tuple(tuple_mix);
-    const t_from_mix = @typeInfo(@TypeOf(tFromMix));
-    try std.testing.expect(t_from_mix == .Struct);
-    try std.testing.expect(t_from_mix.Struct.is_tuple);
-    try std.testing.expect(t_from_mix.Struct.fields.len == 3);
-    try std.testing.expect(@field(tFromMix, "0") == Astruct);
-    try std.testing.expect(@field(tFromMix, "1") == Bstruct);
-    try std.testing.expect(@field(tFromMix, "2") == Cstruct);
-
-    // Tuple with dedup
-    const tuple_dedup = .{ Cstruct, Astruct, tFromStructs, Bstruct };
-    const tFromDedup = Tuple(tuple_dedup);
-    const t_from_dedup = @typeInfo(@TypeOf(tFromDedup));
-    try std.testing.expect(t_from_dedup == .Struct);
-    try std.testing.expect(t_from_dedup.Struct.is_tuple);
-    try std.testing.expect(t_from_dedup.Struct.fields.len == 3);
-    try std.testing.expect(@field(tFromDedup, "0") == Cstruct);
-    try std.testing.expect(@field(tFromDedup, "1") == Astruct);
-    try std.testing.expect(@field(tFromDedup, "2") == Bstruct);
-
-    std.debug.print("Generate Tuple: OK\n", .{});
 }

src/html/document.zig

@@ -1,6 +1,24 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
 
 const jsruntime = @import("jsruntime");
 const Case = jsruntime.test_utils.Case;
@@ -10,6 +28,7 @@ const Node = @import("../dom/node.zig").Node;
 const Document = @import("../dom/document.zig").Document;
 const NodeList = @import("../dom/nodelist.zig").NodeList;
 const HTMLElem = @import("elements.zig");
+const Location = @import("location.zig").Location;
 
 const collection = @import("../dom/html_collection.zig");
 const Walker = @import("../dom/walker.zig").WalkerDepthFirst;
@@ -80,7 +99,7 @@ pub const HTMLDocument = struct {
     }
 
     pub fn _getElementsByName(self: *parser.DocumentHTML, alloc: std.mem.Allocator, name: []const u8) !NodeList {
-        var list = try NodeList.init();
+        var list = NodeList.init();
         errdefer list.deinit(alloc);
 
         if (name.len == 0) return list;
@@ -135,8 +154,12 @@ pub const HTMLDocument = struct {
         return try collection.HTMLCollectionAll(parser.documentHTMLToNode(self), true);
     }
 
-    pub fn get_currentScript(_: *parser.DocumentHTML) !?*parser.Element {
-        return null;
+    pub fn get_currentScript(self: *parser.DocumentHTML) !?*parser.Script {
+        return try parser.documentHTMLGetCurrentScript(self);
+    }
+
+    pub fn get_location(self: *parser.DocumentHTML) !?*Location {
+        return try parser.documentHTMLGetLocation(Location, self);
     }
 
     pub fn get_designMode(_: *parser.DocumentHTML) []const u8 {

src/html/elements.zig

@@ -1,7 +1,32 @@
-const parser = @import("../netsurf.zig");
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+const std = @import("std");
+
+const parser = @import("netsurf");
 const generate = @import("../generate.zig");
 
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
 const Element = @import("../dom/element.zig").Element;
+const URL = @import("../url/url.zig").URL;
+const Node = @import("../dom/node.zig").Node;
 
 // HTMLElement interfaces
 pub const Interfaces = .{
@@ -72,18 +97,45 @@ pub const Interfaces = .{
     HTMLTrackElement,
     HTMLUListElement,
     HTMLVideoElement,
+    CSSProperties,
 };
-const Generated = generate.Union.compile(Interfaces);
-pub const Union = Generated._union;
-pub const Tags = Generated._enum;
+
+pub const Union = generate.Union(Interfaces);
 
 // Abstract class
 // --------------
 
+const CSSProperties = struct {
+    pub const mem_guarantied = true;
+};
+
 pub const HTMLElement = struct {
     pub const Self = parser.ElementHTML;
     pub const prototype = *Element;
     pub const mem_guarantied = true;
+
+    pub fn get_style(_: *parser.ElementHTML) CSSProperties {
+        return .{};
+    }
+
+    pub fn get_innerText(e: *parser.ElementHTML) ![]const u8 {
+        const n = @as(*parser.Node, @ptrCast(e));
+        return try parser.nodeTextContent(n) orelse "";
+    }
+
+    pub fn set_innerText(e: *parser.ElementHTML, s: []const u8) !void {
+        const n = @as(*parser.Node, @ptrCast(e));
+
+        // create text node.
+        const doc = try parser.nodeOwnerDocument(n) orelse return error.NoDocument;
+        const t = try parser.documentCreateTextNode(doc, s);
+
+        // remove existing children.
+        try Node.removeChildren(n);
+
+        // attach the text node.
+        _ = try parser.nodeAppendChild(n, @as(*parser.Node, @ptrCast(t)));
+    }
 };
 
 // Deprecated HTMLElements in Chrome (2023/03/15)
@@ -108,10 +160,286 @@ pub const HTMLUnknownElement = struct {
     pub const mem_guarantied = true;
 };
 
+// https://html.spec.whatwg.org/#the-a-element
 pub const HTMLAnchorElement = struct {
     pub const Self = parser.Anchor;
     pub const prototype = *HTMLElement;
     pub const mem_guarantied = true;
+
+    pub fn get_target(self: *parser.Anchor) ![]const u8 {
+        return try parser.anchorGetTarget(self);
+    }
+
+    pub fn set_target(self: *parser.Anchor, href: []const u8) !void {
+        return try parser.anchorSetTarget(self, href);
+    }
+
+    pub fn get_download(_: *parser.Anchor) ![]const u8 {
+        return ""; // TODO
+    }
+
+    pub fn get_href(self: *parser.Anchor) ![]const u8 {
+        return try parser.anchorGetHref(self);
+    }
+
+    pub fn set_href(self: *parser.Anchor, href: []const u8) !void {
+        return try parser.anchorSetHref(self, href);
+    }
+
+    pub fn get_hreflang(self: *parser.Anchor) ![]const u8 {
+        return try parser.anchorGetHrefLang(self);
+    }
+
+    pub fn set_hreflang(self: *parser.Anchor, href: []const u8) !void {
+        return try parser.anchorSetHrefLang(self, href);
+    }
+
+    pub fn get_type(self: *parser.Anchor) ![]const u8 {
+        return try parser.anchorGetType(self);
+    }
+
+    pub fn set_type(self: *parser.Anchor, t: []const u8) !void {
+        return try parser.anchorSetType(self, t);
+    }
+
+    pub fn get_rel(self: *parser.Anchor) ![]const u8 {
+        return try parser.anchorGetRel(self);
+    }
+
+    pub fn set_rel(self: *parser.Anchor, t: []const u8) !void {
+        return try parser.anchorSetRel(self, t);
+    }
+
+    pub fn get_text(self: *parser.Anchor) !?[]const u8 {
+        return try parser.nodeTextContent(parser.anchorToNode(self));
+    }
+
+    pub fn set_text(self: *parser.Anchor, v: []const u8) !void {
+        return try parser.nodeSetTextContent(parser.anchorToNode(self), v);
+    }
+
+    inline fn url(self: *parser.Anchor, alloc: std.mem.Allocator) !URL {
+        const href = try parser.anchorGetHref(self);
+        return URL.constructor(alloc, href, null); // TODO inject base url
+    }
+
+    // TODO return a disposable string
+    pub fn get_origin(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try u.get_origin(alloc);
+    }
+
+    // TODO return a disposable string
+    pub fn get_protocol(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return u.get_protocol(alloc);
+    }
+
+    pub fn set_protocol(self: *parser.Anchor, alloc: std.mem.Allocator, v: []const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        u.uri.scheme = v;
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_host(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try u.get_host(alloc);
+    }
+
+    pub fn set_host(self: *parser.Anchor, alloc: std.mem.Allocator, v: []const u8) !void {
+        // search : separator
+        var p: ?u16 = null;
+        var h: []const u8 = undefined;
+        for (v, 0..) |c, i| {
+            if (c == ':') {
+                h = v[0..i];
+                p = try std.fmt.parseInt(u16, v[i + 1 ..], 10);
+                break;
+            }
+        }
+
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (p) |pp| {
+            u.uri.host = .{ .raw = h };
+            u.uri.port = pp;
+        } else {
+            u.uri.host = .{ .raw = v };
+            u.uri.port = null;
+        }
+
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_hostname(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try alloc.dupe(u8, u.get_hostname());
+    }
+
+    pub fn set_hostname(self: *parser.Anchor, alloc: std.mem.Allocator, v: []const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        u.uri.host = .{ .raw = v };
+        const href = try u.format(alloc);
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_port(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try u.get_port(alloc);
+    }
+
+    pub fn set_port(self: *parser.Anchor, alloc: std.mem.Allocator, v: ?[]const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (v != null and v.?.len > 0) {
+            u.uri.port = try std.fmt.parseInt(u16, v.?, 10);
+        } else {
+            u.uri.port = null;
+        }
+
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_username(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try alloc.dupe(u8, u.get_username());
+    }
+
+    pub fn set_username(self: *parser.Anchor, alloc: std.mem.Allocator, v: ?[]const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (v) |vv| {
+            u.uri.user = .{ .raw = vv };
+        } else {
+            u.uri.user = null;
+        }
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_password(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try alloc.dupe(u8, u.get_password());
+    }
+
+    pub fn set_password(self: *parser.Anchor, alloc: std.mem.Allocator, v: ?[]const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (v) |vv| {
+            u.uri.password = .{ .raw = vv };
+        } else {
+            u.uri.password = null;
+        }
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_pathname(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try alloc.dupe(u8, u.get_pathname());
+    }
+
+    pub fn set_pathname(self: *parser.Anchor, alloc: std.mem.Allocator, v: []const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        u.uri.path = .{ .raw = v };
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_search(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try u.get_search(alloc);
+    }
+
+    pub fn set_search(self: *parser.Anchor, alloc: std.mem.Allocator, v: ?[]const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (v) |vv| {
+            u.uri.query = .{ .raw = vv };
+        } else {
+            u.uri.query = null;
+        }
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    // TODO return a disposable string
+    pub fn get_hash(self: *parser.Anchor, alloc: std.mem.Allocator) ![]const u8 {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        return try u.get_hash(alloc);
+    }
+
+    pub fn set_hash(self: *parser.Anchor, alloc: std.mem.Allocator, v: ?[]const u8) !void {
+        var u = try url(self, alloc);
+        defer u.deinit(alloc);
+
+        if (v) |vv| {
+            u.uri.fragment = .{ .raw = vv };
+        } else {
+            u.uri.fragment = null;
+        }
+        const href = try u.format(alloc);
+        defer alloc.free(href);
+
+        try parser.anchorSetHref(self, href);
+    }
+
+    pub fn deinit(_: *parser.Anchor, _: std.mem.Allocator) void {}
 };
 
 pub const HTMLAppletElement = struct {
@@ -390,10 +718,120 @@ pub const HTMLQuoteElement = struct {
     pub const mem_guarantied = true;
 };
 
+// https://html.spec.whatwg.org/#the-script-element
 pub const HTMLScriptElement = struct {
     pub const Self = parser.Script;
     pub const prototype = *HTMLElement;
     pub const mem_guarantied = true;
+
+    pub fn get_src(self: *parser.Script) !?[]const u8 {
+        return try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "src",
+        ) orelse "";
+    }
+
+    pub fn set_src(self: *parser.Script, v: []const u8) !void {
+        return try parser.elementSetAttribute(
+            parser.scriptToElt(self),
+            "src",
+            v,
+        );
+    }
+
+    pub fn get_type(self: *parser.Script) !?[]const u8 {
+        return try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "type",
+        ) orelse "";
+    }
+
+    pub fn set_type(self: *parser.Script, v: []const u8) !void {
+        return try parser.elementSetAttribute(
+            parser.scriptToElt(self),
+            "type",
+            v,
+        );
+    }
+
+    pub fn get_text(self: *parser.Script) !?[]const u8 {
+        return try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "text",
+        ) orelse "";
+    }
+
+    pub fn set_text(self: *parser.Script, v: []const u8) !void {
+        return try parser.elementSetAttribute(
+            parser.scriptToElt(self),
+            "text",
+            v,
+        );
+    }
+
+    pub fn get_integrity(self: *parser.Script) !?[]const u8 {
+        return try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "integrity",
+        ) orelse "";
+    }
+
+    pub fn set_integrity(self: *parser.Script, v: []const u8) !void {
+        return try parser.elementSetAttribute(
+            parser.scriptToElt(self),
+            "integrity",
+            v,
+        );
+    }
+
+    pub fn get_async(self: *parser.Script) !bool {
+        _ = try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "async",
+        ) orelse return false;
+
+        return true;
+    }
+
+    pub fn set_async(self: *parser.Script, v: bool) !void {
+        if (v) {
+            return try parser.elementSetAttribute(parser.scriptToElt(self), "async", "");
+        }
+
+        return try parser.elementRemoveAttribute(parser.scriptToElt(self), "async");
+    }
+
+    pub fn get_defer(self: *parser.Script) !bool {
+        _ = try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "defer",
+        ) orelse false;
+        return true;
+    }
+
+    pub fn set_defer(self: *parser.Script, v: bool) !void {
+        if (v) {
+            return try parser.elementSetAttribute(parser.scriptToElt(self), "defer", "");
+        }
+
+        return try parser.elementRemoveAttribute(parser.scriptToElt(self), "defer");
+    }
+
+    pub fn get_noModule(self: *parser.Script) !bool {
+        _ = try parser.elementGetAttribute(
+            parser.scriptToElt(self),
+            "nomodule",
+        ) orelse false;
+        return true;
+    }
+
+    pub fn set_noModule(self: *parser.Script, v: bool) !void {
+        if (v) {
+            return try parser.elementSetAttribute(parser.scriptToElt(self), "nomodule", "");
+        }
+
+        return try parser.elementRemoveAttribute(parser.scriptToElt(self), "nomodule");
+    }
 };
 
 pub const HTMLSelectElement = struct {
@@ -571,3 +1009,90 @@ pub fn toInterface(comptime T: type, e: *parser.Element) !T {
         .undef => .{ .HTMLUnknownElement = @as(*parser.Unknown, @ptrCast(elem)) },
     };
 }
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var anchor = [_]Case{
+        .{ .src = "let a = document.getElementById('link')", .ex = "undefined" },
+        .{ .src = "a.target", .ex = "" },
+        .{ .src = "a.target = '_blank'", .ex = "_blank" },
+        .{ .src = "a.target", .ex = "_blank" },
+        .{ .src = "a.target = ''", .ex = "" },
+
+        .{ .src = "a.href", .ex = "foo" },
+        .{ .src = "a.href = 'https://lightpanda.io/'", .ex = "https://lightpanda.io/" },
+        .{ .src = "a.href", .ex = "https://lightpanda.io/" },
+
+        .{ .src = "a.origin", .ex = "https://lightpanda.io" },
+
+        .{ .src = "a.host = 'lightpanda.io:443'", .ex = "lightpanda.io:443" },
+        .{ .src = "a.host", .ex = "lightpanda.io:443" },
+        .{ .src = "a.port", .ex = "443" },
+        .{ .src = "a.hostname", .ex = "lightpanda.io" },
+
+        .{ .src = "a.host = 'lightpanda.io'", .ex = "lightpanda.io" },
+        .{ .src = "a.host", .ex = "lightpanda.io" },
+        .{ .src = "a.port", .ex = "" },
+        .{ .src = "a.hostname", .ex = "lightpanda.io" },
+
+        .{ .src = "a.host", .ex = "lightpanda.io" },
+        .{ .src = "a.hostname", .ex = "lightpanda.io" },
+        .{ .src = "a.hostname = 'foo.bar'", .ex = "foo.bar" },
+        .{ .src = "a.href", .ex = "https://foo.bar/" },
+
+        .{ .src = "a.search", .ex = "" },
+        .{ .src = "a.search = 'q=bar'", .ex = "q=bar" },
+        .{ .src = "a.search", .ex = "?q=bar" },
+        .{ .src = "a.href", .ex = "https://foo.bar/?q=bar" },
+
+        .{ .src = "a.hash", .ex = "" },
+        .{ .src = "a.hash = 'frag'", .ex = "frag" },
+        .{ .src = "a.hash", .ex = "#frag" },
+        .{ .src = "a.href", .ex = "https://foo.bar/?q=bar#frag" },
+
+        .{ .src = "a.port", .ex = "" },
+        .{ .src = "a.port = '443'", .ex = "443" },
+        .{ .src = "a.host", .ex = "foo.bar:443" },
+        .{ .src = "a.hostname", .ex = "foo.bar" },
+        .{ .src = "a.href", .ex = "https://foo.bar:443/?q=bar#frag" },
+        .{ .src = "a.port = null", .ex = "null" },
+        .{ .src = "a.href", .ex = "https://foo.bar/?q=bar#frag" },
+
+        .{ .src = "a.href = 'foo'", .ex = "foo" },
+
+        .{ .src = "a.type", .ex = "" },
+        .{ .src = "a.type = 'text/html'", .ex = "text/html" },
+        .{ .src = "a.type", .ex = "text/html" },
+        .{ .src = "a.type = ''", .ex = "" },
+
+        .{ .src = "a.text", .ex = "OK" },
+        .{ .src = "a.text = 'foo'", .ex = "foo" },
+        .{ .src = "a.text", .ex = "foo" },
+        .{ .src = "a.text = 'OK'", .ex = "OK" },
+    };
+    try checkCases(js_env, &anchor);
+
+    var script = [_]Case{
+        .{ .src = "let script = document.createElement('script')", .ex = "undefined" },
+        .{ .src = "script.src = 'foo.bar'", .ex = "foo.bar" },
+
+        .{ .src = "script.async = true", .ex = "true" },
+        .{ .src = "script.async", .ex = "true" },
+        .{ .src = "script.async = false", .ex = "false" },
+        .{ .src = "script.async", .ex = "false" },
+    };
+    try checkCases(js_env, &script);
+
+    var innertext = [_]Case{
+        .{ .src = "const backup = document.getElementById('content')", .ex = "undefined" },
+        .{ .src = "document.getElementById('content').innerText = 'foo';", .ex = "foo" },
+        .{ .src = "document.getElementById('content').innerText", .ex = "foo" },
+        .{ .src = "document.getElementById('content').innerHTML = backup; true;", .ex = "true" },
+    };
+    try checkCases(js_env, &innertext);
+}

src/html/history.zig

@@ -0,0 +1,128 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const builtin = @import("builtin");
+const jsruntime = @import("jsruntime");
+
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+// https://html.spec.whatwg.org/multipage/nav-history-apis.html#the-history-interface
+pub const History = struct {
+    pub const mem_guarantied = true;
+
+    const ScrollRestorationMode = enum {
+        auto,
+        manual,
+    };
+
+    scrollRestoration: ScrollRestorationMode = .auto,
+    state: std.json.Value = .null,
+
+    // count tracks the history length until we implement correctly pushstate.
+    count: u32 = 0,
+
+    pub fn get_length(self: *History) u32 {
+        // TODO return the real history length value.
+        return self.count;
+    }
+
+    pub fn get_scrollRestoration(self: *History) []const u8 {
+        return switch (self.scrollRestoration) {
+            .auto => "auto",
+            .manual => "manual",
+        };
+    }
+
+    pub fn set_scrollRestoration(self: *History, mode: []const u8) void {
+        if (std.mem.eql(u8, "manual", mode)) self.scrollRestoration = .manual;
+        if (std.mem.eql(u8, "auto", mode)) self.scrollRestoration = .auto;
+    }
+
+    pub fn get_state(self: *History) std.json.Value {
+        return self.state;
+    }
+
+    // TODO implement the function
+    // data must handle any argument. We could expect a std.json.Value but
+    // https://github.com/lightpanda-io/zig-js-runtime/issues/267 is missing.
+    pub fn _pushState(self: *History, data: []const u8, _: ?[]const u8, url: ?[]const u8) void {
+        self.count += 1;
+        _ = url;
+        _ = data;
+    }
+
+    // TODO implement the function
+    // data must handle any argument. We could expect a std.json.Value but
+    // https://github.com/lightpanda-io/zig-js-runtime/issues/267 is missing.
+    pub fn _replaceState(self: *History, data: []const u8, _: ?[]const u8, url: ?[]const u8) void {
+        _ = self;
+        _ = url;
+        _ = data;
+    }
+
+    // TODO implement the function
+    pub fn _go(self: *History, delta: ?i32) void {
+        _ = self;
+        _ = delta;
+    }
+
+    // TODO implement the function
+    pub fn _back(self: *History) void {
+        _ = self;
+    }
+
+    // TODO implement the function
+    pub fn _forward(self: *History) void {
+        _ = self;
+    }
+};
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var history = [_]Case{
+        .{ .src = "history.scrollRestoration", .ex = "auto" },
+        .{ .src = "history.scrollRestoration = 'manual'", .ex = "manual" },
+        .{ .src = "history.scrollRestoration = 'foo'", .ex = "foo" },
+        .{ .src = "history.scrollRestoration", .ex = "manual" },
+        .{ .src = "history.scrollRestoration = 'auto'", .ex = "auto" },
+        .{ .src = "history.scrollRestoration", .ex = "auto" },
+
+        .{ .src = "history.state", .ex = "null" },
+
+        .{ .src = "history.pushState({}, null, '')", .ex = "undefined" },
+
+        .{ .src = "history.replaceState({}, null, '')", .ex = "undefined" },
+
+        .{ .src = "history.go()", .ex = "undefined" },
+        .{ .src = "history.go(1)", .ex = "undefined" },
+        .{ .src = "history.go(-1)", .ex = "undefined" },
+
+        .{ .src = "history.forward()", .ex = "undefined" },
+
+        .{ .src = "history.back()", .ex = "undefined" },
+    };
+    try checkCases(js_env, &history);
+}

src/html/html.zig

@@ -1,13 +1,37 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const generate = @import("../generate.zig");
 
 const HTMLDocument = @import("document.zig").HTMLDocument;
 const HTMLElem = @import("elements.zig");
 const Window = @import("window.zig").Window;
+const Navigator = @import("navigator.zig").Navigator;
+const History = @import("history.zig").History;
+const Location = @import("location.zig").Location;
 
