Stowage/ladybird
2
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird.git synced 2025-10-19 15:43:20 +00:00
Code Issues Projects Releases Packages Wiki Activity
ladybird/Libraries/LibWeb/CSS/Parser/ValueParsing.cpp

3992 lines
166 KiB
C++
Raw Normal View History

LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
/*
* Copyright (c) 2018-2024, Andreas Kling <andreas@ladybird.org>
* Copyright (c) 2020-2021, the SerenityOS developers.
* Copyright (c) 2021-2025, Sam Atkins <sam@ladybird.org>
* Copyright (c) 2021, Tobias Christiansen <tobyase@serenityos.org>
* Copyright (c) 2022, MacDue <macdue@dueutil.tech>
* Copyright (c) 2024, Shannon Booth <shannon@serenityos.org>
* Copyright (c) 2024, Tommy van der Vorst <tommy@pixelspark.nl>
* Copyright (c) 2024, Matthew Olsson <mattco@serenityos.org>
* Copyright (c) 2024, Glenn Skrzypczak <glenn.skrzypczak@gmail.com>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <AK/Debug.h>
#include <AK/GenericLexer.h>
#include <AK/TemporaryChange.h>
LibWeb/CSS: Treat 'mask' as a longhand property Before this change, an element masked with 'mask-image: url(...)' would show the mask, but 'mask: url(...)' would not. On e.g. dialogic.nl it would show white boxes instead of the actual images in the top navigation bar. We still do not support many of the other mask properties, but with this change at least the masks show up in both cases.
2025-03-01 11:27:55 +01:00
#include <LibURL/URL.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/CSS/Parser/Parser.h>
#include <LibWeb/CSS/PropertyName.h>
#include <LibWeb/CSS/StyleValues/AngleStyleValue.h>
#include <LibWeb/CSS/StyleValues/BackgroundRepeatStyleValue.h>
#include <LibWeb/CSS/StyleValues/BackgroundSizeStyleValue.h>
#include <LibWeb/CSS/StyleValues/BasicShapeStyleValue.h>
#include <LibWeb/CSS/StyleValues/CSSColor.h>
#include <LibWeb/CSS/StyleValues/CSSColorValue.h>
#include <LibWeb/CSS/StyleValues/CSSHSL.h>
#include <LibWeb/CSS/StyleValues/CSSHWB.h>
#include <LibWeb/CSS/StyleValues/CSSKeywordValue.h>
#include <LibWeb/CSS/StyleValues/CSSLCHLike.h>
#include <LibWeb/CSS/StyleValues/CSSLabLike.h>
#include <LibWeb/CSS/StyleValues/CSSLightDark.h>
#include <LibWeb/CSS/StyleValues/CSSRGB.h>
LibWeb/CSS: Clarify types of image-parsing methods parse_image_value() always returns some kind of AbstractImageStyleValue.
2025-02-17 17:11:02 +00:00
#include <LibWeb/CSS/StyleValues/ConicGradientStyleValue.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/CSS/StyleValues/CounterDefinitionsStyleValue.h>
#include <LibWeb/CSS/StyleValues/CounterStyleValue.h>
#include <LibWeb/CSS/StyleValues/CustomIdentStyleValue.h>
#include <LibWeb/CSS/StyleValues/EasingStyleValue.h>
#include <LibWeb/CSS/StyleValues/EdgeStyleValue.h>
LibWeb: Parse CSS fit-content(<length-percentage>) values Before this change, we only parsed fit-content as a standalone keyword, but CSS-SIZING-3 added it as a function as well. I don't know of anything else in CSS that is overloaded like this, so it ends up looking a little awkward in the implementation. Note that a lot of code had already been prepped for fit-content values to have an argument, we just weren't parsing it.
2025-02-26 18:16:36 +01:00
#include <LibWeb/CSS/StyleValues/FitContentStyleValue.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/CSS/StyleValues/FlexStyleValue.h>
#include <LibWeb/CSS/StyleValues/FrequencyStyleValue.h>
#include <LibWeb/CSS/StyleValues/GridTrackPlacementStyleValue.h>
#include <LibWeb/CSS/StyleValues/GridTrackSizeListStyleValue.h>
#include <LibWeb/CSS/StyleValues/ImageStyleValue.h>
#include <LibWeb/CSS/StyleValues/IntegerStyleValue.h>
#include <LibWeb/CSS/StyleValues/LengthStyleValue.h>
LibWeb/CSS: Clarify types of image-parsing methods parse_image_value() always returns some kind of AbstractImageStyleValue.
2025-02-17 17:11:02 +00:00
#include <LibWeb/CSS/StyleValues/LinearGradientStyleValue.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/CSS/StyleValues/NumberStyleValue.h>
#include <LibWeb/CSS/StyleValues/PercentageStyleValue.h>
#include <LibWeb/CSS/StyleValues/PositionStyleValue.h>
LibWeb/CSS: Clarify types of image-parsing methods parse_image_value() always returns some kind of AbstractImageStyleValue.
2025-02-17 17:11:02 +00:00
#include <LibWeb/CSS/StyleValues/RadialGradientStyleValue.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/CSS/StyleValues/RatioStyleValue.h>
#include <LibWeb/CSS/StyleValues/RectStyleValue.h>
#include <LibWeb/CSS/StyleValues/ResolutionStyleValue.h>
#include <LibWeb/CSS/StyleValues/StringStyleValue.h>
#include <LibWeb/CSS/StyleValues/TimeStyleValue.h>
#include <LibWeb/CSS/StyleValues/URLStyleValue.h>
#include <LibWeb/CSS/StyleValues/UnresolvedStyleValue.h>
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
#include <LibWeb/DOM/Element.h>
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
#include <LibWeb/Dump.h>
#include <LibWeb/Infra/CharacterTypes.h>
namespace Web::CSS::Parser {
Optional<Dimension> Parser::parse_dimension(ComponentValue const& component_value)
{
if (component_value.is(Token::Type::Dimension)) {
auto numeric_value = component_value.token().dimension_value();
auto unit_string = component_value.token().dimension_unit();
if (auto length_type = Length::unit_from_name(unit_string); length_type.has_value())
return Length { numeric_value, length_type.release_value() };
if (auto angle_type = Angle::unit_from_name(unit_string); angle_type.has_value())
return Angle { numeric_value, angle_type.release_value() };
if (auto flex_type = Flex::unit_from_name(unit_string); flex_type.has_value())
return Flex { numeric_value, flex_type.release_value() };
if (auto frequency_type = Frequency::unit_from_name(unit_string); frequency_type.has_value())
return Frequency { numeric_value, frequency_type.release_value() };
if (auto resolution_type = Resolution::unit_from_name(unit_string); resolution_type.has_value())
return Resolution { numeric_value, resolution_type.release_value() };
if (auto time_type = Time::unit_from_name(unit_string); time_type.has_value())
return Time { numeric_value, time_type.release_value() };
}
if (component_value.is(Token::Type::Percentage))
return Percentage { component_value.token().percentage() };
if (component_value.is(Token::Type::Number)) {
auto numeric_value = component_value.token().number_value();
if (numeric_value == 0)
return Length::make_px(0);
if (context_allows_quirky_length())
return Length::make_px(CSSPixels::nearest_value_for(numeric_value));
}
return {};
}
Optional<AngleOrCalculated> Parser::parse_angle(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_angle_value(tokens)) {
if (value->is_angle())
return value->as_angle().angle();
if (value->is_calculated())
return AngleOrCalculated { value->as_calculated() };
}
return {};
}
Optional<AnglePercentage> Parser::parse_angle_percentage(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_angle_percentage_value(tokens)) {
if (value->is_angle())
return value->as_angle().angle();
if (value->is_percentage())
return value->as_percentage().percentage();
if (value->is_calculated())
return AnglePercentage { value->as_calculated() };
}
return {};
}
Optional<FlexOrCalculated> Parser::parse_flex(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_flex_value(tokens)) {
if (value->is_flex())
return value->as_flex().flex();
if (value->is_calculated())
return FlexOrCalculated { value->as_calculated() };
}
return {};
}
Optional<FrequencyOrCalculated> Parser::parse_frequency(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_frequency_value(tokens)) {
if (value->is_frequency())
return value->as_frequency().frequency();
if (value->is_calculated())
return FrequencyOrCalculated { value->as_calculated() };
}
return {};
}
Optional<FrequencyPercentage> Parser::parse_frequency_percentage(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_frequency_percentage_value(tokens)) {
if (value->is_frequency())
return value->as_frequency().frequency();
if (value->is_percentage())
return value->as_percentage().percentage();
if (value->is_calculated())
return FrequencyPercentage { value->as_calculated() };
}
return {};
}
Optional<IntegerOrCalculated> Parser::parse_integer(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_integer_value(tokens)) {
if (value->is_integer())
return value->as_integer().integer();
if (value->is_calculated())
return IntegerOrCalculated { value->as_calculated() };
}
return {};
}
Optional<LengthOrCalculated> Parser::parse_length(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_length_value(tokens)) {
if (value->is_length())
return value->as_length().length();
if (value->is_calculated())
return LengthOrCalculated { value->as_calculated() };
}
return {};
}
Optional<LengthPercentage> Parser::parse_length_percentage(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_length_percentage_value(tokens)) {
if (value->is_length())
return value->as_length().length();
if (value->is_percentage())
return value->as_percentage().percentage();
if (value->is_calculated())
return LengthPercentage { value->as_calculated() };
}
return {};
}
Optional<NumberOrCalculated> Parser::parse_number(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_number_value(tokens)) {
if (value->is_number())
return value->as_number().number();
if (value->is_calculated())
return NumberOrCalculated { value->as_calculated() };
}
return {};
}
Optional<NumberPercentage> Parser::parse_number_percentage(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_number_percentage_value(tokens)) {
if (value->is_number())
return Number { Number::Type::Number, value->as_number().number() };
if (value->is_percentage())
return value->as_percentage().percentage();
if (value->is_calculated())
return NumberPercentage { value->as_calculated() };
}
return {};
}
Optional<ResolutionOrCalculated> Parser::parse_resolution(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_resolution_value(tokens)) {
if (value->is_resolution())
return value->as_resolution().resolution();
if (value->is_calculated())
return ResolutionOrCalculated { value->as_calculated() };
}
return {};
}
Optional<TimeOrCalculated> Parser::parse_time(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_time_value(tokens)) {
if (value->is_time())
return value->as_time().time();
if (value->is_calculated())
return TimeOrCalculated { value->as_calculated() };
}
return {};
}
Optional<TimePercentage> Parser::parse_time_percentage(TokenStream<ComponentValue>& tokens)
{
if (auto value = parse_time_percentage_value(tokens)) {
if (value->is_time())
return value->as_time().time();
if (value->is_percentage())
return value->as_percentage().percentage();
if (value->is_calculated())
return TimePercentage { value->as_calculated() };
}
return {};
}
Optional<Ratio> Parser::parse_ratio(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
// FIXME: It seems like `calc(...) / calc(...)` is a valid <ratio>, but this case is neither mentioned in a spec,
// nor tested in WPT, as far as I can tell.
// Still, we should probably support it. That means not assuming we can resolve the calculation immediately.
auto read_number_value = [this](ComponentValue const& component_value) -> Optional<double> {
if (component_value.is(Token::Type::Number))
return component_value.token().number_value();
if (component_value.is_function()) {
auto maybe_calc = parse_calculated_value(component_value);
if (!maybe_calc)
return {};
if (maybe_calc->is_number())
return maybe_calc->as_number().value();
if (!maybe_calc->is_calculated() || !maybe_calc->as_calculated().resolves_to_number())
return {};
if (auto resolved_number = maybe_calc->as_calculated().resolve_number({}); resolved_number.has_value() && resolved_number.value() >= 0) {
return resolved_number.value();
}
}
return {};
};
// `<ratio> = <number [0,∞]> [ / <number [0,∞]> ]?`
auto maybe_numerator = read_number_value(tokens.consume_a_token());
if (!maybe_numerator.has_value() || maybe_numerator.value() < 0)
return {};
auto numerator = maybe_numerator.value();
{
auto two_value_transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto const& solidus = tokens.consume_a_token();
tokens.discard_whitespace();
auto maybe_denominator = read_number_value(tokens.consume_a_token());
if (solidus.is_delim('/') && maybe_denominator.has_value() && maybe_denominator.value() >= 0) {
auto denominator = maybe_denominator.value();
// Two-value ratio
two_value_transaction.commit();
transaction.commit();
return Ratio { numerator, denominator };
}
}
// Single-value ratio
transaction.commit();
return Ratio { numerator };
}
// https://www.w3.org/TR/css-syntax-3/#urange-syntax
Optional<Gfx::UnicodeRange> Parser::parse_unicode_range(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
// <urange> =
// u '+' <ident-token> '?'* |
// u <dimension-token> '?'* |
// u <number-token> '?'* |
// u <number-token> <dimension-token> |
// u <number-token> <number-token> |
// u '+' '?'+
// (All with no whitespace in between tokens.)
// NOTE: Parsing this is different from usual. We take these steps:
// 1. Match the grammar above against the tokens, concatenating them into a string using their original representation.
// 2. Then, parse that string according to the spec algorithm.
// Step 2 is performed by calling the other parse_unicode_range() overload.
auto is_ending_token = [](ComponentValue const& component_value) {
return component_value.is(Token::Type::EndOfFile)
|| component_value.is(Token::Type::Comma)
|| component_value.is(Token::Type::Semicolon)
|| component_value.is(Token::Type::Whitespace);
};
auto create_unicode_range = [&](StringView text, auto& local_transaction) -> Optional<Gfx::UnicodeRange> {
auto maybe_unicode_range = parse_unicode_range(text);
if (maybe_unicode_range.has_value()) {
local_transaction.commit();
transaction.commit();
}
return maybe_unicode_range;
};
// All options start with 'u'/'U'.
auto const& u = tokens.consume_a_token();
if (!u.is_ident("u"sv)) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> does not start with 'u'");
return {};
}
auto const& second_token = tokens.consume_a_token();
// u '+' <ident-token> '?'* |
// u '+' '?'+
if (second_token.is_delim('+')) {
auto local_transaction = tokens.begin_transaction();
StringBuilder string_builder;
string_builder.append(second_token.token().original_source_text());
auto const& third_token = tokens.consume_a_token();
if (third_token.is(Token::Type::Ident) || third_token.is_delim('?')) {
string_builder.append(third_token.token().original_source_text());
while (tokens.next_token().is_delim('?'))
string_builder.append(tokens.consume_a_token().token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
}
}
// u <dimension-token> '?'*
if (second_token.is(Token::Type::Dimension)) {
auto local_transaction = tokens.begin_transaction();
StringBuilder string_builder;
string_builder.append(second_token.token().original_source_text());
while (tokens.next_token().is_delim('?'))
string_builder.append(tokens.consume_a_token().token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
}
// u <number-token> '?'* |
// u <number-token> <dimension-token> |
// u <number-token> <number-token>
if (second_token.is(Token::Type::Number)) {
auto local_transaction = tokens.begin_transaction();
StringBuilder string_builder;
string_builder.append(second_token.token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
auto const& third_token = tokens.consume_a_token();
if (third_token.is_delim('?')) {
string_builder.append(third_token.token().original_source_text());
while (tokens.next_token().is_delim('?'))
string_builder.append(tokens.consume_a_token().token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
} else if (third_token.is(Token::Type::Dimension)) {
string_builder.append(third_token.token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
} else if (third_token.is(Token::Type::Number)) {
string_builder.append(third_token.token().original_source_text());
if (is_ending_token(tokens.next_token()))
return create_unicode_range(string_builder.string_view(), local_transaction);
}
}
if constexpr (CSS_PARSER_DEBUG) {
dbgln("CSSParser: Tokens did not match <urange> grammar.");
tokens.dump_all_tokens();
}
return {};
}
Optional<Gfx::UnicodeRange> Parser::parse_unicode_range(StringView text)
{
auto make_valid_unicode_range = [&](u32 start_value, u32 end_value) -> Optional<Gfx::UnicodeRange> {
// https://www.w3.org/TR/css-syntax-3/#maximum-allowed-code-point
constexpr u32 maximum_allowed_code_point = 0x10FFFF;
// To determine what codepoints the <urange> represents:
// 1. If end value is greater than the maximum allowed code point,
// the <urange> is invalid and a syntax error.
if (end_value > maximum_allowed_code_point) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: Invalid <urange>: end_value ({}) > maximum ({})", end_value, maximum_allowed_code_point);
return {};
}
// 2. If start value is greater than end value, the <urange> is invalid and a syntax error.
if (start_value > end_value) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: Invalid <urange>: start_value ({}) > end_value ({})", start_value, end_value);
return {};
}
// 3. Otherwise, the <urange> represents a contiguous range of codepoints from start value to end value, inclusive.
return Gfx::UnicodeRange { start_value, end_value };
};
// 1. Skipping the first u token, concatenate the representations of all the tokens in the production together.
// Let this be text.
// NOTE: The concatenation is already done by the caller.
GenericLexer lexer { text };
// 2. If the first character of text is U+002B PLUS SIGN, consume it.
// Otherwise, this is an invalid <urange>, and this algorithm must exit.
if (lexer.next_is('+')) {
lexer.consume();
} else {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: Second character of <urange> was not '+'; got: '{}'", lexer.consume());
return {};
}
// 3. Consume as many hex digits from text as possible.
