RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-26 01:47:34 +00:00
Code Issues Activity Actions

LibWeb/MimeSniff: Add match an archive type pattern algorithm

Browse source
This commit is contained in:
Kemal Zebari 2023-10-14 20:02:30 -07:00 • committed by Andrew Kaster
parent d5e5a1138f
commit f6d3ea33fa
2 changed files with 35 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
Tests/LibWeb/TestMimeSniff.cpp
View file

@ -76,6 +76,9 @@ TEST_CASE(compute_unknown_mime_type)
mime_type_to_headers_map.set("audio/midi"sv, { "MThd\x00\x00\x00\x06"sv });
mime_type_to_headers_map.set("video/avi"sv, { "RIFF\x00\x00\x00\x00\x41\x56\x49\x20"sv });
mime_type_to_headers_map.set("audio/wave"sv, { "RIFF\x00\x00\x00\x00WAVE"sv });
mime_type_to_headers_map.set("application/x-gzip"sv, { "\x1F\x8B\x08"sv });
mime_type_to_headers_map.set("application/zip"sv, { "PK\x03\x04"sv });
mime_type_to_headers_map.set("application/x-rar-compressed"sv, { "Rar\x20\x1A\x07\x00"sv });
for (auto const& mime_type_to_headers : mime_type_to_headers_map) {
auto mime_type = mime_type_to_headers.key;

33
Userland/Libraries/LibWeb/MimeSniff/Resource.cpp
View file

@ -182,6 +182,36 @@ ErrorOr<Optional<MimeType>> match_an_audio_or_video_type_pattern(ReadonlyBytes i
return OptionalNone {};
}
// https://mimesniff.spec.whatwg.org/#matching-an-archive-type-pattern
ErrorOr<Optional<MimeType>> match_an_archive_type_pattern(ReadonlyBytes input)
{
// 1. Execute the following steps for each row row in the following table:
static Array<BytePatternTableRow, 3> constexpr pattern_table {
// The GZIP archive signature.
BytePatternTableRow { "\x1F\x8B\x08"sv, "\xFF\xFF\xFF"sv, no_ignored_bytes, "application/x-gzip"sv },
// The string "PK" followed by ETX EOT, the ZIP archive signature.
BytePatternTableRow { "\x50\x4B\x03\x04"sv, "\xFF\xFF\xFF\xFF"sv, no_ignored_bytes, "application/zip"sv },
// The string "Rar " followed by SUB BEL NUL, the RAR archive signature.
BytePatternTableRow { "\x52\x61\x72\x20\x1A\x07\x00"sv, "\xFF\xFF\xFF\xFF\xFF\xFF\xFF"sv, no_ignored_bytes, "application/x-rar-compressed"sv },
};
for (auto const& row : pattern_table) {
// 1. Let patternMatched be the result of the pattern matching algorithm given input, the
// value in the first column of row, the value in the second column of row, and the
// value in the third column of row.
auto pattern_matched = pattern_matching_algorithm(input, row.byte_pattern.bytes(), row.pattern_mask.bytes(), row.ignored_leading_bytes);
// 2. If patternMatched is true, return the value in the fourth column of row.
if (pattern_matched)
return MimeType::parse(row.mime_type);
}
// 2. Return undefined.
return OptionalNone {};
}
// https://mimesniff.spec.whatwg.org/#rules-for-identifying-an-unknown-mime-type
ErrorOr<MimeType> rules_for_identifying_an_unknown_mime_type(Resource const& resource, bool sniff_scriptable = false)
{
@ -314,7 +344,8 @@ ErrorOr<MimeType> rules_for_identifying_an_unknown_mime_type(Resource const& res
if (matched_type.has_value())
return matched_type.release_value();
// FIXME: 7. Set matchedType to the result of executing the archive type pattern matching algorithm given resource’s resource header.
// 7. Set matchedType to the result of executing the archive type pattern matching algorithm given resource’s resource header.
matched_type = TRY(match_an_archive_type_pattern(resource.resource_header()));
// 8. If matchedType is not undefined, return matchedType.
if (matched_type.has_value())