-pub const Interfaces = generate.Tuple(.{
+pub const Interfaces = .{
     HTMLDocument,
     HTMLElem.HTMLElement,
     HTMLElem.HTMLMediaElement,
     HTMLElem.Interfaces,
     Window,
-});
+    Navigator,
+    History,
+    Location,
+};

src/html/location.zig

@@ -0,0 +1,129 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const builtin = @import("builtin");
+const jsruntime = @import("jsruntime");
+
+const URL = @import("../url/url.zig").URL;
+
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+// https://html.spec.whatwg.org/multipage/nav-history-apis.html#the-location-interface
+pub const Location = struct {
+    pub const mem_guarantied = true;
+
+    url: ?*URL = null,
+
+    pub fn deinit(_: *Location, _: std.mem.Allocator) void {}
+
+    pub fn get_href(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_href(alloc);
+
+        return "";
+    }
+
+    pub fn get_protocol(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_protocol(alloc);
+
+        return "";
+    }
+
+    pub fn get_host(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_host(alloc);
+
+        return "";
+    }
+
+    pub fn get_hostname(self: *Location) []const u8 {
+        if (self.url) |u| return u.get_hostname();
+
+        return "";
+    }
+
+    pub fn get_port(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_port(alloc);
+
+        return "";
+    }
+
+    pub fn get_pathname(self: *Location) []const u8 {
+        if (self.url) |u| return u.get_pathname();
+
+        return "";
+    }
+
+    pub fn get_search(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_search(alloc);
+
+        return "";
+    }
+
+    pub fn get_hash(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_hash(alloc);
+
+        return "";
+    }
+
+    pub fn get_origin(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        if (self.url) |u| return u.get_origin(alloc);
+
+        return "";
+    }
+
+    // TODO
+    pub fn _assign(_: *Location, url: []const u8) !void {
+        _ = url;
+    }
+
+    // TODO
+    pub fn _replace(_: *Location, url: []const u8) !void {
+        _ = url;
+    }
+
+    // TODO
+    pub fn _reload(_: *Location) !void {}
+
+    pub fn _toString(self: *Location, alloc: std.mem.Allocator) ![]const u8 {
+        return try self.get_href(alloc);
+    }
+};
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var location = [_]Case{
+        .{ .src = "location.href", .ex = "https://lightpanda.io/opensource-browser/" },
+        .{ .src = "document.location.href", .ex = "https://lightpanda.io/opensource-browser/" },
+
+        .{ .src = "location.host", .ex = "lightpanda.io" },
+        .{ .src = "location.hostname", .ex = "lightpanda.io" },
+        .{ .src = "location.origin", .ex = "https://lightpanda.io" },
+        .{ .src = "location.pathname", .ex = "/opensource-browser/" },
+        .{ .src = "location.hash", .ex = "" },
+        .{ .src = "location.port", .ex = "" },
+        .{ .src = "location.search", .ex = "" },
+    };
+    try checkCases(js_env, &location);
+}

src/html/navigator.zig

@@ -0,0 +1,102 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const builtin = @import("builtin");
+const jsruntime = @import("jsruntime");
+
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+// https://html.spec.whatwg.org/multipage/system-state.html#navigator
+pub const Navigator = struct {
+    pub const mem_guarantied = true;
+
+    agent: []const u8 = "Lightpanda/1.0",
+    version: []const u8 = "1.0",
+    vendor: []const u8 = "",
+    platform: []const u8 = std.fmt.comptimePrint("{any} {any}", .{ builtin.os.tag, builtin.cpu.arch }),
+
+    language: []const u8 = "en-US",
+
+    pub fn get_userAgent(self: *Navigator) []const u8 {
+        return self.agent;
+    }
+    pub fn get_appCodeName(_: *Navigator) []const u8 {
+        return "Mozilla";
+    }
+    pub fn get_appName(_: *Navigator) []const u8 {
+        return "Netscape";
+    }
+    pub fn get_appVersion(self: *Navigator) []const u8 {
+        return self.version;
+    }
+    pub fn get_platform(self: *Navigator) []const u8 {
+        return self.platform;
+    }
+    pub fn get_product(_: *Navigator) []const u8 {
+        return "Gecko";
+    }
+    pub fn get_productSub(_: *Navigator) []const u8 {
+        return "20030107";
+    }
+    pub fn get_vendor(self: *Navigator) []const u8 {
+        return self.vendor;
+    }
+    pub fn get_vendorSub(_: *Navigator) []const u8 {
+        return "";
+    }
+    pub fn get_language(self: *Navigator) []const u8 {
+        return self.language;
+    }
+    // TODO wait for arrays.
+    //pub fn get_languages(self: *Navigator) [][]const u8 {
+    //    return .{self.language};
+    //}
+    pub fn get_online(_: *Navigator) bool {
+        return true;
+    }
+    pub fn _registerProtocolHandler(_: *Navigator, scheme: []const u8, url: []const u8) void {
+        _ = scheme;
+        _ = url;
+    }
+    pub fn _unregisterProtocolHandler(_: *Navigator, scheme: []const u8, url: []const u8) void {
+        _ = scheme;
+        _ = url;
+    }
+
+    pub fn get_cookieEnabled(_: *Navigator) bool {
+        return true;
+    }
+};
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var navigator = [_]Case{
+        .{ .src = "navigator.userAgent", .ex = "Lightpanda/1.0" },
+        .{ .src = "navigator.appVersion", .ex = "1.0" },
+        .{ .src = "navigator.language", .ex = "en-US" },
+    };
+    try checkCases(js_env, &navigator);
+}

src/html/window.zig

@@ -1,35 +1,98 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
-const parser = @import("../netsurf.zig");
+const parser = @import("netsurf");
+const jsruntime = @import("jsruntime");
+const Callback = jsruntime.Callback;
+const CallbackArg = jsruntime.CallbackArg;
+const Loop = jsruntime.Loop;
 
 const EventTarget = @import("../dom/event_target.zig").EventTarget;
+const Navigator = @import("navigator.zig").Navigator;
+const History = @import("history.zig").History;
+const Location = @import("location.zig").Location;
+
+const storage = @import("../storage/storage.zig");
+
+var emptyLocation = Location{};
 
 // https://dom.spec.whatwg.org/#interface-window-extensions
 // https://html.spec.whatwg.org/multipage/nav-history-apis.html#window
 pub const Window = struct {
     pub const prototype = *EventTarget;
     pub const mem_guarantied = true;
+    pub const global_type = true;
 
     // Extend libdom event target for pure zig struct.
     base: parser.EventTargetTBase = parser.EventTargetTBase{},
 
-    document: ?*parser.Document = null,
+    document: ?*parser.DocumentHTML = null,
     target: []const u8,
+    history: History = .{},
+    location: *Location = &emptyLocation,
 
-    pub fn create(target: ?[]const u8) Window {
+    storageShelf: ?*storage.Shelf = null,
+
+    // store a map between internal timeouts ids and pointers to uint.
+    // the maximum number of possible timeouts is fixed.
+    timeoutid: u32 = 0,
+    timeoutids: [512]u64 = undefined,
+
+    navigator: Navigator,
+
+    pub fn create(target: ?[]const u8, navigator: ?Navigator) Window {
         return Window{
             .target = target orelse "",
+            .navigator = navigator orelse .{},
         };
     }
 
-    pub fn replaceDocument(self: *Window, doc: *parser.Document) void {
+    pub fn replaceLocation(self: *Window, loc: *Location) !void {
+        self.location = loc;
+
+        if (self.document != null) {
+            try parser.documentHTMLSetLocation(Location, self.document.?, self.location);
+        }
+    }
+
+    pub fn replaceDocument(self: *Window, doc: *parser.DocumentHTML) !void {
         self.document = doc;
+        try parser.documentHTMLSetLocation(Location, doc, self.location);
+    }
+
+    pub fn setStorageShelf(self: *Window, shelf: *storage.Shelf) void {
+        self.storageShelf = shelf;
     }
 
     pub fn get_window(self: *Window) *Window {
         return self;
     }
 
+    pub fn get_navigator(self: *Window) *Navigator {
+        return &self.navigator;
+    }
+
+    pub fn get_location(self: *Window) *Location {
+        return self.location;
+    }
+
     pub fn get_self(self: *Window) *Window {
         return self;
     }
@@ -38,11 +101,47 @@ pub const Window = struct {
         return self;
     }
 
-    pub fn get_document(self: *Window) ?*parser.Document {
+    pub fn get_document(self: *Window) ?*parser.DocumentHTML {
         return self.document;
     }
 
+    pub fn get_history(self: *Window) *History {
+        return &self.history;
+    }
+
     pub fn get_name(self: *Window) []const u8 {
         return self.target;
     }
+
+    pub fn get_localStorage(self: *Window) !*storage.Bottle {
+        if (self.storageShelf == null) return parser.DOMError.NotSupported;
+        return &self.storageShelf.?.bucket.local;
+    }
+
+    pub fn get_sessionStorage(self: *Window) !*storage.Bottle {
+        if (self.storageShelf == null) return parser.DOMError.NotSupported;
+        return &self.storageShelf.?.bucket.session;
+    }
+
+    // TODO handle callback arguments.
+    pub fn _setTimeout(self: *Window, loop: *Loop, cbk: Callback, delay: ?u32) !u32 {
+        if (self.timeoutid >= self.timeoutids.len) return error.TooMuchTimeout;
+
+        const ddelay: u63 = delay orelse 0;
+        const id = loop.timeout(ddelay * std.time.ns_per_ms, cbk);
+
+        self.timeoutids[self.timeoutid] = id;
+        defer self.timeoutid += 1;
+
+        return self.timeoutid;
+    }
+
+    pub fn _clearTimeout(self: *Window, loop: *Loop, id: u32) void {
+        // I do would prefer return an error in this case, but it seems some JS
+        // uses invalid id, in particular id 0.
+        // So we silently ignore invalid id for now.
+        if (id >= self.timeoutid) return;
+
+        loop.cancel(self.timeoutids[id], null);
+    }
 };

src/html_test.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 pub const html: []const u8 =
     \\<main id='content'>
     \\<a href='foo'>OK</a>

src/iterator/iterator.zig

@@ -0,0 +1,53 @@
+const std = @import("std");
+
+pub const Interfaces = .{
+    U32Iterator,
+};
+
+pub const U32Iterator = struct {
+    pub const mem_guarantied = true;
+
+    length: u32,
+    index: u32 = 0,
+
+    pub const Return = struct {
+        value: u32,
+        done: bool,
+    };
+
+    pub fn _next(self: *U32Iterator) Return {
+        const i = self.index;
+        if (i >= self.length) {
+            return .{
+                .value = 0,
+                .done = true,
+            };
+        }
+
+        self.index = i + 1;
+        return .{
+            .value = i,
+            .done = false,
+        };
+    }
+};
+
+const testing = std.testing;
+test "U32Iterator" {
+    const Return = U32Iterator.Return;
+
+    {
+        var it = U32Iterator{ .length = 0 };
+        try testing.expectEqual(Return{ .value = 0, .done = true }, it._next());
+        try testing.expectEqual(Return{ .value = 0, .done = true }, it._next());
+    }
+
+    {
+        var it = U32Iterator{ .length = 3 };
+        try testing.expectEqual(Return{ .value = 0, .done = false }, it._next());
+        try testing.expectEqual(Return{ .value = 1, .done = false }, it._next());
+        try testing.expectEqual(Return{ .value = 2, .done = false }, it._next());
+        try testing.expectEqual(Return{ .value = 0, .done = true }, it._next());
+        try testing.expectEqual(Return{ .value = 0, .done = true }, it._next());
+    }
+}

