ladybird/Libraries/LibJS/Lexer.cpp

623 lines
23 KiB
C++
Raw Normal View History

/*
* Copyright (c) 2020, Stephan Unverwerth <s.unverwerth@gmx.de>
* Copyright (c) 2020, Linus Groh <mail@linusgroh.de>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* 1. Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include "Lexer.h"
#include <AK/HashMap.h>
#include <AK/StringBuilder.h>
#include <ctype.h>
#include <stdio.h>
//#define LEXER_DEBUG
namespace JS {
HashMap<String, TokenType> Lexer::s_keywords;
HashMap<String, TokenType> Lexer::s_three_char_tokens;
HashMap<String, TokenType> Lexer::s_two_char_tokens;
HashMap<char, TokenType> Lexer::s_single_char_tokens;
Lexer::Lexer(StringView source)
: m_source(source)
, m_current_token(TokenType::Eof, StringView(nullptr), StringView(nullptr), 0, 0)
{
if (s_keywords.is_empty()) {
s_keywords.set("await", TokenType::Await);
s_keywords.set("break", TokenType::Break);
s_keywords.set("case", TokenType::Case);
s_keywords.set("catch", TokenType::Catch);
s_keywords.set("class", TokenType::Class);
s_keywords.set("const", TokenType::Const);
s_keywords.set("continue", TokenType::Continue);
s_keywords.set("debugger", TokenType::Debugger);
s_keywords.set("default", TokenType::Default);
s_keywords.set("delete", TokenType::Delete);
s_keywords.set("do", TokenType::Do);
s_keywords.set("else", TokenType::Else);
s_keywords.set("enum", TokenType::Enum);
s_keywords.set("export", TokenType::Export);
s_keywords.set("extends", TokenType::Extends);
s_keywords.set("false", TokenType::BoolLiteral);
s_keywords.set("finally", TokenType::Finally);
2020-03-12 12:12:12 +00:00
s_keywords.set("for", TokenType::For);
s_keywords.set("function", TokenType::Function);
s_keywords.set("if", TokenType::If);
s_keywords.set("import", TokenType::Import);
s_keywords.set("in", TokenType::In);
s_keywords.set("instanceof", TokenType::Instanceof);
s_keywords.set("let", TokenType::Let);
s_keywords.set("new", TokenType::New);
s_keywords.set("null", TokenType::NullLiteral);
s_keywords.set("return", TokenType::Return);
s_keywords.set("super", TokenType::Super);
s_keywords.set("switch", TokenType::Switch);
2020-04-12 22:42:14 +00:00
s_keywords.set("this", TokenType::This);
s_keywords.set("throw", TokenType::Throw);
s_keywords.set("true", TokenType::BoolLiteral);
s_keywords.set("try", TokenType::Try);
s_keywords.set("typeof", TokenType::Typeof);
s_keywords.set("var", TokenType::Var);
s_keywords.set("void", TokenType::Void);
s_keywords.set("while", TokenType::While);
s_keywords.set("with", TokenType::With);
s_keywords.set("yield", TokenType::Yield);
}
if (s_three_char_tokens.is_empty()) {
s_three_char_tokens.set("===", TokenType::EqualsEqualsEquals);
s_three_char_tokens.set("!==", TokenType::ExclamationMarkEqualsEquals);
s_three_char_tokens.set("**=", TokenType::DoubleAsteriskEquals);
s_three_char_tokens.set("<<=", TokenType::ShiftLeftEquals);
s_three_char_tokens.set(">>=", TokenType::ShiftRightEquals);
s_three_char_tokens.set("&&=", TokenType::DoubleAmpersandEquals);
s_three_char_tokens.set("||=", TokenType::DoublePipeEquals);
s_three_char_tokens.set("\?\?=", TokenType::DoubleQuestionMarkEquals);
s_three_char_tokens.set(">>>", TokenType::UnsignedShiftRight);
s_three_char_tokens.set("...", TokenType::TripleDot);
}
if (s_two_char_tokens.is_empty()) {
2020-03-30 13:26:09 +00:00
s_two_char_tokens.set("=>", TokenType::Arrow);
s_two_char_tokens.set("+=", TokenType::PlusEquals);
s_two_char_tokens.set("-=", TokenType::MinusEquals);