// then consume as many U+003F QUESTION MARK (?) code points as possible.
auto start_position = lexer.tell();
auto hex_digits = lexer.consume_while(is_ascii_hex_digit);
auto question_marks = lexer.consume_while([](auto it) { return it == '?'; });
// If zero code points were consumed, or more than six code points were consumed,
// this is an invalid <urange>, and this algorithm must exit.
size_t consumed_code_points = hex_digits.length() + question_marks.length();
if (consumed_code_points == 0 || consumed_code_points > 6) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> start value had {} digits/?s, expected between 1 and 6.", consumed_code_points);
return {};
}
StringView start_value_code_points = text.substring_view(start_position, consumed_code_points);
// If any U+003F QUESTION MARK (?) code points were consumed, then:
if (question_marks.length() > 0) {
// 1. If there are any code points left in text, this is an invalid <urange>,
// and this algorithm must exit.
if (lexer.tell_remaining() != 0) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> invalid; had {} code points left over.", lexer.tell_remaining());
return {};
}
// 2. Interpret the consumed code points as a hexadecimal number,
// with the U+003F QUESTION MARK (?) code points replaced by U+0030 DIGIT ZERO (0) code points.
// This is the start value.
auto start_value_string = start_value_code_points.replace("?"sv, "0"sv, ReplaceMode::All);
auto maybe_start_value = AK::StringUtils::convert_to_uint_from_hex<u32>(start_value_string);
if (!maybe_start_value.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> ?-converted start value did not parse as hex number.");
return {};
}
u32 start_value = maybe_start_value.release_value();
// 3. Interpret the consumed code points as a hexadecimal number again,
// with the U+003F QUESTION MARK (?) code points replaced by U+0046 LATIN CAPITAL LETTER F (F) code points.
// This is the end value.
auto end_value_string = start_value_code_points.replace("?"sv, "F"sv, ReplaceMode::All);
auto maybe_end_value = AK::StringUtils::convert_to_uint_from_hex<u32>(end_value_string);
if (!maybe_end_value.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> ?-converted end value did not parse as hex number.");
return {};
}
u32 end_value = maybe_end_value.release_value();
// 4. Exit this algorithm.
return make_valid_unicode_range(start_value, end_value);
}
// Otherwise, interpret the consumed code points as a hexadecimal number. This is the start value.
auto maybe_start_value = AK::StringUtils::convert_to_uint_from_hex<u32>(start_value_code_points);
if (!maybe_start_value.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> start value did not parse as hex number.");
return {};
}
u32 start_value = maybe_start_value.release_value();
// 4. If there are no code points left in text, The end value is the same as the start value.
// Exit this algorithm.
if (lexer.tell_remaining() == 0)
return make_valid_unicode_range(start_value, start_value);
// 5. If the next code point in text is U+002D HYPHEN-MINUS (-), consume it.
if (lexer.next_is('-')) {
lexer.consume();
}
// Otherwise, this is an invalid <urange>, and this algorithm must exit.
else {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> start and end values not separated by '-'.");
return {};
}
// 6. Consume as many hex digits as possible from text.
auto end_hex_digits = lexer.consume_while(is_ascii_hex_digit);
// If zero hex digits were consumed, or more than 6 hex digits were consumed,
// this is an invalid <urange>, and this algorithm must exit.
if (end_hex_digits.length() == 0 || end_hex_digits.length() > 6) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> end value had {} digits, expected between 1 and 6.", end_hex_digits.length());
return {};
}
// If there are any code points left in text, this is an invalid <urange>, and this algorithm must exit.
if (lexer.tell_remaining() != 0) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> invalid; had {} code points left over.", lexer.tell_remaining());
return {};
}
// 7. Interpret the consumed code points as a hexadecimal number. This is the end value.
auto maybe_end_value = AK::StringUtils::convert_to_uint_from_hex<u32>(end_hex_digits);
if (!maybe_end_value.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: <urange> end value did not parse as hex number.");
return {};
}
u32 end_value = maybe_end_value.release_value();
return make_valid_unicode_range(start_value, end_value);
}
Vector<Gfx::UnicodeRange> Parser::parse_unicode_ranges(TokenStream<ComponentValue>& tokens)
{
Vector<Gfx::UnicodeRange> unicode_ranges;
auto range_token_lists = parse_a_comma_separated_list_of_component_values(tokens);
for (auto& range_tokens : range_token_lists) {
TokenStream range_token_stream { range_tokens };
auto maybe_unicode_range = parse_unicode_range(range_token_stream);
if (!maybe_unicode_range.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: unicode-range format invalid; discarding.");
return {};
}
unicode_ranges.append(maybe_unicode_range.release_value());
}
return unicode_ranges;
}
RefPtr<CSSStyleValue> Parser::parse_integer_value(TokenStream<ComponentValue>& tokens)
{
auto const& peek_token = tokens.next_token();
if (peek_token.is(Token::Type::Number) && peek_token.token().number().is_integer()) {
tokens.discard_a_token(); // integer
return IntegerStyleValue::create(peek_token.token().number().integer_value());
}
if (auto calc = parse_calculated_value(peek_token); calc
&& (calc->is_integer() || (calc->is_calculated() && calc->as_calculated().resolves_to_number()))) {
tokens.discard_a_token(); // calc
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_number_value(TokenStream<ComponentValue>& tokens)
{
auto const& peek_token = tokens.next_token();
if (peek_token.is(Token::Type::Number)) {
tokens.discard_a_token(); // number
return NumberStyleValue::create(peek_token.token().number().value());
}
if (auto calc = parse_calculated_value(peek_token); calc
&& (calc->is_number() || (calc->is_calculated() && calc->as_calculated().resolves_to_number()))) {
tokens.discard_a_token(); // calc
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_number_percentage_value(TokenStream<ComponentValue>& tokens)
{
// Parses [<percentage> | <number>] (which is equivalent to [<alpha-value>])
if (auto value = parse_number_value(tokens))
return value;
if (auto value = parse_percentage_value(tokens))
return value;
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_number_percentage_none_value(TokenStream<ComponentValue>& tokens)
{
// Parses [<percentage> | <number> | none] (which is equivalent to [<alpha-value> | none])
if (auto value = parse_number_value(tokens))
return value;
if (auto value = parse_percentage_value(tokens))
return value;
if (tokens.next_token().is_ident("none"sv)) {
tokens.discard_a_token(); // keyword none
return CSSKeywordValue::create(Keyword::None);
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_percentage_value(TokenStream<ComponentValue>& tokens)
{
auto const& peek_token = tokens.next_token();
if (peek_token.is(Token::Type::Percentage)) {
tokens.discard_a_token(); // percentage
return PercentageStyleValue::create(Percentage(peek_token.token().percentage()));
}
if (auto calc = parse_calculated_value(peek_token); calc
&& (calc->is_percentage() || (calc->is_calculated() && calc->as_calculated().resolves_to_percentage()))) {
tokens.discard_a_token(); // calc
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_angle_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto angle_type = Angle::unit_from_name(dimension_token.dimension_unit()); angle_type.has_value()) {
transaction.commit();
return AngleStyleValue::create(Angle { (dimension_token.dimension_value()), angle_type.release_value() });
}
return nullptr;
}
// https://svgwg.org/svg2-draft/types.html#presentation-attribute-css-value
// When parsing an SVG attribute, an angle is allowed without a unit.
// FIXME: How should these numbers be interpreted? https://github.com/w3c/svgwg/issues/792
// For now: Convert to an angle in degrees.
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
if (tokens.next_token().is(Token::Type::Number) && is_parsing_svg_presentation_attribute()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
auto numeric_value = tokens.consume_a_token().token().number_value();
return AngleStyleValue::create(Angle::make_degrees(numeric_value));
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_angle() || (calc->is_calculated() && calc->as_calculated().resolves_to_angle()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_angle_percentage_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto angle_type = Angle::unit_from_name(dimension_token.dimension_unit()); angle_type.has_value()) {
transaction.commit();
return AngleStyleValue::create(Angle { (dimension_token.dimension_value()), angle_type.release_value() });
}
return nullptr;
}
if (tokens.next_token().is(Token::Type::Percentage))
return PercentageStyleValue::create(Percentage { tokens.consume_a_token().token().percentage() });
// https://svgwg.org/svg2-draft/types.html#presentation-attribute-css-value
// When parsing an SVG attribute, an angle is allowed without a unit.
// FIXME: How should these numbers be interpreted? https://github.com/w3c/svgwg/issues/792
// For now: Convert to an angle in degrees.
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
if (tokens.next_token().is(Token::Type::Number) && is_parsing_svg_presentation_attribute()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
auto numeric_value = tokens.consume_a_token().token().number_value();
return AngleStyleValue::create(Angle::make_degrees(numeric_value));
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_angle() || calc->is_percentage() || (calc->is_calculated() && calc->as_calculated().resolves_to_angle_percentage()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_flex_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto flex_type = Flex::unit_from_name(dimension_token.dimension_unit()); flex_type.has_value()) {
transaction.commit();
return FlexStyleValue::create(Flex { (dimension_token.dimension_value()), flex_type.release_value() });
}
return nullptr;
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_flex() || (calc->is_calculated() && calc->as_calculated().resolves_to_flex()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_frequency_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto frequency_type = Frequency::unit_from_name(dimension_token.dimension_unit()); frequency_type.has_value()) {
transaction.commit();
return FrequencyStyleValue::create(Frequency { (dimension_token.dimension_value()), frequency_type.release_value() });
}
return nullptr;
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_frequency() || (calc->is_calculated() && calc->as_calculated().resolves_to_frequency()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_frequency_percentage_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto frequency_type = Frequency::unit_from_name(dimension_token.dimension_unit()); frequency_type.has_value()) {
transaction.commit();
return FrequencyStyleValue::create(Frequency { (dimension_token.dimension_value()), frequency_type.release_value() });
}
return nullptr;
}
if (tokens.next_token().is(Token::Type::Percentage))
return PercentageStyleValue::create(Percentage { tokens.consume_a_token().token().percentage() });
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_frequency() || calc->is_percentage() || (calc->is_calculated() && calc->as_calculated().resolves_to_frequency_percentage()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_length_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto length_type = Length::unit_from_name(dimension_token.dimension_unit()); length_type.has_value()) {
transaction.commit();
return LengthStyleValue::create(Length { (dimension_token.dimension_value()), length_type.release_value() });
}
return nullptr;
}
if (tokens.next_token().is(Token::Type::Number)) {
auto transaction = tokens.begin_transaction();
auto numeric_value = tokens.consume_a_token().token().number_value();
if (numeric_value == 0) {
transaction.commit();
return LengthStyleValue::create(Length::make_px(0));
}
if (context_allows_quirky_length()) {
transaction.commit();
return LengthStyleValue::create(Length::make_px(CSSPixels::nearest_value_for(numeric_value)));
}
// https://svgwg.org/svg2-draft/types.html#presentation-attribute-css-value
// When parsing an SVG attribute, a length is allowed without a unit.
// FIXME: How should these numbers be interpreted? https://github.com/w3c/svgwg/issues/792
// For now: Convert to a length in pixels.
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
if (is_parsing_svg_presentation_attribute()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
transaction.commit();
return LengthStyleValue::create(Length::make_px(CSSPixels::nearest_value_for(numeric_value)));
}
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_length() || (calc->is_calculated() && calc->as_calculated().resolves_to_length()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_length_percentage_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto length_type = Length::unit_from_name(dimension_token.dimension_unit()); length_type.has_value()) {
transaction.commit();
return LengthStyleValue::create(Length { (dimension_token.dimension_value()), length_type.release_value() });
}
return nullptr;
}
if (tokens.next_token().is(Token::Type::Percentage))
return PercentageStyleValue::create(Percentage { tokens.consume_a_token().token().percentage() });
if (tokens.next_token().is(Token::Type::Number)) {
auto transaction = tokens.begin_transaction();
auto numeric_value = tokens.consume_a_token().token().number_value();
if (numeric_value == 0) {
transaction.commit();
return LengthStyleValue::create(Length::make_px(0));
}
if (context_allows_quirky_length()) {
transaction.commit();
return LengthStyleValue::create(Length::make_px(CSSPixels::nearest_value_for(numeric_value)));
}
// https://svgwg.org/svg2-draft/types.html#presentation-attribute-css-value
// When parsing an SVG attribute, a length is allowed without a unit.
// FIXME: How should these numbers be interpreted? https://github.com/w3c/svgwg/issues/792
// For now: Convert to a length in pixels.
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
if (is_parsing_svg_presentation_attribute()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
transaction.commit();
return LengthStyleValue::create(Length::make_px(CSSPixels::nearest_value_for(numeric_value)));
}
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_length() || calc->is_percentage() || (calc->is_calculated() && calc->as_calculated().resolves_to_length_percentage()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_resolution_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto resolution_type = Resolution::unit_from_name(dimension_token.dimension_unit()); resolution_type.has_value()) {
transaction.commit();
return ResolutionStyleValue::create(Resolution { (dimension_token.dimension_value()), resolution_type.release_value() });
}
return nullptr;
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_resolution() || (calc->is_calculated() && calc->as_calculated().resolves_to_resolution()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_time_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto time_type = Time::unit_from_name(dimension_token.dimension_unit()); time_type.has_value()) {
transaction.commit();
return TimeStyleValue::create(Time { (dimension_token.dimension_value()), time_type.release_value() });
}
return nullptr;
}
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_time() || (calc->is_calculated() && calc->as_calculated().resolves_to_time()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_time_percentage_value(TokenStream<ComponentValue>& tokens)
{
if (tokens.next_token().is(Token::Type::Dimension)) {
auto transaction = tokens.begin_transaction();
auto& dimension_token = tokens.consume_a_token().token();
if (auto time_type = Time::unit_from_name(dimension_token.dimension_unit()); time_type.has_value()) {
transaction.commit();
return TimeStyleValue::create(Time { (dimension_token.dimension_value()), time_type.release_value() });
}
return nullptr;
}
if (tokens.next_token().is(Token::Type::Percentage))
return PercentageStyleValue::create(Percentage { tokens.consume_a_token().token().percentage() });
auto transaction = tokens.begin_transaction();
if (auto calc = parse_calculated_value(tokens.consume_a_token()); calc
&& (calc->is_time() || calc->is_percentage() || (calc->is_calculated() && calc->as_calculated().resolves_to_time_percentage()))) {
transaction.commit();
return calc;
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_keyword_value(TokenStream<ComponentValue>& tokens)
{
auto const& peek_token = tokens.next_token();
if (peek_token.is(Token::Type::Ident)) {
auto keyword = keyword_from_string(peek_token.token().ident());
if (keyword.has_value()) {
tokens.discard_a_token(); // ident
return CSSKeywordValue::create(keyword.value());
}
}
return nullptr;
}
// https://www.w3.org/TR/CSS2/visufx.html#value-def-shape
RefPtr<CSSStyleValue> Parser::parse_rect_value(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
auto const& function_token = tokens.consume_a_token();
if (!function_token.is_function("rect"sv))
return nullptr;
auto context_guard = push_temporary_value_parsing_context(FunctionContext { "rect"sv });
Vector<Length, 4> params;
auto argument_tokens = TokenStream { function_token.function().value };
enum class CommaRequirement {
Unknown,
RequiresCommas,
RequiresNoCommas
};
enum class Side {
Top = 0,
Right = 1,
Bottom = 2,
Left = 3
};
auto comma_requirement = CommaRequirement::Unknown;
// In CSS 2.1, the only valid <shape> value is: rect(<top>, <right>, <bottom>, <left>) where
// <top> and <bottom> specify offsets from the top border edge of the box, and <right>, and
// <left> specify offsets from the left border edge of the box.
for (size_t side = 0; side < 4; side++) {
argument_tokens.discard_whitespace();
// <top>, <right>, <bottom>, and <left> may either have a <length> value or 'auto'.
// Negative lengths are permitted.
if (argument_tokens.next_token().is_ident("auto"sv)) {
(void)argument_tokens.consume_a_token(); // `auto`
params.append(Length::make_auto());
} else {
auto maybe_length = parse_length(argument_tokens);
if (!maybe_length.has_value())
return nullptr;
if (maybe_length.value().is_calculated()) {
dbgln("FIXME: Support calculated lengths in rect(): {}", maybe_length.value().calculated()->to_string(CSS::CSSStyleValue::SerializationMode::Normal));
return nullptr;
}
params.append(maybe_length.value().value());
}
argument_tokens.discard_whitespace();
// The last side, should be no more tokens following it.
if (static_cast<Side>(side) == Side::Left) {
if (argument_tokens.has_next_token())
return nullptr;
break;
}
bool next_is_comma = argument_tokens.next_token().is(Token::Type::Comma);
// Authors should separate offset values with commas. User agents must support separation
// with commas, but may also support separation without commas (but not a combination),
// because a previous revision of this specification was ambiguous in this respect.
if (comma_requirement == CommaRequirement::Unknown)
comma_requirement = next_is_comma ? CommaRequirement::RequiresCommas : CommaRequirement::RequiresNoCommas;
if (comma_requirement == CommaRequirement::RequiresCommas) {
if (next_is_comma)
argument_tokens.discard_a_token();
else
return nullptr;
} else if (comma_requirement == CommaRequirement::RequiresNoCommas) {
if (next_is_comma)
return nullptr;
} else {
VERIFY_NOT_REACHED();
}
}
transaction.commit();
return RectStyleValue::create(EdgeRect { params[0], params[1], params[2], params[3] });
}
// https://www.w3.org/TR/css-color-4/#typedef-hue
RefPtr<CSSStyleValue> Parser::parse_hue_none_value(TokenStream<ComponentValue>& tokens)
{
// Parses [<hue> | none]
// <hue> = <number> | <angle>
if (auto angle = parse_angle_value(tokens))
return angle;
if (auto number = parse_number_value(tokens))
return number;
if (tokens.next_token().is_ident("none"sv)) {
tokens.discard_a_token(); // keyword none
return CSSKeywordValue::create(Keyword::None);
}
return nullptr;
}
// https://www.w3.org/TR/css-color-4/#typedef-color-alpha-value
RefPtr<CSSStyleValue> Parser::parse_solidus_and_alpha_value(TokenStream<ComponentValue>& tokens)
{
// [ / [<alpha-value> | none] ]?
