YourWishes/jerryscript
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Improve parse_identifier (#4691)

Browse Source 
Ascii string length is no longer computed during string allocation.

JerryScript-DCO-1.0-Signed-off-by: Daniel Batiz batizjob@gmail.com
This commit is contained in:
batizdaniel
2021-08-17 12:16:58 +02:00
committed by GitHub
parent e7ffb70ae0
commit 3bcd48f72d
19 changed files with 172 additions and 92 deletions
Download Patch File Download Diff File Expand all files Collapse all files
jerry-core/ecma/base/ecma-helpers-string.c
+28 -2
View File
@@ -335,6 +335,33 @@ ecma_find_special_string (const lit_utf8_byte_t *string_p, /**< utf8 string */
return NULL;
} /* ecma_find_special_string */
/**
* Allocate new ecma-string and fill it with characters from ascii characters
*
* @return pointer to ecma-string descriptor
*/
ecma_string_t *
ecma_new_ecma_string_from_ascii (const lit_utf8_byte_t *string_p, /**< ascii string */
lit_utf8_size_t string_size) /**< string size */
{
JERRY_ASSERT (string_p != NULL || string_size == 0);
ecma_string_t *string_desc_p = ecma_find_special_string (string_p, string_size);
if (string_desc_p != NULL)
{
return string_desc_p;
}
lit_utf8_byte_t *data_p;
string_desc_p = ecma_new_ecma_string_from_utf8_buffer (string_size, string_size, &data_p);
string_desc_p->u.hash = lit_utf8_string_calc_hash (string_p, string_size);
memcpy (data_p, string_p, string_size);
return string_desc_p;
} /* ecma_new_ecma_string_from_ascii */
/**
* Allocate new ecma-string and fill it with characters from the utf8 string
*
@@ -2449,8 +2476,7 @@ ecma_string_substr (const ecma_string_t *string_p, /**< pointer to an ecma strin
if (string_length == buffer_size)
{
ecma_string_p = ecma_new_ecma_string_from_utf8 (start_p + start_pos,
(lit_utf8_size_t) end_pos);
ecma_string_p = ecma_new_ecma_string_from_utf8 (start_p + start_pos, (lit_utf8_size_t) end_pos);
}
else
{
jerry-core/ecma/base/ecma-helpers.h
+4 -1
View File
@@ -299,7 +299,10 @@ ecma_length_t ecma_op_advance_string_index (ecma_string_t *str_p, ecma_length_t
ecma_string_t *ecma_new_map_key_string (ecma_value_t value);
bool ecma_prop_name_is_map_key (ecma_string_t *string_p);
#endif /* JERRY_BUILTIN_CONTAINER */
ecma_string_t *ecma_new_ecma_string_from_utf8 (const lit_utf8_byte_t *string_p, lit_utf8_size_t string_size);
ecma_string_t *ecma_new_ecma_string_from_ascii (const lit_utf8_byte_t *string_p,
lit_utf8_size_t string_size);
ecma_string_t *ecma_new_ecma_string_from_utf8 (const lit_utf8_byte_t *string_p,
lit_utf8_size_t string_size);
ecma_string_t *ecma_new_ecma_string_from_utf8_converted_to_cesu8 (const lit_utf8_byte_t *string_p,
lit_utf8_size_t string_size);
ecma_string_t *ecma_new_ecma_external_string_from_cesu8 (const lit_utf8_byte_t *string_p, lit_utf8_size_t string_size,
jerry-core/ecma/base/ecma-literal-storage.c
+5 -3
View File
@@ -165,9 +165,11 @@ ecma_finalize_lit_storage (void)
*/
ecma_value_t
ecma_find_or_create_literal_string (const lit_utf8_byte_t *chars_p, /**< string to be searched */
lit_utf8_size_t size) /**< size of the string */
lit_utf8_size_t size, /**< size of the string */
bool is_ascii) /**< encode of the string */
{
ecma_string_t *string_p = ecma_new_ecma_string_from_utf8 (chars_p, size);
ecma_string_t *string_p = (is_ascii ? ecma_new_ecma_string_from_ascii (chars_p, size)
: ecma_new_ecma_string_from_utf8 (chars_p, size));
if (ECMA_IS_DIRECT_STRING (string_p))
{
@@ -702,7 +704,7 @@ ecma_snapshot_get_literal (const uint8_t *literal_base_p, /**< literal start */
uint16_t length = *(const uint16_t *) literal_p;
return ecma_find_or_create_literal_string (literal_p + sizeof (uint16_t), length);
return ecma_find_or_create_literal_string (literal_p + sizeof (uint16_t), length, false);
} /* ecma_snapshot_get_literal */
/**
jerry-core/ecma/base/ecma-literal-storage.h
+1 -1
View File
@@ -40,7 +40,7 @@ typedef struct
void ecma_finalize_lit_storage (void);
ecma_value_t ecma_find_or_create_literal_string (const lit_utf8_byte_t *chars_p, lit_utf8_size_t size);
ecma_value_t ecma_find_or_create_literal_string (const lit_utf8_byte_t *chars_p, lit_utf8_size_t size, bool is_ascii);
ecma_value_t ecma_find_or_create_literal_number (ecma_number_t number_arg);
#if JERRY_BUILTIN_BIGINT
ecma_value_t ecma_find_or_create_literal_bigint (ecma_value_t bigint);
jerry-core/ecma/builtin-objects/ecma-builtin-helpers-date.c
+1 -1
View File
