LibWeb/CSS: Bring TokenStream in line with spec

When the TokenStream code was originally written, there was no such
concept in the CSS Syntax spec. But since then, it's been officially
added, (https://drafts.csswg.org/css-syntax/#css-token-stream) and the
parsing algorithms are described in terms of it. This patch brings our
implementation in line with the spec. A few deprecated TokenStream
methods are left around until their users are also updated to match the
newer spec.

There are a few differences:

- They name things differently. The main confusing one is we had
  `next_token()` which consumed a token and returned it, but the spec
  has a `next_token()` which peeks the next token. The spec names are
  honestly better than what I'd come up with. (`discard_a_token()` is a
  nice addition too!)

- We used to store the index of the token that was just consumed, and
  they instead store the index of the token that will be consumed next.
  This is a perfect breeding ground for off-by-one errors, so I've
  finally added a test suite for TokenStream itself.

- We use a transaction system for rewinding, and the spec uses a stack
  of "marks", which can be manually rewound to. These should be able to
  coexist as long as we stick with marks in the parser spec algorithms,
  and stick with transactions elsewhere.
This commit is contained in:
Sam Atkins 2024-10-09 12:29:29 +01:00 committed by Sam Atkins
parent 5df6c6eecf
commit b645e26e9b
Notes: github-actions[bot] 2024-10-09 16:30:23 +00:00
8 changed files with 763 additions and 603 deletions

View file

@ -97,15 +97,15 @@ namespace Web::CSS::Parser {
static Optional<RoundingStrategy> parse_rounding_strategy(Vector<ComponentValue> const& tokens)
{
auto stream = TokenStream { tokens };
stream.skip_whitespace();
stream.discard_whitespace();
if (!stream.has_next_token())
return {};
auto& ident = stream.next_token();
auto& ident = stream.consume_a_token();
if (!ident.is(Token::Type::Ident))
return {};
stream.skip_whitespace();
stream.discard_whitespace();
if (stream.has_next_token())
return {};

View file

@ -1,6 +1,7 @@
set(TEST_SOURCES
TestCSSIDSpeed.cpp
TestCSSPixels.cpp
TestCSSTokenStream.cpp
TestFetchInfrastructure.cpp
TestFetchURL.cpp
TestHTMLTokenizer.cpp

View file

@ -0,0 +1,98 @@
/*
* Copyright (c) 2024, Sam Atkins <sam@ladybird.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <AK/FlyString.h>
#include <AK/Vector.h>
#include <LibTest/TestCase.h>
#include <LibWeb/CSS/Parser/TokenStream.h>
namespace Web::CSS::Parser {
TEST_CASE(basic)
{
Vector<Token> tokens {
Token::create_ident("hello"_fly_string),
};
TokenStream stream { tokens };
EXPECT(!stream.is_empty());
EXPECT(stream.has_next_token());
EXPECT_EQ(stream.remaining_token_count(), 1u);
// next_token() doesn't consume it
auto const& next = stream.next_token();
EXPECT(!stream.is_empty());
EXPECT(stream.has_next_token());
EXPECT_EQ(stream.remaining_token_count(), 1u);
// Check what the token is
EXPECT(next.is(Token::Type::Ident));
EXPECT_EQ(next.ident(), "hello"_fly_string);
// consume_a_token() does consume it
auto const& consumed = stream.consume_a_token();
EXPECT(stream.is_empty());
EXPECT(!stream.has_next_token());
EXPECT_EQ(stream.remaining_token_count(), 0u);
// Check what the token is
EXPECT(consumed.is(Token::Type::Ident));
EXPECT_EQ(consumed.ident(), "hello"_fly_string);
// Now, any further tokens should be EOF
EXPECT(stream.next_token().is(Token::Type::EndOfFile));
EXPECT(stream.consume_a_token().is(Token::Type::EndOfFile));
}
TEST_CASE(marks)
{
Vector<Token> tokens {
Token::create_ident("a"_fly_string),
Token::create_ident("b"_fly_string),
Token::create_ident("c"_fly_string),
Token::create_ident("d"_fly_string),
Token::create_ident("e"_fly_string),
Token::create_ident("f"_fly_string),
Token::create_ident("g"_fly_string),
};
TokenStream stream { tokens };
stream.mark(); // 0
EXPECT_EQ(stream.remaining_token_count(), 7u);
stream.discard_a_token();
stream.discard_a_token();
stream.discard_a_token();
EXPECT_EQ(stream.remaining_token_count(), 4u);
stream.mark(); // 3
stream.discard_a_token();
EXPECT_EQ(stream.remaining_token_count(), 3u);
stream.restore_a_mark(); // Back to 3
EXPECT_EQ(stream.remaining_token_count(), 4u);
stream.discard_a_token();
stream.discard_a_token();
stream.discard_a_token();
EXPECT_EQ(stream.remaining_token_count(), 1u);
stream.mark(); // 6
stream.discard_a_mark();
EXPECT_EQ(stream.remaining_token_count(), 1u);
stream.restore_a_mark(); // Back to 0
EXPECT_EQ(stream.remaining_token_count(), 7u);
}
}