// <alpha-value> = <number> | <percentage>
// Common to the modern-syntax color functions.
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
if (!tokens.consume_a_token().is_delim('/'))
return {};
tokens.discard_whitespace();
auto alpha = parse_number_percentage_none_value(tokens);
if (!alpha)
return {};
tokens.discard_whitespace();
transaction.commit();
return alpha;
}
// https://www.w3.org/TR/css-color-4/#funcdef-rgb
RefPtr<CSSStyleValue> Parser::parse_rgb_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// rgb() = [ <legacy-rgb-syntax> | <modern-rgb-syntax> ]
// rgba() = [ <legacy-rgba-syntax> | <modern-rgba-syntax> ]
// <legacy-rgb-syntax> = rgb( <percentage>#{3} , <alpha-value>? ) |
// rgb( <number>#{3} , <alpha-value>? )
// <legacy-rgba-syntax> = rgba( <percentage>#{3} , <alpha-value>? ) |
// rgba( <number>#{3} , <alpha-value>? )
// <modern-rgb-syntax> = rgb(
// [ <number> | <percentage> | none]{3}
// [ / [<alpha-value> | none] ]? )
// <modern-rgba-syntax> = rgba(
// [ <number> | <percentage> | none]{3}
// [ / [<alpha-value> | none] ]? )
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function("rgb"sv) && !function_token.is_function("rgba"sv))
return {};
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function_token.function().name });
RefPtr<CSSStyleValue> red;
RefPtr<CSSStyleValue> green;
RefPtr<CSSStyleValue> blue;
RefPtr<CSSStyleValue> alpha;
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
red = parse_number_percentage_none_value(inner_tokens);
if (!red)
return {};
inner_tokens.discard_whitespace();
bool legacy_syntax = inner_tokens.next_token().is(Token::Type::Comma);
if (legacy_syntax) {
// Legacy syntax
// <percentage>#{3} , <alpha-value>?
// | <number>#{3} , <alpha-value>?
// So, r/g/b can be numbers or percentages, as long as they're all the same type.
// We accepted the 'none' keyword when parsing the red value, but it's not allowed in the legacy syntax.
if (red->is_keyword())
return {};
inner_tokens.discard_a_token(); // comma
inner_tokens.discard_whitespace();
green = parse_number_percentage_value(inner_tokens);
if (!green)
return {};
inner_tokens.discard_whitespace();
if (!inner_tokens.consume_a_token().is(Token::Type::Comma))
return {};
inner_tokens.discard_whitespace();
blue = parse_number_percentage_value(inner_tokens);
if (!blue)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
// Try and read comma and alpha
if (!inner_tokens.consume_a_token().is(Token::Type::Comma))
return {};
inner_tokens.discard_whitespace();
alpha = parse_number_percentage_value(inner_tokens);
if (!alpha)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token())
return {};
}
// Verify we're all percentages or all numbers
auto is_percentage = [](CSSStyleValue const& style_value) {
return style_value.is_percentage()
|| (style_value.is_calculated() && style_value.as_calculated().resolves_to_percentage());
};
bool red_is_percentage = is_percentage(*red);
bool green_is_percentage = is_percentage(*green);
bool blue_is_percentage = is_percentage(*blue);
if (red_is_percentage != green_is_percentage || red_is_percentage != blue_is_percentage)
return {};
} else {
// Modern syntax
// [ <number> | <percentage> | none]{3} [ / [<alpha-value> | none] ]?
green = parse_number_percentage_none_value(inner_tokens);
if (!green)
return {};
inner_tokens.discard_whitespace();
blue = parse_number_percentage_none_value(inner_tokens);
if (!blue)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return {};
}
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
LibWeb: Implement css gradient-interpolation-method
2025-02-19 21:02:12 +11:00
return CSSRGB::create(red.release_nonnull(), green.release_nonnull(), blue.release_nonnull(), alpha.release_nonnull(), legacy_syntax ? ColorSyntax::Legacy : ColorSyntax::Modern);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
// https://www.w3.org/TR/css-color-4/#funcdef-hsl
RefPtr<CSSStyleValue> Parser::parse_hsl_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// hsl() = [ <legacy-hsl-syntax> | <modern-hsl-syntax> ]
// hsla() = [ <legacy-hsla-syntax> | <modern-hsla-syntax> ]
// <modern-hsl-syntax> = hsl(
// [<hue> | none]
// [<percentage> | <number> | none]
// [<percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
// <modern-hsla-syntax> = hsla(
// [<hue> | none]
// [<percentage> | <number> | none]
// [<percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
// <legacy-hsl-syntax> = hsl( <hue>, <percentage>, <percentage>, <alpha-value>? )
// <legacy-hsla-syntax> = hsla( <hue>, <percentage>, <percentage>, <alpha-value>? )
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function("hsl"sv) && !function_token.is_function("hsla"sv))
return {};
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function_token.function().name });
RefPtr<CSSStyleValue> h;
RefPtr<CSSStyleValue> s;
RefPtr<CSSStyleValue> l;
RefPtr<CSSStyleValue> alpha;
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
h = parse_hue_none_value(inner_tokens);
if (!h)
return {};
inner_tokens.discard_whitespace();
bool legacy_syntax = inner_tokens.next_token().is(Token::Type::Comma);
if (legacy_syntax) {
// Legacy syntax
// <hue>, <percentage>, <percentage>, <alpha-value>?
// We accepted the 'none' keyword when parsing the h value, but it's not allowed in the legacy syntax.
if (h->is_keyword())
return {};
(void)inner_tokens.consume_a_token(); // comma
inner_tokens.discard_whitespace();
s = parse_percentage_value(inner_tokens);
if (!s)
return {};
inner_tokens.discard_whitespace();
if (!inner_tokens.consume_a_token().is(Token::Type::Comma))
return {};
inner_tokens.discard_whitespace();
l = parse_percentage_value(inner_tokens);
if (!l)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
// Try and read comma and alpha
if (!inner_tokens.consume_a_token().is(Token::Type::Comma))
return {};
inner_tokens.discard_whitespace();
alpha = parse_number_percentage_value(inner_tokens);
// The parser has consumed a comma, so the alpha value is now required
if (!alpha)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token())
return {};
}
} else {
// Modern syntax
// [<hue> | none]
// [<percentage> | <number> | none]
// [<percentage> | <number> | none]
// [ / [<alpha-value> | none] ]?
s = parse_number_percentage_none_value(inner_tokens);
if (!s)
return {};
inner_tokens.discard_whitespace();
l = parse_number_percentage_none_value(inner_tokens);
if (!l)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return {};
}
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
LibWeb: Implement css gradient-interpolation-method
2025-02-19 21:02:12 +11:00
return CSSHSL::create(h.release_nonnull(), s.release_nonnull(), l.release_nonnull(), alpha.release_nonnull(), legacy_syntax ? ColorSyntax::Legacy : ColorSyntax::Modern);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
// https://www.w3.org/TR/css-color-4/#funcdef-hwb
RefPtr<CSSStyleValue> Parser::parse_hwb_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// hwb() = hwb(
// [<hue> | none]
// [<percentage> | <number> | none]
// [<percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function("hwb"sv))
return {};
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function_token.function().name });
RefPtr<CSSStyleValue> h;
RefPtr<CSSStyleValue> w;
RefPtr<CSSStyleValue> b;
RefPtr<CSSStyleValue> alpha;
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
h = parse_hue_none_value(inner_tokens);
if (!h)
return {};
inner_tokens.discard_whitespace();
w = parse_number_percentage_none_value(inner_tokens);
if (!w)
return {};
inner_tokens.discard_whitespace();
b = parse_number_percentage_none_value(inner_tokens);
if (!b)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return {};
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
return CSSHWB::create(h.release_nonnull(), w.release_nonnull(), b.release_nonnull(), alpha.release_nonnull());
}
Optional<Array<RefPtr<CSSStyleValue>, 4>> Parser::parse_lab_like_color_value(TokenStream<ComponentValue>& outer_tokens, StringView function_name)
{
// This helper is designed to be compatible with lab and oklab and parses a function with a form like:
// f() = f( [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function(function_name))
return OptionalNone {};
RefPtr<CSSStyleValue> l;
RefPtr<CSSStyleValue> a;
RefPtr<CSSStyleValue> b;
RefPtr<CSSStyleValue> alpha;
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
l = parse_number_percentage_none_value(inner_tokens);
if (!l)
return OptionalNone {};
inner_tokens.discard_whitespace();
a = parse_number_percentage_none_value(inner_tokens);
if (!a)
return OptionalNone {};
inner_tokens.discard_whitespace();
b = parse_number_percentage_none_value(inner_tokens);
if (!b)
return OptionalNone {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return OptionalNone {};
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
return Array { move(l), move(a), move(b), move(alpha) };
}
// https://www.w3.org/TR/css-color-4/#funcdef-lab
RefPtr<CSSStyleValue> Parser::parse_lab_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// lab() = lab( [<percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
auto maybe_color_values = parse_lab_like_color_value(outer_tokens, "lab"sv);
if (!maybe_color_values.has_value())
return {};
auto& color_values = *maybe_color_values;
return CSSLabLike::create<CSSLab>(color_values[0].release_nonnull(),
color_values[1].release_nonnull(),
color_values[2].release_nonnull(),
color_values[3].release_nonnull());
}
// https://www.w3.org/TR/css-color-4/#funcdef-oklab
RefPtr<CSSStyleValue> Parser::parse_oklab_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// oklab() = oklab( [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ / [<alpha-value> | none] ]? )
auto maybe_color_values = parse_lab_like_color_value(outer_tokens, "oklab"sv);
if (!maybe_color_values.has_value())
return {};
auto& color_values = *maybe_color_values;
return CSSLabLike::create<CSSOKLab>(color_values[0].release_nonnull(),
color_values[1].release_nonnull(),
color_values[2].release_nonnull(),
color_values[3].release_nonnull());
}
Optional<Array<RefPtr<CSSStyleValue>, 4>> Parser::parse_lch_like_color_value(TokenStream<ComponentValue>& outer_tokens, StringView function_name)
{
// This helper is designed to be compatible with lch and oklch and parses a function with a form like:
// f() = f( [<percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <hue> | none]
// [ / [<alpha-value> | none] ]? )
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto const& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function(function_name))
return OptionalNone {};
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
auto l = parse_number_percentage_none_value(inner_tokens);
if (!l)
return OptionalNone {};
inner_tokens.discard_whitespace();
auto c = parse_number_percentage_none_value(inner_tokens);
if (!c)
return OptionalNone {};
inner_tokens.discard_whitespace();
auto h = parse_hue_none_value(inner_tokens);
if (!h)
return OptionalNone {};
inner_tokens.discard_whitespace();
RefPtr<CSSStyleValue> alpha;
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return OptionalNone {};
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
return Array { move(l), move(c), move(h), move(alpha) };
}
// https://www.w3.org/TR/css-color-4/#funcdef-lch
RefPtr<CSSStyleValue> Parser::parse_lch_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// lch() = lch( [<percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <hue> | none]
// [ / [<alpha-value> | none] ]? )
auto maybe_color_values = parse_lch_like_color_value(outer_tokens, "lch"sv);
if (!maybe_color_values.has_value())
return {};
auto& color_values = *maybe_color_values;
return CSSLCHLike::create<CSSLCH>(color_values[0].release_nonnull(),
color_values[1].release_nonnull(),
color_values[2].release_nonnull(),
color_values[3].release_nonnull());
}
// https://www.w3.org/TR/css-color-4/#funcdef-oklch
RefPtr<CSSStyleValue> Parser::parse_oklch_color_value(TokenStream<ComponentValue>& outer_tokens)
{
// oklch() = oklch( [ <percentage> | <number> | none]
// [ <percentage> | <number> | none]
// [ <hue> | none]
// [ / [<alpha-value> | none] ]? )
auto maybe_color_values = parse_lch_like_color_value(outer_tokens, "oklch"sv);
if (!maybe_color_values.has_value())
return {};
auto& color_values = *maybe_color_values;
return CSSLCHLike::create<CSSOKLCH>(color_values[0].release_nonnull(),
color_values[1].release_nonnull(),
color_values[2].release_nonnull(),
color_values[3].release_nonnull());
}
// https://www.w3.org/TR/css-color-4/#funcdef-color
RefPtr<CSSStyleValue> Parser::parse_color_function(TokenStream<ComponentValue>& outer_tokens)
{
// color() = color( <colorspace-params> [ / [ <alpha-value> | none ] ]? )
// <colorspace-params> = [ <predefined-rgb-params> | <xyz-params>]
// <predefined-rgb-params> = <predefined-rgb> [ <number> | <percentage> | none ]{3}
// <predefined-rgb> = srgb | srgb-linear | display-p3 | a98-rgb | prophoto-rgb | rec2020
// <xyz-params> = <xyz-space> [ <number> | <percentage> | none ]{3}
// <xyz-space> = xyz | xyz-d50 | xyz-d65
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto const& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function("color"sv))
return {};
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function_token.function().name });
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
auto const& maybe_color_space = inner_tokens.consume_a_token();
inner_tokens.discard_whitespace();
if (!any_of(CSSColor::s_supported_color_space, [&](auto supported) { return maybe_color_space.is_ident(supported); }))
return {};
auto const& color_space = maybe_color_space.token().ident();
auto c1 = parse_number_percentage_none_value(inner_tokens);
if (!c1)
return {};
inner_tokens.discard_whitespace();
auto c2 = parse_number_percentage_none_value(inner_tokens);
if (!c2)
return {};
inner_tokens.discard_whitespace();
auto c3 = parse_number_percentage_none_value(inner_tokens);
if (!c3)
return {};
inner_tokens.discard_whitespace();
RefPtr<CSSStyleValue> alpha;
if (inner_tokens.has_next_token()) {
alpha = parse_solidus_and_alpha_value(inner_tokens);
if (!alpha || inner_tokens.has_next_token())
return {};
}
if (!alpha)
alpha = NumberStyleValue::create(1);
transaction.commit();
return CSSColor::create(color_space.to_ascii_lowercase(),
c1.release_nonnull(),
c2.release_nonnull(),
c3.release_nonnull(),
alpha.release_nonnull());
}
// https://drafts.csswg.org/css-color-5/#funcdef-light-dark
RefPtr<CSSStyleValue> Parser::parse_light_dark_color_value(TokenStream<ComponentValue>& outer_tokens)
{
auto transaction = outer_tokens.begin_transaction();
outer_tokens.discard_whitespace();
auto const& function_token = outer_tokens.consume_a_token();
if (!function_token.is_function("light-dark"sv))
return {};
auto inner_tokens = TokenStream { function_token.function().value };
inner_tokens.discard_whitespace();
auto light = parse_color_value(inner_tokens);
if (!light)
return {};
inner_tokens.discard_whitespace();
if (!inner_tokens.consume_a_token().is(Token::Type::Comma))
return {};
inner_tokens.discard_whitespace();
auto dark = parse_color_value(inner_tokens);
if (!dark)
return {};
inner_tokens.discard_whitespace();
if (inner_tokens.has_next_token())
return {};
transaction.commit();
return CSSLightDark::create(light.release_nonnull(), dark.release_nonnull());
}
// https://www.w3.org/TR/css-color-4/#color-syntax
RefPtr<CSSStyleValue> Parser::parse_color_value(TokenStream<ComponentValue>& tokens)
{
// Keywords: <system-color> | <deprecated-color> | currentColor
{
auto transaction = tokens.begin_transaction();
if (auto keyword = parse_keyword_value(tokens); keyword && keyword->has_color()) {
transaction.commit();
return keyword;
}
}
// Functions
if (auto color = parse_color_function(tokens))
return color;
if (auto rgb = parse_rgb_color_value(tokens))
return rgb;
if (auto hsl = parse_hsl_color_value(tokens))
return hsl;
if (auto hwb = parse_hwb_color_value(tokens))
return hwb;
if (auto lab = parse_lab_color_value(tokens))
return lab;
if (auto lch = parse_lch_color_value(tokens))
return lch;
if (auto oklab = parse_oklab_color_value(tokens))
return oklab;
if (auto oklch = parse_oklch_color_value(tokens))
return oklch;
if (auto light_dark = parse_light_dark_color_value(tokens))
return light_dark;
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto const& component_value = tokens.consume_a_token();
if (component_value.is(Token::Type::Ident)) {
auto ident = component_value.token().ident();
auto color = Color::from_string(ident);
if (color.has_value()) {
transaction.commit();
LibWeb: Implement css gradient-interpolation-method
2025-02-19 21:02:12 +11:00
return CSSColorValue::create_from_color(color.release_value(), ColorSyntax::Legacy, ident);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
// Otherwise, fall through to the hashless-hex-color case
}
if (component_value.is(Token::Type::Hash)) {
auto color = Color::from_string(MUST(String::formatted("#{}", component_value.token().hash_value())));
if (color.has_value()) {
transaction.commit();
LibWeb: Implement css gradient-interpolation-method
2025-02-19 21:02:12 +11:00
return CSSColorValue::create_from_color(color.release_value(), ColorSyntax::Legacy);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
return {};
}
// https://drafts.csswg.org/css-color-4/#quirky-color
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
if (in_quirks_mode()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// "When CSS is being parsed in quirks mode, <quirky-color> is a type of <color> that is only valid in certain properties:"
// (NOTE: List skipped for brevity; quirks data is assigned in Properties.json)
// "It is not valid in properties that include or reference these properties, such as the background shorthand,
// or inside functional notations such as color-mix()"
bool quirky_color_allowed = false;
if (!m_value_context.is_empty()) {
quirky_color_allowed = m_value_context.first().visit(
[](PropertyID const& property_id) { return property_has_quirk(property_id, Quirk::HashlessHexColor); },
[](FunctionContext const&) { return false; });
}
for (auto i = 1u; i < m_value_context.size() && quirky_color_allowed; i++) {
quirky_color_allowed = m_value_context[i].visit(
[](PropertyID const& property_id) { return property_has_quirk(property_id, Quirk::UnitlessLength); },
[](FunctionContext const&) {
return false;
});
}
if (quirky_color_allowed) {
// NOTE: This algorithm is no longer in the spec, since the concept got moved and renamed. However, it works,
// and so we might as well keep using it.
