fergalmoran/ladybird
1
0
Fork 0
mirror of https://github.com/fergalmoran/ladybird.git synced 2026-02-05 23:35:33 +00:00
Code Issues Packages Projects Releases Wiki Activity

LibWeb: Use text encoding from DOM when parsing URLs

Browse Source 
This passes the DOM encoding down to the URL parser, so the correct
encoder can be used.
This commit is contained in:
BenJilks
2024-08-05 21:47:45 +01:00
committed by Tim Ledbetter
parent d80575a410
commit c1958437f9
5 changed files with 11 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
Userland/Libraries/LibURL/Parser.cpp
View File

@@ -806,8 +806,7 @@ ErrorOr<String> Parser::percent_encode_after_encoding(TextCodec::Encoder& encode
}
// https://url.spec.whatwg.org/#concept-basic-url-parser
// NOTE: This parser assumes a UTF-8 encoding.
URL Parser::basic_parse(StringView raw_input, Optional<URL> const& base_url, Optional<URL> url, Optional<State> state_override)
URL Parser::basic_parse(StringView raw_input, Optional<URL> const& base_url, Optional<URL> url, Optional<State> state_override, Optional<StringView> encoding)
{
dbgln_if(URL_PARSER_DEBUG, "URL::Parser::basic_parse: Parsing '{}'", raw_input);
@@ -855,7 +854,11 @@ URL Parser::basic_parse(StringView raw_input, Optional<URL> const& base_url, Opt
State state = state_override.value_or(State::SchemeStart);
// 5. Set encoding to the result of getting an output encoding from encoding.
auto encoder = TextCodec::encoder_for("utf-8"sv);
Optional<TextCodec::Encoder&> encoder = {};
if (encoding.has_value())
encoder = TextCodec::encoder_for(TextCodec::get_output_encoding(*encoding));
if (!encoder.has_value())
encoder = TextCodec::encoder_for("utf-8"sv);
VERIFY(encoder.has_value());
// 6. Let buffer be the empty string.

2
Userland/Libraries/LibURL/Parser.h
View File

@@ -58,7 +58,7 @@ public:
}
// https://url.spec.whatwg.org/#concept-basic-url-parser
static URL basic_parse(StringView input, Optional<URL> const& base_url = {}, Optional<URL> url = {}, Optional<State> state_override = {});
static URL basic_parse(StringView input, Optional<URL> const& base_url = {}, Optional<URL> url = {}, Optional<State> state_override = {}, Optional<StringView> encoding = {});
// https://url.spec.whatwg.org/#string-percent-encode-after-encoding
static ErrorOr<String> percent_encode_after_encoding(TextCodec::Encoder&, StringView input, PercentEncodeSet percent_encode_set, bool space_as_plus = false);

2
Userland/Libraries/LibWeb/DOM/Document.cpp
View File

@@ -1034,7 +1034,7 @@ URL::URL Document::parse_url(StringView url) const
auto base_url = this->base_url();
// 2. Return the result of applying the URL parser to url, with baseURL.
return DOMURL::parse(url, base_url);
return DOMURL::parse(url, base_url, Optional<StringView> { m_encoding });
}
void Document::set_needs_layout()

4
Userland/Libraries/LibWeb/DOMURL/DOMURL.cpp
View File

@@ -585,12 +585,12 @@ void strip_trailing_spaces_from_an_opaque_path(DOMURL& url)
}
// https://url.spec.whatwg.org/#concept-url-parser
URL::URL parse(StringView input, Optional<URL::URL> const& base_url)
URL::URL parse(StringView input, Optional<URL::URL> const& base_url, Optional<StringView> encoding)
{
// FIXME: We should probably have an extended version of URL::URL for LibWeb instead of standalone functions like this.
// 1. Let url be the result of running the basic URL parser on input with base and encoding.
auto url = URL::Parser::basic_parse(input, base_url);
auto url = URL::Parser::basic_parse(input, base_url, {}, {}, encoding);
// 2. If url is failure, return failure.
if (!url.is_valid())

2
Userland/Libraries/LibWeb/DOMURL/DOMURL.h
View File

@@ -99,6 +99,6 @@ bool host_is_domain(URL::Host const&);
void strip_trailing_spaces_from_an_opaque_path(DOMURL& url);
// https://url.spec.whatwg.org/#concept-url-parser
URL::URL parse(StringView input, Optional<URL::URL> const& base_url = {});
URL::URL parse(StringView input, Optional<URL::URL> const& base_url = {}, Optional<StringView> encoding = {});
}