RGBCube/serenity
1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-05-31 21:08:12 +00:00
Code Issues Activity Actions

LibWeb: Stop treating EOF as a valid part of an identifier

Browse source 
This was specifically causing the string "0" to be parsed as an invalid
Dimension token with no units, instead of as a Number. That then caused
out generated `property_initial_value()` function to fail for those
values.
This commit is contained in:
Sam Atkins 2021-09-17 16:49:45 +01:00 • committed by Andreas Kling
parent 2a3675f0c1
commit bb1cc99750
1 changed files with 6 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

8
Userland/Libraries/LibWeb/CSS/Parser/Tokenizer.cpp
View file

@ -11,7 +11,7 @@
#include <LibTextCodec/Decoder.h>
#include <LibWeb/CSS/Parser/Tokenizer.h>
//U+FFFD REPLACEMENT CHARACTER (<28>)
// U+FFFD REPLACEMENT CHARACTER (<28>)
#define REPLACEMENT_CHARACTER 0xFFFD
static const u32 TOKENIZER_EOF = 0xFFFFFFFF;
@ -42,7 +42,10 @@ static inline bool is_low_line(u32 code_point)
static inline bool is_name_start_code_point(u32 code_point)
{
return is_ascii_alpha(code_point) || !is_ascii(code_point) || is_low_line(code_point);
// FIXME: We use !is_ascii() for "non-ASCII code point" in the spec, but it's not quite right -
// it treats EOF as a valid! The spec also lacks a definition of code point. For now, the
// !is_eof() check is a hack, but it should work.
return !is_eof(code_point) && (is_ascii_alpha(code_point) || !is_ascii(code_point) || is_low_line(code_point));
}
static inline bool is_hyphen_minus(u32 code_point)
@ -585,6 +588,7 @@ Token Tokenizer::consume_a_numeric_token()
token.m_number_type = number.type;
auto unit = consume_a_name();
VERIFY(!unit.is_empty() && !unit.is_whitespace());
token.m_unit.append(unit);
return token;