View file

@ -27,20 +27,20 @@ Optional<Vector<TElement>> Parser::parse_color_stop_list(TokenStream<ComponentVa
};
auto parse_color_stop_list_element = [&](TElement& element) -> ElementType {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return ElementType::Garbage;
RefPtr<CSSStyleValue> color;
Optional<typename TElement::PositionType> position;
Optional<typename TElement::PositionType> second_position;
if (auto dimension = parse_dimension(tokens.peek_token()); dimension.has_value() && is_position(*dimension)) {
if (auto dimension = parse_dimension(tokens.next_token()); dimension.has_value() && is_position(*dimension)) {
// [<T-percentage> <color>] or [<T-percentage>]
position = get_position(*dimension);
(void)tokens.next_token(); // dimension
tokens.skip_whitespace();
tokens.discard_a_token(); // dimension
tokens.discard_whitespace();
// <T-percentage>
if (!tokens.has_next_token() || tokens.peek_token().is(Token::Type::Comma)) {
if (!tokens.has_next_token() || tokens.next_token().is(Token::Type::Comma)) {
element.transition_hint = typename TElement::ColorHint { *position };
return ElementType::ColorHint;
}
@ -55,16 +55,16 @@ Optional<Vector<TElement>> Parser::parse_color_stop_list(TokenStream<ComponentVa
if (!maybe_color)
return ElementType::Garbage;
color = maybe_color.release_nonnull();
tokens.skip_whitespace();
tokens.discard_whitespace();
// Allow up to [<color> <T-percentage> <T-percentage>] (double-position color stops)
// Note: Double-position color stops only appear to be valid in this order.
for (auto stop_position : Array { &position, &second_position }) {
if (tokens.has_next_token() && !tokens.peek_token().is(Token::Type::Comma)) {
auto dimension = parse_dimension(tokens.next_token());
if (tokens.has_next_token() && !tokens.next_token().is(Token::Type::Comma)) {
auto dimension = parse_dimension(tokens.consume_a_token());
if (!dimension.has_value() || !is_position(*dimension))
return ElementType::Garbage;
*stop_position = get_position(*dimension);
tokens.skip_whitespace();
tokens.discard_whitespace();
}
}
}
@ -83,14 +83,14 @@ Optional<Vector<TElement>> Parser::parse_color_stop_list(TokenStream<ComponentVa
Vector<TElement> color_stops { first_element };
while (tokens.has_next_token()) {
TElement list_element {};
tokens.skip_whitespace();
if (!tokens.next_token().is(Token::Type::Comma))
tokens.discard_whitespace();
if (!tokens.consume_a_token().is(Token::Type::Comma))
return {};
auto element_type = parse_color_stop_list_element(list_element);
if (element_type == ElementType::ColorHint) {
// <color-hint>, <color-stop>
tokens.skip_whitespace();
if (!tokens.next_token().is(Token::Type::Comma))
tokens.discard_whitespace();
if (!tokens.consume_a_token().is(Token::Type::Comma))
return {};
// Note: This fills in the color stop on the same list_element as the color hint (it does not overwrite it).
if (parse_color_stop_list_element(list_element) != ElementType::ColorStop)
@ -140,7 +140,7 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
using GradientType = LinearGradientStyleValue::GradientType;
auto transaction = outer_tokens.begin_transaction();
auto& component_value = outer_tokens.next_token();
auto& component_value = outer_tokens.consume_a_token();
if (!component_value.is_function())
return nullptr;
@ -164,7 +164,7 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
// linear-gradient() = linear-gradient([ <angle> | to <side-or-corner> ]?, <color-stop-list>)
TokenStream tokens { component_value.function().values() };
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
@ -194,10 +194,10 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
return token.token().ident().equals_ignoring_ascii_case("to"sv);
};
auto const& first_param = tokens.peek_token();
auto const& first_param = tokens.next_token();
if (first_param.is(Token::Type::Dimension)) {
// <angle>
tokens.next_token();
tokens.discard_a_token();
auto angle_value = first_param.token().dimension_value();
auto unit_string = first_param.token().dimension_unit();
auto angle_type = Angle::unit_from_name(unit_string);
@ -211,23 +211,23 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
// Note: -webkit-linear-gradient does not include to the "to" prefix on the side or corner
if (gradient_type == GradientType::Standard) {
tokens.next_token();
tokens.skip_whitespace();
tokens.discard_a_token();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
}
// [left | right] || [top | bottom]
auto const& first_side = tokens.next_token();
auto const& first_side = tokens.consume_a_token();
if (!first_side.is(Token::Type::Ident))
return nullptr;
auto side_a = to_side(first_side.token().ident());
tokens.skip_whitespace();
tokens.discard_whitespace();
Optional<SideOrCorner> side_b;
if (tokens.has_next_token() && tokens.peek_token().is(Token::Type::Ident))
side_b = to_side(tokens.next_token().token().ident());
if (tokens.has_next_token() && tokens.next_token().is(Token::Type::Ident))
side_b = to_side(tokens.consume_a_token().token().ident());
if (side_a.has_value() && !side_b.has_value()) {
gradient_direction = *side_a;
@ -252,11 +252,11 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
has_direction_param = false;
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
if (has_direction_param && !tokens.next_token().is(Token::Type::Comma))
if (has_direction_param && !tokens.consume_a_token().is(Token::Type::Comma))
return nullptr;
auto color_stops = parse_linear_color_stop_list(tokens);
@ -270,7 +270,7 @@ RefPtr<CSSStyleValue> Parser::parse_linear_gradient_function(TokenStream<Compone
RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<ComponentValue>& outer_tokens)
{
auto transaction = outer_tokens.begin_transaction();
auto& component_value = outer_tokens.next_token();
auto& component_value = outer_tokens.consume_a_token();
if (!component_value.is_function())
return nullptr;
@ -287,7 +287,7 @@ RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<Componen
return nullptr;
TokenStream tokens { component_value.function().values() };
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
@ -297,7 +297,7 @@ RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<Componen
// conic-gradient( [ [ from <angle> ]? [ at <position> ]? ] ||
// <color-interpolation-method> , <angular-color-stop-list> )
auto token = tokens.peek_token();
auto token = tokens.next_token();
bool got_from_angle = false;
bool got_color_interpolation_method = false;
bool got_at_position = false;
@ -305,8 +305,8 @@ RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<Componen
auto consume_identifier = [&](auto identifier) {
auto token_string = token.token().ident();
if (token_string.equals_ignoring_ascii_case(identifier)) {
(void)tokens.next_token();
tokens.skip_whitespace();
tokens.discard_a_token();
tokens.discard_whitespace();
return true;
}
return false;
@ -319,7 +319,7 @@ RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<Componen
if (!tokens.has_next_token())
return nullptr;
auto angle_token = tokens.next_token();
auto angle_token = tokens.consume_a_token();
if (!angle_token.is(Token::Type::Dimension))
return nullptr;
auto angle = angle_token.token().dimension_value();
@ -348,16 +348,16 @@ RefPtr<CSSStyleValue> Parser::parse_conic_gradient_function(TokenStream<Componen
} else {
break;
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
token = tokens.peek_token();
token = tokens.next_token();
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
if ((got_from_angle || got_at_position || got_color_interpolation_method) && !tokens.next_token().is(Token::Type::Comma))
if ((got_from_angle || got_at_position || got_color_interpolation_method) && !tokens.consume_a_token().is(Token::Type::Comma))
return nullptr;
auto color_stops = parse_angular_color_stop_list(tokens);
@ -380,7 +380,7 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
using Size = RadialGradientStyleValue::Size;
auto transaction = outer_tokens.begin_transaction();
auto& component_value = outer_tokens.next_token();
auto& component_value = outer_tokens.consume_a_token();
if (!component_value.is_function())
return nullptr;
@ -397,7 +397,7 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
return nullptr;
TokenStream tokens { component_value.function().values() };
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
@ -416,8 +416,8 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
auto parse_ending_shape = [&]() -> Optional<EndingShape> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
auto& token = tokens.next_token();
tokens.discard_whitespace();
auto& token = tokens.consume_a_token();
if (!token.is(Token::Type::Ident))
return {};
auto ident = token.token().ident();
@ -446,11 +446,11 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
// <length [0,∞]> |
// <length-percentage [0,∞]>{2}
auto transaction_size = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return {};
if (tokens.peek_token().is(Token::Type::Ident)) {
auto extent = parse_extent_keyword(tokens.next_token().token().ident());
if (tokens.next_token().is(Token::Type::Ident)) {
auto extent = parse_extent_keyword(tokens.consume_a_token().token().ident());
if (!extent.has_value())
return {};
return commit_value(*extent, transaction_size);
@ -459,7 +459,7 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
if (!first_radius.has_value())
return {};
auto transaction_second_dimension = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (tokens.has_next_token()) {
auto second_radius = parse_length_percentage(tokens);
if (second_radius.has_value())
@ -494,13 +494,13 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
}
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
auto& token = tokens.peek_token();
auto& token = tokens.next_token();
if (token.is_ident("at"sv)) {
(void)tokens.next_token();
tokens.discard_a_token();
auto position = parse_position_value(tokens);
if (!position)
return nullptr;
@ -508,10 +508,10 @@ RefPtr<CSSStyleValue> Parser::parse_radial_gradient_function(TokenStream<Compone
expect_comma = true;
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return nullptr;
if (expect_comma && !tokens.next_token().is(Token::Type::Comma))
if (expect_comma && !tokens.consume_a_token().is(Token::Type::Comma))
return nullptr;
// <color-stop-list>