s_two_char_tokens.set("*=", TokenType::AsteriskEquals);
s_two_char_tokens.set("/=", TokenType::SlashEquals);
s_two_char_tokens.set("%=", TokenType::PercentEquals);
s_two_char_tokens.set("&=", TokenType::AmpersandEquals);
s_two_char_tokens.set("|=", TokenType::PipeEquals);
s_two_char_tokens.set("^=", TokenType::CaretEquals);
s_two_char_tokens.set("&&", TokenType::DoubleAmpersand);
s_two_char_tokens.set("||", TokenType::DoublePipe);
s_two_char_tokens.set("??", TokenType::DoubleQuestionMark);
s_two_char_tokens.set("**", TokenType::DoubleAsterisk);
s_two_char_tokens.set("==", TokenType::EqualsEquals);
s_two_char_tokens.set("<=", TokenType::LessThanEquals);
s_two_char_tokens.set(">=", TokenType::GreaterThanEquals);
s_two_char_tokens.set("!=", TokenType::ExclamationMarkEquals);
s_two_char_tokens.set("--", TokenType::MinusMinus);
s_two_char_tokens.set("++", TokenType::PlusPlus);
s_two_char_tokens.set("<<", TokenType::ShiftLeft);
s_two_char_tokens.set(">>", TokenType::ShiftRight);
s_two_char_tokens.set("?.", TokenType::QuestionMarkPeriod);
}
if (s_single_char_tokens.is_empty()) {
s_single_char_tokens.set('&', TokenType::Ampersand);
s_single_char_tokens.set('*', TokenType::Asterisk);
s_single_char_tokens.set('[', TokenType::BracketOpen);
s_single_char_tokens.set(']', TokenType::BracketClose);
s_single_char_tokens.set('^', TokenType::Caret);
2020-03-21 00:29:00 +00:00
s_single_char_tokens.set(':', TokenType::Colon);
s_single_char_tokens.set(',', TokenType::Comma);
s_single_char_tokens.set('{', TokenType::CurlyOpen);
s_single_char_tokens.set('}', TokenType::CurlyClose);
s_single_char_tokens.set('=', TokenType::Equals);
s_single_char_tokens.set('!', TokenType::ExclamationMark);
s_single_char_tokens.set('-', TokenType::Minus);
s_single_char_tokens.set('(', TokenType::ParenOpen);
s_single_char_tokens.set(')', TokenType::ParenClose);
s_single_char_tokens.set('%', TokenType::Percent);
s_single_char_tokens.set('.', TokenType::Period);
s_single_char_tokens.set('|', TokenType::Pipe);
s_single_char_tokens.set('+', TokenType::Plus);
s_single_char_tokens.set('?', TokenType::QuestionMark);
s_single_char_tokens.set(';', TokenType::Semicolon);
s_single_char_tokens.set('/', TokenType::Slash);
s_single_char_tokens.set('~', TokenType::Tilde);
s_single_char_tokens.set('<', TokenType::LessThan);
s_single_char_tokens.set('>', TokenType::GreaterThan);
}
consume();
}
void Lexer::consume()
{
auto did_reach_eof = [this] {
if (m_position != m_source.length())
return false;
m_position++;
m_line_column++;
m_current_char = EOF;
return true;
};
if (m_position > m_source.length())
return;
if (did_reach_eof())
return;
if (is_line_terminator()) {
#ifdef LEXER_DEBUG
String type;
if (m_current_char == '\n')
type = "LINE FEED";
else if (m_current_char == '\r')
type = "CARRIAGE RETURN";
else if (m_source[m_position + 1] == (char)0xa8)
type = "LINE SEPARATOR";
else
type = "PARAGRAPH SEPARATOR";
dbg() << "Found a line terminator: " << type;
#endif
// This is a three-char line terminator, we need to increase m_position some more.
// We might reach EOF and need to check again.
if (m_current_char != '\n' && m_current_char != '\r') {
m_position += 2;
if (did_reach_eof())
return;
}
// If the previous character is \r and the current one \n we already updated line number
// and column - don't do it again. From https://tc39.es/ecma262/#sec-line-terminators:
// The sequence <CR><LF> is commonly used as a line terminator.