// The value of a quirky color is obtained from the possible component values using the following algorithm,
// aborting on the first step that returns a value:
// 1. Let cv be the component value.
auto const& cv = component_value;
String serialization;
// 2. If cv is a <number-token> or a <dimension-token>, follow these substeps:
if (cv.is(Token::Type::Number) || cv.is(Token::Type::Dimension)) {
// 1. If cv’s type flag is not "integer", return an error.
// This means that values that happen to use scientific notation, e.g., 5e5e5e, will fail to parse.
if (!cv.token().number().is_integer())
return {};
// 2. If cv’s value is less than zero, return an error.
auto value = cv.is(Token::Type::Number) ? cv.token().to_integer() : cv.token().dimension_value_int();
if (value < 0)
return {};
// 3. Let serialization be the serialization of cv’s value, as a base-ten integer using digits 0-9 (U+0030 to U+0039) in the shortest form possible.
StringBuilder serialization_builder;
serialization_builder.appendff("{}", value);
// 4. If cv is a <dimension-token>, append the unit to serialization.
if (cv.is(Token::Type::Dimension))
serialization_builder.append(cv.token().dimension_unit());
// 5. If serialization consists of fewer than six characters, prepend zeros (U+0030) so that it becomes six characters.
serialization = MUST(serialization_builder.to_string());
if (serialization_builder.length() < 6) {
StringBuilder builder;
for (size_t i = 0; i < (6 - serialization_builder.length()); i++)
builder.append('0');
builder.append(serialization_builder.string_view());
serialization = MUST(builder.to_string());
}
}
// 3. Otherwise, cv is an <ident-token>; let serialization be cv’s value.
else {
if (!cv.is(Token::Type::Ident))
return {};
serialization = cv.token().ident().to_string();
}
// 4. If serialization does not consist of three or six characters, return an error.
if (serialization.bytes().size() != 3 && serialization.bytes().size() != 6)
return {};
// 5. If serialization contains any characters not in the range [0-9A-Fa-f] (U+0030 to U+0039, U+0041 to U+0046, U+0061 to U+0066), return an error.
for (auto c : serialization.bytes_as_string_view()) {
if (!((c >= '0' && c <= '9') || (c >= 'A' && c <= 'F') || (c >= 'a' && c <= 'f')))
return {};
}
// 6. Return the concatenation of "#" (U+0023) and serialization.
auto color = Color::from_string(MUST(String::formatted("#{}", serialization)));
if (color.has_value()) {
transaction.commit();
LibWeb: Implement css gradient-interpolation-method
2025-02-19 21:02:12 +11:00
return CSSColorValue::create_from_color(color.release_value(), ColorSyntax::Legacy);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
}
}
return {};
}
// https://drafts.csswg.org/css-lists-3/#counter-functions
RefPtr<CSSStyleValue> Parser::parse_counter_value(TokenStream<ComponentValue>& tokens)
{
auto parse_counter_name = [this](TokenStream<ComponentValue>& tokens) -> Optional<FlyString> {
// https://drafts.csswg.org/css-lists-3/#typedef-counter-name
// Counters are referred to in CSS syntax using the <counter-name> type, which represents
// their name as a <custom-ident>. A <counter-name> name cannot match the keyword none;
// such an identifier is invalid as a <counter-name>.
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
LibWeb/CSS: Take custom-ident blacklist as a Span Using std::initializer_list here was a bit of a hack, and makes it awkward to pass those blacklists around.
2025-02-25 11:46:52 +00:00
auto counter_name = parse_custom_ident_value(tokens, { { "none"sv } });
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (!counter_name)
return {};
tokens.discard_whitespace();
if (tokens.has_next_token())
return {};
transaction.commit();
return counter_name->custom_ident();
};
auto parse_counter_style = [this](TokenStream<ComponentValue>& tokens) -> RefPtr<CSSStyleValue> {
// https://drafts.csswg.org/css-counter-styles-3/#typedef-counter-style
// <counter-style> = <counter-style-name> | <symbols()>
// For now we just support <counter-style-name>, found here:
// https://drafts.csswg.org/css-counter-styles-3/#typedef-counter-style-name
// <counter-style-name> is a <custom-ident> that is not an ASCII case-insensitive match for none.
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
LibWeb/CSS: Take custom-ident blacklist as a Span Using std::initializer_list here was a bit of a hack, and makes it awkward to pass those blacklists around.
2025-02-25 11:46:52 +00:00
auto counter_style_name = parse_custom_ident_value(tokens, { { "none"sv } });
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (!counter_style_name)
return {};
tokens.discard_whitespace();
if (tokens.has_next_token())
return {};
transaction.commit();
return counter_style_name.release_nonnull();
};
auto transaction = tokens.begin_transaction();
auto const& token = tokens.consume_a_token();
if (token.is_function("counter"sv)) {
// counter() = counter( <counter-name>, <counter-style>? )
auto& function = token.function();
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function.name });
TokenStream function_tokens { function.value };
auto function_values = parse_a_comma_separated_list_of_component_values(function_tokens);
if (function_values.is_empty() || function_values.size() > 2)
return nullptr;
TokenStream name_tokens { function_values[0] };
auto counter_name = parse_counter_name(name_tokens);
if (!counter_name.has_value())
return nullptr;
RefPtr<CSSStyleValue> counter_style;
if (function_values.size() > 1) {
TokenStream counter_style_tokens { function_values[1] };
counter_style = parse_counter_style(counter_style_tokens);
if (!counter_style)
return nullptr;
} else {
// In both cases, if the <counter-style> argument is omitted it defaults to `decimal`.
counter_style = CustomIdentStyleValue::create("decimal"_fly_string);
}
transaction.commit();
return CounterStyleValue::create_counter(counter_name.release_value(), counter_style.release_nonnull());
}
if (token.is_function("counters"sv)) {
// counters() = counters( <counter-name>, <string>, <counter-style>? )
auto& function = token.function();
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function.name });
TokenStream function_tokens { function.value };
auto function_values = parse_a_comma_separated_list_of_component_values(function_tokens);
if (function_values.size() < 2 || function_values.size() > 3)
return nullptr;
TokenStream name_tokens { function_values[0] };
auto counter_name = parse_counter_name(name_tokens);
if (!counter_name.has_value())
return nullptr;
TokenStream string_tokens { function_values[1] };
string_tokens.discard_whitespace();
auto join_string = parse_string_value(string_tokens);
string_tokens.discard_whitespace();
if (!join_string || string_tokens.has_next_token())
return nullptr;
RefPtr<CSSStyleValue> counter_style;
if (function_values.size() > 2) {
TokenStream counter_style_tokens { function_values[2] };
counter_style = parse_counter_style(counter_style_tokens);
if (!counter_style)
return nullptr;
} else {
// In both cases, if the <counter-style> argument is omitted it defaults to `decimal`.
counter_style = CustomIdentStyleValue::create("decimal"_fly_string);
}
transaction.commit();
return CounterStyleValue::create_counters(counter_name.release_value(), join_string->string_value(), counter_style.release_nonnull());
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_ratio_value(TokenStream<ComponentValue>& tokens)
{
if (auto ratio = parse_ratio(tokens); ratio.has_value())
return RatioStyleValue::create(ratio.release_value());
return nullptr;
}
RefPtr<StringStyleValue> Parser::parse_string_value(TokenStream<ComponentValue>& tokens)
{
auto const& peek = tokens.next_token();
if (peek.is(Token::Type::String)) {
tokens.discard_a_token();
return StringStyleValue::create(peek.token().string());
}
return nullptr;
}
LibWeb/CSS: Clarify types of image-parsing methods parse_image_value() always returns some kind of AbstractImageStyleValue.
2025-02-17 17:11:02 +00:00
RefPtr<AbstractImageStyleValue> Parser::parse_image_value(TokenStream<ComponentValue>& tokens)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
LibWeb/CSS: Treat 'mask' as a longhand property Before this change, an element masked with 'mask-image: url(...)' would show the mask, but 'mask: url(...)' would not. On e.g. dialogic.nl it would show white boxes instead of the actual images in the top navigation bar. We still do not support many of the other mask properties, but with this change at least the masks show up in both cases.
2025-03-01 11:27:55 +01:00
tokens.mark();
auto url = parse_url_function(tokens);
if (url.has_value()) {
// If the value is a 'url(..)' parse as image, but if it is just a reference 'url(#xx)', leave it alone,
// so we can parse as URL further on. These URLs are used as references inside SVG documents for masks.
if (!url.value().equals(m_url, URL::ExcludeFragment::Yes)) {
tokens.discard_a_mark();
return ImageStyleValue::create(url.value());
}
tokens.restore_a_mark();
return nullptr;
}
tokens.discard_a_mark();
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (auto linear_gradient = parse_linear_gradient_function(tokens))
return linear_gradient;
if (auto conic_gradient = parse_conic_gradient_function(tokens))
return conic_gradient;
if (auto radial_gradient = parse_radial_gradient_function(tokens))
return radial_gradient;
return nullptr;
}
// https://svgwg.org/svg2-draft/painting.html#SpecifyingPaint
RefPtr<CSSStyleValue> Parser::parse_paint_value(TokenStream<ComponentValue>& tokens)
{
// `<paint> = none | <color> | <url> [none | <color>]? | context-fill | context-stroke`
auto parse_color_or_none = [&]() -> Optional<RefPtr<CSSStyleValue>> {
if (auto color = parse_color_value(tokens))
return color;
// NOTE: <color> also accepts identifiers, so we do this identifier check last.
if (tokens.next_token().is(Token::Type::Ident)) {
auto maybe_keyword = keyword_from_string(tokens.next_token().token().ident());
if (maybe_keyword.has_value()) {
// FIXME: Accept `context-fill` and `context-stroke`
switch (*maybe_keyword) {
case Keyword::None:
tokens.discard_a_token();
return CSSKeywordValue::create(*maybe_keyword);
default:
return nullptr;
}
}
}
return OptionalNone {};
};
// FIMXE: Allow context-fill/context-stroke here
if (auto color_or_none = parse_color_or_none(); color_or_none.has_value())
return *color_or_none;
if (auto url = parse_url_value(tokens)) {
tokens.discard_whitespace();
if (auto color_or_none = parse_color_or_none(); color_or_none == nullptr) {
// Fail to parse if the fallback is invalid, but otherwise ignore it.
// FIXME: Use fallback color
return nullptr;
}
return url;
}
return nullptr;
}
// https://www.w3.org/TR/css-values-4/#position
RefPtr<PositionStyleValue> Parser::parse_position_value(TokenStream<ComponentValue>& tokens, PositionParsingMode position_parsing_mode)
{
auto parse_position_edge = [](TokenStream<ComponentValue>& tokens) -> Optional<PositionEdge> {
auto transaction = tokens.begin_transaction();
auto& token = tokens.consume_a_token();
if (!token.is(Token::Type::Ident))
return {};
auto keyword = keyword_from_string(token.token().ident());
if (!keyword.has_value())
return {};
transaction.commit();
return keyword_to_position_edge(*keyword);
};
auto is_horizontal = [](PositionEdge edge, bool accept_center) -> bool {
switch (edge) {
case PositionEdge::Left:
case PositionEdge::Right:
return true;
case PositionEdge::Center:
return accept_center;
default:
return false;
}
};
auto is_vertical = [](PositionEdge edge, bool accept_center) -> bool {
switch (edge) {
case PositionEdge::Top:
case PositionEdge::Bottom:
return true;
case PositionEdge::Center:
return accept_center;
default:
return false;
}
};
// <position> = [
// [ left | center | right | top | bottom | <length-percentage> ]
// |
// [ left | center | right ] && [ top | center | bottom ]
// |
// [ left | center | right | <length-percentage> ]
// [ top | center | bottom | <length-percentage> ]
// |
// [ [ left | right ] <length-percentage> ] &&
// [ [ top | bottom ] <length-percentage> ]
// ]
// [ left | center | right | top | bottom | <length-percentage> ]
auto alternative_1 = [&]() -> RefPtr<PositionStyleValue> {
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
// [ left | center | right | top | bottom ]
if (auto maybe_edge = parse_position_edge(tokens); maybe_edge.has_value()) {
auto edge = maybe_edge.release_value();
transaction.commit();
// [ left | right ]
if (is_horizontal(edge, false))
return PositionStyleValue::create(EdgeStyleValue::create(edge, {}), EdgeStyleValue::create(PositionEdge::Center, {}));
// [ top | bottom ]
if (is_vertical(edge, false))
return PositionStyleValue::create(EdgeStyleValue::create(PositionEdge::Center, {}), EdgeStyleValue::create(edge, {}));
// [ center ]
VERIFY(edge == PositionEdge::Center);
return PositionStyleValue::create(EdgeStyleValue::create(PositionEdge::Center, {}), EdgeStyleValue::create(PositionEdge::Center, {}));
}
// [ <length-percentage> ]
if (auto maybe_percentage = parse_length_percentage(tokens); maybe_percentage.has_value()) {
transaction.commit();
return PositionStyleValue::create(EdgeStyleValue::create({}, *maybe_percentage), EdgeStyleValue::create(PositionEdge::Center, {}));
}
return nullptr;
};
// [ left | center | right ] && [ top | center | bottom ]
auto alternative_2 = [&]() -> RefPtr<PositionStyleValue> {
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
// Parse out two position edges
auto maybe_first_edge = parse_position_edge(tokens);
if (!maybe_first_edge.has_value())
return nullptr;
auto first_edge = maybe_first_edge.release_value();
tokens.discard_whitespace();
auto maybe_second_edge = parse_position_edge(tokens);
if (!maybe_second_edge.has_value())
return nullptr;
auto second_edge = maybe_second_edge.release_value();
// If 'left' or 'right' is given, that position is X and the other is Y.
// Conversely -
// If 'top' or 'bottom' is given, that position is Y and the other is X.
if (is_vertical(first_edge, false) || is_horizontal(second_edge, false))
swap(first_edge, second_edge);
// [ left | center | right ] [ top | bottom | center ]
if (is_horizontal(first_edge, true) && is_vertical(second_edge, true)) {
transaction.commit();
return PositionStyleValue::create(EdgeStyleValue::create(first_edge, {}), EdgeStyleValue::create(second_edge, {}));
}
return nullptr;
};
// [ left | center | right | <length-percentage> ]
// [ top | center | bottom | <length-percentage> ]
auto alternative_3 = [&]() -> RefPtr<PositionStyleValue> {
auto transaction = tokens.begin_transaction();
auto parse_position_or_length = [&](bool as_horizontal) -> RefPtr<EdgeStyleValue> {
tokens.discard_whitespace();
if (auto maybe_position = parse_position_edge(tokens); maybe_position.has_value()) {
auto position = maybe_position.release_value();
bool valid = as_horizontal ? is_horizontal(position, true) : is_vertical(position, true);
if (!valid)
return nullptr;
return EdgeStyleValue::create(position, {});
}
auto maybe_length = parse_length_percentage(tokens);
if (!maybe_length.has_value())
return nullptr;
return EdgeStyleValue::create({}, maybe_length);
};
// [ left | center | right | <length-percentage> ]
auto horizontal_edge = parse_position_or_length(true);
if (!horizontal_edge)
return nullptr;
// [ top | center | bottom | <length-percentage> ]
auto vertical_edge = parse_position_or_length(false);
if (!vertical_edge)
return nullptr;
transaction.commit();
return PositionStyleValue::create(horizontal_edge.release_nonnull(), vertical_edge.release_nonnull());
};
// [ [ left | right ] <length-percentage> ] &&
// [ [ top | bottom ] <length-percentage> ]
auto alternative_4 = [&]() -> RefPtr<PositionStyleValue> {
struct PositionAndLength {
PositionEdge position;
LengthPercentage length;
};
auto parse_position_and_length = [&]() -> Optional<PositionAndLength> {
tokens.discard_whitespace();
auto maybe_position = parse_position_edge(tokens);
if (!maybe_position.has_value())
return {};
tokens.discard_whitespace();
auto maybe_length = parse_length_percentage(tokens);
if (!maybe_length.has_value())
return {};
return PositionAndLength {
.position = maybe_position.release_value(),
.length = maybe_length.release_value(),
};
};
auto transaction = tokens.begin_transaction();
auto maybe_group1 = parse_position_and_length();
if (!maybe_group1.has_value())
return nullptr;
auto maybe_group2 = parse_position_and_length();
if (!maybe_group2.has_value())
return nullptr;
auto group1 = maybe_group1.release_value();
auto group2 = maybe_group2.release_value();
// [ [ left | right ] <length-percentage> ] [ [ top | bottom ] <length-percentage> ]
if (is_horizontal(group1.position, false) && is_vertical(group2.position, false)) {
transaction.commit();
return PositionStyleValue::create(EdgeStyleValue::create(group1.position, group1.length), EdgeStyleValue::create(group2.position, group2.length));
}
// [ [ top | bottom ] <length-percentage> ] [ [ left | right ] <length-percentage> ]
if (is_vertical(group1.position, false) && is_horizontal(group2.position, false)) {
transaction.commit();
return PositionStyleValue::create(EdgeStyleValue::create(group2.position, group2.length), EdgeStyleValue::create(group1.position, group1.length));
}
return nullptr;
};
// The extra 3-value syntax that's allowed for background-position:
// [ center | [ left | right ] <length-percentage>? ] &&
// [ center | [ top | bottom ] <length-percentage>? ]
auto alternative_5_for_background_position = [&]() -> RefPtr<PositionStyleValue> {
auto transaction = tokens.begin_transaction();
struct PositionAndMaybeLength {
PositionEdge position;
Optional<LengthPercentage> length;
};
// [ <position> <length-percentage>? ]
auto parse_position_and_maybe_length = [&]() -> Optional<PositionAndMaybeLength> {
auto inner_transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto maybe_position = parse_position_edge(tokens);
if (!maybe_position.has_value())
return {};
tokens.discard_whitespace();
auto maybe_length = parse_length_percentage(tokens);
if (maybe_length.has_value()) {
// 'center' cannot be followed by a <length-percentage>
if (maybe_position.value() == PositionEdge::Center && maybe_length.has_value())
return {};
}
inner_transaction.commit();
return PositionAndMaybeLength {
.position = maybe_position.release_value(),
.length = move(maybe_length),
};
};
auto maybe_group1 = parse_position_and_maybe_length();
if (!maybe_group1.has_value())
return nullptr;
auto maybe_group2 = parse_position_and_maybe_length();
if (!maybe_group2.has_value())
return nullptr;
auto group1 = maybe_group1.release_value();
auto group2 = maybe_group2.release_value();
// 2-value or 4-value if both <length-percentage>s are present or missing.
if (group1.length.has_value() == group2.length.has_value())
return nullptr;
// If 'left' or 'right' is given, that position is X and the other is Y.
