store charset value directly in Mime

src/browser/mime.zig

@@ -22,13 +22,15 @@ const Allocator = std.mem.Allocator;
 pub const Mime = struct {
     content_type: ContentType,
     params: []const u8 = "",
-    charset: ?[:0]const u8 = null,
+    // IANA defines max. charset value length as 40.
+    // We keep 41 for null-termination since HTML parser expects in this format.
+    charset: [41]u8 = default_charset,
 
-    pub const unknown = Mime{
+    /// String "UTF-8" continued by null characters.
-        .params = "",
+    pub const default_charset = .{ 'U', 'T', 'F', '-', '8' } ++ .{0} ** 36;
-        .charset = null,
+
-        .content_type = .{ .unknown = {} },
+    /// Mime with unknown Content-Type, empty params and empty charset.
-    };
+    pub const unknown = Mime{ .content_type = .{ .unknown = {} } };
 
     pub const ContentTypeEnum = enum {
         text_xml,
@@ -52,6 +54,11 @@ pub const Mime = struct {
         other: struct { type: []const u8, sub_type: []const u8 },
     };
 
+    /// Returns the null-terminated charset value.
+    pub inline fn charsetString(mime: *const Mime) [:0]const u8 {
+        return @ptrCast(&mime.charset);
+    }
+
     /// Removes quotes of value if quotes are given.
     ///
     /// Currently we don't validate the charset.
@@ -158,7 +165,7 @@ pub const Mime = struct {
 
         const params = trimLeft(normalized[type_len..]);
 
-        var charset: ?[:0]const u8 = null;
+        var charset: [41]u8 = undefined;
 
         var it = std.mem.splitScalar(u8, params, ';');
         while (it.next()) |attr| {
@@ -176,35 +183,14 @@ pub const Mime = struct {
 
             switch (attribute_name) {
                 .charset => {
-                    // We used to have a proper value parser, but we currently
-                    // only care about the charset attribute, plus only about
-                    // the UTF-8 value. It's a lot easier to do it this way,
-                    // and it doesn't require an allocation to (a) unescape the
-                    // value or (b) ensure the correct lifetime.
                     if (value.len == 0) {
                         break;
                     }
-                    var attribute_value = value;
-                    if (value[0] == '"') {
-                        if (value.len < 3 or value[value.len - 1] != '"') {
-                            return error.Invalid;
-                        }
-                        attribute_value = value[1 .. value.len - 1];
-                    }
 
-                    if (std.ascii.eqlIgnoreCase(attribute_value, "utf-8")) {
+                    const attribute_value = try parseCharset(value);
-                        charset = "UTF-8";
+                    @memcpy(charset[0..attribute_value.len], attribute_value);
-                    } else if (std.ascii.eqlIgnoreCase(attribute_value, "iso-8859-1")) {
+                    // Fill the rest with zeroes.
-                        charset = "ISO-8859-1";
+                    @memset(charset[attribute_value.len..], 0);
-                    } else {
-                        // we only care about null (which we default to UTF-8)
-                        // or UTF-8. If this is actually set (i.e. not null)
-                        // and isn't UTF-8, we'll just put a dummy value. If
-                        // we want to capture the actual value, we'll need to
-                        // dupe/allocate it. Since, for now, we don't need that
-                        // we can avoid the allocation.
-                        charset = "lightpanda:UNSUPPORTED";
-                    }
                 },
             }
         }

src/browser/page.zig

@@ -672,14 +672,14 @@ pub const Page = struct {
             log.debug(.http, "navigate first chunk", .{ .content_type = mime.content_type, .len = data.len });
 
             self.mode = switch (mime.content_type) {
-                .text_html => .{ .html = try parser.Parser.init(mime.charset orelse "UTF-8") },
+                .text_html => .{ .html = try parser.Parser.init(mime.charsetString()) },
 
                 .application_json,
                 .text_javascript,
                 .text_css,
                 .text_plain,
                 => blk: {
-                    var p = try parser.Parser.init(mime.charset orelse "UTF-8");
+                    var p = try parser.Parser.init(mime.charsetString());
                     try p.process("<html><head><meta charset=\"utf-8\"></head><body><pre>");
                     break :blk .{ .text = p };
                 },

src/browser/xhr/xhr.zig

@@ -679,7 +679,7 @@ pub const XMLHttpRequest = struct {
         }
 
         var fbs = std.io.fixedBufferStream(self.response_bytes.items);
-        const doc = parser.documentHTMLParse(fbs.reader(), mime.charset orelse "UTF-8") catch {
+        const doc = parser.documentHTMLParse(fbs.reader(), mime.charsetString()) catch {
             self.response_obj = .{ .Failure = {} };
             return;
         };