// It should be considered a single SourceCharacter for the purpose of reporting line numbers.
auto second_char_of_crlf = m_position > 1 && m_source[m_position - 2] == '\r' && m_current_char == '\n';
if (!second_char_of_crlf) {
m_line_number++;
m_line_column = 1;
#ifdef LEXER_DEBUG
dbg() << "Incremented line number, now at: line " << m_line_number << ", column 1";
} else {
dbg() << "Previous was CR, this is LF - not incrementing line number again.";
#endif
}
} else {
m_line_column++;
}
m_current_char = m_source[m_position++];
}
bool Lexer::consume_exponent()
{
consume();
if (m_current_char == '-' || m_current_char == '+')
consume();
if (!isdigit(m_current_char))
return false;
while (isdigit(m_current_char)) {
consume();
}
return true;
}
bool Lexer::consume_octal_number()
{
consume();
if (!(m_current_char >= '0' && m_current_char <= '7'))
return false;
while (m_current_char >= '0' && m_current_char <= '7') {
consume();
}
return true;
}
bool Lexer::consume_hexadecimal_number()
{
consume();
if (!isxdigit(m_current_char))
return false;
while (isxdigit(m_current_char))
consume();
return true;
}
bool Lexer::consume_binary_number()
{
consume();
if (!(m_current_char == '0' || m_current_char == '1'))
return false;
while (m_current_char == '0' || m_current_char == '1')
consume();
return true;
}
2020-04-13 17:50:58 +00:00
bool Lexer::match(char a, char b) const
{
if (m_position >= m_source.length())
return false;
return m_current_char == a
&& m_source[m_position] == b;
}
bool Lexer::match(char a, char b, char c) const
{
if (m_position + 1 >= m_source.length())
return false;
return m_current_char == a
&& m_source[m_position] == b
&& m_source[m_position + 1] == c;
}
bool Lexer::match(char a, char b, char c, char d) const
{
if (m_position + 2 >= m_source.length())
return false;
return m_current_char == a
&& m_source[m_position] == b
&& m_source[m_position + 1] == c
&& m_source[m_position + 2] == d;
}
bool Lexer::is_eof() const
{
return m_current_char == EOF;
}
bool Lexer::is_line_terminator() const
{
if (m_current_char == '\n' || m_current_char == '\r')
return true;
if (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);
}
return false;
}
bool Lexer::is_identifier_start() const
{
return isalpha(m_current_char) || m_current_char == '_' || m_current_char == '$';
}
bool Lexer::is_identifier_middle() const
{
return is_identifier_start() || isdigit(m_current_char);
}
bool Lexer::is_line_comment_start() const
{
2020-04-13 17:50:58 +00:00
return match('/', '/') || match('<', '!', '-', '-') || match('-', '-', '>');
}
bool Lexer::is_block_comment_start() const
{
2020-04-13 17:50:58 +00:00
return match('/', '*');
}
bool Lexer::is_block_comment_end() const
{
2020-04-13 17:50:58 +00:00
return match('*', '/');
}
bool Lexer::is_numeric_literal_start() const
{
return isdigit(m_current_char) || (m_current_char == '.' && m_position < m_source.length() && isdigit(m_source[m_position]));
}
bool Lexer::slash_means_division() const
{
auto type = m_current_token.type();
return type == TokenType::BigIntLiteral
|| type == TokenType::BoolLiteral
|| type == TokenType::BracketClose
|| type == TokenType::CurlyClose
|| type == TokenType::Identifier
|| type == TokenType::NullLiteral
|| type == TokenType::NumericLiteral
|| type == TokenType::ParenClose
|| type == TokenType::RegexLiteral
|| type == TokenType::StringLiteral
|| type == TokenType::TemplateLiteralEnd
|| type == TokenType::This;
}
Token Lexer::next()
{
size_t trivia_start = m_position;
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
auto in_template = !m_template_states.is_empty();
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
if (!in_template || m_template_states.last().in_expr) {
// consume whitespace and comments
while (true) {
if (isspace(m_current_char) || is_line_terminator()) {
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
do {
consume();
} while (isspace(m_current_char) || is_line_terminator());
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
} else if (is_line_comment_start()) {
consume();
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
do {
consume();
} while (!is_eof() && !is_line_terminator());
