This Pr largely tightens up a lot of the code. 'v8' is no longer imported
outside of js. A number of helper functions have been moved to the js.Context.
For example, js.Function.getName used to call:
```zig
return js.valueToString(allocator, name, self.context.isolate, self.context.v8_context);
```
It now calls:
```zig
return self.context.valueToString(name, .{ .allocator = allocator });
```
Page.main_context has been renamed to `Page.js`. This, in combination with new
promise helpers, turns:
```zig
const resolver = page.main_context.createPromiseResolver();
try resolver.resolve({});
return resolver.promise();
```
into:
```zig
return page.js.resolvePromise({});
```
Back in the zig-js-runtime days, globals were used for the state and webapi
declarations. This caused problems largely because it was done across
compilation units (using @import("root")...).
The generic Env(S, WebApi) was used to solve these problems, while still making
it work for different States and WebApis.
This change removes the generics and hard-codes the *Page as the state and
only supports our WebApis for the class declarations.
To accommodate this change, the runtime/*tests* have been removed. I don't
consider this a huge loss - whatever behavior these were testing, already
exists in the browser/**/*.zig web api.
As we write more complex/complete WebApis, we're seeing more and more cases
that need to rely on js objects directly (JsObject, Function, Promises, etc...).
The goal is to make these easier to use. Rather than using Env.JsObject, you
now import "js.zig" and use js.JsObject (TODO: rename JsObject to Object).
Everything is just a plain Zig struct, rather than being nested in a generic.
After this change, I plan on:
1 - Renaming the js objects, JsObject -> Object. These should be referenced in
the webapi as js.Object, js.This, ...
2 - Splitting the code across multiple files (Env.zig, Context.zig,
Caller.zig, ...)
There is some risk to this change. The first is that I made a mistake. The
other is that one of the APIs that doesn't currently return an error changes
in the future.
- Continue to reuse the Browser/Env/Isolate, but no start a new session per test.
- Test http server now properly closes the sendFile fd
- Run WPT in ReleaseMode
- Add --quiet option to WPT and some commented out debug code for dumping v8
memory stats
Further reducing bouncing between page and server for loop polling. If there is
a page, the page polls. If there isn't a page, the server polls. Simpler.
Depends on https://github.com/lightpanda-io/browser/pull/993
There's currently 3 ways to execute a page:
1 - page.navigate (as used in both the 'fetch' and 'serve' commands)
2 - jsRunner as used in unit tests
3 - main_wpt as used in the WPT runner
Both jsRunner and main_wpt replicate the page.navigate code, but in their own
hack-ish way. main_wpt re-implements the DOM walking in order to extract and
execute <script> tags, as well as the needed page lifecycle events.
This PR replaces the existing main_wpt loader with a call to page.navigate. To
support this, a test HTTP server was added. (The test HTTP server is extracted
from the existing unit test test server, and re-used between the two).
There are benefits to this approach:
1 - The code is simpler
2 - More of the actual code and flow is tested
3 - There's 1 way to do things (page.navigate)
4 - Having an HTTP server might unlock some WPT tests
Technically, we're replacing file IO with network IO i.e. http requests). This
has potential downsides:
1 - The tests might be more brittle
2 - The tests might be slower
I think we need to run it for a while to see if we get flaky behavior.
The goal for following PRs is to bring this unification to the jsRunner.
Removes optional platform, which only existed for tests.
There is now a global `@import("testing.zig").test_app` available. This is setup
when the test runner starts, and cleaned up at the end of tests. Individual
tests don't have to worry about creating app, which I assume was the reason I
Platform optional, since that woul dhave been something else that needed to be
setup.
Previously, the IO loop was doing three things:
1 - Managing timeouts (either from scripts or for our own needs)
2 - Handling browser IO events (page/script/xhr)
3 - Handling CDP events (accept, read, write, timeout)
With the libcurl merge, 1 was moved to an in-process scheduler and 2 was moved
to libcurl's own event loop. That means the entire loop code, including
the dependency on tigerbeetle-io existed for handling a single TCP client.