src/lexbor.zig

@@ -1,458 +0,0 @@
-const std = @import("std");
-
-const c = @cImport({
-    @cInclude("lexbor/html/html.h");
-});
-
-// Public API
-// ----------
-
-// Tag
-
-pub const Tag = enum(u8) {
-    a = c.LXB_TAG_A,
-    area = c.LXB_TAG_AREA,
-    audio = c.LXB_TAG_AUDIO,
-    br = c.LXB_TAG_BR,
-    base = c.LXB_TAG_BASE,
-    body = c.LXB_TAG_BODY,
-    button = c.LXB_TAG_BUTTON,
-    canvas = c.LXB_TAG_CANVAS,
-    dl = c.LXB_TAG_DL,
-    dialog = c.LXB_TAG_DIALOG,
-    data = c.LXB_TAG_DATA,
-    div = c.LXB_TAG_DIV,
-    embed = c.LXB_TAG_EMBED,
-    fieldset = c.LXB_TAG_FIELDSET,
-    form = c.LXB_TAG_FORM,
-    frameset = c.LXB_TAG_FRAMESET,
-    hr = c.LXB_TAG_HR,
-    head = c.LXB_TAG_HEAD,
-    h1 = c.LXB_TAG_H1,
-    h2 = c.LXB_TAG_H2,
-    h3 = c.LXB_TAG_H3,
-    h4 = c.LXB_TAG_H4,
-    h5 = c.LXB_TAG_H5,
-    h6 = c.LXB_TAG_H6,
-    html = c.LXB_TAG_HTML,
-    iframe = c.LXB_TAG_IFRAME,
-    img = c.LXB_TAG_IMG,
-    input = c.LXB_TAG_INPUT,
-    li = c.LXB_TAG_LI,
-    label = c.LXB_TAG_LABEL,
-    legend = c.LXB_TAG_LEGEND,
-    link = c.LXB_TAG_LINK,
-    map = c.LXB_TAG_MAP,
-    meta = c.LXB_TAG_META,
-    meter = c.LXB_TAG_METER,
-    ins = c.LXB_TAG_INS,
-    del = c.LXB_TAG_DEL,
-    ol = c.LXB_TAG_OL,
-    object = c.LXB_TAG_OBJECT,
-    optgroup = c.LXB_TAG_OPTGROUP,
-    option = c.LXB_TAG_OPTION,
-    output = c.LXB_TAG_OUTPUT,
-    p = c.LXB_TAG_P,
-    picture = c.LXB_TAG_PICTURE,
-    pre = c.LXB_TAG_PRE,
-    progress = c.LXB_TAG_PROGRESS,
-    blockquote = c.LXB_TAG_BLOCKQUOTE,
-    q = c.LXB_TAG_Q,
-    script = c.LXB_TAG_SCRIPT,
-    select = c.LXB_TAG_SELECT,
-    source = c.LXB_TAG_SOURCE,
-    span = c.LXB_TAG_SPAN,
-    style = c.LXB_TAG_STYLE,
-    table = c.LXB_TAG_TABLE,
-    caption = c.LXB_TAG_CAPTION,
-    th = c.LXB_TAG_TH,
-    td = c.LXB_TAG_TD,
-    col = c.LXB_TAG_COL,
-    tr = c.LXB_TAG_TR,
-    thead = c.LXB_TAG_THEAD,
-    tbody = c.LXB_TAG_TBODY,
-    tfoot = c.LXB_TAG_TFOOT,
-    template = c.LXB_TAG_TEMPLATE,
-    textarea = c.LXB_TAG_TEXTAREA,
-    time = c.LXB_TAG_TIME,
-    title = c.LXB_TAG_TITLE,
-    track = c.LXB_TAG_TRACK,
-    ul = c.LXB_TAG_UL,
-    video = c.LXB_TAG_VIDEO,
-    undef = c.LXB_TAG__UNDEF,
-
-    pub fn all() []Tag {
-        comptime {
-            const info = @typeInfo(Tag).Enum;
-            comptime var l: [info.fields.len]Tag = undefined;
-            inline for (info.fields, 0..) |field, i| {
-                l[i] = @as(Tag, @enumFromInt(field.value));
-            }
-            return &l;
-        }
-    }
-
-    pub fn allElements() [][]const u8 {
-        comptime {
-            const tags = all();
-            var names: [tags.len][]const u8 = undefined;
-            inline for (tags, 0..) |tag, i| {
-                names[i] = tag.elementName();
-            }
-            return &names;
-        }
-    }
-
-    fn upperName(comptime name: []const u8) []const u8 {
-        comptime {
-            var upper_name: [name.len]u8 = undefined;
-            for (name, 0..) |char, i| {
-                var to_upper = false;
-                if (i == 0) {
-                    to_upper = true;
-                } else if (i == 1 and name.len == 2) {
-                    to_upper = true;
-                }
-                if (to_upper) {
-                    upper_name[i] = std.ascii.toUpper(char);
-                } else {
-                    upper_name[i] = char;
-                }
-            }
-            return &upper_name;
-        }
-    }
-
-    fn elementName(comptime tag: Tag) []const u8 {
-        return switch (tag) {
-            .a => "Anchor",
-            .dl => "DList",
-            .fieldset => "FieldSet",
-            .frameset => "FrameSet",
-            .h1, .h2, .h3, .h4, .h5, .h6 => "Heading",
-            .iframe => "IFrame",
-            .img => "Image",
-            .ins, .del => "Mod",
-            .ol => "OList",
-            .optgroup => "OptGroup",
-            .p => "Paragraph",
-            .blockquote, .q => "Quote",
-            .caption => "TableCaption",
-            .th, .td => "TableCell",
-            .col => "TableCol",
-            .tr => "TableRow",
-            .thead, .tbody, .tfoot => "TableSection",
-            .textarea => "TextArea",
-            .ul => "UList",
-            .undef => "Unknown",
-            else => upperName(@tagName(tag)),
-        };
-    }
-};
-
-// EventTarget
-
-pub const EventTarget = c.lxb_dom_event_target_t;
-
-// Node
-
-pub const Node = c.lxb_dom_node_t;
-
-pub const NodeType = enum(u4) {
-    undef,
-    element,
-    attribute,
-    text,
-    cdata_section,
-    entity_reference,
-    entity,
-    processing_instruction,
-    comment,
-    document,
-    document_type,
-    document_fragment,
-    notation,
-    last_entry,
-};
-
-pub inline fn nodeEventTarget(node: *Node) *EventTarget {
-    return c.lxb_dom_interface_event_target(node);
-}
-
-pub inline fn nodeTag(node: *Node) Tag {
-    // FIXME: lxb_dom_node_tag_id returns a big number if element is unknwon
-    // while it should return 0 (value of enum LXB_TAG__UNDEF).
-    // This fix the problem by assuming that a value greater than an u8 (the basis
-    // of Tag enum) is 0.
-    var val = c.lxb_dom_node_tag_id(node);
-    if (val > 256) {
-        val = 0;
-    }
-    return @as(Tag, @enumFromInt(val));
-}
-
-pub const nodeWalker = (fn (node: ?*Node, _: ?*anyopaque) callconv(.C) Action);
-
-pub inline fn nodeName(node: *Node) [*c]const u8 {
-    var s: usize = undefined;
-    return c.lxb_dom_node_name(node, &s);
-}
-
-pub inline fn nodeType(node: *Node) NodeType {
-    return @as(NodeType, @enumFromInt(node.*.type));
-}
-
-pub inline fn nodeWalk(node: *Node, comptime walker: nodeWalker) !void {
-    c.lxb_dom_node_simple_walk(node, walker, null);
-}
-
-// Element
-
-pub const Element = c.lxb_dom_element_t;
-
-pub inline fn elementNode(element: *Element) *Node {
-    return c.lxb_dom_interface_node(element);
-}
-
-pub inline fn elementLocalName(element: *Element) []const u8 {
-    var size: usize = undefined;
-    const local_name = c.lxb_dom_element_local_name(element, &size);
-    return std.mem.sliceTo(local_name, 0);
-}
-
-pub inline fn elementsByAttr(
-    element: *Element,
-    collection: *Collection,
-    attr: []const u8,
-    value: []const u8,
-    case_sensitve: bool,
-) !void {
-    const status = c.lxb_dom_elements_by_attr(
-        element,
-        collection,
-        attr.ptr,
-        attr.len,
-        value.ptr,
-        value.len,
-        case_sensitve,
-    );
-    if (status != 0) {
-        return error.ElementsByAttr;
-    }
-}
-
-// DocumentHTML
-
-pub const DocumentHTML = c.lxb_html_document_t;
-
-pub inline fn documentHTMLInit() *DocumentHTML {
-    return c.lxb_html_document_create();
-}
-
-pub inline fn documentHTMLDeinit(document_html: *DocumentHTML) void {
-    _ = c.lxb_html_document_destroy(document_html);
-}
-
-pub inline fn documentHTMLParse(document_html: *DocumentHTML, html: []const u8) !void {
-    const status = c.lxb_html_document_parse(document_html, html.ptr, html.len - 1);
-    if (status != 0) {
-        return error.DocumentHTMLParse;
-    }
-}
-
-pub inline fn documentHTMLToNode(document_html: *DocumentHTML) *Node {
-    return c.lxb_dom_interface_node(document_html);
-}
-
-pub inline fn documentHTMLToDocument(document_html: *DocumentHTML) *Document {
-    return &document_html.dom_document;
-}
-
-pub inline fn documentHTMLBody(document_html: *DocumentHTML) *Body {
-    return document_html.body;
-}
-
-// Document
-
-pub const Document = c.lxb_dom_document_t;
-
-pub inline fn documentCreateElement(document: *Document, tag_name: []const u8) *Element {
-    return c.lxb_dom_document_create_element(document, tag_name.ptr, tag_name.len, null);
-}
-
-// Collection
-
-pub const Collection = c.lxb_dom_collection_t;
-
-pub inline fn collectionInit(document: *Document, size: usize) *Collection {
-    return c.lxb_dom_collection_make(document, size);
-}
-
-pub inline fn collectionDeinit(collection: *Collection) void {
-    _ = c.lxb_dom_collection_destroy(collection, true);
-}
-
-pub inline fn collectionElement(collection: *Collection, index: usize) *Element {
-    return c.lxb_dom_collection_element(collection, index);
-}
-
-// HTML Elements
-
-pub const HTMLElement = c.lxb_html_element_t;
-pub const MediaElement = c.lxb_html_media_element_t;
-
-pub const Unknown = c.lxb_html_unknown_element_t;
-pub const Anchor = c.lxb_html_anchor_element_t;
-pub const Area = c.lxb_html_area_element_t;
-pub const Audio = c.lxb_html_audio_element_t;
-pub const BR = c.lxb_html_br_element_t;
-pub const Base = c.lxb_html_base_element_t;
-pub const Body = c.lxb_html_body_element_t;
-pub const Button = c.lxb_html_button_element_t;
-pub const Canvas = c.lxb_html_canvas_element_t;
-pub const DList = c.lxb_html_d_list_element_t;
-pub const Data = c.lxb_html_data_element_t;
-pub const Dialog = c.lxb_html_dialog_element_t;
-pub const Div = c.lxb_html_div_element_t;
-pub const Embed = c.lxb_html_embed_element_t;
-pub const FieldSet = c.lxb_html_field_set_element_t;
-pub const Form = c.lxb_html_form_element_t;
-pub const FrameSet = c.lxb_html_frame_set_element_t;
-pub const HR = c.lxb_html_hr_element_t;
-pub const Head = c.lxb_html_head_element_t;
-pub const Heading = c.lxb_html_heading_element_t;
-pub const Html = c.lxb_html_html_element_t;
-pub const IFrame = c.lxb_html_iframe_element_t;
-pub const Image = c.lxb_html_image_element_t;
-pub const Input = c.lxb_html_input_element_t;
-pub const LI = c.lxb_html_li_element_t;
-pub const Label = c.lxb_html_label_element_t;
-pub const Legend = c.lxb_html_legend_element_t;
-pub const Link = c.lxb_html_link_element_t;
-pub const Map = c.lxb_html_map_element_t;
-pub const Meta = c.lxb_html_meta_element_t;
-pub const Meter = c.lxb_html_meter_element_t;
-pub const Mod = c.lxb_html_mod_element_t;
-pub const OList = c.lxb_html_o_list_element_t;
-pub const Object = c.lxb_html_object_element_t;
-pub const OptGroup = c.lxb_html_opt_group_element_t;
-pub const Option = c.lxb_html_option_element_t;
-pub const Output = c.lxb_html_output_element_t;
-pub const Paragraph = c.lxb_html_paragraph_element_t;
-pub const Picture = c.lxb_html_picture_element_t;
-pub const Pre = c.lxb_html_pre_element_t;
-pub const Progress = c.lxb_html_progress_element_t;
-pub const Quote = c.lxb_html_quote_element_t;
-pub const Script = c.lxb_html_script_element_t;
-pub const Select = c.lxb_html_select_element_t;
-pub const Source = c.lxb_html_source_element_t;
-pub const Span = c.lxb_html_span_element_t;
-pub const Style = c.lxb_html_style_element_t;
-pub const Table = c.lxb_html_table_element_t;
-pub const TableCaption = c.lxb_html_table_caption_element_t;
-pub const TableCell = c.lxb_html_table_cell_element_t;
-pub const TableCol = c.lxb_html_table_col_element_t;
-pub const TableRow = c.lxb_html_table_row_element_t;
-pub const TableSection = c.lxb_html_table_section_element_t;
-pub const Template = c.lxb_html_template_element_t;
-pub const TextArea = c.lxb_html_text_area_element_t;
-pub const Time = c.lxb_html_time_element_t;
-pub const Title = c.lxb_html_title_element_t;
-pub const Track = c.lxb_html_track_element_t;
-pub const UList = c.lxb_html_u_list_element_t;
-pub const Video = c.lxb_html_video_element_t;
-
-// Base
-
-pub const Action = c.lexbor_action_t;
-
-// TODO: use enum?
-pub const ActionStop = c.LEXBOR_ACTION_STOP;
-pub const ActionNext = c.LEXBOR_ACTION_NEXT;
-pub const ActionOk = c.LEXBOR_ACTION_OK;
-
-// Playground
-// ----------
-
-fn serialize_callback(_: [*c]const u8, _: usize, _: ?*anyopaque) callconv(.C) c_uint {
-    return 0;
-}
-
-fn walker_play(nn: ?*c.lxb_dom_node_t, _: ?*anyopaque) callconv(.C) c.lexbor_action_t {
-    if (nn == null) {
-        return c.LEXBOR_ACTION_STOP;
-    }
-    const n = nn.?;
-
-    var s: usize = undefined;
-    const name = c.lxb_dom_node_name(n, &s);
-
-    std.debug.print("type: {d}, name: {s}\n", .{ n.*.type, name });
-    if (n.*.local_name == c.LXB_TAG_A) {
-        const element = c.lxb_dom_interface_element(n);
-        const attr = element.*.first_attr;
-        std.debug.print("link, attr: {any}\n", .{attr.*.upper_name});
-    }
-    return c.LEXBOR_ACTION_OK;
-}
-
-pub fn parse_document() void {
-    const html = "<div><a href='foo'>OK</a><p>blah-blah-blah</p></div>";
-    const html_len = html.len - 1;
-
-    // parse
-    const doc = c.lxb_html_document_create();
-    const status_parse = c.lxb_html_document_parse(doc, html, html_len);
-    std.debug.print("status parse: {any}\n", .{status_parse});
-
-    // tree
-    const document_node = c.lxb_dom_interface_node(doc);
-    std.debug.print("document node is empty: {any}\n", .{c.lxb_dom_node_is_empty(document_node)});
-    std.debug.print("document node type: {any}\n", .{document_node.*.type});
-    std.debug.print("document node name: {any}\n", .{document_node.*.local_name});
-
-    c.lxb_dom_node_simple_walk(document_node, walker_play, null);
-
-    const first_child = c.lxb_dom_node_last_child(document_node);
-    if (first_child == null) {
-        std.debug.print("hummm is null\n", .{});
-    }
-    std.debug.print("first child type: {any}\n", .{first_child.*.type});
-    std.debug.print("first child name: {any}\n", .{first_child.*.local_name});
-
-    const tt = c.lxb_dom_node_first_child(first_child);
-    std.debug.print("tt type: {any}\n", .{tt.*.type});
-    std.debug.print("tt name: {any}\n", .{tt.*.local_name});
-    std.debug.print("{any}\n", .{c.LXB_DOM_NODE_TYPE_TEXT});
-
-    var s: usize = undefined;
-    const tt_name = c.lxb_dom_node_name(tt, &s);
-    std.debug.print("tt name: {s}\n", .{tt_name});
-
-    const nn = tt.*.first_child;
-    if (nn == null) {
-        std.debug.print("is null\n", .{});
-    }
-
-    // text
-    var text_len: usize = undefined;
-    var text = c.lxb_dom_node_text_content(tt, &text_len);
-    std.debug.print("size: {d}\n", .{text_len});
-    std.debug.print("text: {s}\n", .{text});
-
-    // serialize
-    const status_serialize = c.lxb_html_serialize_pretty_tree_cb(
-        document_node,
-        c.LXB_HTML_SERIALIZE_OPT_UNDEF,
-        0,
-        serialize_callback,
-        null,
-    );
-    std.debug.print("status serialize: {any}\n", .{status_serialize});
-
-    // destroy
-    _ = c.lxb_html_document_destroy(doc);
-    // _ = c.lxb_dom_document_destroy_text(first_child.*.owner_document, &text);
-    // _ = c.lxb_dom_document_destroy_text(c.lxb_dom_interface_document(document), text);
-    std.debug.print("text2: {s}\n", .{text}); // should not work
-}

src/main.zig

@@ -1,88 +1,307 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
+const builtin = @import("builtin");
 
 const jsruntime = @import("jsruntime");
 
-const parser = @import("netsurf.zig");
+const Browser = @import("browser/browser.zig").Browser;
+const server = @import("server.zig");
+
+const parser = @import("netsurf");
 const apiweb = @import("apiweb.zig");
 
 pub const Types = jsruntime.reflect(apiweb.Interfaces);
+pub const UserContext = apiweb.UserContext;
+pub const IO = @import("asyncio").Wrapper(jsruntime.Loop);
 
-const socket_path = "/tmp/browsercore-server.sock";
+// Simple blocking websocket connection model
+// ie. 1 thread per ws connection without thread pool and epoll/kqueue
+pub const websocket_blocking = true;
 
-var doc: *parser.DocumentHTML = undefined;
-var server: std.net.StreamServer = undefined;
+const log = std.log.scoped(.cli);
 
-fn execJS(
-    alloc: std.mem.Allocator,
-    js_env: *jsruntime.Env,
-) anyerror!void {
+pub const std_options = .{
+    // Set the log level to info
+    .log_level = .debug,
 
-    // start JS env
-    try js_env.start(alloc);
-    defer js_env.stop();
+    // Define logFn to override the std implementation
+    .logFn = logFn,
+};
 
-    // alias global as self and window
-    try js_env.attachObject(try js_env.getGlobal(), "self", null);
-    try js_env.attachObject(try js_env.getGlobal(), "window", null);
+const usage =
+    \\usage: {s} [options] [URL]
+    \\
+    \\  start Lightpanda browser
+    \\
+    \\  * if an url is provided the browser will fetch the page and exit
+    \\  * otherwhise the browser starts a CDP server
+    \\
+    \\  -h, --help      Print this help message and exit.
+    \\  --verbose       Display all logs. By default only info, warn and err levels are displayed.
+    \\  --host          Host of the CDP server (default "127.0.0.1")
+    \\  --port          Port of the CDP server (default "9222")
+    \\  --timeout       Timeout for incoming connections of the CDP server (in seconds, default "3")
+    \\  --dump          Dump document in stdout (fetch mode only)
+    \\
+;
 
-    // add document object
-    try js_env.addObject(doc, "document");
-
-    while (true) {
-
-        // read cmd
-        const conn = try server.accept();
-        var buf: [100]u8 = undefined;
-        const read = try conn.stream.read(&buf);
-        const cmd = buf[0..read];
-        std.debug.print("<- {s}\n", .{cmd});
-        if (std.mem.eql(u8, cmd, "exit")) {
-            break;
-        }
-
-        const res = try js_env.execTryCatch(alloc, cmd, "cdp");
-        if (res.success) {
-            std.debug.print("-> {s}\n", .{res.result});
-        }
-        _ = try conn.stream.write(res.result);
-    }
+fn printUsageExit(execname: []const u8, res: u8) anyerror {
+    std.io.getStdErr().writer().print(usage, .{execname}) catch |err| {
+        std.log.err("Print usage error: {any}", .{err});
+        return error.Cli;
+    };
+    if (res == 1) return error.Usage;
+    return error.NoError;
 }
 
+const CliModeTag = enum {
+    server,
+    fetch,
+};
+
+const CliMode = union(CliModeTag) {
+    server: Server,
+    fetch: Fetch,
+
+    const Server = struct {
+        execname: []const u8 = undefined,
+        args: *std.process.ArgIterator = undefined,
+        host: []const u8 = Host,
+        port: u16 = Port,
+        timeout: u8 = Timeout,
+
+        // default options
+        const Host = "127.0.0.1";
+        const Port = 9222;
+        const Timeout = 3; // in seconds
+    };
+
+    const Fetch = struct {
+        execname: []const u8 = undefined,
+        args: *std.process.ArgIterator = undefined,
+        url: []const u8 = "",
+        dump: bool = false,
+    };
+
+    fn init(alloc: std.mem.Allocator, args: *std.process.ArgIterator) !CliMode {
+        args.* = try std.process.argsWithAllocator(alloc);
+        errdefer args.deinit();
+
+        const execname = args.next().?;
+        var default_mode: CliModeTag = .server;
+
+        var _server = Server{};
+        var _fetch = Fetch{};
+
+        while (args.next()) |opt| {
+            if (std.mem.eql(u8, "-h", opt) or std.mem.eql(u8, "--help", opt)) {
+                return printUsageExit(execname, 0);
+            }
+            if (std.mem.eql(u8, "--verbose", opt)) {
+                verbose = true;
+                continue;
+            }
+            if (std.mem.eql(u8, "--dump", opt)) {
+                _fetch.dump = true;
+                continue;
+            }
+            if (std.mem.eql(u8, "--host", opt)) {
+                if (args.next()) |arg| {
+                    _server.host = arg;
+                    continue;
+                } else {
+                    std.log.err("--host not provided\n", .{});
+                    return printUsageExit(execname, 1);
+                }
+            }
+            if (std.mem.eql(u8, "--port", opt)) {
+                if (args.next()) |arg| {
+                    _server.port = std.fmt.parseInt(u16, arg, 10) catch |err| {
+                        log.err("--port {any}\n", .{err});
+                        return printUsageExit(execname, 1);
+                    };
+                    continue;
+                } else {
+                    log.err("--port not provided\n", .{});
+                    return printUsageExit(execname, 1);
+                }
+            }
+            if (std.mem.eql(u8, "--timeout", opt)) {
+                if (args.next()) |arg| {
+                    _server.timeout = std.fmt.parseInt(u8, arg, 10) catch |err| {
+                        log.err("--timeout {any}\n", .{err});
+                        return printUsageExit(execname, 1);
+                    };
+                    continue;
+                } else {
+                    log.err("--timeout not provided\n", .{});
+                    return printUsageExit(execname, 1);
+                }
+            }
+
+            // unknown option
+            if (std.mem.startsWith(u8, opt, "--")) {
+                log.err("unknown option\n", .{});
+                return printUsageExit(execname, 1);
+            }
+
+            // other argument is considered to be an URL, ie. fetch mode
+            default_mode = .fetch;
+
+            // allow only one url
+            if (_fetch.url.len != 0) {
+                log.err("more than 1 url provided\n", .{});
+                return printUsageExit(execname, 1);
+            }
+
+            _fetch.url = opt;
+        }
+
+        if (default_mode == .server) {
+
+            // server mode
+            _server.execname = execname;
+            _server.args = args;
+            return CliMode{ .server = _server };
+        } else {
+
+            // fetch mode
+            _fetch.execname = execname;
+            _fetch.args = args;
+            return CliMode{ .fetch = _fetch };
+        }
+    }
+
+    fn deinit(self: CliMode) void {
+        switch (self) {
+            inline .server, .fetch => |*_mode| {
+                _mode.args.deinit();
+            },
+        }
+    }
+};
+
 pub fn main() !void {
 
-    // create v8 vm
+    // allocator
+    // - in Debug mode we use the General Purpose Allocator to detect memory leaks
+    // - in Release mode we use the page allocator
+    var alloc: std.mem.Allocator = undefined;
+    var _gpa: ?std.heap.GeneralPurposeAllocator(.{}) = null;
+    if (builtin.mode == .Debug) {
+        _gpa = std.heap.GeneralPurposeAllocator(.{}){};
+        alloc = _gpa.?.allocator();
+    } else {
+        alloc = std.heap.page_allocator;
+    }
+    defer {
+        if (_gpa) |*gpa| {
+            switch (gpa.deinit()) {
+                .ok => std.debug.print("No memory leaks\n", .{}),
+                .leak => @panic("Memory leak"),
+            }
+        }
+    }
+
+    // args
+    var args: std.process.ArgIterator = undefined;
+    const cli_mode = CliMode.init(alloc, &args) catch |err| {
+        if (err == error.NoError) {
+            std.posix.exit(0);
+        } else {
+            std.posix.exit(1);
+        }
+        return;
+    };
+    defer cli_mode.deinit();
+
+    switch (cli_mode) {
+        .server => |opts| {
+            const address = std.net.Address.parseIp4(opts.host, opts.port) catch |err| {
+                log.err("address (host:port) {any}\n", .{err});
+                return printUsageExit(opts.execname, 1);
+            };
+
+            var loop = try jsruntime.Loop.init(alloc);
+            defer loop.deinit();
+
+            const timeout = std.time.ns_per_s * @as(u64, opts.timeout);
+            server.run(alloc, address, timeout, &loop) catch |err| {
+                log.err("Server error", .{});
+                return err;
+            };
+        },
+
+        .fetch => |opts| {
+            log.debug("Fetch mode: url {s}, dump {any}", .{ opts.url, opts.dump });
+
+            // vm
             const vm = jsruntime.VM.init();
             defer vm.deinit();
 
-    // alloc
-    var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
-    defer arena.deinit();
+            // loop
+            var loop = try jsruntime.Loop.init(alloc);
+            defer loop.deinit();
 
-    // document
-    const file = try std.fs.cwd().openFile("test.html", .{});
-    defer file.close();
+            // browser
+            var browser = Browser{};
+            try Browser.init(&browser, alloc, &loop, vm);
+            defer browser.deinit();
 
-    doc = try parser.documentHTMLParse(file.reader(), "UTF-8");
-    defer parser.documentHTMLClose(doc) catch |err| {
-        std.debug.print("documentHTMLClose error: {s}\n", .{@errorName(err)});
+            // page
+            const page = try browser.session.createPage();
+            try page.start(null);
+            defer page.end();
+
+            _ = page.navigate(opts.url, null) catch |err| switch (err) {
+                error.UnsupportedUriScheme, error.UriMissingHost => {
+                    log.err("'{s}' is not a valid URL ({any})\n", .{ opts.url, err });
+                    return printUsageExit(opts.execname, 1);
+                },
+                else => {
+                    log.err("'{s}' fetching error ({any})s\n", .{ opts.url, err });
+                    return printUsageExit(opts.execname, 1);
+                },
             };
 
-    // remove socket file of internal server
-    // reuse_address (SO_REUSEADDR flag) does not seems to work on unix socket
-    // see: https://gavv.net/articles/unix-socket-reuse/
-    // TODO: use a lock file instead
-    std.os.unlink(socket_path) catch |err| {
-        if (err != error.FileNotFound) {
-            return err;
+            try page.wait();
+
+            // dump
+            if (opts.dump) {
+                try page.dump(std.io.getStdOut());
+            }
+        },
     }
-    };
-
-    // server
-    const addr = try std.net.Address.initUnix(socket_path);
-    server = std.net.StreamServer.init(.{});
-    defer server.deinit();
-    try server.listen(addr);
-    std.debug.print("Listening on: {s}...\n", .{socket_path});
-
-    try jsruntime.loadEnv(&arena, execJS);
+}
+
+var verbose: bool = builtin.mode == .Debug; // In debug mode, force verbose.
+fn logFn(
+    comptime level: std.log.Level,
+    comptime scope: @Type(.EnumLiteral),
+    comptime format: []const u8,
+    args: anytype,
+) void {
+    if (!verbose) {
+        // hide all messages with level greater of equal to debug level.
+        if (@intFromEnum(level) >= @intFromEnum(std.log.Level.debug)) return;
+    }
+    // default std log function.
+    std.log.defaultLog(level, scope, format, args);
 }

src/main_get.zig

@@ -1,73 +0,0 @@
-const std = @import("std");
-const Browser = @import("browser/browser.zig").Browser;
-
-const jsruntime = @import("jsruntime");
-const apiweb = @import("apiweb.zig");
-pub const Types = jsruntime.reflect(apiweb.Interfaces);
-
-pub const std_options = struct {
-    pub const log_level = .debug;
-};
-
-const usage =
-    \\usage: {s} [options] <url>
-    \\  request the url with the browser
-    \\
-    \\  -h, --help      Print this help message and exit.
-    \\  --dump          Dump document in stdout
-    \\
-;
-
-pub fn main() !void {
-    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
-    defer {
-        const check = gpa.deinit();
-        if (check == .leak) {
-            std.log.warn("leaks detected\n", .{});
-        }
-    }
-    const allocator = gpa.allocator();
-
-    var args = try std.process.argsWithAllocator(allocator);
-    defer args.deinit();
-
-    const execname = args.next().?;
-    var url: []const u8 = "";
-    var dump: bool = false;
-
-    while (args.next()) |arg| {
-        if (std.mem.eql(u8, "-h", arg) or std.mem.eql(u8, "--help", arg)) {
-            try std.io.getStdErr().writer().print(usage, .{execname});
-            std.os.exit(0);
-        }
-        if (std.mem.eql(u8, "--dump", arg)) {
-            dump = true;
-            continue;
-        }
-        // allow only one url
-        if (url.len != 0) {
-            try std.io.getStdErr().writer().print(usage, .{execname});
-            std.os.exit(1);
-        }
-        url = arg;
-    }
-
-    if (url.len == 0) {
-        try std.io.getStdErr().writer().print(usage, .{execname});
-        std.os.exit(1);
-    }
-
-    const vm = jsruntime.VM.init();
-    defer vm.deinit();
-
-    var browser = try Browser.init(allocator, vm);
-    defer browser.deinit();
-
-    var page = try browser.currentSession().createPage();
-    defer page.end();
-    try page.navigate(url);
-
-    if (dump) {
-        try page.dump(std.io.getStdOut());
-    }
-}

src/main_shell.zig

@@ -1,13 +1,36 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
 
-const parser = @import("netsurf.zig");
+const parser = @import("netsurf");
 const apiweb = @import("apiweb.zig");
+const Window = @import("html/window.zig").Window;
+const storage = @import("storage/storage.zig");
+const Client = @import("asyncio").Client;
 
 const html_test = @import("html_test.zig").html;
 
 pub const Types = jsruntime.reflect(apiweb.Interfaces);
+pub const UserContext = apiweb.UserContext;
+pub const IO = @import("asyncio").Wrapper(jsruntime.Loop);
 
 var doc: *parser.DocumentHTML = undefined;
 