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
} else if (is_block_comment_start()) {
consume();
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
do {
consume();
} while (!is_eof() && !is_block_comment_end());
consume(); // consume *
consume(); // consume /
} else {
break;
}
}
}
size_t value_start = m_position;
size_t value_start_line_number = m_line_number;
size_t value_start_column_number = m_line_column;
auto token_type = TokenType::Invalid;
if (m_current_token.type() == TokenType::RegexLiteral && !is_eof() && isalpha(m_current_char)) {
token_type = TokenType::RegexFlags;
while (!is_eof() && isalpha(m_current_char))
consume();
} else if (m_current_char == '`') {
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
consume();
if (!in_template) {
token_type = TokenType::TemplateLiteralStart;
m_template_states.append({ false, 0 });
} else {
if (m_template_states.last().in_expr) {
m_template_states.append({ false, 0 });
token_type = TokenType::TemplateLiteralStart;
} else {
m_template_states.take_last();
token_type = TokenType::TemplateLiteralEnd;
}
}
} else if (in_template && m_template_states.last().in_expr && m_template_states.last().open_bracket_count == 0 && m_current_char == '}') {
consume();
token_type = TokenType::TemplateLiteralExprEnd;
m_template_states.last().in_expr = false;
} else if (in_template && !m_template_states.last().in_expr) {
if (is_eof()) {
token_type = TokenType::UnterminatedTemplateLiteral;
m_template_states.take_last();
} else if (match('$', '{')) {
token_type = TokenType::TemplateLiteralExprStart;
consume();
consume();
m_template_states.last().in_expr = true;
} else {
while (!match('$', '{') && m_current_char != '`' && !is_eof()) {
if (match('\\', '$') || match('\\', '`'))
consume();
consume();
}
token_type = TokenType::TemplateLiteralString;
}
} else if (is_identifier_start()) {
// identifier or keyword
do {
consume();
} while (is_identifier_middle());
StringView value = m_source.substring_view(value_start - 1, m_position - value_start);
auto it = s_keywords.find(value.hash(), [&](auto& entry) { return entry.key == value; });
if (it == s_keywords.end()) {
token_type = TokenType::Identifier;
} else {
token_type = it->value;
}
} else if (is_numeric_literal_start()) {
2020-06-06 00:14:10 +00:00
token_type = TokenType::NumericLiteral;
bool is_invalid_numeric_literal = false;
if (m_current_char == '0') {
consume();
if (m_current_char == '.') {
// decimal
consume();
2020-06-06 00:14:10 +00:00
while (isdigit(m_current_char))
consume();
2020-06-06 00:14:10 +00:00
if (m_current_char == 'e' || m_current_char == 'E')
is_invalid_numeric_literal = !consume_exponent();
} else if (m_current_char == 'e' || m_current_char == 'E') {
is_invalid_numeric_literal = !consume_exponent();
} else if (m_current_char == 'o' || m_current_char == 'O') {
// octal
is_invalid_numeric_literal = !consume_octal_number();
} else if (m_current_char == 'b' || m_current_char == 'B') {
// binary
is_invalid_numeric_literal = !consume_binary_number();
} else if (m_current_char == 'x' || m_current_char == 'X') {
// hexadecimal
is_invalid_numeric_literal = !consume_hexadecimal_number();
2020-06-06 00:14:10 +00:00
} else if (m_current_char == 'n') {
consume();
token_type = TokenType::BigIntLiteral;
} else if (isdigit(m_current_char)) {
// octal without '0o' prefix. Forbidden in 'strict mode'
do {
consume();
} while (isdigit(m_current_char));
}
} else {
// 1...9 or period
2020-06-06 00:14:10 +00:00
while (isdigit(m_current_char))
consume();
2020-06-06 00:14:10 +00:00
if (m_current_char == 'n') {
consume();
2020-06-06 00:14:10 +00:00
token_type = TokenType::BigIntLiteral;
} else {
if (m_current_char == '.') {
consume();
2020-06-06 00:14:10 +00:00
while (isdigit(m_current_char))
consume();
}
2020-06-06 00:14:10 +00:00
if (m_current_char == 'e' || m_current_char == 'E')
is_invalid_numeric_literal = !consume_exponent();
}
}
if (is_invalid_numeric_literal)
token_type = TokenType::Invalid;
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
} else if (m_current_char == '"' || m_current_char == '\'') {
char stop_char = m_current_char;
consume();
// Note: LS/PS line terminators are allowed in string literals.