Not only is that a lot of code, there was also friction between the two loops
(the libcurl one and our IO loop), which would result in latency - while one
loop is waiting for the events, any events on the other loop go un-processed.
This PR removes our IO loop. To accomplish this:
1 - The main accept loop is blocking. This is simpler and works perfectly well,
given we only allow 1 active connection.
2 - The client socket is passed to libcurl - yes, libcurl's loop can take
arbitrary FDs and poll them along with its own.
In addition to having one less dependency, the CDP code is quite a bit simpler,
especially around shutdowns and writes. This also removes _some_ of the latency
caused by the friction between page process and CDP processing. Specifically,
when CDP now blocks for input, http page events (script loading, xhr, ...) will
still be processed.
There's still friction. For one, the reverse isn't true: when the page is
waiting for events, CDP events aren't going to be processed. But the page.wait
already have some sensitivity to this (e.g. the page.request_intercepted flag).
Also, when CDP waits, while we will process network events, page timeouts are
still not processed. Because of both these remaining issues, we still need to
jump between the two loops - but being able to block on CDP (even for a short
time) WITHOUT stopping the page's network I/O, should reduce some latency.
Minor improvement to correctness of TreeWalker.
Fun fact, this is the first time, that I've run into, where we have to default
null and undefined to different values.
Also, tweaked the WPT test runner. WPT test results use | as a field delimiter.
But a WPT test (and, I assume a message) can contain a |. So we had at least
a few tests that were being reported as failed, only because the result line
was weird / unexpected. No great robust way to parse this, but I changed it
to look explicitly for |Pass or |Fail and use those positions as anchor points.
This is always run, but only the full webcomponents polyfill, it's very
small and isn't intrusive. This introduces a layer of indirection so that,
if the full polyfill is loaded, its monkeypatched constructor will be called
When the browser microtask was added, zig-specific timeout functions were
added to the loop. This was necessary for two reasons:
1 - The existing functions were JS specific
2 - We wanted a different reset counter for JS and Zig
Like we did in https://github.com/lightpanda-io/browser/pull/577, the loop is
now JS-agnostic. It gets a Zig callback, and the Zig callback can execute JS
(or do whatever). An intrusive node, like with events, is used to minimize
allocations.
Also, because the microtask was recently moved to the page, there is no longer
a need for separate event counters. All timeouts are scoped to the page.
The new timeout callback can now be used to efficiently reschedule a task. This
reuses the IO.completion and Context, avoiding 2 allocations. More importantly
it makes the internal timer_id static for the lifetime of an "interval". This
is important for window.setInterval, where the callback can itself clear the
interval, which we would need to detect in the callback handler to avoid
re-scheduling. With the stable timer_id, the existing cancel mechanism works
as expected.
The loop no longer has a cbk_error. Callback code is expected to try/catch
callbacks (or use callback.tryCall) and handle errors accordingly.
* Rework WPT runner
We have no crashing tests, remove safe mode. Allows better re-use of arenas,
and if we do introduce a crash, it won't be easy to ignore. Could allow for
re-using the environment across tests to further improve performance.
Remove console now that we have a working console api.
* Update workflows, add summary
Remove --safe option from WPT workflows (it's no longer valid)
Include a total test/case summary when --summary or --text (default) is used.
* remove wpt --safe flag from Makefile
* handle tests in the root of the test folder
* Fix a couple possible segfaults base on strange usage (WPT stuff)
* generate proper JSON
* generate proper JSON (for real this time?)
* fix tag type check
Currently, we treat every .html file in tests/wpt as-if it's a test. That
isn't always the case. wpt has a manifest tool to generate a manifest (in
JSON) of the tests, we should probably use that (but it's quite large).
This PR filters out two groups. First, everything in resources/ appears to
be things to _run_ the tests, not actual tests.
Second, any file without a "testharness.js" doesn't appear to be a test
either. Note that WPT's own manifest generator looks at this:
43a0615361/tools/manifest/sourcefile.py (L676)
(which is used later in the file to determine the type of file).