@@ -15,17 +38,26 @@ fn execJS(
     alloc: std.mem.Allocator,
     js_env: *jsruntime.Env,
 ) anyerror!void {
-
     // start JS env
-    try js_env.start(alloc);
+    try js_env.start();
     defer js_env.stop();
 
-    // alias global as self and window
-    try js_env.attachObject(try js_env.getGlobal(), "self", null);
-    try js_env.attachObject(try js_env.getGlobal(), "window", null);
+    var cli = Client{ .allocator = alloc };
+    defer cli.deinit();
 
-    // add document object
-    try js_env.addObject(doc, "document");
+    try js_env.setUserContext(UserContext{
+        .document = doc,
+        .httpClient = &cli,
+    });
+
+    var storageShelf = storage.Shelf.init(alloc);
+    defer storageShelf.deinit();
+
+    // alias global as self and window
+    var window = Window.create(null, null);
+    try window.replaceDocument(doc);
+    window.setStorageShelf(&storageShelf);
+    try js_env.bindGlobal(window);
 
     // launch shellExec
     try jsruntime.shellExec(alloc, js_env);
@@ -39,6 +71,9 @@ pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(gpa.allocator());
     defer arena.deinit();
 
+    try parser.init();
+    defer parser.deinit();
+
     // document
     const file = try std.fs.cwd().openFile("test.html", .{});
     defer file.close();
@@ -53,5 +88,5 @@ pub fn main() !void {
     defer vm.deinit();
 
     // launch shell
-    try jsruntime.shell(&arena, execJS, .{ .app_name = "browsercore" });
+    try jsruntime.shell(&arena, execJS, .{ .app_name = "lightpanda-shell" });
 }

src/main_tests.zig

@@ -0,0 +1,415 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const builtin = @import("builtin");
+
+const jsruntime = @import("jsruntime");
+const generate = @import("generate.zig");
+const pretty = @import("pretty");
+
+const parser = @import("netsurf");
+const apiweb = @import("apiweb.zig");
+const Window = @import("html/window.zig").Window;
+const xhr = @import("xhr/xhr.zig");
+const storage = @import("storage/storage.zig");
+const url = @import("url/url.zig");
+const URL = url.URL;
+const urlquery = @import("url/query.zig");
+const Client = @import("asyncio").Client;
+const Location = @import("html/location.zig").Location;
+
+const documentTestExecFn = @import("dom/document.zig").testExecFn;
+const HTMLDocumentTestExecFn = @import("html/document.zig").testExecFn;
+const nodeTestExecFn = @import("dom/node.zig").testExecFn;
+const characterDataTestExecFn = @import("dom/character_data.zig").testExecFn;
+const textTestExecFn = @import("dom/text.zig").testExecFn;
+const elementTestExecFn = @import("dom/element.zig").testExecFn;
+const HTMLCollectionTestExecFn = @import("dom/html_collection.zig").testExecFn;
+const DOMExceptionTestExecFn = @import("dom/exceptions.zig").testExecFn;
+const DOMImplementationExecFn = @import("dom/implementation.zig").testExecFn;
+const NamedNodeMapExecFn = @import("dom/namednodemap.zig").testExecFn;
+const DOMTokenListExecFn = @import("dom/token_list.zig").testExecFn;
+const NodeListTestExecFn = @import("dom/nodelist.zig").testExecFn;
+const AttrTestExecFn = @import("dom/attribute.zig").testExecFn;
+const EventTargetTestExecFn = @import("dom/event_target.zig").testExecFn;
+const ProcessingInstructionTestExecFn = @import("dom/processing_instruction.zig").testExecFn;
+const CommentTestExecFn = @import("dom/comment.zig").testExecFn;
+const DocumentFragmentTestExecFn = @import("dom/document_fragment.zig").testExecFn;
+const EventTestExecFn = @import("events/event.zig").testExecFn;
+const XHRTestExecFn = xhr.testExecFn;
+const ProgressEventTestExecFn = @import("xhr/progress_event.zig").testExecFn;
+const StorageTestExecFn = storage.testExecFn;
+const URLTestExecFn = url.testExecFn;
+const HTMLElementTestExecFn = @import("html/elements.zig").testExecFn;
+const MutationObserverTestExecFn = @import("dom/mutation_observer.zig").testExecFn;
+
+pub const Types = jsruntime.reflect(apiweb.Interfaces);
+pub const UserContext = @import("user_context.zig").UserContext;
+pub const IO = @import("asyncio").Wrapper(jsruntime.Loop);
+
+var doc: *parser.DocumentHTML = undefined;
+
+fn testExecFn(
+    alloc: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+    comptime execFn: jsruntime.ContextExecFn,
+) anyerror!void {
+    try parser.init();
+    defer parser.deinit();
+
+    // start JS env
+    try js_env.start();
+    defer js_env.stop();
+
+    var storageShelf = storage.Shelf.init(alloc);
+    defer storageShelf.deinit();
+
+    // document
+    const file = try std.fs.cwd().openFile("test.html", .{});
+    defer file.close();
+
+    doc = try parser.documentHTMLParse(file.reader(), "UTF-8");
+    defer parser.documentHTMLClose(doc) catch |err| {
+        std.debug.print("documentHTMLClose error: {s}\n", .{@errorName(err)});
+    };
+
+    var cli = Client{ .allocator = alloc };
+    defer cli.deinit();
+
+    try js_env.setUserContext(.{
+        .document = doc,
+        .httpClient = &cli,
+    });
+
+    // alias global as self and window
+    var window = Window.create(null, null);
+
+    var u = try URL.constructor(alloc, "https://lightpanda.io/opensource-browser/", null);
+    defer u.deinit(alloc);
+    var location = Location{ .url = &u };
+    try window.replaceLocation(&location);
+
+    try window.replaceDocument(doc);
+    window.setStorageShelf(&storageShelf);
+
+    try js_env.bindGlobal(window);
+
+    // run test
+    try execFn(alloc, js_env);
+}
+
+fn testsAllExecFn(
+    alloc: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    const testFns = [_]jsruntime.ContextExecFn{
+        documentTestExecFn,
+        HTMLDocumentTestExecFn,
+        nodeTestExecFn,
+        characterDataTestExecFn,
+        textTestExecFn,
+        elementTestExecFn,
+        HTMLCollectionTestExecFn,
+        DOMExceptionTestExecFn,
+        DOMImplementationExecFn,
+        NamedNodeMapExecFn,
+        DOMTokenListExecFn,
+        NodeListTestExecFn,
+        AttrTestExecFn,
+        CommentTestExecFn,
+        DocumentFragmentTestExecFn,
+        EventTargetTestExecFn,
+        EventTestExecFn,
+        XHRTestExecFn,
+        ProgressEventTestExecFn,
+        ProcessingInstructionTestExecFn,
+        StorageTestExecFn,
+        URLTestExecFn,
+        HTMLElementTestExecFn,
+        MutationObserverTestExecFn,
+        @import("polyfill/fetch.zig").testExecFn,
+        @import("html/navigator.zig").testExecFn,
+        @import("html/history.zig").testExecFn,
+        @import("html/location.zig").testExecFn,
+        @import("xmlserializer/xmlserializer.zig").testExecFn,
+    };
+
+    inline for (testFns) |testFn| {
+        try testExecFn(alloc, js_env, testFn);
+    }
+}
+
+const usage =
+    \\usage: test [options]
+    \\  Run the tests. By default the command will run both js and unit tests.
+    \\
+    \\  -h, --help       Print this help message and exit.
+    \\  --browser        run only browser js tests
+    \\  --unit           run only js unit tests
+    \\  --json           bench result is formatted in JSON.
+    \\                   only browser tests are benchmarked.
+    \\
+;
+
+// Out list all the ouputs handled by benchmark result and written on stdout.
+const Out = enum {
+    text,
+    json,
+};
+
+// Which tests must be run.
+const Run = enum {
+    all,
+    browser,
+    unit,
+};
+
+pub fn main() !void {
+    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
+    defer _ = gpa.deinit();
+    const gpa_alloc = gpa.allocator();
+
+    var args = try std.process.argsWithAllocator(gpa_alloc);
+    defer args.deinit();
+
+    // ignore the exec name.
+    _ = args.next().?;
+
+    var out: Out = .text;
+    var run: Run = .all;
+
+    while (args.next()) |arg| {
+        if (std.mem.eql(u8, "-h", arg) or std.mem.eql(u8, "--help", arg)) {
+            try std.io.getStdErr().writer().print(usage, .{});
+            std.posix.exit(0);
+        }
+        if (std.mem.eql(u8, "--json", arg)) {
+            out = .json;
+            continue;
+        }
+        if (std.mem.eql(u8, "--browser", arg)) {
+            run = .browser;
+            continue;
+        }
+        if (std.mem.eql(u8, "--unit", arg)) {
+            run = .unit;
+            continue;
+        }
+    }
+
+    // run js tests
+    if (run == .all or run == .browser) try run_js(out);
+
+    // run standard unit tests.
+    if (run == .all or run == .unit) {
+        std.debug.print("\n", .{});
+        for (builtin.test_functions) |test_fn| {
+            try parser.init();
+            defer parser.deinit();
+
+            std.testing.allocator_instance = .{};
+            try test_fn.func();
+
+            if (std.testing.allocator_instance.deinit() == .leak) {
+                std.debug.print("======Memory Leak: {s}======\n", .{test_fn.name});
+            } else {
+                std.debug.print("{s}\tOK\n", .{test_fn.name});
+            }
+        }
+    }
+}
+
+// Run js test and display the output depending of the output parameter.
+fn run_js(out: Out) !void {
+    var bench_alloc = jsruntime.bench_allocator(std.testing.allocator);
+
+    const start = try std.time.Instant.now();
+
+    // run js exectuion tests
+    try testJSRuntime(bench_alloc.allocator());
+
+    const duration = std.time.Instant.since(try std.time.Instant.now(), start);
+    const stats = bench_alloc.stats();
+
+    // get and display the results
+    if (out == .json) {
+        const res = [_]struct {
+            name: []const u8,
+            bench: struct {
+                duration: u64,
+
+                alloc_nb: usize,
+                realloc_nb: usize,
+                alloc_size: usize,
+            },
+        }{
+            .{ .name = "browser", .bench = .{
+                .duration = duration,
+                .alloc_nb = stats.alloc_nb,
+                .realloc_nb = stats.realloc_nb,
+                .alloc_size = stats.alloc_size,
+            } },
+            // TODO get libdom bench info.
+            .{ .name = "libdom", .bench = .{
+                .duration = duration,
+                .alloc_nb = 0,
+                .realloc_nb = 0,
+                .alloc_size = 0,
+            } },
+            // TODO get v8 bench info.
+            .{ .name = "v8", .bench = .{
+                .duration = duration,
+                .alloc_nb = 0,
+                .realloc_nb = 0,
+                .alloc_size = 0,
+            } },
+            // TODO get main bench info.
+            .{ .name = "main", .bench = .{
+                .duration = duration,
+                .alloc_nb = 0,
+                .realloc_nb = 0,
+                .alloc_size = 0,
+            } },
+        };
+
+        try std.json.stringify(res, .{ .whitespace = .indent_2 }, std.io.getStdOut().writer());
+        return;
+    }
+
+    // display console result by default
+    const dur = pretty.Measure{ .unit = "ms", .value = duration / ms };
+    const size = pretty.Measure{ .unit = "kb", .value = stats.alloc_size / kb };
+
+    const zerosize = pretty.Measure{ .unit = "kb", .value = 0 };
+
+    // benchmark table
+    const row_shape = .{ []const u8, pretty.Measure, u64, u64, pretty.Measure };
+    const table = try pretty.GenerateTable(4, row_shape, pretty.TableConf{ .margin_left = "  " });
+    const header = .{ "FUNCTION", "DURATION", "ALLOCATIONS (nb)", "RE-ALLOCATIONS (nb)", "HEAP SIZE" };
+    var t = table.init("Benchmark lightpanda 🚀", header);
+    try t.addRow(.{ "browser", dur, stats.alloc_nb, stats.realloc_nb, size });
+    try t.addRow(.{ "libdom", dur, 0, 0, zerosize }); // TODO get libdom bench info.
+    try t.addRow(.{ "v8", dur, 0, 0, zerosize }); // TODO get v8 bench info.
+    try t.addRow(.{ "main", dur, 0, 0, zerosize }); // TODO get main bench info.
+    try t.render(std.io.getStdOut().writer());
+}
+
+const kb = 1024;
+const ms = std.time.ns_per_ms;
+
+test {
+    const dumpTest = @import("browser/dump.zig");
+    std.testing.refAllDecls(dumpTest);
+
+    const mimeTest = @import("browser/mime.zig");
+    std.testing.refAllDecls(mimeTest);
+
+    const cssTest = @import("css/css.zig");
+    std.testing.refAllDecls(cssTest);
+
+    const cssParserTest = @import("css/parser.zig");
+    std.testing.refAllDecls(cssParserTest);
+
+    const cssMatchTest = @import("css/match_test.zig");
+    std.testing.refAllDecls(cssMatchTest);
+
+    const cssLibdomTest = @import("css/libdom_test.zig");
+    std.testing.refAllDecls(cssLibdomTest);
+
+    const queryTest = @import("url/query.zig");
+    std.testing.refAllDecls(queryTest);
+
+    std.testing.refAllDecls(@import("generate.zig"));
+    std.testing.refAllDecls(@import("cdp/msg.zig"));
+
+    // Don't use refAllDecls, as this will pull in the entire project
+    // and break the test build.
+    // We should fix this. See this branch & the commit message for details:
+    // https://github.com/karlseguin/browser/commit/193ab5ceab3d3758ea06db04f7690460d79eb79e
+    _ = @import("server.zig");
+}
+
+fn testJSRuntime(alloc: std.mem.Allocator) !void {
+    // create JS vm
+    const vm = jsruntime.VM.init();
+    defer vm.deinit();
+
+    var arena_alloc = std.heap.ArenaAllocator.init(alloc);
+    defer arena_alloc.deinit();
+
+    try jsruntime.loadEnv(&arena_alloc, null, testsAllExecFn);
+}
+
+test "DocumentHTMLParseFromStr" {
+    const file = try std.fs.cwd().openFile("test.html", .{});
+    defer file.close();
+
+    const str = try file.readToEndAlloc(std.testing.allocator, std.math.maxInt(u32));
+    defer std.testing.allocator.free(str);
+
+    doc = try parser.documentHTMLParseFromStr(str);
+    parser.documentHTMLClose(doc) catch {};
+}
+
+// https://github.com/lightpanda-io/libdom/issues/4
+test "bug document html parsing #4" {
+    const file = try std.fs.cwd().openFile("tests/html/bug-html-parsing-4.html", .{});
+    defer file.close();
+
+    doc = try parser.documentHTMLParse(file.reader(), "UTF-8");
+    parser.documentHTMLClose(doc) catch {};
+}
+
+test "Window is a libdom event target" {
+    var window = Window.create(null, null);
+
+    const event = try parser.eventCreate();
+    try parser.eventInit(event, "foo", .{});
+
+    const et = parser.toEventTarget(Window, &window);
+    _ = try parser.eventTargetDispatchEvent(et, event);
+}
+
+test "DocumentHTML is a libdom event target" {
+    doc = try parser.documentHTMLParseFromStr("<body></body>");
+    parser.documentHTMLClose(doc) catch {};
+
+    const event = try parser.eventCreate();
+    try parser.eventInit(event, "foo", .{});
+
+    const et = parser.toEventTarget(parser.DocumentHTML, doc);
+    _ = try parser.eventTargetDispatchEvent(et, event);
+}
+
+test "XMLHttpRequest.validMethod" {
+    // valid methods
+    for ([_][]const u8{ "get", "GET", "head", "HEAD" }) |tc| {
+        _ = try xhr.XMLHttpRequest.validMethod(tc);
+    }
+
+    // forbidden
+    for ([_][]const u8{ "connect", "CONNECT" }) |tc| {
+        try std.testing.expectError(parser.DOMError.Security, xhr.XMLHttpRequest.validMethod(tc));
+    }
+
+    // syntax
+    for ([_][]const u8{ "foo", "BAR" }) |tc| {
+        try std.testing.expectError(parser.DOMError.Syntax, xhr.XMLHttpRequest.validMethod(tc));
+    }
+}

src/main_wpt.zig

@@ -1,3 +1,21 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const jsruntime = @import("jsruntime");
@@ -30,11 +48,14 @@ const Out = enum {
 };
 
 pub const Types = jsruntime.reflect(apiweb.Interfaces);
+pub const GlobalType = apiweb.GlobalType;
+pub const UserContext = apiweb.UserContext;
+pub const IO = @import("asyncio").Wrapper(jsruntime.Loop);
 
 // TODO For now the WPT tests run is specific to WPT.
 // It manually load js framwork libs, and run the first script w/ js content in
 // the HTML page.
-// Once browsercore will have the html loader, it would be useful to refacto
+// Once lightpanda will have the html loader, it would be useful to refacto
 // this test to use it.
 pub fn main() !void {
     var gpa = std.heap.GeneralPurposeAllocator(.{}){};
@@ -56,7 +77,7 @@ pub fn main() !void {
     while (args.next()) |arg| {
         if (std.mem.eql(u8, "-h", arg) or std.mem.eql(u8, "--help", arg)) {
             try std.io.getStdErr().writer().print(usage, .{execname});
-            std.os.exit(0);
+            std.posix.exit(0);
         }
         if (std.mem.eql(u8, "--json", arg)) {
             out = .json;
@@ -122,7 +143,7 @@ pub fn main() !void {
         defer arena.deinit();
 
         const res = wpt.run(&arena, wpt_dir, tc, &loader) catch |err| {
-            const suite = try Suite.init(alloc, tc, false, @errorName(err), null);
+            const suite = try Suite.init(alloc, tc, false, @errorName(err));
             try results.append(suite);
 
             if (out == .text) {
@@ -131,9 +152,9 @@ pub fn main() !void {
             failures += 1;
             continue;
         };
-        // no need to call res.deinit() thanks to the arena allocator.
+        defer res.deinit(arena.allocator());
 
-        const suite = try Suite.init(alloc, tc, res.success, res.result, res.stack);
+        const suite = try Suite.init(alloc, tc, res.ok, res.msg orelse "");
         try results.append(suite);
 
         if (out == .json) {
@@ -176,7 +197,7 @@ pub fn main() !void {
                 try cases.append(Case{
                     .pass = suite.pass,
                     .name = suite.name,
-                    .message = suite.stack orelse suite.message,
+                    .message = suite.message,
                 });
             }
 
@@ -194,12 +215,12 @@ pub fn main() !void {
         }
 
         try std.json.stringify(output.items, .{ .whitespace = .indent_2 }, std.io.getStdOut().writer());
-        std.os.exit(0);
+        std.posix.exit(0);
     }
 
     if (out == .text and failures > 0) {
         std.debug.print("{d}/{d} tests suites failures\n", .{ failures, run });
-        std.os.exit(1);
+        std.posix.exit(1);
     }
 }
 
@@ -268,7 +289,7 @@ fn runSafe(
         argv.appendAssumeCapacity(tc);
         defer _ = argv.pop();
 