while (m_current_char != stop_char && m_current_char != '\r' && m_current_char != '\n' && !is_eof()) {
if (m_current_char == '\\') {
consume();
}
consume();
}
if (m_current_char != stop_char) {
token_type = TokenType::UnterminatedStringLiteral;
} else {
consume();
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
token_type = TokenType::StringLiteral;
}
} else if (m_current_char == '/' && !slash_means_division()) {
consume();
token_type = TokenType::RegexLiteral;
while (!is_eof()) {
if (m_current_char == '[') {
m_regex_is_in_character_class = true;
} else if (m_current_char == ']') {
m_regex_is_in_character_class = false;
} else if (!m_regex_is_in_character_class && m_current_char == '/') {
break;
}
if (match('\\', '/') || match('\\', '[') || match('\\', '\\') || (m_regex_is_in_character_class && match('\\', ']')))
consume();
consume();
}
if (is_eof()) {
token_type = TokenType::UnterminatedRegexLiteral;
} else {
consume();
}
} else if (m_current_char == EOF) {
token_type = TokenType::Eof;
} else {
// There is only one four-char operator: >>>=
bool found_four_char_token = false;
2020-04-13 17:50:58 +00:00
if (match('>', '>', '>', '=')) {
found_four_char_token = true;
consume();
consume();
consume();
consume();
token_type = TokenType::UnsignedShiftRightEquals;
}
bool found_three_char_token = false;
if (!found_four_char_token && m_position + 1 < m_source.length()) {
auto three_chars_view = m_source.substring_view(m_position - 1, 3);
auto it = s_three_char_tokens.find(three_chars_view.hash(), [&](auto& entry) { return entry.key == three_chars_view; });
if (it != s_three_char_tokens.end()) {
found_three_char_token = true;
consume();
consume();
consume();
token_type = it->value;
}
}
bool found_two_char_token = false;
if (!found_four_char_token && !found_three_char_token && m_position < m_source.length()) {
auto two_chars_view = m_source.substring_view(m_position - 1, 2);
auto it = s_two_char_tokens.find(two_chars_view.hash(), [&](auto& entry) { return entry.key == two_chars_view; });
if (it != s_two_char_tokens.end()) {
found_two_char_token = true;
consume();
consume();
token_type = it->value;
}
}
bool found_one_char_token = false;
if (!found_four_char_token && !found_three_char_token && !found_two_char_token) {
auto it = s_single_char_tokens.find(m_current_char);
if (it != s_single_char_tokens.end()) {
found_one_char_token = true;
consume();
token_type = it->value;
}
}
if (!found_four_char_token && !found_three_char_token && !found_two_char_token && !found_one_char_token) {
consume();
token_type = TokenType::Invalid;
}
}
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
if (!m_template_states.is_empty() && m_template_states.last().in_expr) {
if (token_type == TokenType::CurlyOpen) {
m_template_states.last().open_bracket_count++;
} else if (token_type == TokenType::CurlyClose) {
m_template_states.last().open_bracket_count--;
}
}
m_current_token = Token(
token_type,
m_source.substring_view(trivia_start - 1, value_start - trivia_start),
m_source.substring_view(value_start - 1, m_position - value_start),
value_start_line_number,
value_start_column_number);
#ifdef LEXER_DEBUG
dbg() << "------------------------------";
dbg() << "Token: " << m_current_token.name();
dbg() << "Trivia: _" << m_current_token.trivia() << "_";
dbg() << "Value: _" << m_current_token.value() << "_";
dbg() << "Line: " << m_current_token.line_number() << ", Column: " << m_current_token.line_column();
dbg() << "------------------------------";
#endif
return m_current_token;
}
LibJS: Add template literals Adds fully functioning template literals. Because template literals contain expressions, most of the work has to be done in the Lexer rather than the Parser. And because of the complexity of template literals (expressions, nesting, escapes, etc), the Lexer needs to have some template-related state. When entering a new template literal, a TemplateLiteralStart token is emitted. When inside a literal, all text will be parsed up until a '${' or '`' (or EOF, but that's a syntax error) is seen, and then a TemplateLiteralExprStart token is emitted. At this point, the Lexer proceeds as normal, however it keeps track of the number of opening and closing curly braces it has seen in order to determine the close of the expression. Once it finds a matching curly brace for the '${', a TemplateLiteralExprEnd token is emitted and the state is updated accordingly. When the Lexer is inside of a template literal, but not an expression, and sees a '`', this must be the closing grave: a TemplateLiteralEnd token is emitted. The state required to correctly parse template strings consists of a vector (for nesting) of two pieces of information: whether or not we are in a template expression (as opposed to a template string); and the count of the number of unmatched open curly braces we have seen (only applicable if the Lexer is currently in a template expression). TODO: Add support for template literal newlines in the JS REPL (this will cause a syntax error currently): > `foo > bar` 'foo bar'
2020-05-03 22:41:14 +00:00
}