View file

@ -29,7 +29,7 @@ Vector<NonnullRefPtr<MediaQuery>> Parser::parse_a_media_query_list(TokenStream<T
// AD-HOC: Ignore whitespace-only queries
// to make `@media {..}` equivalent to `@media all {..}`
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token())
return {};
@ -64,8 +64,8 @@ NonnullRefPtr<MediaQuery> Parser::parse_media_query(TokenStream<ComponentValue>&
// `[ not | only ]?`, Returns whether to negate the query
auto parse_initial_modifier = [](auto& tokens) -> Optional<bool> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
auto& token = tokens.next_token();
tokens.discard_whitespace();
auto& token = tokens.consume_a_token();
if (!token.is(Token::Type::Ident))
return {};
@ -92,11 +92,11 @@ NonnullRefPtr<MediaQuery> Parser::parse_media_query(TokenStream<ComponentValue>&
};
auto media_query = MediaQuery::create();
tokens.skip_whitespace();
tokens.discard_whitespace();
// `<media-condition>`
if (auto media_condition = parse_media_condition(tokens, MediaCondition::AllowOr::Yes)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (tokens.has_next_token())
return invalid_media_query();
media_query->m_media_condition = move(media_condition);
@ -106,13 +106,13 @@ NonnullRefPtr<MediaQuery> Parser::parse_media_query(TokenStream<ComponentValue>&
// `[ not | only ]?`
if (auto modifier = parse_initial_modifier(tokens); modifier.has_value()) {
media_query->m_negated = modifier.value();
tokens.skip_whitespace();
tokens.discard_whitespace();
}
// `<media-type>`
if (auto media_type = parse_media_type(tokens); media_type.has_value()) {
media_query->m_media_type = media_type.value();
tokens.skip_whitespace();
tokens.discard_whitespace();
} else {
return invalid_media_query();
}
@ -121,9 +121,9 @@ NonnullRefPtr<MediaQuery> Parser::parse_media_query(TokenStream<ComponentValue>&
return media_query;
// `[ and <media-condition-without-or> ]?`
if (auto maybe_and = tokens.next_token(); maybe_and.is_ident("and"sv)) {
if (auto maybe_and = tokens.consume_a_token(); maybe_and.is_ident("and"sv)) {
if (auto media_condition = parse_media_condition(tokens, MediaCondition::AllowOr::No)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (tokens.has_next_token())
return invalid_media_query();
media_query->m_media_condition = move(media_condition);
@ -142,14 +142,14 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
{
// `<media-not> | <media-in-parens> [ <media-and>* | <media-or>* ]`
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
// `<media-not> = not <media-in-parens>`
auto parse_media_not = [&](auto& tokens) -> OwnPtr<MediaCondition> {
auto local_transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
auto& first_token = tokens.next_token();
auto& first_token = tokens.consume_a_token();
if (first_token.is_ident("not"sv)) {
if (auto child_condition = parse_media_condition(tokens, MediaCondition::AllowOr::Yes)) {
local_transaction.commit();
@ -162,11 +162,11 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
auto parse_media_with_combinator = [&](auto& tokens, StringView combinator) -> OwnPtr<MediaCondition> {
auto local_transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
auto& first = tokens.next_token();
auto& first = tokens.consume_a_token();
if (first.is_ident(combinator)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto media_in_parens = parse_media_in_parens(tokens)) {
local_transaction.commit();
return media_in_parens;
@ -189,7 +189,7 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
// `<media-in-parens> [ <media-and>* | <media-or>* ]`
if (auto maybe_media_in_parens = parse_media_in_parens(tokens)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
// Only `<media-in-parens>`
if (!tokens.has_next_token()) {
transaction.commit();
@ -203,11 +203,11 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
if (auto media_and = parse_media_and(tokens)) {
child_conditions.append(media_and.release_nonnull());
tokens.skip_whitespace();
tokens.discard_whitespace();
while (tokens.has_next_token()) {
if (auto next_media_and = parse_media_and(tokens)) {
child_conditions.append(next_media_and.release_nonnull());
tokens.skip_whitespace();
tokens.discard_whitespace();
continue;
}
// We failed - invalid syntax!
@ -223,11 +223,11 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
if (auto media_or = parse_media_or(tokens)) {
child_conditions.append(media_or.release_nonnull());
tokens.skip_whitespace();
tokens.discard_whitespace();
while (tokens.has_next_token()) {
if (auto next_media_or = parse_media_or(tokens)) {
child_conditions.append(next_media_or.release_nonnull());
tokens.skip_whitespace();
tokens.discard_whitespace();
continue;
}
// We failed - invalid syntax!
@ -247,7 +247,7 @@ OwnPtr<MediaCondition> Parser::parse_media_condition(TokenStream<ComponentValue>
Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>& tokens)
{
// `[ <mf-plain> | <mf-boolean> | <mf-range> ]`
tokens.skip_whitespace();
tokens.discard_whitespace();
// `<mf-name> = <ident>`
struct MediaFeatureName {
@ -260,7 +260,7 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
};
auto parse_mf_name = [](auto& tokens, bool allow_min_max_prefix) -> Optional<MediaFeatureName> {
auto transaction = tokens.begin_transaction();
auto& token = tokens.next_token();
auto& token = tokens.consume_a_token();
if (token.is(Token::Type::Ident)) {
auto name = token.token().ident();
if (auto id = media_feature_id_from_string(name); id.has_value()) {
@ -285,10 +285,10 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
// `<mf-boolean> = <mf-name>`
auto parse_mf_boolean = [&](auto& tokens) -> Optional<MediaFeature> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_name = parse_mf_name(tokens, false); maybe_name.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token()) {
transaction.commit();
return MediaFeature::boolean(maybe_name->id);
@ -301,14 +301,14 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
// `<mf-plain> = <mf-name> : <mf-value>`
auto parse_mf_plain = [&](auto& tokens) -> Optional<MediaFeature> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_name = parse_mf_name(tokens, true); maybe_name.has_value()) {
tokens.skip_whitespace();
if (tokens.next_token().is(Token::Type::Colon)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (tokens.consume_a_token().is(Token::Type::Colon)) {
tokens.discard_whitespace();
if (auto maybe_value = parse_media_feature_value(maybe_name->id, tokens); maybe_value.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token()) {
transaction.commit();
switch (maybe_name->type) {
@ -333,9 +333,9 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
// <mf-comparison> = <mf-lt> | <mf-gt> | <mf-eq>`
auto parse_comparison = [](auto& tokens) -> Optional<MediaFeature::Comparison> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
auto& first = tokens.next_token();
auto& first = tokens.consume_a_token();
if (first.is(Token::Type::Delim)) {
auto first_delim = first.token().delim();
if (first_delim == '=') {
@ -343,9 +343,9 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
return MediaFeature::Comparison::Equal;
}
if (first_delim == '<') {