-        const run = try std.ChildProcess.run(.{
+        const run = try std.process.Child.run(.{
             .allocator = alloc,
             .argv = argv.items,
             .max_output_bytes = 1024 * 1024,
@@ -303,7 +324,7 @@ fn runSafe(
                     if (c.pass) pass += 1;
                 }
             }
-            const status = if (pass == all) "Pass" else "Fail";
+            const status = if (all > 0 and pass == all) "Pass" else "Fail";
             std.debug.print("{s} {d}/{d}", .{ status, pass, all });
 
             continue;
@@ -346,7 +367,8 @@ fn runSafe(
                 if (c.pass) pass += 1;
             }
         }
-        std.debug.print("{d}/{d}\n\n", .{ pass, all });
+        const status = if (all > 0 and pass == all) "Pass" else "Fail";
+        std.debug.print("{s} {d}/{d}\n\n", .{ status, pass, all });
     }
 
     if (out == .json) {

src/mimalloc/mimalloc.zig

@@ -0,0 +1,76 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+// This file makes the glue between mimalloc heap allocation and libdom memory
+// management.
+// We replace the libdom default usage of allocations with mimalloc heap
+// allocation to be able to free all memory used at once, like an arena usage.
+
+const std = @import("std");
+const c = @cImport({
+    @cInclude("mimalloc.h");
+});
+
+const Error = error{
+    HeapNotNull,
+    HeapNull,
+};
+
+var heap: ?*c.mi_heap_t = null;
+
+pub fn create() Error!void {
+    if (heap != null) return Error.HeapNotNull;
+    heap = c.mi_heap_new();
+    if (heap == null) return Error.HeapNull;
+}
+
+pub fn destroy() void {
+    if (heap == null) return;
+    c.mi_heap_destroy(heap.?);
+    heap = null;
+}
+
+pub export fn m_alloc(size: usize) callconv(.C) ?*anyopaque {
+    if (heap == null) return null;
+    return c.mi_heap_malloc(heap.?, size);
+}
+
+pub export fn re_alloc(ptr: ?*anyopaque, size: usize) callconv(.C) ?*anyopaque {
+    if (heap == null) return null;
+    return c.mi_heap_realloc(heap.?, ptr, size);
+}
+
+pub export fn c_alloc(nmemb: usize, size: usize) callconv(.C) ?*anyopaque {
+    if (heap == null) return null;
+    return c.mi_heap_calloc(heap.?, nmemb, size);
+}
+
+pub export fn str_dup(s: [*c]const u8) callconv(.C) [*c]u8 {
+    if (heap == null) return null;
+    return c.mi_heap_strdup(heap.?, s);
+}
+
+pub export fn strn_dup(s: [*c]const u8, size: usize) callconv(.C) [*c]u8 {
+    if (heap == null) return null;
+    return c.mi_heap_strndup(heap.?, s, size);
+}
+
+// NOOP, use destroy to clear all the memory allocated at once.
+pub export fn f_ree(_: ?*anyopaque) callconv(.C) void {
+    return;
+}

src/netsurf/netsurf.zig

@@ -1,14 +1,52 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 const std = @import("std");
 
 const c = @cImport({
     @cInclude("dom/dom.h");
+    @cInclude("core/pi.h");
     @cInclude("dom/bindings/hubbub/parser.h");
     @cInclude("events/event_target.h");
     @cInclude("events/event.h");
 });
 
+const mimalloc = @import("mimalloc");
+
 const Callback = @import("jsruntime").Callback;
-const EventToInterface = @import("events/event.zig").Event.toInterface;
+
+// init initializes netsurf lib.
+// init starts a mimalloc heap arena for the netsurf session. The caller must
+// call deinit() to free the arena memory.
+pub fn init() !void {
+    try mimalloc.create();
+}
+
+// deinit frees the mimalloc heap arena memory.
+// It also clean dom namespaces and lwc strings.
+pub fn deinit() void {
+    _ = c.dom_namespace_finalise();
+
+    // destroy all lwc strings.
+    c.lwc_deinit_strings();
+
+    mimalloc.destroy();
+}
 
 // Vtable
 // ------
@@ -223,8 +261,8 @@ pub const Tag = enum(u8) {
     pub fn all() []Tag {
         comptime {
             const info = @typeInfo(Tag).Enum;
-            comptime var l: [info.fields.len]Tag = undefined;
-            inline for (info.fields, 0..) |field, i| {
+            var l: [info.fields.len]Tag = undefined;
+            for (info.fields, 0..) |field, i| {
                 l[i] = @as(Tag, @enumFromInt(field.value));
             }
             return &l;
@@ -235,7 +273,7 @@ pub const Tag = enum(u8) {
         comptime {
             const tags = all();
             var names: [tags.len][]const u8 = undefined;
-            inline for (tags, 0..) |tag, i| {
+            for (tags, 0..) |tag, i| {
                 names[i] = tag.elementName();
             }
             return &names;
@@ -318,6 +356,12 @@ pub const DOMError = error{
     Timeout,
     InvalidNodeType,
     DataClone,
+
+    // custom netsurf error
+    UnspecifiedEventType,
+    DispatchRequest,
+    NoMemory,
+    AttributeWrongType,
 };
 
 const DOMException = c.dom_exception;
@@ -342,6 +386,13 @@ fn DOMErr(except: DOMException) DOMError!void {
         c.DOM_INVALID_ACCESS_ERR => DOMError.InvalidAccess,
         c.DOM_VALIDATION_ERR => DOMError.Validation,
         c.DOM_TYPE_MISMATCH_ERR => DOMError.TypeMismatch,
+
+        // custom netsurf error
+        c.DOM_UNSPECIFIED_EVENT_TYPE_ERR => DOMError.UnspecifiedEventType,
+        c.DOM_DISPATCH_REQUEST_ERR => DOMError.DispatchRequest,
+        c.DOM_NO_MEM_ERR => DOMError.NoMemory,
+        c.DOM_ATTR_WRONG_TYPE_ERR => DOMError.AttributeWrongType,
+
         else => unreachable,
     };
 }
@@ -376,6 +427,10 @@ pub fn eventType(evt: *Event) ![]const u8 {
     var s: ?*String = undefined;
     const err = c._dom_event_get_type(evt, &s);
     try DOMErr(err);
+
+    // if the event type is null, return a empty string.
+    if (s == null) return "";
+
     return strToData(s.?);
 }
 
@@ -467,30 +522,34 @@ pub const EventType = enum(u8) {
     progress_event = 1,
 };
 
-// EventHandler
-fn event_handler_cbk(data: *anyopaque) *Callback {
-    const ptr: *align(@alignOf(*Callback)) anyopaque = @alignCast(data);
-    return @as(*Callback, @ptrCast(ptr));
+pub const MutationEvent = c.dom_mutation_event;
+
+pub fn eventToMutationEvent(evt: *Event) *MutationEvent {
+    return @as(*MutationEvent, @ptrCast(evt));
 }
 
-const event_handler = struct {
-    fn handle(event: ?*Event, data: ?*anyopaque) callconv(.C) void {
-        if (data) |d| {
-            const func = event_handler_cbk(d);
+pub fn mutationEventAttributeName(evt: *MutationEvent) ![]const u8 {
+    var s: ?*String = undefined;
+    const err = c._dom_mutation_event_get_attr_name(evt, &s);
+    try DOMErr(err);
+    return strToData(s.?);
+}
 
-            if (event) |evt| {
-                func.call(.{
-                    EventToInterface(evt) catch unreachable,
-                }) catch unreachable;
-            } else {
-                func.call(.{event}) catch unreachable;
-            }
-            // NOTE: we can not call func.deinit here
-            // b/c the handler can be called several times
-            // either on this dispatch event or in anoter one
-        }
-    }
-}.handle;
+pub fn mutationEventPrevValue(evt: *MutationEvent) !?[]const u8 {
+    var s: ?*String = undefined;
+    const err = c._dom_mutation_event_get_prev_value(evt, &s);
+    try DOMErr(err);
+    if (s == null) return null;
+    return strToData(s.?);
+}
+
+pub fn mutationEventRelatedNode(evt: *MutationEvent) !?*Node {
+    var n: NodeExternal = undefined;
+    const err = c._dom_mutation_event_get_related_node(evt, &n);
+    try DOMErr(err);
+    if (n == null) return null;
+    return @as(*Node, @ptrCast(n));
+}
 
 // EventListener
 pub const EventListener = c.dom_event_listener;
@@ -503,6 +562,10 @@ fn eventListenerGetData(lst: *EventListener) ?*anyopaque {
 // EventTarget
 pub const EventTarget = c.dom_event_target;
 
+pub fn eventTargetToNode(et: *EventTarget) *Node {
+    return @as(*Node, @ptrCast(et));
+}
+
 fn eventTargetVtable(et: *EventTarget) c.dom_event_target_vtable {
     // retrieve the vtable
     const vtable = et.*.vtable.?;
@@ -551,10 +614,9 @@ pub fn eventTargetHasListener(
             // and capture property,
             // let's check if the callback handler is the same
             defer c.dom_event_listener_unref(listener);
-            const data = eventListenerGetData(listener);
-            if (data) |d| {
-                const cbk = event_handler_cbk(d);
-                if (cbk_id == cbk.id()) {
+            const ehd = EventHandlerDataInternal.fromListener(listener);
+            if (ehd) |d| {
+                if (cbk_id == d.data.cbk.id()) {
                     return lst;
                 }
             }
@@ -572,21 +634,99 @@ pub fn eventTargetHasListener(
     return null;
 }
 
+// EventHandlerFunc is a zig function called when the event is dispatched to a
+// listener.
+// The EventHandlerFunc is responsible to call the callback included into the
+// EventHandlerData.
+pub const EventHandlerFunc = *const fn (event: ?*Event, data: EventHandlerData) void;
+
+// EventHandler implements the function exposed in C and called by libdom.
+// It retrieves the EventHandlerInternalData and call the EventHandlerFunc with
+// the EventHandlerData in parameter.
+const EventHandler = struct {
+    fn handle(event: ?*Event, data: ?*anyopaque) callconv(.C) void {
+        if (data) |d| {
+            const ehd = EventHandlerDataInternal.get(d);
+            ehd.handler(event, ehd.data);
+
+            // NOTE: we can not call func.deinit here
+            // b/c the handler can be called several times
+            // either on this dispatch event or in anoter one
+        }
+    }
+}.handle;
+
+// EventHandlerData contains a JS callback and the data associated to the
+// handler.
+// If given, deinitFunc is called with the data pointer to allow the creator to
+// clean memory.
+// The callback is deinit by EventHandlerDataInternal. It must NOT be deinit
+// into deinitFunc.
+pub const EventHandlerData = struct {
+    cbk: Callback,
+    data: ?*anyopaque = null,
+    // deinitFunc implements the data deinitialization.
+    deinitFunc: ?DeinitFunc = null,
+
+    pub const DeinitFunc = *const fn (data: ?*anyopaque, alloc: std.mem.Allocator) void;
+};
+
+// EventHandlerDataInternal groups the EventHandlerFunc and the EventHandlerData.
+const EventHandlerDataInternal = struct {
+    data: EventHandlerData,
+    handler: EventHandlerFunc,
+
+    fn init(alloc: std.mem.Allocator, handler: EventHandlerFunc, data: EventHandlerData) !*EventHandlerDataInternal {
+        const ptr = try alloc.create(EventHandlerDataInternal);
+        ptr.* = .{
+            .data = data,
+            .handler = handler,
+        };
+        return ptr;
+    }
+
+    fn deinit(self: *EventHandlerDataInternal, alloc: std.mem.Allocator) void {
+        if (self.data.deinitFunc) |d| d(self.data.data, alloc);
+        self.data.cbk.deinit(alloc);
+        alloc.destroy(self);
+    }
+
+    fn get(data: *anyopaque) *EventHandlerDataInternal {
+        const ptr: *align(@alignOf(*EventHandlerDataInternal)) anyopaque = @alignCast(data);
+        return @as(*EventHandlerDataInternal, @ptrCast(ptr));
+    }
+
+    // retrieve a EventHandlerDataInternal from a listener.
+    fn fromListener(lst: *EventListener) ?*EventHandlerDataInternal {
+        const data = eventListenerGetData(lst);
+        // free cbk allocation made on eventTargetAddEventListener
+        if (data == null) return null;
+
+        return get(data.?);
+    }
+};
+
 pub fn eventTargetAddEventListener(
     et: *EventTarget,
     alloc: std.mem.Allocator,
     typ: []const u8,
-    cbk: Callback,
+    handlerFunc: EventHandlerFunc,
+    data: EventHandlerData,
     capture: bool,
 ) !void {
     // this allocation will be removed either on
     // eventTargetRemoveEventListener or eventTargetRemoveAllEventListeners
-    const cbk_ptr = try alloc.create(Callback);
-    cbk_ptr.* = cbk;
+    const ehd = try EventHandlerDataInternal.init(alloc, handlerFunc, data);
+    errdefer ehd.deinit(alloc);
 
-    const ctx = @as(*anyopaque, @ptrCast(cbk_ptr));
+    // When a function is used as an event handler, its this parameter is bound
+    // to the DOM element on which the listener is placed.
+    // https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/this#this_in_dom_event_handlers
+    try ehd.data.cbk.setThisArg(et);
+
+    const ctx = @as(*anyopaque, @ptrCast(ehd));
     var listener: ?*EventListener = undefined;
-    const errLst = c.dom_event_listener_create(event_handler, ctx, &listener);
+    const errLst = c.dom_event_listener_create(EventHandler, ctx, &listener);
     try DOMErr(errLst);
     defer c.dom_event_listener_unref(listener);
 
@@ -602,13 +742,9 @@ pub fn eventTargetRemoveEventListener(
     lst: *EventListener,
     capture: bool,
 ) !void {
-    const data = eventListenerGetData(lst);
-    // free cbk allocation made on eventTargetAddEventListener
-    if (data) |d| {
-        const cbk_ptr = event_handler_cbk(d);
-        cbk_ptr.deinit(alloc);
-        alloc.destroy(cbk_ptr);
-    }
+    // free data allocation made on eventTargetAddEventListener
+    const ehd = EventHandlerDataInternal.fromListener(lst);
+    if (ehd) |d| d.deinit(alloc);
 
     const s = try strFromData(typ);
     const err = eventTargetVtable(et).remove_event_listener.?(et, s, lst, capture);
@@ -636,13 +772,10 @@ pub fn eventTargetRemoveAllEventListeners(
 
         if (lst) |listener| {
             defer c.dom_event_listener_unref(listener);
-            const data = eventListenerGetData(listener);
-            if (data) |d| {
-                // free cbk allocation made on eventTargetAddEventListener
-                const cbk = event_handler_cbk(d);
-                cbk.deinit(alloc);
-                alloc.destroy(cbk);
-            }
+
+            const ehd = EventHandlerDataInternal.fromListener(listener);
+            if (ehd) |d| d.deinit(alloc);
+
             const err = eventTargetVtable(et).remove_event_listener.?(
                 et,
                 null,
@@ -875,6 +1008,7 @@ pub fn nodeLocalName(node: *Node) ![]const u8 {
     var s: ?*String = undefined;
     const err = nodeVtable(node).dom_node_get_local_name.?(node, &s);
     try DOMErr(err);
+    if (s == null) return "";
     var s_lower: ?*String = undefined;
     const errStr = c.dom_string_tolower(s, true, &s_lower);
     try DOMErr(errStr);
@@ -965,6 +1099,7 @@ pub fn nodeName(node: *Node) ![]const u8 {
     var s: ?*String = undefined;
     const err = nodeVtable(node).dom_node_get_node_name.?(node, &s);
     try DOMErr(err);
+    if (s == null) return "";
     return strToData(s.?);
 }
 
@@ -1235,6 +1370,18 @@ pub const Comment = c.dom_comment;
 // ProcessingInstruction
 pub const ProcessingInstruction = c.dom_processing_instruction;
 
+// processingInstructionToNode is an helper to convert an ProcessingInstruction to a node.
+pub inline fn processingInstructionToNode(pi: *ProcessingInstruction) *Node {
+    return @as(*Node, @ptrCast(pi));
+}
+
+pub fn processInstructionCopy(pi: *ProcessingInstruction) !*ProcessingInstruction {
+    var res: ?*Node = undefined;
+    const err = c._dom_pi_copy(processingInstructionToNode(pi), &res);
+    try DOMErr(err);
+    return @as(*ProcessingInstruction, @ptrCast(res.?));
+}
+
 // Attribute
 pub const Attribute = c.dom_attr;
 
@@ -1299,6 +1446,20 @@ pub fn elementGetAttribute(elem: *Element, name: []const u8) !?[]const u8 {
     return strToData(s.?);
 }
 
+pub fn elementGetAttributeNS(elem: *Element, ns: []const u8, name: []const u8) !?[]const u8 {
+    var s: ?*String = undefined;
+    const err = elementVtable(elem).dom_element_get_attribute_ns.?(
+        elem,
+        try strFromData(ns),
+        try strFromData(name),
+        &s,
+    );
+    try DOMErr(err);
+    if (s == null) return null;
+
+    return strToData(s.?);
+}
+
 pub fn elementSetAttribute(elem: *Element, qname: []const u8, value: []const u8) !void {
     const err = elementVtable(elem).dom_element_set_attribute.?(
         elem,
@@ -1308,11 +1469,35 @@ pub fn elementSetAttribute(elem: *Element, qname: []const u8, value: []const u8)
     try DOMErr(err);
 }
 
+pub fn elementSetAttributeNS(
+    elem: *Element,
+    ns: []const u8,
+    qname: []const u8,
+    value: []const u8,
+) !void {
+    const err = elementVtable(elem).dom_element_set_attribute_ns.?(
+        elem,
+        try strFromData(ns),
+        try strFromData(qname),
+        try strFromData(value),
+    );
+    try DOMErr(err);
+}
+
 pub fn elementRemoveAttribute(elem: *Element, qname: []const u8) !void {
     const err = elementVtable(elem).dom_element_remove_attribute.?(elem, try strFromData(qname));
     try DOMErr(err);
 }
 
+pub fn elementRemoveAttributeNS(elem: *Element, ns: []const u8, qname: []const u8) !void {
+    const err = elementVtable(elem).dom_element_remove_attribute_ns.?(
+        elem,
+        try strFromData(ns),
+        try strFromData(qname),
+    );
+    try DOMErr(err);
+}
+
 pub fn elementHasAttribute(elem: *Element, qname: []const u8) !bool {
     var res: bool = undefined;
     const err = elementVtable(elem).dom_element_has_attribute.?(elem, try strFromData(qname), &res);
@@ -1440,6 +1625,85 @@ pub fn elementHTMLGetTagType(elem_html: *ElementHTML) !Tag {
     return @as(Tag, @enumFromInt(tag_type));
 }
 
+// HTMLScriptElement
+
+// scriptToElt is an helper to convert an script to an element.
+pub inline fn scriptToElt(s: *Script) *Element {
+    return @as(*Element, @ptrCast(s));
+}
+
+// HTMLAnchorElement
+
+// anchorToNode is an helper to convert an anchor to a node.
+pub inline fn anchorToNode(a: *Anchor) *Node {
+    return @as(*Node, @ptrCast(a));
+}
+
+pub fn anchorGetTarget(a: *Anchor) ![]const u8 {
+    var res: ?*String = undefined;
+    const err = c.dom_html_anchor_element_get_target(a, &res);
+    try DOMErr(err);
+    if (res == null) return "";
+    return strToData(res.?);
+}
+
+pub fn anchorSetTarget(a: *Anchor, target: []const u8) !void {
+    const err = c.dom_html_anchor_element_set_target(a, try strFromData(target));
+    try DOMErr(err);
+}
+
+pub fn anchorGetHref(a: *Anchor) ![]const u8 {
+    var res: ?*String = undefined;
+    const err = c.dom_html_anchor_element_get_href(a, &res);
+    try DOMErr(err);
+    if (res == null) return "";
+    return strToData(res.?);
+}
+
+pub fn anchorSetHref(a: *Anchor, href: []const u8) !void {
+    const err = c.dom_html_anchor_element_set_href(a, try strFromData(href));
+    try DOMErr(err);
+}
+
+pub fn anchorGetHrefLang(a: *Anchor) ![]const u8 {
+    var res: ?*String = undefined;
+    const err = c.dom_html_anchor_element_get_hreflang(a, &res);
+    try DOMErr(err);
+    if (res == null) return "";
+    return strToData(res.?);
+}
+
+pub fn anchorSetHrefLang(a: *Anchor, href: []const u8) !void {
+    const err = c.dom_html_anchor_element_set_hreflang(a, try strFromData(href));
+    try DOMErr(err);
+}
+
+pub fn anchorGetType(a: *Anchor) ![]const u8 {
+    var res: ?*String = undefined;
+    const err = c.dom_html_anchor_element_get_type(a, &res);
+    try DOMErr(err);
+    if (res == null) return "";
+    return strToData(res.?);
+}
+
+pub fn anchorSetType(a: *Anchor, t: []const u8) !void {
+    const err = c.dom_html_anchor_element_set_type(a, try strFromData(t));
+    try DOMErr(err);
+}
+
+pub fn anchorGetRel(a: *Anchor) ![]const u8 {
+    var res: ?*String = undefined;
+    const err = c.dom_html_anchor_element_get_rel(a, &res);
+    try DOMErr(err);
+    if (res == null) return "";
+    return strToData(res.?);
+}
+
+pub fn anchorSetRel(a: *Anchor, rel: []const u8) !void {
+    const err = c.dom_html_anchor_element_set_rel(a, try strFromData(rel));
+    try DOMErr(err);
+}
+
 // ElementsHTML
 
 pub const MediaElement = struct { base: *c.dom_html_element };
@@ -1514,9 +1778,25 @@ pub const Video = struct { base: *c.dom_html_element };
 // Document Fragment
 pub const DocumentFragment = c.dom_document_fragment;
 