// Conversely -
// If 'top' or 'bottom' is given, that position is Y and the other is X.
if (is_vertical(group1.position, false) || is_horizontal(group2.position, false))
swap(group1, group2);
// [ center | [ left | right ] ]
if (!is_horizontal(group1.position, true))
return nullptr;
// [ center | [ top | bottom ] ]
if (!is_vertical(group2.position, true))
return nullptr;
auto to_style_value = [&](PositionAndMaybeLength const& group) -> NonnullRefPtr<EdgeStyleValue> {
if (group.position == PositionEdge::Center)
return EdgeStyleValue::create(PositionEdge::Center, {});
return EdgeStyleValue::create(group.position, group.length);
};
transaction.commit();
return PositionStyleValue::create(to_style_value(group1), to_style_value(group2));
};
// Note: The alternatives must be attempted in this order since shorter alternatives can match a prefix of longer ones.
if (auto position = alternative_4())
return position;
if (position_parsing_mode == PositionParsingMode::BackgroundPosition) {
if (auto position = alternative_5_for_background_position())
return position;
}
if (auto position = alternative_3())
return position;
if (auto position = alternative_2())
return position;
if (auto position = alternative_1())
return position;
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_easing_value(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto const& part = tokens.consume_a_token();
if (part.is(Token::Type::Ident)) {
auto name = part.token().ident();
auto maybe_simple_easing = [&] -> RefPtr<EasingStyleValue> {
if (name.equals_ignoring_ascii_case("linear"sv))
return EasingStyleValue::create(EasingStyleValue::Linear::identity());
if (name.equals_ignoring_ascii_case("ease"sv))
return EasingStyleValue::create(EasingStyleValue::CubicBezier::ease());
if (name.equals_ignoring_ascii_case("ease-in"sv))
return EasingStyleValue::create(EasingStyleValue::CubicBezier::ease_in());
if (name.equals_ignoring_ascii_case("ease-out"sv))
return EasingStyleValue::create(EasingStyleValue::CubicBezier::ease_out());
if (name.equals_ignoring_ascii_case("ease-in-out"sv))
return EasingStyleValue::create(EasingStyleValue::CubicBezier::ease_in_out());
if (name.equals_ignoring_ascii_case("step-start"sv))
return EasingStyleValue::create(EasingStyleValue::Steps::step_start());
if (name.equals_ignoring_ascii_case("step-end"sv))
return EasingStyleValue::create(EasingStyleValue::Steps::step_end());
return {};
}();
if (!maybe_simple_easing)
return nullptr;
transaction.commit();
return maybe_simple_easing;
}
if (!part.is_function())
return nullptr;
TokenStream argument_tokens { part.function().value };
auto comma_separated_arguments = parse_a_comma_separated_list_of_component_values(argument_tokens);
// Remove whitespace
for (auto& argument : comma_separated_arguments)
argument.remove_all_matching([](auto& value) { return value.is(Token::Type::Whitespace); });
auto name = part.function().name;
auto context_guard = push_temporary_value_parsing_context(FunctionContext { name });
if (name.equals_ignoring_ascii_case("linear"sv)) {
// linear() = linear( [ <number> && <percentage>{0,2} ]# )
Vector<EasingStyleValue::Linear::Stop> stops;
for (auto const& argument : comma_separated_arguments) {
TokenStream argument_tokens { argument };
Optional<double> output;
Optional<double> first_input;
Optional<double> second_input;
if (argument_tokens.next_token().is(Token::Type::Number))
output = argument_tokens.consume_a_token().token().number_value();
if (argument_tokens.next_token().is(Token::Type::Percentage)) {
first_input = argument_tokens.consume_a_token().token().percentage() / 100;
if (argument_tokens.next_token().is(Token::Type::Percentage)) {
second_input = argument_tokens.consume_a_token().token().percentage() / 100;
}
}
if (argument_tokens.next_token().is(Token::Type::Number)) {
if (output.has_value())
return nullptr;
output = argument_tokens.consume_a_token().token().number_value();
}
if (argument_tokens.has_next_token() || !output.has_value())
return nullptr;
stops.append({ output.value(), first_input, first_input.has_value() });
if (second_input.has_value())
stops.append({ output.value(), second_input, true });
}
if (stops.is_empty())
return nullptr;
transaction.commit();
return EasingStyleValue::create(EasingStyleValue::Linear { move(stops) });
}
if (name.equals_ignoring_ascii_case("cubic-bezier"sv)) {
if (comma_separated_arguments.size() != 4)
return nullptr;
for (auto const& argument : comma_separated_arguments) {
if (argument.size() != 1)
return nullptr;
if (!argument[0].is(Token::Type::Number))
return nullptr;
}
EasingStyleValue::CubicBezier bezier {
comma_separated_arguments[0][0].token().number_value(),
comma_separated_arguments[1][0].token().number_value(),
comma_separated_arguments[2][0].token().number_value(),
comma_separated_arguments[3][0].token().number_value(),
};
if (bezier.x1 < 0.0 || bezier.x1 > 1.0 || bezier.x2 < 0.0 || bezier.x2 > 1.0)
return nullptr;
transaction.commit();
return EasingStyleValue::create(bezier);
}
if (name.equals_ignoring_ascii_case("steps"sv)) {
if (comma_separated_arguments.is_empty() || comma_separated_arguments.size() > 2)
return nullptr;
for (auto const& argument : comma_separated_arguments) {
if (argument.size() != 1)
return nullptr;
}
EasingStyleValue::Steps steps;
auto const& intervals_argument = comma_separated_arguments[0][0];
if (!intervals_argument.is(Token::Type::Number))
return nullptr;
if (!intervals_argument.token().number().is_integer())
return nullptr;
auto intervals = intervals_argument.token().to_integer();
if (comma_separated_arguments.size() == 2) {
TokenStream identifier_stream { comma_separated_arguments[1] };
auto keyword_value = parse_keyword_value(identifier_stream);
if (!keyword_value)
return nullptr;
switch (keyword_value->to_keyword()) {
case Keyword::JumpStart:
steps.position = EasingStyleValue::Steps::Position::JumpStart;
break;
case Keyword::JumpEnd:
steps.position = EasingStyleValue::Steps::Position::JumpEnd;
break;
case Keyword::JumpBoth:
steps.position = EasingStyleValue::Steps::Position::JumpBoth;
break;
case Keyword::JumpNone:
steps.position = EasingStyleValue::Steps::Position::JumpNone;
break;
case Keyword::Start:
steps.position = EasingStyleValue::Steps::Position::Start;
break;
case Keyword::End:
steps.position = EasingStyleValue::Steps::Position::End;
break;
default:
return nullptr;
}
}
// Perform extra validation
// https://drafts.csswg.org/css-easing/#step-easing-functions
// If the <step-position> is jump-none, the <integer> must be at least 2, or the function is invalid.
// Otherwise, the <integer> must be at least 1, or the function is invalid.
if (steps.position == EasingStyleValue::Steps::Position::JumpNone) {
if (intervals <= 1)
return nullptr;
} else if (intervals <= 0) {
return nullptr;
}
steps.number_of_intervals = intervals;
transaction.commit();
return EasingStyleValue::create(steps);
}
return nullptr;
}
Optional<URL::URL> Parser::parse_url_function(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
auto& component_value = tokens.consume_a_token();
auto convert_string_to_url = [&](StringView url_string) -> Optional<URL::URL> {
LibWeb/CSS: Integrate ParsingContext into the Parser This is not really a context, but more of a set of parameters for creating a Parser. So, treat it as such: Rename it to ParsingParams, and store its values and methods directly in the Parser instead of keeping the ParsingContext around. This has a nice side-effect of not including DOM/Document.h everywhere that needs a Parser.
2025-02-05 12:08:27 +00:00
auto url = complete_url(url_string);
LibURL+LibWeb: Port URL::complete_url to Optional Removing one more source of the URL::is_valid API.
2025-02-15 22:55:46 +13:00
if (url.has_value()) {
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
transaction.commit();
return url;
}
return {};
};
if (component_value.is(Token::Type::Url)) {
auto url_string = component_value.token().url();
return convert_string_to_url(url_string);
}
if (component_value.is_function("url"sv)) {
auto const& function_values = component_value.function().value;
// FIXME: Handle url-modifiers. https://www.w3.org/TR/css-values-4/#url-modifiers
for (size_t i = 0; i < function_values.size(); ++i) {
auto const& value = function_values[i];
if (value.is(Token::Type::Whitespace))
continue;
if (value.is(Token::Type::String)) {
auto url_string = value.token().string();
return convert_string_to_url(url_string);
}
break;
}
}
return {};
}
RefPtr<CSSStyleValue> Parser::parse_url_value(TokenStream<ComponentValue>& tokens)
{
auto url = parse_url_function(tokens);
if (!url.has_value())
return nullptr;
return URLStyleValue::create(*url);
}
// https://www.w3.org/TR/css-shapes-1/#typedef-shape-radius
Optional<ShapeRadius> Parser::parse_shape_radius(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto maybe_radius = parse_length_percentage(tokens);
if (maybe_radius.has_value()) {
// Negative radius is invalid.
auto radius = maybe_radius.value();
if ((radius.is_length() && radius.length().raw_value() < 0) || (radius.is_percentage() && radius.percentage().value() < 0))
return {};
transaction.commit();
return radius;
}
if (tokens.next_token().is_ident("closest-side"sv)) {
tokens.discard_a_token();
transaction.commit();
return FitSide::ClosestSide;
}
if (tokens.next_token().is_ident("farthest-side"sv)) {
tokens.discard_a_token();
transaction.commit();
return FitSide::FarthestSide;
}
return {};
}
LibWeb: Parse CSS fit-content(<length-percentage>) values Before this change, we only parsed fit-content as a standalone keyword, but CSS-SIZING-3 added it as a function as well. I don't know of anything else in CSS that is overloaded like this, so it ends up looking a little awkward in the implementation. Note that a lot of code had already been prepped for fit-content values to have an argument, we just weren't parsing it.
2025-02-26 18:16:36 +01:00
RefPtr<FitContentStyleValue> Parser::parse_fit_content_value(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
auto& component_value = tokens.consume_a_token();
if (component_value.is_ident("fit-content"sv)) {
transaction.commit();
return FitContentStyleValue::create();
return nullptr;
}
if (!component_value.is_function())
return nullptr;
auto const& function = component_value.function();
if (function.name != "fit-content"sv)
return nullptr;
TokenStream argument_tokens { function.value };
argument_tokens.discard_whitespace();
auto maybe_length = parse_length_percentage(argument_tokens);
if (!maybe_length.has_value())
return nullptr;
argument_tokens.discard_whitespace();
if (argument_tokens.has_next_token())
return nullptr;
transaction.commit();
return FitContentStyleValue::create(maybe_length.release_value());
}
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
RefPtr<CSSStyleValue> Parser::parse_basic_shape_value(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
auto& component_value = tokens.consume_a_token();
if (!component_value.is_function())
return nullptr;
auto function_name = component_value.function().name.bytes_as_string_view();
// FIXME: Implement path(). See: https://www.w3.org/TR/css-shapes-1/#basic-shape-functions
if (function_name.equals_ignoring_ascii_case("inset"sv)) {
// inset() = inset( <length-percentage>{1,4} [ round <'border-radius'> ]? )
// FIXME: Parse the border-radius.
auto arguments_tokens = TokenStream { component_value.function().value };
// If less than four <length-percentage> values are provided,
// the omitted values default in the same way as the margin shorthand:
// an omitted second or third value defaults to the first, and an omitted fourth value defaults to the second.
// The four <length-percentage>s define the position of the top, right, bottom, and left edges of a rectangle.
arguments_tokens.discard_whitespace();
auto top = parse_length_percentage(arguments_tokens);
if (!top.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
auto right = parse_length_percentage(arguments_tokens);
if (!right.has_value())
right = top;
arguments_tokens.discard_whitespace();
auto bottom = parse_length_percentage(arguments_tokens);
if (!bottom.has_value())
bottom = top;
arguments_tokens.discard_whitespace();
auto left = parse_length_percentage(arguments_tokens);
if (!left.has_value())
left = right;
arguments_tokens.discard_whitespace();
if (arguments_tokens.has_next_token())
return nullptr;
transaction.commit();
return BasicShapeStyleValue::create(Inset { LengthBox(top.value(), right.value(), bottom.value(), left.value()) });
}
if (function_name.equals_ignoring_ascii_case("xywh"sv)) {
// xywh() = xywh( <length-percentage>{2} <length-percentage [0,∞]>{2} [ round <'border-radius'> ]? )
// FIXME: Parse the border-radius.
auto arguments_tokens = TokenStream { component_value.function().value };
arguments_tokens.discard_whitespace();
auto x = parse_length_percentage(arguments_tokens);
if (!x.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
auto y = parse_length_percentage(arguments_tokens);
if (!y.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
auto width = parse_length_percentage(arguments_tokens);
if (!width.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
auto height = parse_length_percentage(arguments_tokens);
if (!height.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
if (arguments_tokens.has_next_token())
return nullptr;
// Negative width or height is invalid.
if ((width->is_length() && width->length().raw_value() < 0)
|| (width->is_percentage() && width->percentage().value() < 0)
|| (height->is_length() && height->length().raw_value() < 0)
|| (height->is_percentage() && height->percentage().value() < 0))
return nullptr;
transaction.commit();
return BasicShapeStyleValue::create(Xywh { x.value(), y.value(), width.value(), height.value() });
}
if (function_name.equals_ignoring_ascii_case("rect"sv)) {
// rect() = rect( [ <length-percentage> | auto ]{4} [ round <'border-radius'> ]? )
// FIXME: Parse the border-radius.