auto& second = tokens.peek_token();
auto& second = tokens.next_token();
if (second.is_delim('=')) {
tokens.next_token();
tokens.discard_a_token();
transaction.commit();
return MediaFeature::Comparison::LessThanOrEqual;
}
@ -353,9 +353,9 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
return MediaFeature::Comparison::LessThan;
}
if (first_delim == '>') {
auto& second = tokens.peek_token();
auto& second = tokens.next_token();
if (second.is_delim('=')) {
tokens.next_token();
tokens.discard_a_token();
transaction.commit();
return MediaFeature::Comparison::GreaterThanOrEqual;
}
@ -403,16 +403,16 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
// | <mf-value> <mf-gt> <mf-name> <mf-gt> <mf-value>`
auto parse_mf_range = [&](auto& tokens) -> Optional<MediaFeature> {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
// `<mf-name> <mf-comparison> <mf-value>`
// NOTE: We have to check for <mf-name> first, since all <mf-name>s will also parse as <mf-value>.
if (auto maybe_name = parse_mf_name(tokens, false); maybe_name.has_value() && media_feature_type_is_range(maybe_name->id)) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_comparison = parse_comparison(tokens); maybe_comparison.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_value = parse_media_feature_value(maybe_name->id, tokens); maybe_value.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token() && !maybe_value->is_ident()) {
transaction.commit();
return MediaFeature::half_range(maybe_value.release_value(), flip(maybe_comparison.release_value()), maybe_name->id);
@ -435,23 +435,23 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
while (tokens.has_next_token() && !maybe_name.has_value()) {
if (auto maybe_comparison = parse_comparison(tokens); maybe_comparison.has_value()) {
// We found a comparison, so the next non-whitespace token should be the <mf-name>
tokens.skip_whitespace();
tokens.discard_whitespace();
maybe_name = parse_mf_name(tokens, false);
break;
}
tokens.next_token();
tokens.skip_whitespace();
tokens.discard_a_token();
tokens.discard_whitespace();
}
}
// Now, we can parse the range properly.
if (maybe_name.has_value() && media_feature_type_is_range(maybe_name->id)) {
if (auto maybe_left_value = parse_media_feature_value(maybe_name->id, tokens); maybe_left_value.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_left_comparison = parse_comparison(tokens); maybe_left_comparison.has_value()) {
tokens.skip_whitespace();
tokens.next_token(); // The <mf-name> which we already parsed above.
tokens.skip_whitespace();
tokens.discard_whitespace();
tokens.discard_a_token(); // The <mf-name> which we already parsed above.
tokens.discard_whitespace();
if (!tokens.has_next_token()) {
transaction.commit();
@ -459,9 +459,9 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
}
if (auto maybe_right_comparison = parse_comparison(tokens); maybe_right_comparison.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto maybe_right_value = parse_media_feature_value(maybe_name->id, tokens); maybe_right_value.has_value()) {
tokens.skip_whitespace();
tokens.discard_whitespace();
// For this to be valid, the following must be true:
// - Comparisons must either both be >/>= or both be </<=.
// - Neither comparison can be `=`.
@ -500,8 +500,8 @@ Optional<MediaFeature> Parser::parse_media_feature(TokenStream<ComponentValue>&
Optional<MediaQuery::MediaType> Parser::parse_media_type(TokenStream<ComponentValue>& tokens)
{
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
auto const& token = tokens.next_token();
tokens.discard_whitespace();
auto const& token = tokens.consume_a_token();
if (!token.is(Token::Type::Ident))
return {};
@ -517,19 +517,19 @@ OwnPtr<MediaCondition> Parser::parse_media_in_parens(TokenStream<ComponentValue>
{
// `<media-in-parens> = ( <media-condition> ) | ( <media-feature> ) | <general-enclosed>`
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
// `( <media-condition> ) | ( <media-feature> )`
auto const& first_token = tokens.peek_token();
auto const& first_token = tokens.next_token();
if (first_token.is_block() && first_token.block().is_paren()) {
TokenStream inner_token_stream { first_token.block().values() };
if (auto maybe_media_condition = parse_media_condition(inner_token_stream, MediaCondition::AllowOr::Yes)) {
tokens.next_token();
tokens.discard_a_token();
transaction.commit();
return maybe_media_condition.release_nonnull();
}
if (auto maybe_media_feature = parse_media_feature(inner_token_stream); maybe_media_feature.has_value()) {
tokens.next_token();
tokens.discard_a_token();
transaction.commit();
return MediaCondition::from_feature(maybe_media_feature.release_value());
}
@ -553,10 +553,10 @@ Optional<MediaFeatureValue> Parser::parse_media_feature_value(MediaFeatureID med
// NOTE: Calculations are not allowed for media feature values, at least in the current spec, so we reject them.
// Identifiers
if (tokens.peek_token().is(Token::Type::Ident)) {
if (tokens.next_token().is(Token::Type::Ident)) {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
auto keyword = keyword_from_string(tokens.next_token().token().ident());
tokens.discard_whitespace();
auto keyword = keyword_from_string(tokens.consume_a_token().token().ident());
if (keyword.has_value() && media_feature_accepts_keyword(media_feature, keyword.value())) {
transaction.commit();
return MediaFeatureValue(keyword.value());
@ -568,7 +568,7 @@ Optional<MediaFeatureValue> Parser::parse_media_feature_value(MediaFeatureID med
// Boolean (<mq-boolean> in the spec: a 1 or 0)
if (media_feature_accepts_type(media_feature, MediaFeatureValueType::Boolean)) {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto integer = parse_integer(tokens); integer.has_value() && !integer->is_calculated()) {
auto integer_value = integer->value();
if (integer_value == 0 || integer_value == 1) {
@ -590,7 +590,7 @@ Optional<MediaFeatureValue> Parser::parse_media_feature_value(MediaFeatureID med
// Length
if (media_feature_accepts_type(media_feature, MediaFeatureValueType::Length)) {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto length = parse_length(tokens); length.has_value() && !length->is_calculated()) {
transaction.commit();
return MediaFeatureValue(length->value());
@ -600,7 +600,7 @@ Optional<MediaFeatureValue> Parser::parse_media_feature_value(MediaFeatureID med
// Ratio
if (media_feature_accepts_type(media_feature, MediaFeatureValueType::Ratio)) {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto ratio = parse_ratio(tokens); ratio.has_value()) {
transaction.commit();
return MediaFeatureValue(ratio.release_value());
@ -610,7 +610,7 @@ Optional<MediaFeatureValue> Parser::parse_media_feature_value(MediaFeatureID med
// Resolution
if (media_feature_accepts_type(media_feature, MediaFeatureValueType::Resolution)) {
auto transaction = tokens.begin_transaction();
tokens.skip_whitespace();
tokens.discard_whitespace();
if (auto resolution = parse_resolution(tokens); resolution.has_value() && !resolution->is_calculated()) {
transaction.commit();
return MediaFeatureValue(resolution->value());

