1
Fork 0
mirror of https://github.com/RGBCube/serenity synced 2025-07-25 19:07:35 +00:00

LibJS: Correctly handle Unicode characters in JS source text

Also recognize additional white space characters.
This commit is contained in:
davidot 2021-08-14 17:07:47 +02:00 committed by Linus Groh
parent 4d6502de42
commit 47bc72bcf6
6 changed files with 100 additions and 16 deletions

View file

@ -9,6 +9,8 @@
#include <AK/CharacterTypes.h>
#include <AK/Debug.h>
#include <AK/HashMap.h>
#include <AK/Utf8View.h>
#include <LibUnicode/CharacterTypes.h>
#include <stdio.h>
namespace JS {
@ -186,6 +188,26 @@ void Lexer::consume()
} else {
dbgln_if(LEXER_DEBUG, "Previous was CR, this is LF - not incrementing line number again.");
}
} else if (is_unicode_character()) {
size_t char_size = 1;
if ((m_current_char & 64) == 0) {
// invalid char
} else if ((m_current_char & 32) == 0) {
char_size = 2;
} else if ((m_current_char & 16) == 0) {
char_size = 3;
} else if ((m_current_char & 8) == 0) {
char_size = 4;
}
VERIFY(char_size > 1);
--char_size;
m_position += char_size;
if (did_reach_eof())
return;
m_line_column++;
} else {
m_line_column++;
}
@ -310,21 +332,67 @@ bool Lexer::is_line_terminator() const
{
if (m_current_char == '\n' || m_current_char == '\r')
return true;
if (m_position > 0 && m_position + 1 < m_source.length()) {
auto three_chars_view = m_source.substring_view(m_position - 1, 3);
return (three_chars_view == LINE_SEPARATOR) || (three_chars_view == PARAGRAPH_SEPARATOR);
}
if (!is_unicode_character())
return false;
auto code_point = current_code_point();
return code_point == LINE_SEPARATOR || code_point == PARAGRAPH_SEPARATOR;
}
bool Lexer::is_unicode_character() const
{
return (m_current_char & 128) != 0;
}
u32 Lexer::current_code_point() const
{
static constexpr const u32 REPLACEMENT_CHARACTER = 0xFFFD;
if (m_position == 0)
return REPLACEMENT_CHARACTER;
Utf8View utf_8_view { m_source.substring_view(m_position - 1) };
return *utf_8_view.begin();
}
bool Lexer::is_whitespace() const
{
if (is_ascii_space(m_current_char))
return true;
if (!is_unicode_character())
return false;
auto code_point = current_code_point();
if (code_point == NO_BREAK_SPACE)
return true;
static auto space_separator_category = Unicode::general_category_from_string("Space_Separator"sv);
if (space_separator_category.has_value())
return Unicode::code_point_has_general_category(code_point, *space_separator_category);
return false;
}
bool Lexer::is_identifier_start() const
{
return is_ascii_alpha(m_current_char) || m_current_char == '_' || m_current_char == '$';
if (!is_unicode_character())
return is_ascii_alpha(m_current_char) || m_current_char == '_' || m_current_char == '$';
auto code_point = current_code_point();
static auto id_start_category = Unicode::property_from_string("ID_Start"sv);
if (id_start_category.has_value())
return Unicode::code_point_has_property(code_point, *id_start_category);
return false;
}
bool Lexer::is_identifier_middle() const
{
return is_identifier_start() || is_ascii_digit(m_current_char);
if (!is_unicode_character())
return is_identifier_start() || is_ascii_digit(m_current_char);
auto code_point = current_code_point();
if (code_point == ZERO_WIDTH_NON_JOINER || code_point == ZERO_WIDTH_JOINER)
return true;
static auto id_continue_category = Unicode::property_from_string("ID_Continue"sv);
if (id_continue_category.has_value())
return Unicode::code_point_has_property(code_point, *id_continue_category);
return false;
}
bool Lexer::is_line_comment_start(bool line_has_token_yet) const
@ -390,10 +458,10 @@ Token Lexer::next()
do {
consume();
} while (is_line_terminator());
} else if (is_ascii_space(m_current_char)) {
} else if (is_whitespace()) {
do {
consume();
} while (is_ascii_space(m_current_char));
} while (is_whitespace());
} else if (is_line_comment_start(line_has_token_yet)) {
consume();
do {