auto arguments_tokens = TokenStream { component_value.function().value };
auto parse_length_percentage_or_auto = [this](TokenStream<ComponentValue>& tokens) -> Optional<LengthPercentage> {
tokens.discard_whitespace();
auto value = parse_length_percentage(tokens);
if (!value.has_value()) {
if (tokens.consume_a_token().is_ident("auto"sv)) {
value = Length::make_auto();
}
}
return value;
};
auto top = parse_length_percentage_or_auto(arguments_tokens);
auto right = parse_length_percentage_or_auto(arguments_tokens);
auto bottom = parse_length_percentage_or_auto(arguments_tokens);
auto left = parse_length_percentage_or_auto(arguments_tokens);
if (!top.has_value() || !right.has_value() || !bottom.has_value() || !left.has_value())
return nullptr;
arguments_tokens.discard_whitespace();
if (arguments_tokens.has_next_token())
return nullptr;
transaction.commit();
return BasicShapeStyleValue::create(Rect { LengthBox(top.value(), right.value(), bottom.value(), left.value()) });
}
if (function_name.equals_ignoring_ascii_case("circle"sv)) {
// circle() = circle( <shape-radius>? [ at <position> ]? )
auto arguments_tokens = TokenStream { component_value.function().value };
auto radius = parse_shape_radius(arguments_tokens).value_or(FitSide::ClosestSide);
auto position = PositionStyleValue::create_center();
arguments_tokens.discard_whitespace();
if (arguments_tokens.next_token().is_ident("at"sv)) {
arguments_tokens.discard_a_token();
arguments_tokens.discard_whitespace();
auto maybe_position = parse_position_value(arguments_tokens);
if (maybe_position.is_null())
return nullptr;
position = maybe_position.release_nonnull();
}
arguments_tokens.discard_whitespace();
if (arguments_tokens.has_next_token())
return nullptr;
transaction.commit();
return BasicShapeStyleValue::create(Circle { radius, position });
}
if (function_name.equals_ignoring_ascii_case("ellipse"sv)) {
// ellipse() = ellipse( [ <shape-radius>{2} ]? [ at <position> ]? )
auto arguments_tokens = TokenStream { component_value.function().value };
Optional<ShapeRadius> radius_x = parse_shape_radius(arguments_tokens);
Optional<ShapeRadius> radius_y = parse_shape_radius(arguments_tokens);
if (radius_x.has_value() && !radius_y.has_value())
return nullptr;
if (!radius_x.has_value()) {
radius_x = FitSide::ClosestSide;
radius_y = FitSide::ClosestSide;
}
auto position = PositionStyleValue::create_center();
arguments_tokens.discard_whitespace();
if (arguments_tokens.next_token().is_ident("at"sv)) {
arguments_tokens.discard_a_token();
arguments_tokens.discard_whitespace();
auto maybe_position = parse_position_value(arguments_tokens);
if (maybe_position.is_null())
return nullptr;
position = maybe_position.release_nonnull();
}
arguments_tokens.discard_whitespace();
if (arguments_tokens.has_next_token())
return nullptr;
transaction.commit();
return BasicShapeStyleValue::create(Ellipse { radius_x.value(), radius_y.value(), position });
}
if (function_name.equals_ignoring_ascii_case("polygon"sv)) {
// polygon() = polygon( <'fill-rule'>? , [<length-percentage> <length-percentage>]# )
auto arguments_tokens = TokenStream { component_value.function().value };
auto arguments = parse_a_comma_separated_list_of_component_values(arguments_tokens);
if (arguments.size() < 1)
return nullptr;
Optional<Gfx::WindingRule> fill_rule;
auto const& first_argument = arguments[0];
TokenStream first_argument_tokens { first_argument };
first_argument_tokens.discard_whitespace();
if (first_argument_tokens.next_token().is_ident("nonzero"sv)) {
fill_rule = Gfx::WindingRule::Nonzero;
} else if (first_argument_tokens.next_token().is_ident("evenodd"sv)) {
fill_rule = Gfx::WindingRule::EvenOdd;
}
if (fill_rule.has_value()) {
first_argument_tokens.discard_a_token();
if (first_argument_tokens.has_next_token())
return nullptr;
arguments.remove(0);
} else {
fill_rule = Gfx::WindingRule::Nonzero;
}
if (arguments.size() < 1)
return nullptr;
Vector<Polygon::Point> points;
for (auto& argument : arguments) {
TokenStream argument_tokens { argument };
argument_tokens.discard_whitespace();
auto x_pos = parse_length_percentage(argument_tokens);
if (!x_pos.has_value())
return nullptr;
argument_tokens.discard_whitespace();
auto y_pos = parse_length_percentage(argument_tokens);
if (!y_pos.has_value())
return nullptr;
argument_tokens.discard_whitespace();
if (argument_tokens.has_next_token())
return nullptr;
points.append(Polygon::Point { *x_pos, *y_pos });
}
transaction.commit();
return BasicShapeStyleValue::create(Polygon { fill_rule.value(), move(points) });
}
return nullptr;
}
RefPtr<CSSStyleValue> Parser::parse_builtin_value(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
auto& component_value = tokens.consume_a_token();
if (component_value.is(Token::Type::Ident)) {
auto ident = component_value.token().ident();
if (ident.equals_ignoring_ascii_case("inherit"sv)) {
transaction.commit();
return CSSKeywordValue::create(Keyword::Inherit);
}
if (ident.equals_ignoring_ascii_case("initial"sv)) {
transaction.commit();
return CSSKeywordValue::create(Keyword::Initial);
}
if (ident.equals_ignoring_ascii_case("unset"sv)) {
transaction.commit();
return CSSKeywordValue::create(Keyword::Unset);
}
if (ident.equals_ignoring_ascii_case("revert"sv)) {
transaction.commit();
return CSSKeywordValue::create(Keyword::Revert);
}
if (ident.equals_ignoring_ascii_case("revert-layer"sv)) {
transaction.commit();
return CSSKeywordValue::create(Keyword::RevertLayer);
}
}
return nullptr;
}
// https://www.w3.org/TR/css-values-4/#custom-idents
LibWeb/CSS: Take custom-ident blacklist as a Span Using std::initializer_list here was a bit of a hack, and makes it awkward to pass those blacklists around.
2025-02-25 11:46:52 +00:00
RefPtr<CustomIdentStyleValue> Parser::parse_custom_ident_value(TokenStream<ComponentValue>& tokens, ReadonlySpan<StringView> blacklist)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
auto transaction = tokens.begin_transaction();
tokens.discard_whitespace();
auto const& token = tokens.consume_a_token();
if (!token.is(Token::Type::Ident))
return nullptr;
auto custom_ident = token.token().ident();
// The CSS-wide keywords are not valid <custom-ident>s.
if (is_css_wide_keyword(custom_ident))
return nullptr;
// The default keyword is reserved and is also not a valid <custom-ident>.
if (custom_ident.equals_ignoring_ascii_case("default"sv))
return nullptr;
// Specifications using <custom-ident> must specify clearly what other keywords are excluded from <custom-ident>,
// if any—for example by saying that any pre-defined keywords in that property’s value definition are excluded.
// Excluded keywords are excluded in all ASCII case permutations.
for (auto& value : blacklist) {
if (custom_ident.equals_ignoring_ascii_case(value))
return nullptr;
}
transaction.commit();
return CustomIdentStyleValue::create(custom_ident);
}
Optional<CSS::GridSize> Parser::parse_grid_size(ComponentValue const& component_value)
{
if (component_value.is_function()) {
if (auto maybe_calculated = parse_calculated_value(component_value)) {
if (maybe_calculated->is_length())
return GridSize(maybe_calculated->as_length().length());
if (maybe_calculated->is_percentage())
return GridSize(maybe_calculated->as_percentage().percentage());
if (maybe_calculated->is_calculated() && maybe_calculated->as_calculated().resolves_to_length_percentage())
return GridSize(LengthPercentage(maybe_calculated->as_calculated()));
// FIXME: Support calculated <flex>
}
return {};
}
if (component_value.is_ident("auto"sv))
return GridSize::make_auto();
if (component_value.is_ident("max-content"sv))
return GridSize(GridSize::Type::MaxContent);
if (component_value.is_ident("min-content"sv))
return GridSize(GridSize::Type::MinContent);
auto dimension = parse_dimension(component_value);
if (!dimension.has_value())
return {};
if (dimension->is_length())
return GridSize(dimension->length());
else if (dimension->is_percentage())
return GridSize(dimension->percentage());
else if (dimension->is_flex())
return GridSize(dimension->flex());
return {};
}
LibWeb: Parse CSS fit-content(<length-percentage>) values Before this change, we only parsed fit-content as a standalone keyword, but CSS-SIZING-3 added it as a function as well. I don't know of anything else in CSS that is overloaded like this, so it ends up looking a little awkward in the implementation. Note that a lot of code had already been prepped for fit-content values to have an argument, we just weren't parsing it.
2025-02-26 18:16:36 +01:00
Optional<CSS::GridFitContent> Parser::parse_grid_fit_content(Vector<ComponentValue> const& component_values)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
// https://www.w3.org/TR/css-grid-2/#valdef-grid-template-columns-fit-content
// 'fit-content( <length-percentage> )'
// Represents the formula max(minimum, min(limit, max-content)), where minimum represents an auto minimum (which is often, but not always,
// equal to a min-content minimum), and limit is the track sizing function passed as an argument to fit-content().
// This is essentially calculated as the smaller of minmax(auto, max-content) and minmax(auto, limit).
auto function_tokens = TokenStream(component_values);
function_tokens.discard_whitespace();
auto maybe_length_percentage = parse_length_percentage(function_tokens);
if (maybe_length_percentage.has_value())
return CSS::GridFitContent(CSS::GridSize(CSS::GridSize::Type::FitContent, maybe_length_percentage.value()));
return {};
}
Optional<CSS::GridMinMax> Parser::parse_min_max(Vector<ComponentValue> const& component_values)
{
// https://www.w3.org/TR/css-grid-2/#valdef-grid-template-columns-minmax
// 'minmax(min, max)'
// Defines a size range greater than or equal to min and less than or equal to max. If the max is
// less than the min, then the max will be floored by the min (essentially yielding minmax(min,
// min)). As a maximum, a <flex> value sets the track’s flex factor; it is invalid as a minimum.
auto function_tokens = TokenStream(component_values);
auto comma_separated_list = parse_a_comma_separated_list_of_component_values(function_tokens);
if (comma_separated_list.size() != 2)
return {};
TokenStream part_one_tokens { comma_separated_list[0] };
part_one_tokens.discard_whitespace();
if (!part_one_tokens.has_next_token())
return {};
NonnullRawPtr<ComponentValue const> current_token = part_one_tokens.consume_a_token();
auto min_grid_size = parse_grid_size(current_token);
TokenStream part_two_tokens { comma_separated_list[1] };
part_two_tokens.discard_whitespace();
if (!part_two_tokens.has_next_token())
return {};
current_token = part_two_tokens.consume_a_token();
auto max_grid_size = parse_grid_size(current_token);
if (min_grid_size.has_value() && max_grid_size.has_value()) {
// https://www.w3.org/TR/css-grid-2/#valdef-grid-template-columns-minmax
// As a maximum, a <flex> value sets the track’s flex factor; it is invalid as a minimum.
if (min_grid_size.value().is_flexible_length())
return {};
return CSS::GridMinMax(min_grid_size.value(), max_grid_size.value());
}
return {};
}
Optional<CSS::GridRepeat> Parser::parse_repeat(Vector<ComponentValue> const& component_values)
{
// https://www.w3.org/TR/css-grid-2/#repeat-syntax
// 7.2.3.1. Syntax of repeat()
// The generic form of the repeat() syntax is, approximately,
// repeat( [ <integer [1,∞]> | auto-fill | auto-fit ] , <track-list> )
auto is_auto_fill = false;
auto is_auto_fit = false;
auto function_tokens = TokenStream(component_values);
auto comma_separated_list = parse_a_comma_separated_list_of_component_values(function_tokens);
if (comma_separated_list.size() != 2)
return {};
// The first argument specifies the number of repetitions.
TokenStream part_one_tokens { comma_separated_list[0] };
part_one_tokens.discard_whitespace();
if (!part_one_tokens.has_next_token())
return {};
auto& current_token = part_one_tokens.consume_a_token();
auto repeat_count = 0;
if (current_token.is(Token::Type::Number) && current_token.token().number().is_integer() && current_token.token().number_value() > 0)
repeat_count = current_token.token().number_value();
else if (current_token.is_ident("auto-fill"sv))
is_auto_fill = true;
else if (current_token.is_ident("auto-fit"sv))
is_auto_fit = true;
// The second argument is a track list, which is repeated that number of times.
TokenStream part_two_tokens { comma_separated_list[1] };
part_two_tokens.discard_whitespace();
if (!part_two_tokens.has_next_token())
return {};
Vector<Variant<ExplicitGridTrack, GridLineNames>> repeat_params;
auto last_object_was_line_names = false;
while (part_two_tokens.has_next_token()) {
auto const& token = part_two_tokens.consume_a_token();
Vector<String> line_names;
if (token.is_block()) {
if (last_object_was_line_names)
return {};
last_object_was_line_names = true;
if (!token.block().is_square())
return {};
TokenStream block_tokens { token.block().value };
while (block_tokens.has_next_token()) {
auto const& current_block_token = block_tokens.consume_a_token();
line_names.append(current_block_token.token().ident().to_string());
block_tokens.discard_whitespace();
}
repeat_params.append(GridLineNames { move(line_names) });
part_two_tokens.discard_whitespace();
} else {
last_object_was_line_names = false;
auto track_sizing_function = parse_track_sizing_function(token);
if (!track_sizing_function.has_value())
return {};
// However, there are some restrictions:
// The repeat() notation can’t be nested.
if (track_sizing_function.value().is_repeat())
return {};
// Automatic repetitions (auto-fill or auto-fit) cannot be combined with intrinsic or flexible sizes.
// Note that 'auto' is also an intrinsic size (and thus not permitted) but we can't use
// track_sizing_function.is_auto(..) to check for it, as it requires AvailableSize, which is why there is
// a separate check for it below.
// https://www.w3.org/TR/css-grid-2/#repeat-syntax
// https://www.w3.org/TR/css-grid-2/#intrinsic-sizing-function
if (track_sizing_function.value().is_default()
&& (track_sizing_function.value().grid_size().is_flexible_length() || token.is_ident("auto"sv))
&& (is_auto_fill || is_auto_fit))
return {};
repeat_params.append(track_sizing_function.value());
part_two_tokens.discard_whitespace();
}
}
// Thus the precise syntax of the repeat() notation has several forms:
// <track-repeat> = repeat( [ <integer [1,∞]> ] , [ <line-names>? <track-size> ]+ <line-names>? )
// <auto-repeat> = repeat( [ auto-fill | auto-fit ] , [ <line-names>? <fixed-size> ]+ <line-names>? )
// <fixed-repeat> = repeat( [ <integer [1,∞]> ] , [ <line-names>? <fixed-size> ]+ <line-names>? )
// <name-repeat> = repeat( [ <integer [1,∞]> | auto-fill ], <line-names>+)
// The <track-repeat> variant can represent the repetition of any <track-size>, but is limited to a
// fixed number of repetitions.
// The <auto-repeat> variant can repeat automatically to fill a space, but requires definite track
// sizes so that the number of repetitions can be calculated. It can only appear once in the track
// list, but the same track list can also contain <fixed-repeat>s.
// The <name-repeat> variant is for adding line names to subgrids. It can only be used with the
// subgrid keyword and cannot specify track sizes, only line names.
// If a repeat() function that is not a <name-repeat> ends up placing two <line-names> adjacent to
// each other, the name lists are merged. For example, repeat(2, [a] 1fr [b]) is equivalent to [a]
// 1fr [b a] 1fr [b].
if (is_auto_fill)
return GridRepeat(GridTrackSizeList(move(repeat_params)), GridRepeat::Type::AutoFill);
else if (is_auto_fit)
return GridRepeat(GridTrackSizeList(move(repeat_params)), GridRepeat::Type::AutoFit);
else
return GridRepeat(GridTrackSizeList(move(repeat_params)), repeat_count);
}
Optional<CSS::ExplicitGridTrack> Parser::parse_track_sizing_function(ComponentValue const& token)
{
if (token.is_function()) {
auto const& function_token = token.function();
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function_token.name });
if (function_token.name.equals_ignoring_ascii_case("repeat"sv)) {
auto maybe_repeat = parse_repeat(function_token.value);
if (maybe_repeat.has_value())
return CSS::ExplicitGridTrack(maybe_repeat.value());
else
return {};
} else if (function_token.name.equals_ignoring_ascii_case("minmax"sv)) {
auto maybe_min_max_value = parse_min_max(function_token.value);
if (maybe_min_max_value.has_value())
return CSS::ExplicitGridTrack(maybe_min_max_value.value());
else
return {};
} else if (function_token.name.equals_ignoring_ascii_case("fit-content"sv)) {
LibWeb: Parse CSS fit-content(<length-percentage>) values Before this change, we only parsed fit-content as a standalone keyword, but CSS-SIZING-3 added it as a function as well. I don't know of anything else in CSS that is overloaded like this, so it ends up looking a little awkward in the implementation. Note that a lot of code had already been prepped for fit-content values to have an argument, we just weren't parsing it.
2025-02-26 18:16:36 +01:00
auto maybe_fit_content_value = parse_grid_fit_content(function_token.value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (maybe_fit_content_value.has_value())
return CSS::ExplicitGridTrack(maybe_fit_content_value.value());
return {};
} else if (auto maybe_calculated = parse_calculated_value(token)) {
if (maybe_calculated->is_length())
return ExplicitGridTrack(GridSize(maybe_calculated->as_length().length()));
if (maybe_calculated->is_percentage())
return ExplicitGridTrack(GridSize(maybe_calculated->as_percentage().percentage()));
if (maybe_calculated->is_calculated() && maybe_calculated->as_calculated().resolves_to_length_percentage())
return ExplicitGridTrack(GridSize(LengthPercentage(maybe_calculated->as_calculated())));
}
return {};
} else if (token.is_ident("auto"sv)) {
return CSS::ExplicitGridTrack(GridSize(Length::make_auto()));
} else if (token.is_block()) {
return {};
} else {
auto grid_size = parse_grid_size(token);
if (!grid_size.has_value())
return {};
return CSS::ExplicitGridTrack(grid_size.value());
}
}
RefPtr<GridTrackPlacementStyleValue> Parser::parse_grid_track_placement(TokenStream<ComponentValue>& tokens)
{
// FIXME: This shouldn't be needed. Right now, the below code returns a CSSStyleValue even if no tokens are consumed!
if (!tokens.has_next_token())
return nullptr;
LibWeb/CSS: Reject invalid grid track placement property values
2025-03-11 11:16:02 +00:00
if (tokens.remaining_token_count() > 3)
return nullptr;
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// https://www.w3.org/TR/css-grid-2/#line-placement
// Line-based Placement: the grid-row-start, grid-column-start, grid-row-end, and grid-column-end properties
// <grid-line> =
// auto |
// <custom-ident> |
// [ <integer> && <custom-ident>? ] |
// [ span && [ <integer> || <custom-ident> ] ]
auto is_valid_integer = [](auto& token) -> bool {
// An <integer> value of zero makes the declaration invalid.
if (token.is(Token::Type::Number) && token.token().number().is_integer() && token.token().number_value() != 0)
return true;
return false;
};
auto parse_custom_ident = [this](auto& tokens) {
// The <custom-ident> additionally excludes the keywords span and auto.
LibWeb/CSS: Take custom-ident blacklist as a Span Using std::initializer_list here was a bit of a hack, and makes it awkward to pass those blacklists around.