File diff suppressed because it is too large Load diff

View file

@ -115,11 +115,11 @@ Parser::ParseErrorOr<NonnullRefPtr<Selector>> Parser::parse_complex_selector(Tok
Parser::ParseErrorOr<Optional<Selector::CompoundSelector>> Parser::parse_compound_selector(TokenStream<ComponentValue>& tokens)
{
tokens.skip_whitespace();
tokens.discard_whitespace();
auto combinator = parse_selector_combinator(tokens).value_or(Selector::Combinator::Descendant);
tokens.skip_whitespace();
tokens.discard_whitespace();
Vector<Selector::SimpleSelector> simple_selectors;
@ -138,7 +138,7 @@ Parser::ParseErrorOr<Optional<Selector::CompoundSelector>> Parser::parse_compoun
Optional<Selector::Combinator> Parser::parse_selector_combinator(TokenStream<ComponentValue>& tokens)
{
auto const& current_value = tokens.next_token();
auto const& current_value = tokens.consume_a_token();
if (current_value.is(Token::Type::Delim)) {
switch (current_value.token().delim()) {
case '>':
@ -148,12 +148,12 @@ Optional<Selector::Combinator> Parser::parse_selector_combinator(TokenStream<Com
case '~':
return Selector::Combinator::SubsequentSibling;
case '|': {
auto const& next = tokens.peek_token();
auto const& next = tokens.next_token();
if (next.is(Token::Type::EndOfFile))
return {};
if (next.is_delim('|')) {
tokens.next_token();
tokens.discard_a_token();
return Selector::Combinator::Column;
}
}
@ -184,11 +184,11 @@ Optional<Selector::SimpleSelector::QualifiedName> Parser::parse_selector_qualifi
auto transaction = tokens.begin_transaction();
auto first_token = tokens.next_token();
auto first_token = tokens.consume_a_token();
if (first_token.is_delim('|')) {
// Case 1: `|<name>`
if (is_name(tokens.peek_token())) {
auto name_token = tokens.next_token();
if (is_name(tokens.next_token())) {
auto name_token = tokens.consume_a_token();
if (allow_wildcard_name == AllowWildcardName::No && name_token.is_delim('*'))
return {};
@ -205,11 +205,11 @@ Optional<Selector::SimpleSelector::QualifiedName> Parser::parse_selector_qualifi
if (!is_name(first_token))
return {};
if (tokens.peek_token().is_delim('|') && is_name(tokens.peek_token(1))) {
if (tokens.next_token().is_delim('|') && is_name(tokens.peek_token(1))) {
// Case 2: `<namespace>|<name>`
(void)tokens.next_token(); // `|`
tokens.discard_a_token(); // `|`
auto namespace_ = get_name(first_token);
auto name = get_name(tokens.next_token());
auto name = get_name(tokens.consume_a_token());
if (allow_wildcard_name == AllowWildcardName::No && name == "*"sv)
return {};
@ -242,7 +242,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
{
auto attribute_tokens = TokenStream { first_value.block().values() };
attribute_tokens.skip_whitespace();
attribute_tokens.discard_whitespace();
if (!attribute_tokens.has_next_token()) {
dbgln_if(CSS_PARSER_DEBUG, "CSS attribute selector is empty!");
@ -251,7 +251,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
auto maybe_qualified_name = parse_selector_qualified_name(attribute_tokens, AllowWildcardName::No);
if (!maybe_qualified_name.has_value()) {
dbgln_if(CSS_PARSER_DEBUG, "Expected qualified-name for attribute name, got: '{}'", attribute_tokens.peek_token().to_debug_string());
dbgln_if(CSS_PARSER_DEBUG, "Expected qualified-name for attribute name, got: '{}'", attribute_tokens.next_token().to_debug_string());
return ParseError::SyntaxError;
}
auto qualified_name = maybe_qualified_name.release_value();
@ -265,11 +265,11 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
}
};
attribute_tokens.skip_whitespace();
attribute_tokens.discard_whitespace();
if (!attribute_tokens.has_next_token())
return simple_selector;
auto const& delim_part = attribute_tokens.next_token();
auto const& delim_part = attribute_tokens.consume_a_token();
if (!delim_part.is(Token::Type::Delim)) {
dbgln_if(CSS_PARSER_DEBUG, "Expected a delim for attribute comparison, got: '{}'", delim_part.to_debug_string());
return ParseError::SyntaxError;
@ -283,7 +283,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
return ParseError::SyntaxError;
}
auto const& delim_second_part = attribute_tokens.next_token();
auto const& delim_second_part = attribute_tokens.consume_a_token();
if (!delim_second_part.is_delim('=')) {
dbgln_if(CSS_PARSER_DEBUG, "Expected a double delim for attribute comparison, got: '{}{}'", delim_part.to_debug_string(), delim_second_part.to_debug_string());
return ParseError::SyntaxError;
@ -309,13 +309,13 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
}
}
attribute_tokens.skip_whitespace();
attribute_tokens.discard_whitespace();
if (!attribute_tokens.has_next_token()) {
dbgln_if(CSS_PARSER_DEBUG, "Attribute selector ended without a value to match.");
return ParseError::SyntaxError;
}
auto const& value_part = attribute_tokens.next_token();
auto const& value_part = attribute_tokens.consume_a_token();
if (!value_part.is(Token::Type::Ident) && !value_part.is(Token::Type::String)) {
dbgln_if(CSS_PARSER_DEBUG, "Expected a string or ident for the value to match attribute against, got: '{}'", value_part.to_debug_string());
return ParseError::SyntaxError;
@ -323,10 +323,10 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
auto const& value_string = value_part.token().is(Token::Type::Ident) ? value_part.token().ident() : value_part.token().string();
simple_selector.attribute().value = value_string.to_string();
attribute_tokens.skip_whitespace();
attribute_tokens.discard_whitespace();
// Handle case-sensitivity suffixes. https://www.w3.org/TR/selectors-4/#attribute-case