+pub inline fn documentFragmentToNode(doc: *DocumentFragment) *Node {
+    return @as(*Node, @ptrCast(doc));
+}
+
+pub fn documentFragmentBodyChildren(doc: *DocumentFragment) !?*NodeList {
+    const node = documentFragmentToNode(doc);
+    const html = try nodeFirstChild(node) orelse return null;
+    // TODO unref
+    const head = try nodeFirstChild(html) orelse return null;
+    // TODO unref
+    const body = try nodeNextSibling(head) orelse return null;
+    // TODO unref
+
+    return try nodeGetChildNodes(body);
+}
+
 // Document Position
 
-pub const DocumentPosition = enum(u2) {
+pub const DocumentPosition = enum(u32) {
     disconnected = c.DOM_DOCUMENT_POSITION_DISCONNECTED,
     preceding = c.DOM_DOCUMENT_POSITION_PRECEDING,
     following = c.DOM_DOCUMENT_POSITION_FOLLOWING,
@@ -1595,21 +1875,29 @@ pub inline fn domImplementationCreateDocumentType(
     return dt.?;
 }
 
-pub inline fn domImplementationCreateHTMLDocument(title: ?[]const u8) !*Document {
-    var doc: ?*Document = undefined;
-    const err = c.dom_implementation_create_document(
-        c.DOM_IMPLEMENTATION_HTML,
-        null,
-        null,
-        null,
-        null,
-        null,
-        &doc,
-    );
-    try DOMErr(err);
-    // TODO set title
-    _ = title;
-    return doc.?;
+pub inline fn domImplementationCreateHTMLDocument(title: ?[]const u8) !*DocumentHTML {
+    const doc_html = try documentCreateDocument(title);
+    const doc = documentHTMLToDocument(doc_html);
+
+    // add hierarchy: html, head, body.
+    const html = try documentCreateElement(doc, "html");
+    _ = try nodeAppendChild(documentToNode(doc), elementToNode(html));
+
+    const head = try documentCreateElement(doc, "head");
+    _ = try nodeAppendChild(elementToNode(html), elementToNode(head));
+
+    if (title) |t| {
+        try documentHTMLSetTitle(doc_html, t);
+        const htitle = try documentCreateElement(doc, "title");
+        const txt = try documentCreateTextNode(doc, t);
+        _ = try nodeAppendChild(elementToNode(htitle), @as(*Node, @ptrCast(txt)));
+        _ = try nodeAppendChild(elementToNode(head), elementToNode(htitle));
+    }
+
+    const body = try documentCreateElement(doc, "body");
+    _ = try nodeAppendChild(elementToNode(html), elementToNode(body));
+
+    return doc_html;
 }
 
 // Document
@@ -1653,6 +1941,11 @@ pub inline fn documentGetDocumentURI(doc: *Document) ![]const u8 {
     return strToData(s.?);
 }
 
+pub fn documentSetDocumentURI(doc: *Document, uri: []const u8) !void {
+    const err = documentVtable(doc).dom_document_set_uri.?(doc, try strFromData(uri));
+    try DOMErr(err);
+}
+
 pub inline fn documentGetInputEncoding(doc: *Document) ![]const u8 {
     var s: ?*String = undefined;
     const err = documentVtable(doc).dom_document_get_input_encoding.?(doc, &s);
@@ -1660,6 +1953,28 @@ pub inline fn documentGetInputEncoding(doc: *Document) ![]const u8 {
     return strToData(s.?);
 }
 
+pub inline fn documentSetInputEncoding(doc: *Document, enc: []const u8) !void {
+    const err = documentVtable(doc).dom_document_set_input_encoding.?(doc, try strFromData(enc));
+    try DOMErr(err);
+}
+
+pub inline fn documentCreateDocument(title: ?[]const u8) !*DocumentHTML {
+    var doc: ?*Document = undefined;
+    const err = c.dom_implementation_create_document(
+        c.DOM_IMPLEMENTATION_HTML,
+        null,
+        null,
+        null,
+        null,
+        null,
+        &doc,
+    );
+    try DOMErr(err);
+    const doc_html = @as(*DocumentHTML, @ptrCast(doc.?));
+    if (title) |t| try documentHTMLSetTitle(doc_html, t);
+    return doc_html;
+}
+
 pub inline fn documentCreateElement(doc: *Document, tag_name: []const u8) !*Element {
     var elem: ?*Element = undefined;
     const err = documentVtable(doc).dom_document_create_element.?(doc, try strFromData(tag_name), &elem);
@@ -1821,9 +2136,40 @@ pub fn documentHTMLParse(reader: anytype, enc: ?[:0]const u8) !*DocumentHTML {
     var parser: ?*c.dom_hubbub_parser = undefined;
     var doc: ?*c.dom_document = undefined;
     var err: c.hubbub_error = undefined;
+    var params = parseParams(enc);
 
-    var params = c.dom_hubbub_parser_params{
-        .enc = null,
+    err = c.dom_hubbub_parser_create(&params, &parser, &doc);
+    try parserErr(err);
+    defer c.dom_hubbub_parser_destroy(parser);
+
+    try parseData(parser.?, reader);
+
+    return @as(*DocumentHTML, @ptrCast(doc.?));
+}
+
+pub fn documentParseFragmentFromStr(self: *Document, str: []const u8) !*DocumentFragment {
+    var fbs = std.io.fixedBufferStream(str);
+    return try documentParseFragment(self, fbs.reader(), "UTF-8");
+}
+
+pub fn documentParseFragment(self: *Document, reader: anytype, enc: ?[:0]const u8) !*DocumentFragment {
+    var parser: ?*c.dom_hubbub_parser = undefined;
+    var fragment: ?*c.dom_document_fragment = undefined;
+    var err: c.hubbub_error = undefined;
+    var params = parseParams(enc);
+
+    err = c.dom_hubbub_fragment_parser_create(&params, self, &parser, &fragment);
+    try parserErr(err);
+    defer c.dom_hubbub_parser_destroy(parser);
+
+    try parseData(parser.?, reader);
+
+    return @as(*DocumentFragment, @ptrCast(fragment.?));
+}
+
+fn parseParams(enc: ?[:0]const u8) c.dom_hubbub_parser_params {
+    return .{
+        .enc = enc orelse null,
         .fix_enc = true,
         .msg = null,
         .script = null,
@@ -1831,13 +2177,10 @@ pub fn documentHTMLParse(reader: anytype, enc: ?[:0]const u8) !*DocumentHTML {
         .ctx = null,
         .daf = null,
     };
+}
 
-    if (enc) |e| params.enc = e;
-
-    err = c.dom_hubbub_parser_create(&params, &parser, &doc);
-    try parserErr(err);
-    defer c.dom_hubbub_parser_destroy(parser);
-
+fn parseData(parser: *c.dom_hubbub_parser, reader: anytype) !void {
+    var err: c.hubbub_error = undefined;
     var buffer: [1024]u8 = undefined;
     var ln = buffer.len;
     while (ln > 0) {
@@ -1855,8 +2198,6 @@ pub fn documentHTMLParse(reader: anytype, enc: ?[:0]const u8) !*DocumentHTML {
 
     err = c.dom_hubbub_parser_completed(parser);
     try parserErr(err);
-
-    return @as(*DocumentHTML, @ptrCast(doc.?));
 }
 
 // documentHTMLClose closes the document.
@@ -1910,3 +2251,35 @@ pub inline fn documentHTMLSetTitle(doc: *DocumentHTML, v: []const u8) !void {
     const err = documentHTMLVtable(doc).set_title.?(doc, try strFromData(v));
     try DOMErr(err);
 }
+
+pub fn documentHTMLSetCurrentScript(doc: *DocumentHTML, script: ?*Script) !void {
+    var s: ?*ElementHTML = null;
+    if (script != null) s = @ptrCast(script.?);
+    const err = documentHTMLVtable(doc).set_current_script.?(doc, s);
+    try DOMErr(err);
+}
+
+pub fn documentHTMLGetCurrentScript(doc: *DocumentHTML) !?*Script {
+    var elem: ?*ElementHTML = undefined;
+    const err = documentHTMLVtable(doc).get_current_script.?(doc, &elem);
+    try DOMErr(err);
+    if (elem == null) return null;
+    return @ptrCast(elem.?);
+}
+
+pub fn documentHTMLSetLocation(T: type, doc: *DocumentHTML, location: *T) !void {
+    const l = @as(*anyopaque, @ptrCast(location));
+    const err = documentHTMLVtable(doc).set_location.?(doc, l);
+    try DOMErr(err);
+}
+
+pub fn documentHTMLGetLocation(T: type, doc: *DocumentHTML) !?*T {
+    var l: ?*anyopaque = undefined;
+    const err = documentHTMLVtable(doc).get_location.?(doc, &l);
+    try DOMErr(err);
+
+    if (l == null) return null;
+
+    const ptr: *align(@alignOf(*T)) anyopaque = @alignCast(l.?);
+    return @as(*T, @ptrCast(ptr));
+}

src/polyfill/fetch.js

@@ -0,0 +1,671 @@
+// fetch.js code comes from
+// https://github.com/JakeChampion/fetch/blob/main/fetch.js
+//
+// The original code source is available in MIT license.
+//
+// The script comes from the built version from npm.
+// You can get the package with the command:
+//
+// wget $(npm view whatwg-fetch dist.tarball)
+//
+// The source is the content of `package/dist/fetch.umd.js` file.
+(function (global, factory) {
+  typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports) :
+  typeof define === 'function' && define.amd ? define(['exports'], factory) :
+  (factory((global.WHATWGFetch = {})));
+}(this, (function (exports) { 'use strict';
+
+  /* eslint-disable no-prototype-builtins */
+  var g =
+    (typeof globalThis !== 'undefined' && globalThis) ||
+    (typeof self !== 'undefined' && self) ||
+    // eslint-disable-next-line no-undef
+    (typeof global !== 'undefined' && global) ||
+    {};
+
+  var support = {
+    searchParams: 'URLSearchParams' in g,
+    iterable: 'Symbol' in g && 'iterator' in Symbol,
+    blob:
+      'FileReader' in g &&
+      'Blob' in g &&
+      (function() {
+        try {
+          new Blob();
+          return true
+        } catch (e) {
+          return false
+        }
+      })(),
+    formData: 'FormData' in g,
+
+    // Arraybuffer is available but xhr doesn't implement it for now.
+    // arrayBuffer: 'ArrayBuffer' in g
+    arrayBuffer: false
+  };
+
+  function isDataView(obj) {
+    return obj && DataView.prototype.isPrototypeOf(obj)
+  }
+
+  if (support.arrayBuffer) {
+    var viewClasses = [
+      '[object Int8Array]',
+      '[object Uint8Array]',
+      '[object Uint8ClampedArray]',
+      '[object Int16Array]',
+      '[object Uint16Array]',
+      '[object Int32Array]',
+      '[object Uint32Array]',
+      '[object Float32Array]',
+      '[object Float64Array]'
+    ];
+
+    var isArrayBufferView =
+      ArrayBuffer.isView ||
+      function(obj) {
+        return obj && viewClasses.indexOf(Object.prototype.toString.call(obj)) > -1
+      };
+  }
+
+  function normalizeName(name) {
+    if (typeof name !== 'string') {
+      name = String(name);
+    }
+    if (/[^a-z0-9\-#$%&'*+.^_`|~!]/i.test(name) || name === '') {
+      throw new TypeError('Invalid character in header field name: "' + name + '"')
+    }
+    return name.toLowerCase()
+  }
+
+  function normalizeValue(value) {
+    if (typeof value !== 'string') {
+      value = String(value);
+    }
+    return value
+  }
+
+  // Build a destructive iterator for the value list
+  function iteratorFor(items) {
+    var iterator = {
+      next: function() {
+        var value = items.shift();
+        return {done: value === undefined, value: value}
+      }
+    };
+
+    if (support.iterable) {
+      iterator[Symbol.iterator] = function() {
+        return iterator
+      };
+    }
+
+    return iterator
+  }
+
+  function Headers(headers) {
+    this.map = {};
+
+    if (headers instanceof Headers) {
+      headers.forEach(function(value, name) {
+        this.append(name, value);
+      }, this);
+    } else if (Array.isArray(headers)) {
+      headers.forEach(function(header) {
+        if (header.length != 2) {
+          throw new TypeError('Headers constructor: expected name/value pair to be length 2, found' + header.length)
+        }
+        this.append(header[0], header[1]);
+      }, this);
+    } else if (headers) {
+      Object.getOwnPropertyNames(headers).forEach(function(name) {
+        this.append(name, headers[name]);
+      }, this);
+    }
+  }
+
+  Headers.prototype.append = function(name, value) {
+    name = normalizeName(name);
+    value = normalizeValue(value);
+    var oldValue = this.map[name];
+    this.map[name] = oldValue ? oldValue + ', ' + value : value;
+  };
+
+  Headers.prototype['delete'] = function(name) {
+    delete this.map[normalizeName(name)];
+  };
+
+  Headers.prototype.get = function(name) {
+    name = normalizeName(name);
+    return this.has(name) ? this.map[name] : null
+  };
+
+  Headers.prototype.has = function(name) {
+    return this.map.hasOwnProperty(normalizeName(name))
+  };
+
+  Headers.prototype.set = function(name, value) {
+    this.map[normalizeName(name)] = normalizeValue(value);
+  };
+
+  Headers.prototype.forEach = function(callback, thisArg) {
+    for (var name in this.map) {
+      if (this.map.hasOwnProperty(name)) {
+        callback.call(thisArg, this.map[name], name, this);
+      }
+    }
+  };
+
+  Headers.prototype.keys = function() {
+    var items = [];
+    this.forEach(function(value, name) {
+      items.push(name);
+    });
+    return iteratorFor(items)
+  };
+
+  Headers.prototype.values = function() {
+    var items = [];
+    this.forEach(function(value) {
+      items.push(value);
+    });
+    return iteratorFor(items)
+  };
+
+  Headers.prototype.entries = function() {
+    var items = [];
+    this.forEach(function(value, name) {
+      items.push([name, value]);
+    });
+    return iteratorFor(items)
+  };
+
+  if (support.iterable) {
+    Headers.prototype[Symbol.iterator] = Headers.prototype.entries;
+  }
+
+  function consumed(body) {
+    if (body._noBody) return
+    if (body.bodyUsed) {
+      return Promise.reject(new TypeError('Already read'))
+    }
+    body.bodyUsed = true;
+  }
+
+  function fileReaderReady(reader) {
+    return new Promise(function(resolve, reject) {
+      reader.onload = function() {
+        resolve(reader.result);
+      };
+      reader.onerror = function() {
+        reject(reader.error);
+      };
+    })
+  }
+
+  function readBlobAsArrayBuffer(blob) {
+    var reader = new FileReader();
+    var promise = fileReaderReady(reader);
+    reader.readAsArrayBuffer(blob);
+    return promise
+  }
+
+  function readBlobAsText(blob) {
+    var reader = new FileReader();
+    var promise = fileReaderReady(reader);
+    var match = /charset=([A-Za-z0-9_-]+)/.exec(blob.type);
+    var encoding = match ? match[1] : 'utf-8';
+    reader.readAsText(blob, encoding);
+    return promise
+  }
+
+  function readArrayBufferAsText(buf) {
+    var view = new Uint8Array(buf);
+    var chars = new Array(view.length);
+
+    for (var i = 0; i < view.length; i++) {
+      chars[i] = String.fromCharCode(view[i]);
+    }
+    return chars.join('')
+  }
+
+  function bufferClone(buf) {
+    if (buf.slice) {
+      return buf.slice(0)
+    } else {
+      var view = new Uint8Array(buf.byteLength);
+      view.set(new Uint8Array(buf));
+      return view.buffer
+    }
+  }
+
+  function Body() {
+    this.bodyUsed = false;
+
+    this._initBody = function(body) {
+      /*
+        fetch-mock wraps the Response object in an ES6 Proxy to
+        provide useful test harness features such as flush. However, on
+        ES5 browsers without fetch or Proxy support pollyfills must be used;
+        the proxy-pollyfill is unable to proxy an attribute unless it exists
+        on the object before the Proxy is created. This change ensures
+        Response.bodyUsed exists on the instance, while maintaining the
+        semantic of setting Request.bodyUsed in the constructor before
+        _initBody is called.
+      */
+      // eslint-disable-next-line no-self-assign
+      this.bodyUsed = this.bodyUsed;
+      this._bodyInit = body;
+      if (!body) {
+        this._noBody = true;
+        this._bodyText = '';
+      } else if (typeof body === 'string') {
+        this._bodyText = body;
+      } else if (support.blob && Blob.prototype.isPrototypeOf(body)) {
+        this._bodyBlob = body;
+      } else if (support.formData && FormData.prototype.isPrototypeOf(body)) {
+        this._bodyFormData = body;
+      } else if (support.searchParams && URLSearchParams.prototype.isPrototypeOf(body)) {
+        this._bodyText = body.toString();
+      } else if (support.arrayBuffer && support.blob && isDataView(body)) {
+        this._bodyArrayBuffer = bufferClone(body.buffer);
+        // IE 10-11 can't handle a DataView body.
+        this._bodyInit = new Blob([this._bodyArrayBuffer]);
+      } else if (support.arrayBuffer && (ArrayBuffer.prototype.isPrototypeOf(body) || isArrayBufferView(body))) {
+        this._bodyArrayBuffer = bufferClone(body);
+      } else {
+        this._bodyText = body = Object.prototype.toString.call(body);
+      }
+
+      if (!this.headers.get('content-type')) {
+        if (typeof body === 'string') {
+          this.headers.set('content-type', 'text/plain;charset=UTF-8');
+        } else if (this._bodyBlob && this._bodyBlob.type) {
+          this.headers.set('content-type', this._bodyBlob.type);
+        } else if (support.searchParams && URLSearchParams.prototype.isPrototypeOf(body)) {
+          this.headers.set('content-type', 'application/x-www-form-urlencoded;charset=UTF-8');
+        }
+      }
+    };
+
+    if (support.blob) {
+      this.blob = function() {
+        var rejected = consumed(this);
+        if (rejected) {
+          return rejected
+        }
+
+        if (this._bodyBlob) {
+          return Promise.resolve(this._bodyBlob)
+        } else if (this._bodyArrayBuffer) {
+          return Promise.resolve(new Blob([this._bodyArrayBuffer]))
+        } else if (this._bodyFormData) {
+          throw new Error('could not read FormData body as blob')
+        } else {
+          return Promise.resolve(new Blob([this._bodyText]))
+        }
+      };
+    }
+
+    this.arrayBuffer = function() {
+      if (this._bodyArrayBuffer) {
+        var isConsumed = consumed(this);
+        if (isConsumed) {
+          return isConsumed
+        } else if (ArrayBuffer.isView(this._bodyArrayBuffer)) {
+          return Promise.resolve(
+            this._bodyArrayBuffer.buffer.slice(
+              this._bodyArrayBuffer.byteOffset,
+              this._bodyArrayBuffer.byteOffset + this._bodyArrayBuffer.byteLength
+            )
+          )
+        } else {
+          return Promise.resolve(this._bodyArrayBuffer)
+        }
+      } else if (support.blob) {
+        return this.blob().then(readBlobAsArrayBuffer)
+      } else {
+        throw new Error('could not read as ArrayBuffer')
+      }
+    };
+
+    this.text = function() {
+      var rejected = consumed(this);
+      if (rejected) {
+        return rejected
+      }
+
+      if (this._bodyBlob) {
+        return readBlobAsText(this._bodyBlob)
+      } else if (this._bodyArrayBuffer) {
+        return Promise.resolve(readArrayBufferAsText(this._bodyArrayBuffer))
+      } else if (this._bodyFormData) {
+        throw new Error('could not read FormData body as text')
+      } else {
+        return Promise.resolve(this._bodyText)
+      }
+    };
+
+    if (support.formData) {
+      this.formData = function() {
+        return this.text().then(decode)
+      };
+    }
+
+    this.json = function() {
+      return this.text().then(JSON.parse)
+    };
+
+    return this
+  }
+
+  // HTTP methods whose capitalization should be normalized
+  var methods = ['CONNECT', 'DELETE', 'GET', 'HEAD', 'OPTIONS', 'PATCH', 'POST', 'PUT', 'TRACE'];
+
+  function normalizeMethod(method) {
+    var upcased = method.toUpperCase();
+    return methods.indexOf(upcased) > -1 ? upcased : method
+  }
+
+  function Request(input, options) {
+    if (!(this instanceof Request)) {
+      throw new TypeError('Please use the "new" operator, this DOM object constructor cannot be called as a function.')
+    }
+
+    options = options || {};
+    var body = options.body;
+
+    if (input instanceof Request) {
+      if (input.bodyUsed) {
+        throw new TypeError('Already read')
+      }
+      this.url = input.url;
+      this.credentials = input.credentials;
+      if (!options.headers) {
+        this.headers = new Headers(input.headers);
+      }
+      this.method = input.method;
+      this.mode = input.mode;
+      this.signal = input.signal;
+      if (!body && input._bodyInit != null) {
+        body = input._bodyInit;
+        input.bodyUsed = true;
+      }
+    } else {
+      this.url = String(input);
+    }
+
+    this.credentials = options.credentials || this.credentials || 'same-origin';
+    if (options.headers || !this.headers) {
+      this.headers = new Headers(options.headers);
+    }
+    this.method = normalizeMethod(options.method || this.method || 'GET');
+    this.mode = options.mode || this.mode || null;
+    this.signal = options.signal || this.signal || (function () {
+      if ('AbortController' in g) {
+        var ctrl = new AbortController();
+        return ctrl.signal;
+      }
+    }());
+    this.referrer = null;
+
+    if ((this.method === 'GET' || this.method === 'HEAD') && body) {
+      throw new TypeError('Body not allowed for GET or HEAD requests')
+    }
+    this._initBody(body);
+
+    if (this.method === 'GET' || this.method === 'HEAD') {
+      if (options.cache === 'no-store' || options.cache === 'no-cache') {
+        // Search for a '_' parameter in the query string
+        var reParamSearch = /([?&])_=[^&]*/;
+        if (reParamSearch.test(this.url)) {
+          // If it already exists then set the value with the current time
+          this.url = this.url.replace(reParamSearch, '$1_=' + new Date().getTime());
+        } else {
+          // Otherwise add a new '_' parameter to the end with the current time
+          var reQueryString = /\?/;
+          this.url += (reQueryString.test(this.url) ? '&' : '?') + '_=' + new Date().getTime();
+        }
+      }
+    }
+  }
+
+  Request.prototype.clone = function() {
+    return new Request(this, {body: this._bodyInit})
+  };
+
+  function decode(body) {
+    var form = new FormData();
+    body
+      .trim()
+      .split('&')
+      .forEach(function(bytes) {
+        if (bytes) {
+          var split = bytes.split('=');
+          var name = split.shift().replace(/\+/g, ' ');
+          var value = split.join('=').replace(/\+/g, ' ');
+          form.append(decodeURIComponent(name), decodeURIComponent(value));
+        }
+      });
+    return form
+  }
+
+  function parseHeaders(rawHeaders) {
+    var headers = new Headers();
+    // Replace instances of \r\n and \n followed by at least one space or horizontal tab with a space
+    // https://tools.ietf.org/html/rfc7230#section-3.2
+    var preProcessedHeaders = rawHeaders.replace(/\r?\n[\t ]+/g, ' ');
+    // Avoiding split via regex to work around a common IE11 bug with the core-js 3.6.0 regex polyfill
+    // https://github.com/github/fetch/issues/748
+    // https://github.com/zloirock/core-js/issues/751
+    preProcessedHeaders
+      .split('\r')
+      .map(function(header) {
+        return header.indexOf('\n') === 0 ? header.substr(1, header.length) : header
+      })
+      .forEach(function(line) {
+        var parts = line.split(':');
+        var key = parts.shift().trim();
+        if (key) {
+          var value = parts.join(':').trim();
+          try {
+            headers.append(key, value);
+          } catch (error) {
+            console.warn('Response ' + error.message);
+          }
+        }
+      });
+    return headers
+  }
+
+  Body.call(Request.prototype);
+
+  function Response(bodyInit, options) {
+    if (!(this instanceof Response)) {
+      throw new TypeError('Please use the "new" operator, this DOM object constructor cannot be called as a function.')
+    }
+    if (!options) {
+      options = {};
+    }
+
+    this.type = 'default';
+    this.status = options.status === undefined ? 200 : options.status;
+    if (this.status < 200 || this.status > 599) {
+      throw new RangeError("Failed to construct 'Response': The status provided (0) is outside the range [200, 599].")
+    }
+    this.ok = this.status >= 200 && this.status < 300;
+    this.statusText = options.statusText === undefined ? '' : '' + options.statusText;
+    this.headers = new Headers(options.headers);
+    this.url = options.url || '';
+    this._initBody(bodyInit);
+  }
+
+  Body.call(Response.prototype);
+
+  Response.prototype.clone = function() {
+    return new Response(this._bodyInit, {
+      status: this.status,
+      statusText: this.statusText,
+      headers: new Headers(this.headers),
+      url: this.url
+    })
+  };
+
+  Response.error = function() {
+    var response = new Response(null, {status: 200, statusText: ''});
+    response.ok = false;
+    response.status = 0;
+    response.type = 'error';
+    return response
+  };
+
+  var redirectStatuses = [301, 302, 303, 307, 308];
+
+  Response.redirect = function(url, status) {
+    if (redirectStatuses.indexOf(status) === -1) {
+      throw new RangeError('Invalid status code')
+    }
+
+    return new Response(null, {status: status, headers: {location: url}})
+  };
+
+  exports.DOMException = g.DOMException;
+  try {
+    new exports.DOMException();
+  } catch (err) {
+    exports.DOMException = function(message, name) {
+      this.message = message;
+      this.name = name;
+      var error = Error(message);
+      this.stack = error.stack;
+    };
+    exports.DOMException.prototype = Object.create(Error.prototype);
+    exports.DOMException.prototype.constructor = exports.DOMException;
+  }
+
+  function fetch(input, init) {
+    return new Promise(function(resolve, reject) {
+      var request = new Request(input, init);
+
+      if (request.signal && request.signal.aborted) {
+        return reject(new exports.DOMException('Aborted', 'AbortError'))
+      }
+
+      var xhr = new XMLHttpRequest();
+
+      function abortXhr() {
+        xhr.abort();
+      }
+
+      xhr.onload = function() {
+        var options = {
+          statusText: xhr.statusText,
+          headers: parseHeaders(xhr.getAllResponseHeaders() || '')
+        };
+        // This check if specifically for when a user fetches a file locally from the file system
+        // Only if the status is out of a normal range
+        if (request.url.indexOf('file://') === 0 && (xhr.status < 200 || xhr.status > 599)) {
+          options.status = 200;
+        } else {
+          options.status = xhr.status;
+        }
+        options.url = 'responseURL' in xhr ? xhr.responseURL : options.headers.get('X-Request-URL');
+        var body = 'response' in xhr ? xhr.response : xhr.responseText;
+        setTimeout(function() {
+          resolve(new Response(body, options));
+        }, 0);
+      };
+
+      xhr.onerror = function() {
+        setTimeout(function() {
+          reject(new TypeError('Network request failed'));
+        }, 0);
+      };
+
+      xhr.ontimeout = function() {
+        setTimeout(function() {
+          reject(new TypeError('Network request timed out'));
+        }, 0);
+      };
+
+      xhr.onabort = function() {
+        setTimeout(function() {
+          reject(new exports.DOMException('Aborted', 'AbortError'));
+        }, 0);
+      };
+
+      function fixUrl(url) {
+        try {
+          return url === '' && g.location.href ? g.location.href : url
+        } catch (e) {
+          return url
+        }
+      }
+
+      xhr.open(request.method, fixUrl(request.url), true);
+
+      if (request.credentials === 'include') {
+        xhr.withCredentials = true;
+      } else if (request.credentials === 'omit') {
+        xhr.withCredentials = false;
+      }
+
+      if ('responseType' in xhr) {
+        if (support.blob) {
+          xhr.responseType = 'blob';
+        } else if (
+          support.arrayBuffer
+        ) {
+          xhr.responseType = 'arraybuffer';
+        }
+      }
+
+      if (init && typeof init.headers === 'object' && !(init.headers instanceof Headers || (g.Headers && init.headers instanceof g.Headers))) {
+        var names = [];
+        Object.getOwnPropertyNames(init.headers).forEach(function(name) {
+          names.push(normalizeName(name));
+          xhr.setRequestHeader(name, normalizeValue(init.headers[name]));
+        });
+        request.headers.forEach(function(value, name) {
+          if (names.indexOf(name) === -1) {
+            xhr.setRequestHeader(name, value);
+          }
+        });
+      } else {
+        request.headers.forEach(function(value, name) {
+          xhr.setRequestHeader(name, value);
+        });
+      }
+
+      if (request.signal) {
+        request.signal.addEventListener('abort', abortXhr);
+
+        xhr.onreadystatechange = function() {
+          // DONE (success or failure)
+          if (xhr.readyState === 4) {
+            request.signal.removeEventListener('abort', abortXhr);
+          }
+        };
+      }
+
+      xhr.send(typeof request._bodyInit === 'undefined' ? null : request._bodyInit);
+    })
+  }
+
+  fetch.polyfill = true;
+
+  if (!g.fetch) {
+    g.fetch = fetch;
+    g.Headers = Headers;
+    g.Request = Request;
+    g.Response = Response;
+  }
+
+  exports.Headers = Headers;
+  exports.Request = Request;
+  exports.Response = Response;
+  exports.fetch = fetch;
+
+  Object.defineProperty(exports, '__esModule', { value: true });
+
+})));