2025-02-25 11:46:52 +00:00
return parse_custom_ident_value(tokens, { { "span"sv, "auto"sv } });
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
};
auto transaction = tokens.begin_transaction();
// FIXME: Handle the single-token case inside the loop instead, so that we can more easily call this from
// `parse_grid_area_shorthand_value()` using a single TokenStream.
if (tokens.remaining_token_count() == 1) {
if (auto custom_ident = parse_custom_ident(tokens)) {
transaction.commit();
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_line({}, custom_ident->custom_ident().to_string()));
}
auto const& token = tokens.consume_a_token();
if (auto maybe_calculated = parse_calculated_value(token)) {
if (maybe_calculated->is_number()) {
transaction.commit();
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_line(static_cast<int>(maybe_calculated->as_number().number()), {}));
}
if (maybe_calculated->is_calculated() && maybe_calculated->as_calculated().resolves_to_number()) {
transaction.commit();
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_line(static_cast<int>(maybe_calculated->as_calculated().resolve_integer({}).value()), {}));
}
}
if (token.is_ident("auto"sv)) {
transaction.commit();
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_auto());
}
if (is_valid_integer(token)) {
transaction.commit();
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_line(static_cast<int>(token.token().number_value()), {}));
}
return nullptr;
}
auto span_value = false;
auto span_or_position_value = 0;
String identifier_value;
while (tokens.has_next_token()) {
auto const& token = tokens.next_token();
if (token.is_ident("auto"sv))
return nullptr;
if (token.is_ident("span"sv)) {
if (span_value)
return nullptr;
tokens.discard_a_token(); // span
LibWeb/CSS: Reject invalid grid track placement property values
2025-03-11 11:16:02 +00:00
if (tokens.has_next_token() && ((span_or_position_value != 0 && identifier_value.is_empty()) || (span_or_position_value == 0 && !identifier_value.is_empty())))
return nullptr;
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
span_value = true;
continue;
}
if (is_valid_integer(token)) {
if (span_or_position_value != 0)
return nullptr;
span_or_position_value = static_cast<int>(tokens.consume_a_token().token().to_integer());
continue;
}
if (auto custom_ident = parse_custom_ident(tokens)) {
if (!identifier_value.is_empty())
return nullptr;
identifier_value = custom_ident->custom_ident().to_string();
continue;
}
break;
}
LibWeb/CSS: Reject invalid grid track placement property values
2025-03-11 11:16:02 +00:00
if (tokens.has_next_token())
return nullptr;
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// Negative integers or zero are invalid.
if (span_value && span_or_position_value < 1)
return nullptr;
// If the <integer> is omitted, it defaults to 1.
if (span_or_position_value == 0)
span_or_position_value = 1;
transaction.commit();
if (!identifier_value.is_empty())
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_line(span_or_position_value, identifier_value));
return GridTrackPlacementStyleValue::create(GridTrackPlacement::make_span(span_or_position_value));
}
RefPtr<CSSStyleValue> Parser::parse_calculated_value(ComponentValue const& component_value)
{
if (!component_value.is_function())
return nullptr;
auto const& function = component_value.function();
CalculationContext context {};
for (auto const& value_context : m_value_context.in_reverse()) {
auto maybe_context = value_context.visit(
[](PropertyID property_id) -> Optional<CalculationContext> {
return CalculationContext {
.percentages_resolve_as = property_resolves_percentages_relative_to(property_id),
.resolve_numbers_as_integers = property_accepts_type(property_id, ValueType::Integer),
};
},
[](FunctionContext const& function) -> Optional<CalculationContext> {
// Gradients resolve percentages as lengths relative to the gradient-box.
if (function.name.is_one_of_ignoring_ascii_case(
"linear-gradient"sv, "repeating-linear-gradient"sv,
"radial-gradient"sv, "repeating-radial-gradient"sv,
"conic-gradient"sv, "repeating-conic-gradient"sv)) {
return CalculationContext { .percentages_resolve_as = ValueType::Length };
}
// FIXME: Add other functions that provide a context for resolving values
return {};
});
if (maybe_context.has_value()) {
context = maybe_context.release_value();
break;
}
}
auto function_node = parse_a_calc_function_node(function, context);
if (!function_node)
return nullptr;
auto function_type = function_node->numeric_type();
if (!function_type.has_value())
return nullptr;
return CalculatedStyleValue::create(function_node.release_nonnull(), function_type.release_value(), context);
}
RefPtr<CalculationNode> Parser::parse_a_calc_function_node(Function const& function, CalculationContext const& context)
{
auto context_guard = push_temporary_value_parsing_context(FunctionContext { function.name });
if (function.name.equals_ignoring_ascii_case("calc"sv))
return parse_a_calculation(function.value, context);
if (auto maybe_function = parse_math_function(function, context))
return maybe_function;
return nullptr;
}
RefPtr<CalculationNode> Parser::convert_to_calculation_node(CalcParsing::Node const& node, CalculationContext const& context)
{
return node.visit(
[this, &context](NonnullOwnPtr<CalcParsing::ProductNode> const& product_node) -> RefPtr<CalculationNode> {
Vector<NonnullRefPtr<CalculationNode>> children;
children.ensure_capacity(product_node->children.size());
for (auto const& child : product_node->children) {
if (auto child_as_node = convert_to_calculation_node(child, context)) {
children.append(child_as_node.release_nonnull());
} else {
return nullptr;
}
}
return ProductCalculationNode::create(move(children));
},
[this, &context](NonnullOwnPtr<CalcParsing::SumNode> const& sum_node) -> RefPtr<CalculationNode> {
Vector<NonnullRefPtr<CalculationNode>> children;
children.ensure_capacity(sum_node->children.size());
for (auto const& child : sum_node->children) {
if (auto child_as_node = convert_to_calculation_node(child, context)) {
children.append(child_as_node.release_nonnull());
} else {
return nullptr;
}
}
return SumCalculationNode::create(move(children));
},
[this, &context](NonnullOwnPtr<CalcParsing::InvertNode> const& invert_node) -> RefPtr<CalculationNode> {
if (auto child_as_node = convert_to_calculation_node(invert_node->child, context))
return InvertCalculationNode::create(child_as_node.release_nonnull());
return nullptr;
},
[this, &context](NonnullOwnPtr<CalcParsing::NegateNode> const& negate_node) -> RefPtr<CalculationNode> {
if (auto child_as_node = convert_to_calculation_node(negate_node->child, context))
return NegateCalculationNode::create(child_as_node.release_nonnull());
return nullptr;
},
[this, &context](NonnullRawPtr<ComponentValue const> const& component_value) -> RefPtr<CalculationNode> {
// NOTE: This is the "process the leaf nodes" part of step 5 of https://drafts.csswg.org/css-values-4/#parse-a-calculation
// We divert a little from the spec: Rather than modify an existing tree of values, we construct a new one from that source tree.
// This lets us make CalculationNodes immutable.
// 1. If leaf is a parenthesized simple block, replace leaf with the result of parsing a calculation from leaf’s contents.
if (component_value->is_block() && component_value->block().is_paren()) {
auto leaf_calculation = parse_a_calculation(component_value->block().value, context);
if (!leaf_calculation)
return nullptr;
return leaf_calculation.release_nonnull();
}
// 2. If leaf is a math function, replace leaf with the internal representation of that math function.
// NOTE: All function tokens at this point should be math functions.
if (component_value->is_function()) {
auto const& function = component_value->function();
auto leaf_calculation = parse_a_calc_function_node(function, context);
if (!leaf_calculation)
return nullptr;
return leaf_calculation.release_nonnull();
}
// AD-HOC: We also need to convert tokens into their numeric types.
if (component_value->is(Token::Type::Ident)) {
LibWeb/CSS: Use NumericCalculationNode for constants Having multiple kinds of node that hold numeric values made things more complicated than they needed to be, and we were already converting ConstantCalculationNodes to NumericCalculationNodes in the first simplification pass that happens at parse-time, so they didn't exist after that. As noted, the spec allows for other contexts to introduce their own numeric keywords, which might be resolved later than parse-time. We'll need a different mechanism to support those, but ConstantCalculationNode could not have done so anyway.
2025-02-25 16:37:10 +00:00
auto maybe_keyword = keyword_from_string(component_value->token().ident());
if (!maybe_keyword.has_value())
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
return nullptr;
LibWeb/CSS: Use NumericCalculationNode for constants Having multiple kinds of node that hold numeric values made things more complicated than they needed to be, and we were already converting ConstantCalculationNodes to NumericCalculationNodes in the first simplification pass that happens at parse-time, so they didn't exist after that. As noted, the spec allows for other contexts to introduce their own numeric keywords, which might be resolved later than parse-time. We'll need a different mechanism to support those, but ConstantCalculationNode could not have done so anyway.
2025-02-25 16:37:10 +00:00
return NumericCalculationNode::from_keyword(*maybe_keyword, context);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
if (component_value->is(Token::Type::Number))
return NumericCalculationNode::create(component_value->token().number(), context);
if (component_value->is(Token::Type::Dimension)) {
auto numeric_value = component_value->token().dimension_value();
auto unit_string = component_value->token().dimension_unit();
if (auto length_type = Length::unit_from_name(unit_string); length_type.has_value())
return NumericCalculationNode::create(Length { numeric_value, length_type.release_value() }, context);
if (auto angle_type = Angle::unit_from_name(unit_string); angle_type.has_value())
return NumericCalculationNode::create(Angle { numeric_value, angle_type.release_value() }, context);
if (auto flex_type = Flex::unit_from_name(unit_string); flex_type.has_value()) {
// https://www.w3.org/TR/css3-grid-layout/#fr-unit
// NOTE: <flex> values are not <length>s (nor are they compatible with <length>s, like some <percentage> values),
// so they cannot be represented in or combined with other unit types in calc() expressions.
// FIXME: Flex is allowed in calc(), so figure out what this spec text means and how to implement it.
dbgln_if(CSS_PARSER_DEBUG, "Rejecting <flex> in calc()");
return nullptr;
}
if (auto frequency_type = Frequency::unit_from_name(unit_string); frequency_type.has_value())
return NumericCalculationNode::create(Frequency { numeric_value, frequency_type.release_value() }, context);
if (auto resolution_type = Resolution::unit_from_name(unit_string); resolution_type.has_value())
return NumericCalculationNode::create(Resolution { numeric_value, resolution_type.release_value() }, context);
if (auto time_type = Time::unit_from_name(unit_string); time_type.has_value())
return NumericCalculationNode::create(Time { numeric_value, time_type.release_value() }, context);
dbgln_if(CSS_PARSER_DEBUG, "Unrecognized dimension type in calc() expression: {}", component_value->to_string());
return nullptr;
}
if (component_value->is(Token::Type::Percentage))
return NumericCalculationNode::create(Percentage { component_value->token().percentage() }, context);
// NOTE: If we get here, then we have a ComponentValue that didn't get replaced with something else,
// so the calc() is invalid.
dbgln_if(CSS_PARSER_DEBUG, "Leftover ComponentValue in calc tree! That probably means the syntax is invalid, but maybe we just didn't implement `{}` yet.", component_value->to_debug_string());
return nullptr;
},
[](CalcParsing::Operator const& op) -> RefPtr<CalculationNode> {
dbgln_if(CSS_PARSER_DEBUG, "Leftover Operator {} in calc tree!", op.delim);
return nullptr;
});
}
// https://drafts.csswg.org/css-values-4/#parse-a-calculation
RefPtr<CalculationNode> Parser::parse_a_calculation(Vector<ComponentValue> const& original_values, CalculationContext const& context)
{
// 1. Discard any <whitespace-token>s from values.
// 2. An item in values is an “operator” if it’s a <delim-token> with the value "+", "-", "*", or "/". Otherwise, it’s a “value”.
Vector<CalcParsing::Node> values;
for (auto const& value : original_values) {
if (value.is(Token::Type::Whitespace))
continue;
if (value.is(Token::Type::Delim)) {
if (first_is_one_of(value.token().delim(), static_cast<u32>('+'), static_cast<u32>('-'), static_cast<u32>('*'), static_cast<u32>('/'))) {
// NOTE: Sequential operators are invalid syntax.
if (!values.is_empty() && values.last().has<CalcParsing::Operator>())
return nullptr;
values.append(CalcParsing::Operator { static_cast<char>(value.token().delim()) });
continue;
}
}
values.append(NonnullRawPtr { value });
}
// If we have no values, the syntax is invalid.
if (values.is_empty())
return nullptr;
// NOTE: If the first or last value is an operator, the syntax is invalid.
if (values.first().has<CalcParsing::Operator>() || values.last().has<CalcParsing::Operator>())
return nullptr;
// 3. Collect children into Product and Invert nodes.
// For every consecutive run of value items in values separated by "*" or "/" operators:
while (true) {
Optional<size_t> first_product_operator = values.find_first_index_if([](auto const& item) {
return item.template has<CalcParsing::Operator>()
&& first_is_one_of(item.template get<CalcParsing::Operator>().delim, '*', '/');
});
if (!first_product_operator.has_value())
break;
auto start_of_run = first_product_operator.value() - 1;
auto end_of_run = first_product_operator.value() + 1;
for (auto i = start_of_run + 1; i < values.size(); i += 2) {
auto& item = values[i];
if (!item.has<CalcParsing::Operator>()) {
end_of_run = i - 1;
break;
}
auto delim = item.get<CalcParsing::Operator>().delim;
if (!first_is_one_of(delim, '*', '/')) {
end_of_run = i - 1;
break;
}
}
// 1. For each "/" operator in the run, replace its right-hand value item rhs with an Invert node containing rhs as its child.
Vector<CalcParsing::Node> run_values;
run_values.append(move(values[start_of_run]));
for (auto i = start_of_run + 1; i <= end_of_run; i += 2) {
auto& operator_ = values[i].get<CalcParsing::Operator>().delim;
auto& rhs = values[i + 1];
if (operator_ == '/') {
run_values.append(make<CalcParsing::InvertNode>(move(rhs)));
continue;
}
VERIFY(operator_ == '*');
run_values.append(move(rhs));
}
// 2. Replace the entire run with a Product node containing the value items of the run as its children.
values.remove(start_of_run, end_of_run - start_of_run + 1);
values.insert(start_of_run, make<CalcParsing::ProductNode>(move(run_values)));
}
// 4. Collect children into Sum and Negate nodes.
Optional<CalcParsing::Node> single_value;
{
// 1. For each "-" operator item in values, replace its right-hand value item rhs with a Negate node containing rhs as its child.
for (auto i = 0u; i < values.size(); ++i) {
auto& maybe_minus_operator = values[i];
if (!maybe_minus_operator.has<CalcParsing::Operator>() || maybe_minus_operator.get<CalcParsing::Operator>().delim != '-')
continue;
auto rhs_index = ++i;
auto negate_node = make<CalcParsing::NegateNode>(move(values[rhs_index]));
values.remove(rhs_index);
values.insert(rhs_index, move(negate_node));
}
// 2. If values has only one item, and it is a Product node or a parenthesized simple block, replace values with that item.
if (values.size() == 1) {
values.first().visit(
[&](ComponentValue const& component_value) {
if (component_value.is_block() && component_value.block().is_paren())
single_value = NonnullRawPtr { component_value };
},
[&](NonnullOwnPtr<CalcParsing::ProductNode>& node) {
single_value = move(node);
},
[](auto&) {});
}
// Otherwise, replace values with a Sum node containing the value items of values as its children.
if (!single_value.has_value()) {
values.remove_all_matching([](CalcParsing::Node& value) { return value.has<CalcParsing::Operator>(); });
single_value = make<CalcParsing::SumNode>(move(values));
}
}
VERIFY(single_value.has_value());
// 5. At this point values is a tree of Sum, Product, Negate, and Invert nodes, with other types of values at the leaf nodes. Process the leaf nodes.
// NOTE: We process leaf nodes as part of this conversion.
auto calculation_tree = convert_to_calculation_node(*single_value, context);
if (!calculation_tree)
return nullptr;
// 6. Return the result of simplifying a calculation tree from values.
return simplify_a_calculation_tree(*calculation_tree, context, CalculationResolutionContext {});
}
// https://drafts.csswg.org/css-fonts/#typedef-opentype-tag
RefPtr<StringStyleValue> Parser::parse_opentype_tag_value(TokenStream<ComponentValue>& tokens)
{
// <opentype-tag> = <string>
// The <opentype-tag> is a case-sensitive OpenType feature tag.
// As specified in the OpenType specification [OPENTYPE], feature tags contain four ASCII characters.
// Tag strings longer or shorter than four characters, or containing characters outside the U+20–7E codepoint range are invalid.
auto transaction = tokens.begin_transaction();
auto string_value = parse_string_value(tokens);
if (string_value == nullptr)
return nullptr;
auto string = string_value->string_value().bytes_as_string_view();
if (string.length() != 4)
return nullptr;
for (char c : string) {
if (c < 0x20 || c > 0x7E)
return nullptr;
}
transaction.commit();
return string_value;
}
LibWeb/CSS: Move and rename PseudoElement types to prep for code gen The upcoming generated types will match those for pseudo-classes: A PseudoElementSelector type, that then holds a PseudoElement enum defining what it is. That enum will be at the top level in the Web::CSS namespace. In order to keep the diffs clearer, this commit renames and moves the types, and then a following one will replace the handwritten enum with a generated one.
2025-03-20 16:56:46 +00:00
NonnullRefPtr<CSSStyleValue> Parser::resolve_unresolved_style_value(ParsingParams const& context, DOM::Element& element, Optional<PseudoElement> pseudo_element, PropertyID property_id, UnresolvedStyleValue const& unresolved)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
// Unresolved always contains a var() or attr(), unless it is a custom property's value, in which case we shouldn't be trying
// to produce a different CSSStyleValue from it.