if (attribute_tokens.has_next_token()) {
auto const& case_sensitivity_part = attribute_tokens.next_token();
auto const& case_sensitivity_part = attribute_tokens.consume_a_token();
if (case_sensitivity_part.is(Token::Type::Ident)) {
auto case_sensitivity = case_sensitivity_part.token().ident();
if (case_sensitivity.equals_ignoring_ascii_case("i"sv)) {
@ -354,7 +354,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_attribute_simple_se
Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selector(TokenStream<ComponentValue>& tokens)
{
auto peek_token_ends_selector = [&]() -> bool {
auto const& value = tokens.peek_token();
auto const& value = tokens.next_token();
return (value.is(Token::Type::EndOfFile) || value.is(Token::Type::Whitespace) || value.is(Token::Type::Comma));
};
@ -362,15 +362,15 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
return ParseError::SyntaxError;
bool is_pseudo = false;
if (tokens.peek_token().is(Token::Type::Colon)) {
if (tokens.next_token().is(Token::Type::Colon)) {
is_pseudo = true;
tokens.next_token();
tokens.discard_a_token();
if (peek_token_ends_selector())
return ParseError::SyntaxError;
}
if (is_pseudo) {
auto const& name_token = tokens.next_token();
auto const& name_token = tokens.consume_a_token();
if (!name_token.is(Token::Type::Ident)) {
dbgln_if(CSS_PARSER_DEBUG, "Expected an ident for pseudo-element, got: '{}'", name_token.to_debug_string());
return ParseError::SyntaxError;
@ -409,7 +409,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
if (peek_token_ends_selector())
return ParseError::SyntaxError;
auto const& pseudo_class_token = tokens.next_token();
auto const& pseudo_class_token = tokens.consume_a_token();
if (pseudo_class_token.is(Token::Type::Ident)) {
auto pseudo_name = pseudo_class_token.token().ident();
@ -461,7 +461,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
return ParseError::SyntaxError;
}
tokens.skip_whitespace();
tokens.discard_whitespace();
if (!tokens.has_next_token()) {
return Selector::SimpleSelector {
.type = Selector::SimpleSelector::Type::PseudoClass,
@ -475,14 +475,14 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
return ParseError::SyntaxError;
// Parse the `of <selector-list>` syntax
auto const& maybe_of = tokens.next_token();
auto const& maybe_of = tokens.consume_a_token();
if (!maybe_of.is_ident("of"sv))
return ParseError::SyntaxError;
tokens.skip_whitespace();
tokens.discard_whitespace();
auto selector_list = TRY(parse_a_selector_list(tokens, SelectorType::Standalone));
tokens.skip_whitespace();
tokens.discard_whitespace();
if (tokens.has_next_token())
return ParseError::SyntaxError;
@ -558,9 +558,9 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
}
case PseudoClassMetadata::ParameterType::Ident: {
auto function_token_stream = TokenStream(pseudo_function.values());
function_token_stream.skip_whitespace();
auto maybe_keyword_token = function_token_stream.next_token();
function_token_stream.skip_whitespace();
function_token_stream.discard_whitespace();
auto maybe_keyword_token = function_token_stream.consume_a_token();
function_token_stream.discard_whitespace();
if (!maybe_keyword_token.is(Token::Type::Ident) || function_token_stream.has_next_token()) {
dbgln_if(CSS_PARSER_DEBUG, "Failed to parse :{}() parameter as a keyword: not an ident", pseudo_function.name());
return ParseError::SyntaxError;
@ -586,8 +586,8 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
for (auto language_token_list : language_token_lists) {
auto language_token_stream = TokenStream(language_token_list);
language_token_stream.skip_whitespace();
auto language_token = language_token_stream.next_token();
language_token_stream.discard_whitespace();
auto language_token = language_token_stream.consume_a_token();
if (!(language_token.is(Token::Type::Ident) || language_token.is(Token::Type::String))) {
dbgln_if(CSS_PARSER_DEBUG, "Invalid language range in :{}() - not a string/ident", pseudo_function.name());
return ParseError::SyntaxError;
@ -596,7 +596,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
auto language_string = language_token.is(Token::Type::String) ? language_token.token().string() : language_token.token().ident();
languages.append(language_string);
language_token_stream.skip_whitespace();
language_token_stream.discard_whitespace();
if (language_token_stream.has_next_token()) {
dbgln_if(CSS_PARSER_DEBUG, "Invalid language range in :{}() - trailing tokens", pseudo_function.name());
return ParseError::SyntaxError;
@ -633,7 +633,7 @@ Parser::ParseErrorOr<Selector::SimpleSelector> Parser::parse_pseudo_simple_selec
Parser::ParseErrorOr<Optional<Selector::SimpleSelector>> Parser::parse_simple_selector(TokenStream<ComponentValue>& tokens)
{
auto peek_token_ends_selector = [&]() -> bool {
auto const& value = tokens.peek_token();
auto const& value = tokens.next_token();
return (value.is(Token::Type::EndOfFile) || value.is(Token::Type::Whitespace) || value.is(Token::Type::Comma));
};
@ -654,7 +654,7 @@ Parser::ParseErrorOr<Optional<Selector::SimpleSelector>> Parser::parse_simple_se
};
}
auto const& first_value = tokens.next_token();
auto const& first_value = tokens.consume_a_token();
if (first_value.is(Token::Type::Delim)) {
u32 delim = first_value.token().delim();
@ -666,7 +666,7 @@ Parser::ParseErrorOr<Optional<Selector::SimpleSelector>> Parser::parse_simple_se
if (peek_token_ends_selector())
return ParseError::SyntaxError;
