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
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

@ -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>