VERIFY(unresolved.contains_var_or_attr());
// If the value is invalid, we fall back to `unset`: https://www.w3.org/TR/css-variables-1/#invalid-at-computed-value-time
auto parser = Parser::create(context, ""sv);
return parser.resolve_unresolved_style_value(element, pseudo_element, property_id, unresolved);
}
class PropertyDependencyNode : public RefCounted<PropertyDependencyNode> {
public:
static NonnullRefPtr<PropertyDependencyNode> create(FlyString name)
{
return adopt_ref(*new PropertyDependencyNode(move(name)));
}
void add_child(NonnullRefPtr<PropertyDependencyNode> new_child)
{
for (auto const& child : m_children) {
if (child->m_name == new_child->m_name)
return;
}
// We detect self-reference already.
VERIFY(new_child->m_name != m_name);
m_children.append(move(new_child));
}
bool has_cycles()
{
if (m_marked)
return true;
TemporaryChange change { m_marked, true };
for (auto& child : m_children) {
if (child->has_cycles())
return true;
}
return false;
}
private:
explicit PropertyDependencyNode(FlyString name)
: m_name(move(name))
{
}
FlyString m_name;
Vector<NonnullRefPtr<PropertyDependencyNode>> m_children;
bool m_marked { false };
};
LibWeb/CSS: Move and rename PseudoElement types to prep for code gen The upcoming generated types will match those for pseudo-classes: A PseudoElementSelector type, that then holds a PseudoElement enum defining what it is. That enum will be at the top level in the Web::CSS namespace. In order to keep the diffs clearer, this commit renames and moves the types, and then a following one will replace the handwritten enum with a generated one.
2025-03-20 16:56:46 +00:00
NonnullRefPtr<CSSStyleValue> Parser::resolve_unresolved_style_value(DOM::Element& element, Optional<PseudoElement> pseudo_element, PropertyID property_id, UnresolvedStyleValue const& unresolved)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
TokenStream unresolved_values_without_variables_expanded { unresolved.values() };
Vector<ComponentValue> values_with_variables_expanded;
HashMap<FlyString, NonnullRefPtr<PropertyDependencyNode>> dependencies;
ScopeGuard mark_element_if_uses_custom_properties = [&] {
for (auto const& name : dependencies.keys()) {
if (is_a_custom_property_name_string(name)) {
element.set_style_uses_css_custom_properties(true);
return;
}
}
};
if (!expand_variables(element, pseudo_element, string_from_property_id(property_id), dependencies, unresolved_values_without_variables_expanded, values_with_variables_expanded))
return CSSKeywordValue::create(Keyword::Unset);
TokenStream unresolved_values_with_variables_expanded { values_with_variables_expanded };
Vector<ComponentValue> expanded_values;
if (!expand_unresolved_values(element, string_from_property_id(property_id), unresolved_values_with_variables_expanded, expanded_values))
return CSSKeywordValue::create(Keyword::Unset);
auto expanded_value_tokens = TokenStream { expanded_values };
if (auto parsed_value = parse_css_value(property_id, expanded_value_tokens); !parsed_value.is_error())
return parsed_value.release_value();
return CSSKeywordValue::create(Keyword::Unset);
}
LibWeb/CSS: Move and rename PseudoElement types to prep for code gen The upcoming generated types will match those for pseudo-classes: A PseudoElementSelector type, that then holds a PseudoElement enum defining what it is. That enum will be at the top level in the Web::CSS namespace. In order to keep the diffs clearer, this commit renames and moves the types, and then a following one will replace the handwritten enum with a generated one.
2025-03-20 16:56:46 +00:00
static RefPtr<CSSStyleValue const> get_custom_property(DOM::Element const& element, Optional<CSS::PseudoElement> pseudo_element, FlyString const& custom_property_name)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
if (pseudo_element.has_value()) {
if (auto it = element.custom_properties(pseudo_element).find(custom_property_name); it != element.custom_properties(pseudo_element).end())
return it->value.value;
}
for (auto const* current_element = &element; current_element; current_element = current_element->parent_or_shadow_host_element()) {
if (auto it = current_element->custom_properties({}).find(custom_property_name); it != current_element->custom_properties({}).end())
return it->value.value;
}
return nullptr;
}
LibWeb/CSS: Move and rename PseudoElement types to prep for code gen The upcoming generated types will match those for pseudo-classes: A PseudoElementSelector type, that then holds a PseudoElement enum defining what it is. That enum will be at the top level in the Web::CSS namespace. In order to keep the diffs clearer, this commit renames and moves the types, and then a following one will replace the handwritten enum with a generated one.
2025-03-20 16:56:46 +00:00
bool Parser::expand_variables(DOM::Element& element, Optional<PseudoElement> pseudo_element, FlyString const& property_name, HashMap<FlyString, NonnullRefPtr<PropertyDependencyNode>>& dependencies, TokenStream<ComponentValue>& source, Vector<ComponentValue>& dest)
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
{
// Arbitrary large value chosen to avoid the billion-laughs attack.
// https://www.w3.org/TR/css-variables-1/#long-variables
size_t const MAX_VALUE_COUNT = 16384;
if (source.remaining_token_count() + dest.size() > MAX_VALUE_COUNT) {
dbgln("Stopped expanding CSS variables: maximum length reached.");
return false;
}
auto get_dependency_node = [&](FlyString const& name) -> NonnullRefPtr<PropertyDependencyNode> {
if (auto existing = dependencies.get(name); existing.has_value())
return *existing.value();
auto new_node = PropertyDependencyNode::create(name);
dependencies.set(name, new_node);
return new_node;
};
while (source.has_next_token()) {
auto const& value = source.consume_a_token();
if (value.is_block()) {
auto const& source_block = value.block();
Vector<ComponentValue> block_values;
TokenStream source_block_contents { source_block.value };
if (!expand_variables(element, pseudo_element, property_name, dependencies, source_block_contents, block_values))
return false;
dest.empend(SimpleBlock { source_block.token, move(block_values) });
continue;
}
if (!value.is_function()) {
dest.empend(value.token());
continue;
}
if (!value.function().name.equals_ignoring_ascii_case("var"sv)) {
auto const& source_function = value.function();
Vector<ComponentValue> function_values;
TokenStream source_function_contents { source_function.value };
if (!expand_variables(element, pseudo_element, property_name, dependencies, source_function_contents, function_values))
return false;
dest.empend(Function { source_function.name, move(function_values) });
continue;
}
TokenStream var_contents { value.function().value };
var_contents.discard_whitespace();
if (!var_contents.has_next_token())
return false;
auto const& custom_property_name_token = var_contents.consume_a_token();
if (!custom_property_name_token.is(Token::Type::Ident))
return false;
auto custom_property_name = custom_property_name_token.token().ident();
if (!custom_property_name.bytes_as_string_view().starts_with("--"sv))
return false;
// Detect dependency cycles. https://www.w3.org/TR/css-variables-1/#cycles
// We do not do this by the spec, since we are not keeping a graph of var dependencies around,
// but rebuilding it every time.
if (custom_property_name == property_name)
return false;
auto parent = get_dependency_node(property_name);
auto child = get_dependency_node(custom_property_name);
parent->add_child(child);
if (parent->has_cycles())
return false;
if (auto custom_property_value = get_custom_property(element, pseudo_element, custom_property_name)) {
VERIFY(custom_property_value->is_unresolved());
TokenStream custom_property_tokens { custom_property_value->as_unresolved().values() };
LibWeb/CSS: Use fallback var() value if custom property is empty If the expansion of a custom property in variable expansion returns tokens, then the custom property is not the initial guaranteed-invalid value. If it didn't return any tokens, then it is the initial guaranteed-invalid value, and thus we should move on to the fallback value. Makes Shopify checkout show the background colours, borders, skeletons, etc.
2025-02-14 20:00:10 +00:00
auto dest_size_before = dest.size();
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (!expand_variables(element, pseudo_element, custom_property_name, dependencies, custom_property_tokens, dest))
return false;
LibWeb/CSS: Use fallback var() value if custom property is empty If the expansion of a custom property in variable expansion returns tokens, then the custom property is not the initial guaranteed-invalid value. If it didn't return any tokens, then it is the initial guaranteed-invalid value, and thus we should move on to the fallback value. Makes Shopify checkout show the background colours, borders, skeletons, etc.
2025-02-14 20:00:10 +00:00
// If the size of dest has increased, then the custom property is not the initial guaranteed-invalid value.
// If it hasn't increased, then it is the initial guaranteed-invalid value, and thus we should move on to the fallback value.
if (dest_size_before < dest.size())
continue;
dbgln_if(CSS_PARSER_DEBUG, "CSSParser: Expanding custom property '{}' did not return any tokens, treating it as invalid and moving on to the fallback value.", custom_property_name);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
}
// Use the provided fallback value, if any.
var_contents.discard_whitespace();
if (var_contents.has_next_token()) {
auto const& comma_token = var_contents.consume_a_token();
if (!comma_token.is(Token::Type::Comma))
return false;
var_contents.discard_whitespace();
if (!expand_variables(element, pseudo_element, property_name, dependencies, var_contents, dest))
return false;
}
}
return true;
}
bool Parser::expand_unresolved_values(DOM::Element& element, FlyString const& property_name, TokenStream<ComponentValue>& source, Vector<ComponentValue>& dest)
{
while (source.has_next_token()) {
auto const& value = source.consume_a_token();
if (value.is_function()) {
if (value.function().name.equals_ignoring_ascii_case("attr"sv)) {
if (!substitute_attr_function(element, property_name, value.function(), dest))
return false;
continue;
}
auto const& source_function = value.function();
Vector<ComponentValue> function_values;
TokenStream source_function_contents { source_function.value };
if (!expand_unresolved_values(element, property_name, source_function_contents, function_values))
return false;
dest.empend(Function { source_function.name, move(function_values) });
continue;
}
if (value.is_block()) {
auto const& source_block = value.block();
TokenStream source_block_values { source_block.value };
Vector<ComponentValue> block_values;
if (!expand_unresolved_values(element, property_name, source_block_values, block_values))
return false;
dest.empend(SimpleBlock { source_block.token, move(block_values) });
continue;
}
dest.empend(value.token());
}
return true;
}
// https://drafts.csswg.org/css-values-5/#attr-substitution
bool Parser::substitute_attr_function(DOM::Element& element, FlyString const& property_name, Function const& attr_function, Vector<ComponentValue>& dest)
{
// First, parse the arguments to attr():
// attr() = attr( <q-name> <attr-type>? , <declaration-value>?)
// <attr-type> = string | url | ident | color | number | percentage | length | angle | time | frequency | flex | <dimension-unit>
TokenStream attr_contents { attr_function.value };
attr_contents.discard_whitespace();
if (!attr_contents.has_next_token())
return false;
// - Attribute name
// FIXME: Support optional attribute namespace
if (!attr_contents.next_token().is(Token::Type::Ident))
return false;
auto attribute_name = attr_contents.consume_a_token().token().ident();
attr_contents.discard_whitespace();
// - Attribute type (optional)
auto attribute_type = "string"_fly_string;
if (attr_contents.next_token().is(Token::Type::Ident)) {
attribute_type = attr_contents.consume_a_token().token().ident();
attr_contents.discard_whitespace();
}
// - Comma, then fallback values (optional)
bool has_fallback_values = false;
if (attr_contents.has_next_token()) {
if (!attr_contents.next_token().is(Token::Type::Comma))
return false;
(void)attr_contents.consume_a_token(); // Comma
has_fallback_values = true;
}
// Then, run the substitution algorithm:
// 1. If the attr() function has a substitution value, replace the attr() function by the substitution value.
// https://drafts.csswg.org/css-values-5/#attr-types
if (element.has_attribute(attribute_name)) {
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto parse_string_as_component_value = [this](String const& string) {
auto tokens = Tokenizer::tokenize(string, "utf-8"sv);
TokenStream stream { tokens };
return parse_a_component_value(stream);
};
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
auto attribute_value = element.get_attribute_value(attribute_name);
if (attribute_type.equals_ignoring_ascii_case("angle"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <dimension-token> whose unit matches the given type, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Dimension)) {
if (Angle::unit_from_name(component_value->token().dimension_unit()).has_value()) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("color"_fly_string)) {
// Parse a component value from the attribute’s value.
// If the result is a <hex-color> or a named color ident, the substitution value is that result as a <color>.
// Otherwise there is no substitution value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (component_value.has_value()) {
if ((component_value->is(Token::Type::Hash)
&& Color::from_string(MUST(String::formatted("#{}", component_value->token().hash_value()))).has_value())
|| (component_value->is(Token::Type::Ident)
&& Color::from_string(component_value->token().ident()).has_value())) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("flex"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <dimension-token> whose unit matches the given type, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Dimension)) {
if (Flex::unit_from_name(component_value->token().dimension_unit()).has_value()) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("frequency"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <dimension-token> whose unit matches the given type, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Dimension)) {
if (Frequency::unit_from_name(component_value->token().dimension_unit()).has_value()) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("ident"_fly_string)) {
// The substitution value is a CSS <custom-ident>, whose value is the literal value of the attribute,
// with leading and trailing ASCII whitespace stripped. (No CSS parsing of the value is performed.)
// If the attribute value, after trimming, is the empty string, there is instead no substitution value.
// If the <custom-ident>’s value is a CSS-wide keyword or `default`, there is instead no substitution value.
auto substitution_value = MUST(attribute_value.trim(Infra::ASCII_WHITESPACE));
if (!substitution_value.is_empty()
&& !substitution_value.equals_ignoring_ascii_case("default"sv)
&& !is_css_wide_keyword(substitution_value)) {
dest.empend(Token::create_ident(substitution_value));
return true;
}
} else if (attribute_type.equals_ignoring_ascii_case("length"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <dimension-token> whose unit matches the given type, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Dimension)) {
if (Length::unit_from_name(component_value->token().dimension_unit()).has_value()) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("number"_fly_string)) {
// Parse a component value from the attribute’s value.
// If the result is a <number-token>, the result is the substitution value.
// Otherwise, there is no substitution value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (component_value.has_value() && component_value->is(Token::Type::Number)) {
dest.append(component_value.release_value());
return true;
}
} else if (attribute_type.equals_ignoring_ascii_case("percentage"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <percentage-token>, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Percentage)) {
dest.append(component_value.release_value());
return true;
}
} else if (attribute_type.equals_ignoring_ascii_case("string"_fly_string)) {
// The substitution value is a CSS string, whose value is the literal value of the attribute.
// (No CSS parsing or "cleanup" of the value is performed.)
// No value triggers fallback.
dest.empend(Token::create_string(attribute_value));
return true;
} else if (attribute_type.equals_ignoring_ascii_case("time"_fly_string)) {
// Parse a component value from the attribute’s value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
// If the result is a <dimension-token> whose unit matches the given type, the result is the substitution value.
// Otherwise, there is no substitution value.
if (component_value.has_value() && component_value->is(Token::Type::Dimension)) {
if (Time::unit_from_name(component_value->token().dimension_unit()).has_value()) {
dest.append(component_value.release_value());
return true;
}
}
} else if (attribute_type.equals_ignoring_ascii_case("url"_fly_string)) {
// The substitution value is a CSS <url> value, whose url is the literal value of the attribute.
// (No CSS parsing or "cleanup" of the value is performed.)
// No value triggers fallback.
dest.empend(Token::create_url(attribute_value));
return true;
} else {
// Dimension units
// Parse a component value from the attribute’s value.
// If the result is a <number-token>, the substitution value is a dimension with the result’s value, and the given unit.
// Otherwise, there is no substitution value.
LibWeb/CSS: Avoid creating a new Parser unnecessarily This attr() code can instead just tokenize the input manually, and then re-use the existing Parser.
2025-02-05 12:01:32 +00:00
auto component_value = parse_string_as_component_value(attribute_value);
LibWeb/CSS: Split up Parser.cpp This file has been a pain to edit for a while, even with the previous splits. So, I've divided it up into 3 parts: - Parser.cpp has the "base" code. It's the algorithms and entry-points defined in the Syntax spec. - ValueParsing.cpp contains code for parsing single values, such as a length, or a color, or a calculation. - PropertyParsing.cpp contains code for parsing an entire property's value. A few of these sit in a grey area between being a property's value and a value in their own right, but the rule I've used is "is this useful outside of a single property and its shorthands?" This only moves code, with as few modifications as possible to make that work. I did add explicit instantiations for the template implementations as part of this, which revealed a few that are actually only compatible with a single type, so I'll clear those up in a subsequent commit.
2025-02-06 14:08:26 +00:00
if (component_value.has_value() && component_value->is(Token::Type::Number)) {
if (attribute_value == "%"sv) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else if (auto angle_unit = Angle::unit_from_name(attribute_type); angle_unit.has_value()) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else if (auto flex_unit = Flex::unit_from_name(attribute_type); flex_unit.has_value()) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else if (auto frequency_unit = Frequency::unit_from_name(attribute_type); frequency_unit.has_value()) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else if (auto length_unit = Length::unit_from_name(attribute_type); length_unit.has_value()) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else if (auto time_unit = Time::unit_from_name(attribute_type); time_unit.has_value()) {
dest.empend(Token::create_dimension(component_value->token().number_value(), attribute_type));
return true;
} else {
// Not a dimension unit.
return false;
}
}
}
}
// 2. Otherwise, if the attr() function has a fallback value as its last argument, replace the attr() function by the fallback value.
// If there are any var() or attr() references in the fallback, substitute them as well.
if (has_fallback_values)
return expand_unresolved_values(element, property_name, attr_contents, dest);
if (attribute_type.equals_ignoring_ascii_case("string"_fly_string)) {
// If the <attr-type> argument is string, defaults to the empty string if omitted
dest.empend(Token::create_string({}));
return true;
}
// 3. Otherwise, the property containing the attr() function is invalid at computed-value time.
return false;
}
}
Reference in a new issue Copy permalink