auto const& class_name_value = tokens.next_token();
auto const& class_name_value = tokens.consume_a_token();
if (!class_name_value.is(Token::Type::Ident)) {
dbgln_if(CSS_PARSER_DEBUG, "Expected an ident after '.', got: {}", class_name_value.to_debug_string());
return ParseError::SyntaxError;
@ -796,8 +796,8 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// https://www.w3.org/TR/css-syntax-3/#the-anb-type
// Unfortunately these can't be in the same order as in the spec.
values.skip_whitespace();
auto const& first_value = values.next_token();
values.discard_whitespace();
auto const& first_value = values.consume_a_token();
// odd | even
if (first_value.is(Token::Type::Ident)) {
@ -822,11 +822,11 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// <n-dimension> ['+' | '-'] <signless-integer>
if (is_n_dimension(first_value)) {
int a = first_value.token().dimension_value_int();
values.skip_whitespace();
values.discard_whitespace();
// <n-dimension> <signed-integer>
if (is_signed_integer(values.peek_token())) {
int b = values.next_token().token().to_integer();
if (is_signed_integer(values.next_token())) {
int b = values.consume_a_token().token().to_integer();
transaction.commit();
return Selector::SimpleSelector::ANPlusBPattern { a, b };
}
@ -834,9 +834,9 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// <n-dimension> ['+' | '-'] <signless-integer>
{
auto child_transaction = transaction.create_child();
auto const& second_value = values.next_token();
values.skip_whitespace();
auto const& third_value = values.next_token();
auto const& second_value = values.consume_a_token();
values.discard_whitespace();
auto const& third_value = values.consume_a_token();
if (is_sign(second_value) && is_signless_integer(third_value)) {
int b = third_value.token().to_integer() * (second_value.is_delim('+') ? 1 : -1);
@ -851,8 +851,8 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
}
// <ndash-dimension> <signless-integer>
if (is_ndash_dimension(first_value)) {
values.skip_whitespace();
auto const& second_value = values.next_token();
values.discard_whitespace();
auto const& second_value = values.consume_a_token();
if (is_signless_integer(second_value)) {
int a = first_value.token().dimension_value_int();
int b = -second_value.token().to_integer();
@ -888,11 +888,11 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// -n <signed-integer>
// -n ['+' | '-'] <signless-integer>
if (first_value.is_ident("-n"sv)) {
values.skip_whitespace();
values.discard_whitespace();
// -n <signed-integer>
if (is_signed_integer(values.peek_token())) {
int b = values.next_token().token().to_integer();
if (is_signed_integer(values.next_token())) {
int b = values.consume_a_token().token().to_integer();
transaction.commit();
return Selector::SimpleSelector::ANPlusBPattern { -1, b };
}
@ -900,9 +900,9 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// -n ['+' | '-'] <signless-integer>
{
auto child_transaction = transaction.create_child();
auto const& second_value = values.next_token();
values.skip_whitespace();
auto const& third_value = values.next_token();
auto const& second_value = values.consume_a_token();
values.discard_whitespace();
auto const& third_value = values.consume_a_token();
if (is_sign(second_value) && is_signless_integer(third_value)) {
int b = third_value.token().to_integer() * (second_value.is_delim('+') ? 1 : -1);
@ -917,8 +917,8 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
}
// -n- <signless-integer>
if (first_value.is_ident("-n-"sv)) {
values.skip_whitespace();
auto const& second_value = values.next_token();
values.discard_whitespace();
auto const& second_value = values.consume_a_token();
if (is_signless_integer(second_value)) {
int b = -second_value.token().to_integer();
transaction.commit();
@ -941,16 +941,16 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// We do *not* skip whitespace here.
}
auto const& first_after_plus = values.next_token();
auto const& first_after_plus = values.consume_a_token();
// '+'?† n
// '+'?† n <signed-integer>
// '+'?† n ['+' | '-'] <signless-integer>
if (first_after_plus.is_ident("n"sv)) {
values.skip_whitespace();
values.discard_whitespace();
// '+'?† n <signed-integer>
if (is_signed_integer(values.peek_token())) {
int b = values.next_token().token().to_integer();
if (is_signed_integer(values.next_token())) {
int b = values.consume_a_token().token().to_integer();
transaction.commit();
return Selector::SimpleSelector::ANPlusBPattern { 1, b };
}
@ -958,9 +958,9 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// '+'?† n ['+' | '-'] <signless-integer>
{
auto child_transaction = transaction.create_child();
auto const& second_value = values.next_token();
values.skip_whitespace();
auto const& third_value = values.next_token();
auto const& second_value = values.consume_a_token();
values.discard_whitespace();
auto const& third_value = values.consume_a_token();
if (is_sign(second_value) && is_signless_integer(third_value)) {
int b = third_value.token().to_integer() * (second_value.is_delim('+') ? 1 : -1);
@ -976,8 +976,8 @@ Optional<Selector::SimpleSelector::ANPlusBPattern> Parser::parse_a_n_plus_b_patt
// '+'?† n- <signless-integer>
if (first_after_plus.is_ident("n-"sv)) {
values.skip_whitespace();
auto const& second_value = values.next_token();
values.discard_whitespace();
auto const& second_value = values.consume_a_token();
if (is_signless_integer(second_value)) {
int b = -second_value.token().to_integer();
transaction.commit();