src/polyfill/fetch.zig

@@ -0,0 +1,55 @@
+const std = @import("std");
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+
+// fetch.js code comes from
+// https://github.com/JakeChampion/fetch/blob/main/fetch.js
+//
+// The original code source is available in MIT license.
+//
+// The script comes from the built version from npm.
+// You can get the package with the command:
+//
+// wget $(npm view whatwg-fetch dist.tarball)
+//
+// The source is the content of `package/dist/fetch.umd.js` file.
+pub const source = @embedFile("fetch.js");
+
+pub fn testExecFn(
+    alloc: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    try @import("polyfill.zig").load(alloc, js_env.*);
+
+    var fetch = [_]Case{
+        .{
+            .src =
+            \\var ok = false;
+            \\const request = new Request("https://httpbin.io/json");
+            \\fetch(request)
+            \\ .then((response) => { ok = response.ok; });
+            \\false;
+            ,
+            .ex = "false",
+        },
+        // all events have been resolved.
+        .{ .src = "ok", .ex = "true" },
+    };
+    try checkCases(js_env, &fetch);
+
+    var fetch2 = [_]Case{
+        .{
+            .src =
+            \\var ok2 = false;
+            \\const request2 = new Request("https://httpbin.io/json");
+            \\(async function () { resp = await fetch(request2); ok2 = resp.ok; }());
+            \\false;
+            ,
+            .ex = "false",
+        },
+        // all events have been resolved.
+        .{ .src = "ok2", .ex = "true" },
+    };
+    try checkCases(js_env, &fetch2);
+}

src/polyfill/polyfill.zig

@@ -0,0 +1,56 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const builtin = @import("builtin");
+
+const jsruntime = @import("jsruntime");
+const Env = jsruntime.Env;
+
+const fetch = @import("fetch.zig").fetch_polyfill;
+
+const log = std.log.scoped(.polyfill);
+
+const modules = [_]struct {
+    name: []const u8,
+    source: []const u8,
+}{
+    .{ .name = "polyfill-fetch", .source = @import("fetch.zig").source },
+};
+
+pub fn load(alloc: std.mem.Allocator, env: Env) !void {
+    var try_catch: jsruntime.TryCatch = undefined;
+    try_catch.init(env);
+    defer try_catch.deinit();
+
+    for (modules) |m| {
+        const res = env.exec(m.source, m.name) catch {
+            if (try try_catch.err(alloc, env)) |msg| {
+                defer alloc.free(msg);
+                log.err("load {s}: {s}", .{ m.name, msg });
+            }
+            return;
+        };
+
+        if (builtin.mode == .Debug) {
+            const msg = try res.toString(alloc, env);
+            defer alloc.free(msg);
+            log.debug("load {s}: {s}", .{ m.name, msg });
+        }
+    }
+}

src/run_tests.zig

@@ -1,186 +0,0 @@
-const std = @import("std");
-const builtin = @import("builtin");
-
-const jsruntime = @import("jsruntime");
-const generate = @import("generate.zig");
-
-const parser = @import("netsurf.zig");
-const apiweb = @import("apiweb.zig");
-const Window = @import("html/window.zig").Window;
-const xhr = @import("xhr/xhr.zig");
-
-const documentTestExecFn = @import("dom/document.zig").testExecFn;
-const HTMLDocumentTestExecFn = @import("html/document.zig").testExecFn;
-const nodeTestExecFn = @import("dom/node.zig").testExecFn;
-const characterDataTestExecFn = @import("dom/character_data.zig").testExecFn;
-const textTestExecFn = @import("dom/text.zig").testExecFn;
-const elementTestExecFn = @import("dom/element.zig").testExecFn;
-const HTMLCollectionTestExecFn = @import("dom/html_collection.zig").testExecFn;
-const DOMExceptionTestExecFn = @import("dom/exceptions.zig").testExecFn;
-const DOMImplementationExecFn = @import("dom/implementation.zig").testExecFn;
-const NamedNodeMapExecFn = @import("dom/namednodemap.zig").testExecFn;
-const DOMTokenListExecFn = @import("dom/token_list.zig").testExecFn;
-const NodeListTestExecFn = @import("dom/nodelist.zig").testExecFn;
-const AttrTestExecFn = @import("dom/attribute.zig").testExecFn;
-const EventTargetTestExecFn = @import("dom/event_target.zig").testExecFn;
-const EventTestExecFn = @import("events/event.zig").testExecFn;
-const XHRTestExecFn = xhr.testExecFn;
-const ProgressEventTestExecFn = @import("xhr/progress_event.zig").testExecFn;
-
-pub const Types = jsruntime.reflect(apiweb.Interfaces);
-
-var doc: *parser.DocumentHTML = undefined;
-
-fn testExecFn(
-    alloc: std.mem.Allocator,
-    js_env: *jsruntime.Env,
-    comptime execFn: jsruntime.ContextExecFn,
-) anyerror!void {
-
-    // start JS env
-    try js_env.start(alloc);
-    defer js_env.stop();
-
-    // alias global as self and window
-    try js_env.attachObject(try js_env.getGlobal(), "self", null);
-    try js_env.attachObject(try js_env.getGlobal(), "window", null);
-
-    // document
-    const file = try std.fs.cwd().openFile("test.html", .{});
-    defer file.close();
-
-    doc = try parser.documentHTMLParse(file.reader(), "UTF-8");
-    defer parser.documentHTMLClose(doc) catch |err| {
-        std.debug.print("documentHTMLClose error: {s}\n", .{@errorName(err)});
-    };
-
-    // add document object
-    try js_env.addObject(doc, "document");
-
-    // run test
-    try execFn(alloc, js_env);
-}
-
-fn testsAllExecFn(
-    alloc: std.mem.Allocator,
-    js_env: *jsruntime.Env,
-) anyerror!void {
-    const testFns = [_]jsruntime.ContextExecFn{
-        documentTestExecFn,
-        HTMLDocumentTestExecFn,
-        nodeTestExecFn,
-        characterDataTestExecFn,
-        textTestExecFn,
-        elementTestExecFn,
-        HTMLCollectionTestExecFn,
-        DOMExceptionTestExecFn,
-        DOMImplementationExecFn,
-        NamedNodeMapExecFn,
-        DOMTokenListExecFn,
-        NodeListTestExecFn,
-        AttrTestExecFn,
-        EventTargetTestExecFn,
-        EventTestExecFn,
-        XHRTestExecFn,
-        ProgressEventTestExecFn,
-    };
-
-    inline for (testFns) |testFn| {
-        try testExecFn(alloc, js_env, testFn);
-    }
-}
-
-pub fn main() !void {
-    std.debug.print("\n", .{});
-    for (builtin.test_functions) |test_fn| {
-        try test_fn.func();
-        std.debug.print("{s}\tOK\n", .{test_fn.name});
-    }
-}
-
-test {
-    const TestAsync = @import("async/test.zig");
-    std.testing.refAllDecls(TestAsync);
-}
-
-test "jsruntime" {
-    // generate tests
-    try generate.tests();
-
-    // create JS vm
-    const vm = jsruntime.VM.init();
-    defer vm.deinit();
-
-    var bench_alloc = jsruntime.bench_allocator(std.testing.allocator);
-    var arena_alloc = std.heap.ArenaAllocator.init(bench_alloc.allocator());
-    defer arena_alloc.deinit();
-
-    try jsruntime.loadEnv(&arena_alloc, testsAllExecFn);
-}
-
-test "DocumentHTMLParseFromStr" {
-    const file = try std.fs.cwd().openFile("test.html", .{});
-    defer file.close();
-
-    const str = try file.readToEndAlloc(std.testing.allocator, std.math.maxInt(u32));
-    defer std.testing.allocator.free(str);
-
-    doc = try parser.documentHTMLParseFromStr(str);
-    parser.documentHTMLClose(doc) catch {};
-}
-
-// https://github.com/lightpanda-io/libdom/issues/4
-test "bug document html parsing #4" {
-    const file = try std.fs.cwd().openFile("tests/html/bug-html-parsing-4.html", .{});
-    defer file.close();
-
-    doc = try parser.documentHTMLParse(file.reader(), "UTF-8");
-    parser.documentHTMLClose(doc) catch {};
-}
-
-const dump = @import("browser/dump.zig");
-test "run browser tests" {
-    // const out = std.io.getStdOut();
-    const out = try std.fs.openFileAbsolute("/dev/null", .{ .mode = .write_only });
-    defer out.close();
-
-    try dump.HTMLFileTestFn(out);
-}
-
-test "Window is a libdom event target" {
-    var window = Window.create(null);
-
-    const event = try parser.eventCreate();
-    try parser.eventInit(event, "foo", .{});
-
-    const et = parser.toEventTarget(Window, &window);
-    _ = try parser.eventTargetDispatchEvent(et, event);
-}
-
-test "DocumentHTML is a libdom event target" {
-    doc = try parser.documentHTMLParseFromStr("<body></body>");
-    parser.documentHTMLClose(doc) catch {};
-
-    const event = try parser.eventCreate();
-    try parser.eventInit(event, "foo", .{});
-
-    const et = parser.toEventTarget(parser.DocumentHTML, doc);
-    _ = try parser.eventTargetDispatchEvent(et, event);
-}
-
-test "XMLHttpRequest.validMethod" {
-    // valid methods
-    for ([_][]const u8{ "get", "GET", "head", "HEAD" }) |tc| {
-        _ = try xhr.XMLHttpRequest.validMethod(tc);
-    }
-
-    // forbidden
-    for ([_][]const u8{ "connect", "CONNECT" }) |tc| {
-        try std.testing.expectError(parser.DOMError.Security, xhr.XMLHttpRequest.validMethod(tc));
-    }
-
-    // syntax
-    for ([_][]const u8{ "foo", "BAR" }) |tc| {
-        try std.testing.expectError(parser.DOMError.Syntax, xhr.XMLHttpRequest.validMethod(tc));
-    }
-}