View file

@ -1,6 +1,6 @@
/*
* Copyright (c) 2020-2021, the SerenityOS developers.
* Copyright (c) 2021-2023, Sam Atkins <atkinssj@serenityos.org>
* Copyright (c) 2021-2024, Sam Atkins <sam@ladybird.org>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
@ -14,6 +14,7 @@
namespace Web::CSS::Parser {
// https://drafts.csswg.org/css-syntax/#css-token-stream
template<typename T>
class TokenStream {
public:
@ -21,14 +22,14 @@ public:
public:
explicit StateTransaction(TokenStream<T>& token_stream)
: m_token_stream(token_stream)
, m_saved_iterator_offset(token_stream.m_iterator_offset)
, m_saved_index(token_stream.m_index)
{
}
~StateTransaction()
{
if (!m_commit)
m_token_stream.m_iterator_offset = m_saved_iterator_offset;
m_token_stream.m_index = m_saved_index;
}
StateTransaction create_child() { return StateTransaction(*this); }
@ -44,13 +45,13 @@ public:
explicit StateTransaction(StateTransaction& parent)
: m_parent(&parent)
, m_token_stream(parent.m_token_stream)
, m_saved_iterator_offset(parent.m_token_stream.m_iterator_offset)
, m_saved_index(parent.m_token_stream.m_index)
{
}
StateTransaction* m_parent { nullptr };
TokenStream<T>& m_token_stream;
int m_saved_iterator_offset { 0 };
size_t m_saved_index { 0 };
bool m_commit { false };
};
@ -74,60 +75,114 @@ public:
TokenStream(TokenStream<T> const&) = delete;
TokenStream(TokenStream<T>&&) = default;
// https://drafts.csswg.org/css-syntax/#token-stream-next-token
[[nodiscard]] T const& next_token() const
{
// The item of tokens at index.
// If that index would be out-of-bounds past the end of the list, its instead an <eof-token>.
if (m_index < m_tokens.size())
return m_tokens[m_index];
return m_eof;
}
// https://drafts.csswg.org/css-syntax/#token-stream-empty
[[nodiscard]] bool is_empty() const
{
// A token stream is empty if the next token is an <eof-token>.
return next_token().is(Token::Type::EndOfFile);
}
// https://drafts.csswg.org/css-syntax/#token-stream-consume-a-token
[[nodiscard]] T const& consume_a_token()
{
// Let token be the next token. Increment index, then return token.
auto& token = next_token();
++m_index;
return token;
}
// https://drafts.csswg.org/css-syntax/#token-stream-discard-a-token
void discard_a_token()
{
// If the token stream is not empty, increment index.
if (!is_empty())
++m_index;
}
// https://drafts.csswg.org/css-syntax/#token-stream-mark
void mark()
{
// Append index to marked indexes.
m_marked_indexes.append(m_index);
}
// https://drafts.csswg.org/css-syntax/#token-stream-restore-a-mark
void restore_a_mark()
{
// Pop from marked indexes, and set index to the popped value.
m_index = m_marked_indexes.take_last();
}
// https://drafts.csswg.org/css-syntax/#token-stream-discard-a-mark
void discard_a_mark()
{
// Pop from marked indexes, and do nothing with the popped value.
m_marked_indexes.take_last();
}
// https://drafts.csswg.org/css-syntax/#token-stream-discard-whitespace
void discard_whitespace()
{
// While the next token is a <whitespace-token>, discard a token.
while (next_token().is(Token::Type::Whitespace))
discard_a_token();
}
bool has_next_token()
{
return (size_t)(m_iterator_offset + 1) < m_tokens.size();
}
T const& next_token()
{
if (!has_next_token())
return m_eof;
++m_iterator_offset;
return m_tokens.at(m_iterator_offset);
}
T const& peek_token(int offset = 0)
{
if (!has_next_token())
return m_eof;
return m_tokens.at(m_iterator_offset + offset + 1);
return !is_empty();
}
// Deprecated, used in older versions of the spec.
T const& current_token()
{
if ((size_t)m_iterator_offset >= m_tokens.size())
if (m_index < 1 || (m_index - 1) >= m_tokens.size())
return m_eof;
return m_tokens.at(m_iterator_offset);
return m_tokens.at(m_index - 1);
}
// Deprecated
T const& peek_token(size_t offset = 0)
{
if (remaining_token_count() <= offset)
return m_eof;
return m_tokens.at(m_index + offset);
}
// Deprecated, was used in older versions of the spec.
void reconsume_current_input_token()
{
if (m_iterator_offset >= 0)
--m_iterator_offset;
if (m_index > 0)
--m_index;
}
StateTransaction begin_transaction() { return StateTransaction(*this); }
void skip_whitespace()
size_t remaining_token_count() const
{
while (peek_token().is(Token::Type::Whitespace))
next_token();
if (m_tokens.size() > m_index)
return m_tokens.size() - m_index;
return 0;
}
size_t token_count() const { return m_tokens.size(); }
size_t remaining_token_count() const { return token_count() - m_iterator_offset - 1; }
void dump_all_tokens()
{
dbgln("Dumping all tokens:");
for (size_t i = 0; i < m_tokens.size(); ++i) {
auto& token = m_tokens[i];
if ((i - 1) == (size_t)m_iterator_offset)
if (i == m_index)
dbgln("-> {}", token.to_debug_string());
else
dbgln(" {}", token.to_debug_string());
@ -136,12 +191,18 @@ public:
void copy_state(Badge<Parser>, TokenStream<T> const& other)
{
m_iterator_offset = other.m_iterator_offset;
m_index = other.m_index;
}
private:
// https://drafts.csswg.org/css-syntax/#token-stream-tokens
Span<T const> m_tokens;
int m_iterator_offset { -1 };
// https://drafts.csswg.org/css-syntax/#token-stream-index
size_t m_index { 0 };
// https://drafts.csswg.org/css-syntax/#token-stream-marked-indexes
Vector<size_t> m_marked_indexes;
T make_eof()
{