src/storage/storage.zig

@@ -0,0 +1,255 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+
+const jsruntime = @import("jsruntime");
+const Case = jsruntime.test_utils.Case;
+const checkCases = jsruntime.test_utils.checkCases;
+const DOMError = @import("netsurf").DOMError;
+
+const log = std.log.scoped(.storage);
+
+pub const Interfaces = .{
+    Bottle,
+};
+
+// See https://storage.spec.whatwg.org/#model for storage hierarchy.
+// A Shed contains map of Shelves. The key is the document's origin.
+// A Shelf contains on default Bucket (it could contain many in the future).
+// A Bucket contains a local and a session Bottle.
+// A Bottle stores a map of strings and is exposed to the JS.
+
+pub const Shed = struct {
+    const Map = std.StringHashMapUnmanaged(Shelf);
+
+    alloc: std.mem.Allocator,
+    map: Map,
+
+    pub fn init(alloc: std.mem.Allocator) Shed {
+        return .{
+            .alloc = alloc,
+            .map = .{},
+        };
+    }
+
+    pub fn deinit(self: *Shed) void {
+        // loop hover each KV and free the memory.
+        var it = self.map.iterator();
+        while (it.next()) |entry| {
+            entry.value_ptr.deinit();
+            self.alloc.free(entry.key_ptr.*);
+        }
+        self.map.deinit(self.alloc);
+    }
+
+    pub fn getOrPut(self: *Shed, origin: []const u8) !*Shelf {
+        const shelf = self.map.getPtr(origin);
+        if (shelf) |s| return s;
+
+        const oorigin = try self.alloc.dupe(u8, origin);
+        try self.map.put(self.alloc, oorigin, Shelf.init(self.alloc));
+        return self.map.getPtr(origin).?;
+    }
+};
+
+pub const Shelf = struct {
+    bucket: Bucket,
+
+    pub fn init(alloc: std.mem.Allocator) Shelf {
+        return .{ .bucket = Bucket.init(alloc) };
+    }
+
+    pub fn deinit(self: *Shelf) void {
+        self.bucket.deinit();
+    }
+};
+
+pub const Bucket = struct {
+    local: Bottle,
+    session: Bottle,
+
+    pub fn init(alloc: std.mem.Allocator) Bucket {
+        return .{
+            .local = Bottle.init(alloc),
+            .session = Bottle.init(alloc),
+        };
+    }
+
+    pub fn deinit(self: *Bucket) void {
+        self.local.deinit();
+        self.session.deinit();
+    }
+};
+
+// https://html.spec.whatwg.org/multipage/webstorage.html#the-storage-interface
+pub const Bottle = struct {
+    pub const mem_guarantied = true;
+    const Map = std.StringHashMapUnmanaged([]const u8);
+
+    // allocator is stored. we don't use the JS env allocator b/c the storage
+    // data could exists longer than a js env lifetime.
+    alloc: std.mem.Allocator,
+    map: Map,
+
+    pub fn init(alloc: std.mem.Allocator) Bottle {
+        return .{
+            .alloc = alloc,
+            .map = .{},
+        };
+    }
+
+    // loop hover each KV and free the memory.
+    fn free(self: *Bottle) void {
+        var it = self.map.iterator();
+        while (it.next()) |entry| {
+            self.alloc.free(entry.key_ptr.*);
+            self.alloc.free(entry.value_ptr.*);
+        }
+    }
+
+    pub fn deinit(self: *Bottle) void {
+        self.free();
+        self.map.deinit(self.alloc);
+    }
+
+    pub fn get_length(self: *Bottle) u32 {
+        return @intCast(self.map.count());
+    }
+
+    pub fn _key(self: *Bottle, idx: u32) ?[]const u8 {
+        if (idx >= self.map.count()) return null;
+
+        var it = self.map.valueIterator();
+        var i: u32 = 0;
+        while (it.next()) |v| {
+            if (i == idx) return v.*;
+            i += 1;
+        }
+        unreachable;
+    }
+
+    pub fn _getItem(self: *Bottle, k: []const u8) ?[]const u8 {
+        return self.map.get(k);
+    }
+
+    pub fn _setItem(self: *Bottle, k: []const u8, v: []const u8) !void {
+        const gop = self.map.getOrPut(self.alloc, k) catch |e| {
+            log.debug("set item: {any}", .{e});
+            return DOMError.QuotaExceeded;
+        };
+
+        if (gop.found_existing == false) {
+            gop.key_ptr.* = try self.alloc.dupe(u8, k);
+            gop.value_ptr.* = try self.alloc.dupe(u8, v);
+            return;
+        }
+
+        if (std.mem.eql(u8, v, gop.value_ptr.*) == false) {
+            self.alloc.free(gop.value_ptr.*);
+            gop.value_ptr.* = try self.alloc.dupe(u8, v);
+        }
+
+        // > Broadcast this with key, oldValue, and value.
+        // https://html.spec.whatwg.org/multipage/webstorage.html#the-storageevent-interface
+        //
+        // > The storage event of the Window interface fires when a storage
+        // > area (localStorage or sessionStorage) has been modified in the
+        // > context of another document.
+        // https://developer.mozilla.org/en-US/docs/Web/API/Window/storage_event
+        //
+        // So for now, we won't implement the feature.
+    }
+
+    pub fn _removeItem(self: *Bottle, k: []const u8) !void {
+        if (self.map.fetchRemove(k)) |kv| {
+            self.alloc.free(kv.key);
+            self.alloc.free(kv.value);
+        }
+
+        // > Broadcast this with key, oldValue, and null.
+        // https://html.spec.whatwg.org/multipage/webstorage.html#the-storageevent-interface
+        //
+        // > The storage event of the Window interface fires when a storage
+        // > area (localStorage or sessionStorage) has been modified in the
+        // > context of another document.
+        // https://developer.mozilla.org/en-US/docs/Web/API/Window/storage_event
+        //
+        // So for now, we won't impement the feature.
+    }
+
+    pub fn _clear(self: *Bottle) void {
+        self.free();
+        self.map.clearRetainingCapacity();
+
+        // > Broadcast this with null, null, and null.
+        // https://html.spec.whatwg.org/multipage/webstorage.html#the-storageevent-interface
+        //
+        // > The storage event of the Window interface fires when a storage
+        // > area (localStorage or sessionStorage) has been modified in the
+        // > context of another document.
+        // https://developer.mozilla.org/en-US/docs/Web/API/Window/storage_event
+        //
+        // So for now, we won't impement the feature.
+    }
+};
+
+// Tests
+// -----
+
+pub fn testExecFn(
+    _: std.mem.Allocator,
+    js_env: *jsruntime.Env,
+) anyerror!void {
+    var storage = [_]Case{
+        .{ .src = "localStorage.length", .ex = "0" },
+
+        .{ .src = "localStorage.setItem('foo', 'bar')", .ex = "undefined" },
+        .{ .src = "localStorage.length", .ex = "1" },
+        .{ .src = "localStorage.getItem('foo')", .ex = "bar" },
+        .{ .src = "localStorage.removeItem('foo')", .ex = "undefined" },
+        .{ .src = "localStorage.length", .ex = "0" },
+
+        // .{ .src = "localStorage['foo'] = 'bar'", .ex = "undefined" },
+        // .{ .src = "localStorage['foo']", .ex = "bar" },
+        // .{ .src = "localStorage.length", .ex = "1" },
+
+        .{ .src = "localStorage.clear()", .ex = "undefined" },
+        .{ .src = "localStorage.length", .ex = "0" },
+    };
+    try checkCases(js_env, &storage);
+}
+
+test "storage bottle" {
+    var bottle = Bottle.init(std.testing.allocator);
+    defer bottle.deinit();
+
+    try std.testing.expectEqual(0, bottle.get_length());
+    try std.testing.expectEqual(null, bottle._getItem("foo"));
+
+    try bottle._setItem("foo", "bar");
+    try std.testing.expectEqualStrings("bar", bottle._getItem("foo").?);
+
+    try bottle._setItem("foo", "other");
+    try std.testing.expectEqualStrings("other", bottle._getItem("foo").?);
+
+    try bottle._removeItem("foo");
+
+    try std.testing.expectEqual(0, bottle.get_length());
+    try std.testing.expectEqual(null, bottle._getItem("foo"));
+}

src/str/parser.zig

@@ -0,0 +1,125 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+// some utils to parser strings.
+const std = @import("std");
+
+pub const Reader = struct {
+    pos: usize = 0,
+    data: []const u8,
+
+    pub fn until(self: *Reader, c: u8) []const u8 {
+        const pos = self.pos;
+        const data = self.data;
+
+        const index = std.mem.indexOfScalarPos(u8, data, pos, c) orelse data.len;
+        self.pos = index;
+        return data[pos..index];
+    }
+
+    pub fn tail(self: *Reader) []const u8 {
+        const pos = self.pos;
+        const data = self.data;
+        if (pos > data.len) {
+            return "";
+        }
+        self.pos = data.len;
+        return data[pos..];
+    }
+
+    pub fn skip(self: *Reader) bool {
+        const pos = self.pos;
+        if (pos >= self.data.len) {
+            return false;
+        }
+        self.pos = pos + 1;
+        return true;
+    }
+};
+
+// converts a comptime-known string (i.e. null terminated) to an uint
+pub fn asUint(comptime string: anytype) AsUintReturn(string) {
+    const byteLength = @bitSizeOf(@TypeOf(string.*)) / 8 - 1;
+    const expectedType = *const [byteLength:0]u8;
+    if (@TypeOf(string) != expectedType) {
+        @compileError("expected : " ++ @typeName(expectedType) ++
+            ", got: " ++ @typeName(@TypeOf(string)));
+    }
+
+    return @bitCast(@as(*const [byteLength]u8, string).*);
+}
+
+fn AsUintReturn(comptime string: anytype) type {
+    return @Type(.{
+        .Int = .{
+            .bits = @bitSizeOf(@TypeOf(string.*)) - 8, // (- 8) to exclude sentinel 0
+            .signedness = .unsigned,
+        },
+    });
+}
+
+const testing = std.testing;
+test "parser.Reader: skip" {
+    var r = Reader{ .data = "foo" };
+    try testing.expectEqual(true, r.skip());
+    try testing.expectEqual(true, r.skip());
+    try testing.expectEqual(true, r.skip());
+    try testing.expectEqual(false, r.skip());
+    try testing.expectEqual(false, r.skip());
+}
+
+test "parser.Reader: tail" {
+    var r = Reader{ .data = "foo" };
+    try testing.expectEqualStrings("foo", r.tail());
+    try testing.expectEqualStrings("", r.tail());
+    try testing.expectEqualStrings("", r.tail());
+}
+
+test "parser.Reader: until" {
+    var r = Reader{ .data = "foo.bar.baz" };
+    try testing.expectEqualStrings("foo", r.until('.'));
+    _ = r.skip();
+    try testing.expectEqualStrings("bar", r.until('.'));
+    _ = r.skip();
+    try testing.expectEqualStrings("baz", r.until('.'));
+
+    r = Reader{ .data = "foo" };
+    try testing.expectEqualStrings("foo", r.until('.'));
+    try testing.expectEqualStrings("", r.tail());
+
+    r = Reader{ .data = "" };
+    try testing.expectEqualStrings("", r.until('.'));
+    try testing.expectEqualStrings("", r.tail());
+}
+
+test "parser: asUint" {
+    const ASCII_x = @as(u8, @bitCast([1]u8{'x'}));
+    const ASCII_ab = @as(u16, @bitCast([2]u8{ 'a', 'b' }));
+    const ASCII_xyz = @as(u24, @bitCast([3]u8{ 'x', 'y', 'z' }));
+    const ASCII_abcd = @as(u32, @bitCast([4]u8{ 'a', 'b', 'c', 'd' }));
+
+    try testing.expectEqual(ASCII_x, asUint("x"));
+    try testing.expectEqual(ASCII_ab, asUint("ab"));
+    try testing.expectEqual(ASCII_xyz, asUint("xyz"));
+    try testing.expectEqual(ASCII_abcd, asUint("abcd"));
+
+    try testing.expectEqual(u8, @TypeOf(asUint("x")));
+    try testing.expectEqual(u16, @TypeOf(asUint("ab")));
+    try testing.expectEqual(u24, @TypeOf(asUint("xyz")));
+    try testing.expectEqual(u32, @TypeOf(asUint("abcd")));
+}

src/test_runner.zig

@@ -1,9 +0,0 @@
-const std = @import("std");
-
-const tests = @import("run_tests.zig");
-
-pub const Types = tests.Types;
-
-pub fn main() !void {
-    try tests.main();
-}

src/unit_tests.zig

@@ -0,0 +1,347 @@
+// Copyright (C) 2023-2024  Lightpanda (Selecy SAS)
+//
+// Francis Bouvier <francis@lightpanda.io>
+// Pierre Tachoire <pierre@lightpanda.io>
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+const std = @import("std");
+const builtin = @import("builtin");
+
+const Allocator = std.mem.Allocator;
+
+pub const std_options = std.Options{
+    .http_disable_tls = true,
+};
+
+const BORDER = "=" ** 80;
+
+// use in custom panic handler
+var current_test: ?[]const u8 = null;
+
+pub fn main() !void {
+    var mem: [8192]u8 = undefined;
+    var fba = std.heap.FixedBufferAllocator.init(&mem);
+
+    const allocator = fba.allocator();
+
+    const env = Env.init(allocator);
+    defer env.deinit(allocator);
+
+    var slowest = SlowTracker.init(allocator, 5);
+    defer slowest.deinit();
+
+    var pass: usize = 0;
+    var fail: usize = 0;
+    var skip: usize = 0;
+    var leak: usize = 0;
+
+    const address = try std.net.Address.parseIp("127.0.0.1", 9582);
+    var listener = try address.listen(.{ .reuse_address = true });
+    defer listener.deinit();
+    const http_thread = try std.Thread.spawn(.{}, serverHTTP, .{&listener});
+    defer http_thread.join();
+
+    const printer = Printer.init();
+    printer.fmt("\r\x1b[0K", .{}); // beginning of line and clear to end of line
+
+    for (builtin.test_functions) |t| {
+        if (std.mem.eql(u8, t.name, "unit_tests.test_0")) {
+            // don't display anything for this test
+            try t.func();
+            continue;
+        }
+
+        var status = Status.pass;
+        slowest.startTiming();
+
+        const is_unnamed_test = isUnnamed(t);
+        if (env.filter) |f| {
+            if (!is_unnamed_test and std.mem.indexOf(u8, t.name, f) == null) {
+                continue;
+            }
+        }
+
+        const friendly_name = blk: {
+            const name = t.name;
+            var it = std.mem.splitScalar(u8, name, '.');
+            while (it.next()) |value| {
+                if (std.mem.eql(u8, value, "test")) {
+                    const rest = it.rest();
+                    break :blk if (rest.len > 0) rest else name;
+                }
+            }
+            break :blk name;
+        };
+
+        current_test = friendly_name;
+        std.testing.allocator_instance = .{};
+        const result = t.func();
+        current_test = null;
+
+        const ns_taken = slowest.endTiming(friendly_name);
+
+        if (std.testing.allocator_instance.deinit() == .leak) {
+            leak += 1;
+            printer.status(.fail, "\n{s}\n\"{s}\" - Memory Leak\n{s}\n", .{ BORDER, friendly_name, BORDER });
+        }
+
+        if (result) |_| {
+            pass += 1;
+        } else |err| switch (err) {
+            error.SkipZigTest => {
+                skip += 1;
+                status = .skip;
+            },
+            else => {
+                status = .fail;
+                fail += 1;
+                printer.status(.fail, "\n{s}\n\"{s}\" - {s}\n{s}\n", .{ BORDER, friendly_name, @errorName(err), BORDER });
+                if (@errorReturnTrace()) |trace| {
+                    std.debug.dumpStackTrace(trace.*);
+                }
+                if (env.fail_first) {
+                    break;
+                }
+            },
+        }
+
+        if (env.verbose) {
+            const ms = @as(f64, @floatFromInt(ns_taken)) / 1_000_000.0;
+            printer.status(status, "{s} ({d:.2}ms)\n", .{ friendly_name, ms });
+        } else {
+            printer.status(status, ".", .{});
+        }
+    }
+
+    const total_tests = pass + fail;
+    const status = if (fail == 0) Status.pass else Status.fail;
+    printer.status(status, "\n{d} of {d} test{s} passed\n", .{ pass, total_tests, if (total_tests != 1) "s" else "" });
+    if (skip > 0) {
+        printer.status(.skip, "{d} test{s} skipped\n", .{ skip, if (skip != 1) "s" else "" });
+    }
+    if (leak > 0) {
+        printer.status(.fail, "{d} test{s} leaked\n", .{ leak, if (leak != 1) "s" else "" });
+    }
+    printer.fmt("\n", .{});
+    try slowest.display(printer);
+    printer.fmt("\n", .{});
+    std.posix.exit(if (fail == 0) 0 else 1);
+}
+
+const Printer = struct {
+    out: std.fs.File.Writer,
+
+    fn init() Printer {
+        return .{
+            .out = std.io.getStdErr().writer(),
+        };
+    }
+
+    fn fmt(self: Printer, comptime format: []const u8, args: anytype) void {
+        std.fmt.format(self.out, format, args) catch unreachable;
+    }
+
+    fn status(self: Printer, s: Status, comptime format: []const u8, args: anytype) void {
+        const color = switch (s) {
+            .pass => "\x1b[32m",
+            .fail => "\x1b[31m",
+            .skip => "\x1b[33m",
+            else => "",
+        };
+        const out = self.out;
+        out.writeAll(color) catch @panic("writeAll failed?!");
+        std.fmt.format(out, format, args) catch @panic("std.fmt.format failed?!");
+        self.fmt("\x1b[0m", .{});
+    }
+};
+
+const Status = enum {
+    pass,
+    fail,
+    skip,
+    text,
+};
+
+const SlowTracker = struct {
+    const SlowestQueue = std.PriorityDequeue(TestInfo, void, compareTiming);
+    max: usize,
+    slowest: SlowestQueue,
+    timer: std.time.Timer,
+
+    fn init(allocator: Allocator, count: u32) SlowTracker {
+        const timer = std.time.Timer.start() catch @panic("failed to start timer");
+        var slowest = SlowestQueue.init(allocator, {});
+        slowest.ensureTotalCapacity(count) catch @panic("OOM");
+        return .{
+            .max = count,
+            .timer = timer,
+            .slowest = slowest,
+        };
+    }
+
+    const TestInfo = struct {
+        ns: u64,
+        name: []const u8,
+    };
+
+    fn deinit(self: SlowTracker) void {
+        self.slowest.deinit();
+    }
+
+    fn startTiming(self: *SlowTracker) void {
+        self.timer.reset();
+    }
+
+    fn endTiming(self: *SlowTracker, test_name: []const u8) u64 {
+        var timer = self.timer;
+        const ns = timer.lap();
+
+        var slowest = &self.slowest;
+
+        if (slowest.count() < self.max) {
+            // Capacity is fixed to the # of slow tests we want to track
+            // If we've tracked fewer tests than this capacity, than always add
+            slowest.add(TestInfo{ .ns = ns, .name = test_name }) catch @panic("failed to track test timing");
+            return ns;
+        }
+
+        {
+            // Optimization to avoid shifting the dequeue for the common case
+            // where the test isn't one of our slowest.
+            const fastest_of_the_slow = slowest.peekMin() orelse unreachable;
+            if (fastest_of_the_slow.ns > ns) {
+                // the test was faster than our fastest slow test, don't add
+                return ns;
+            }
+        }
+
+        // the previous fastest of our slow tests, has been pushed off.
+        _ = slowest.removeMin();
+        slowest.add(TestInfo{ .ns = ns, .name = test_name }) catch @panic("failed to track test timing");
+        return ns;
+    }
+
+    fn display(self: *SlowTracker, printer: Printer) !void {
+        var slowest = self.slowest;
+        const count = slowest.count();
+        printer.fmt("Slowest {d} test{s}: \n", .{ count, if (count != 1) "s" else "" });
+        while (slowest.removeMinOrNull()) |info| {
+            const ms = @as(f64, @floatFromInt(info.ns)) / 1_000_000.0;
+            printer.fmt("  {d:.2}ms\t{s}\n", .{ ms, info.name });
+        }
+    }
+
+    fn compareTiming(context: void, a: TestInfo, b: TestInfo) std.math.Order {
+        _ = context;
+        return std.math.order(a.ns, b.ns);
+    }
+};
+
+const Env = struct {
+    verbose: bool,
+    fail_first: bool,
+    filter: ?[]const u8,
+
+    fn init(allocator: Allocator) Env {
+        return .{
+            .verbose = readEnvBool(allocator, "TEST_VERBOSE", true),
+            .fail_first = readEnvBool(allocator, "TEST_FAIL_FIRST", false),
+            .filter = readEnv(allocator, "TEST_FILTER"),
+        };
+    }
+
+    fn deinit(self: Env, allocator: Allocator) void {
+        if (self.filter) |f| {
+            allocator.free(f);
+        }
+    }
+
+    fn readEnv(allocator: Allocator, key: []const u8) ?[]const u8 {
+        const v = std.process.getEnvVarOwned(allocator, key) catch |err| {
+            if (err == error.EnvironmentVariableNotFound) {
+                return null;
+            }
+            std.log.warn("failed to get env var {s} due to err {}", .{ key, err });
+            return null;
+        };
+        return v;
+    }
+
+    fn readEnvBool(allocator: Allocator, key: []const u8, deflt: bool) bool {
+        const value = readEnv(allocator, key) orelse return deflt;
+        defer allocator.free(value);
+        return std.ascii.eqlIgnoreCase(value, "true");
+    }
+};
+
+fn isUnnamed(t: std.builtin.TestFn) bool {
+    const marker = ".test_";
+    const test_name = t.name;
+    const index = std.mem.indexOf(u8, test_name, marker) orelse return false;
+    _ = std.fmt.parseInt(u32, test_name[index + marker.len ..], 10) catch return false;
+    return true;
+}
+
+fn serverHTTP(listener: *std.net.Server) !void {
+    var read_buffer: [1024]u8 = undefined;
+    ACCEPT: while (true) {
+        var conn = try listener.accept();
+        defer conn.stream.close();
+        var server = std.http.Server.init(conn, &read_buffer);
+
+        while (server.state == .ready) {
+            var request = server.receiveHead() catch |err| switch (err) {
+                error.HttpConnectionClosing => continue :ACCEPT,
+                else => {
+                    std.debug.print("Test HTTP Server error: {}\n", .{err});
+                    return err;
+                },
+            };
+
+            const path = request.head.target;
+            if (std.mem.eql(u8, path, "/loader")) {
+                try writeResponse(&request, .{
+                    .body = "Hello!",
+                });
+            }
+        }
+    }
+}
+
+const Response = struct {
+    body: []const u8 = "",
+    status: std.http.Status = .ok,
+};
+
+fn writeResponse(req: *std.http.Server.Request, res: Response) !void {
+    try req.respond(res.body, .{ .status = res.status });
+}
+
+test {
+    std.testing.refAllDecls(@import("url/query.zig"));
+    std.testing.refAllDecls(@import("browser/dump.zig"));
+    std.testing.refAllDecls(@import("browser/loader.zig"));
+    std.testing.refAllDecls(@import("browser/mime.zig"));
+    std.testing.refAllDecls(@import("cdp/msg.zig"));
+    std.testing.refAllDecls(@import("css/css.zig"));
+    std.testing.refAllDecls(@import("css/libdom_test.zig"));
+    std.testing.refAllDecls(@import("css/match_test.zig"));
+    std.testing.refAllDecls(@import("css/parser.zig"));
+    std.testing.refAllDecls(@import("generate.zig"));
+    std.testing.refAllDecls(@import("http/Client.zig"));
+    std.testing.refAllDecls(@import("storage/storage.zig"));
+    std.testing.refAllDecls(@import("iterator/iterator.zig"));
+    std.testing.refAllDecls(@import("server.zig"));
+}