@@ -680,7 +680,7 @@ ecma_date_to_string_format (ecma_number_t datetime_number, /**< datetime */
JERRY_ASSERT (dest_p <= date_buffer + date_buffer_length);
return ecma_make_string_value (ecma_new_ecma_string_from_utf8 (date_buffer,
return ecma_make_string_value (ecma_new_ecma_string_from_ascii (date_buffer,
(lit_utf8_size_t) (dest_p - date_buffer)));
} /* ecma_date_to_string_format */
jerry-core/ecma/operations/ecma-bigint.c
+1 -1
View File
@@ -250,7 +250,7 @@ ecma_bigint_to_string (ecma_value_t value, /**< BigInt value */
}
ecma_string_t *string_p;
string_p = ecma_new_ecma_string_from_utf8 (string_buffer_p + char_start_p, char_size_p - char_start_p);
string_p = ecma_new_ecma_string_from_ascii (string_buffer_p + char_start_p, char_size_p - char_start_p);
jmem_heap_free_block (string_buffer_p, char_size_p);
return string_p;
jerry-core/parser/js/common.h
+2 -1
View File
@@ -74,8 +74,9 @@ typedef enum
LEXER_FLAG_SOURCE_PTR = (1 << 2), /**< the literal is directly referenced in the source code
* (no need to allocate memory) */
LEXER_FLAG_LATE_INIT = (1 << 3), /**< initialize this variable after the byte code is freed */
LEXER_FLAG_ASCII = (1 << 4), /**< the literal contains only ascii characters */
#if JERRY_ESNEXT
LEXER_FLAG_GLOBAL = (1 << 4), /**< this local identifier is not a let or const declaration */
LEXER_FLAG_GLOBAL = (1 << 5), /**< this local identifier is not a let or const declaration */
#endif /* JERRY_ESNEXT */
} lexer_literal_status_flags_t;
jerry-core/parser/js/js-lexer.c
+26 -19
View File
@@ -642,7 +642,7 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
parser_line_counter_t column = context_p->column;
const uint8_t *source_end_p = context_p->source_end_p;
size_t length = 0;
uint8_t has_escape = false;
lexer_lit_location_flags_t status_flags = LEXER_LIT_LOCATION_IS_ASCII;
do
{
@@ -657,7 +657,7 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
return true;
}
has_escape = true;
status_flags = LEXER_LIT_LOCATION_HAS_ESCAPE;
#if JERRY_ESNEXT
if (source_p + 5 <= source_end_p && source_p[1] == LIT_CHAR_LOWERCASE_U)
@@ -711,6 +711,8 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
if (JERRY_UNLIKELY (code_point >= LIT_UTF8_2_BYTE_MARKER))
{
status_flags &= (uint32_t) ~LEXER_LIT_LOCATION_IS_ASCII;
#if JERRY_ESNEXT
utf8_length = lit_read_code_point_from_utf8 (source_p,
(lit_utf8_size_t) (source_end_p - source_p),
@@ -738,7 +740,7 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
else if (source_p[0] >= LIT_UTF8_4_BYTE_MARKER)
{
decoded_length = 2 * 3;
has_escape = true;
status_flags = LEXER_LIT_LOCATION_HAS_ESCAPE;
}
#else /* !JERRY_ESNEXT */
if (code_point < LIT_UTF8_4_BYTE_MARKER)
@@ -789,7 +791,7 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
context_p->token.type = LEXER_LITERAL;
context_p->token.lit_location.type = LEXER_IDENT_LITERAL;
context_p->token.lit_location.has_escape = has_escape;
context_p->token.lit_location.status_flags = (uint8_t) status_flags;
context_p->token.column = context_p->column;
context_p->token.lit_location.char_p = context_p->source_p;
@@ -807,7 +809,7 @@ lexer_parse_identifier (parser_context_t *context_p, /**< context */
const uint8_t *ident_start_p = context_p->source_p;
uint8_t buffer_p[LEXER_KEYWORD_MAX_LENGTH];
if (JERRY_UNLIKELY (context_p->token.lit_location.has_escape))
if (JERRY_UNLIKELY (context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
lexer_convert_ident_to_cesu8 (buffer_p, ident_start_p, (prop_length_t) length);
ident_start_p = buffer_p;
@@ -953,7 +955,7 @@ lexer_parse_string (parser_context_t *context_p, /**< context */
parser_line_counter_t original_line = line;
parser_line_counter_t original_column = column;
size_t length = 0;
uint8_t has_escape = false;
lexer_lit_location_flags_t status_flags = LEXER_LIT_LOCATION_IS_ASCII;
#if JERRY_ESNEXT
if (str_end_character == LIT_CHAR_RIGHT_BRACE)
@@ -986,7 +988,7 @@ lexer_parse_string (parser_context_t *context_p, /**< context */
continue;
}
has_escape = true;
status_flags = LEXER_LIT_LOCATION_HAS_ESCAPE;
/* Newline is ignored. */
if (*source_p == LIT_CHAR_CR)
@@ -1163,7 +1165,7 @@ lexer_parse_string (parser_context_t *context_p, /**< context */
* after a backslash). Always converted to two 3 byte
* long sequence. */
length += 2 * 3;
has_escape = true;
status_flags = LEXER_LIT_LOCATION_HAS_ESCAPE;
source_p += 4;
#if JERRY_ESNEXT
raw_length_adjust += 2;
@@ -1192,7 +1194,7 @@ lexer_parse_string (parser_context_t *context_p, /**< context */
Note: ECMAScript v6, 11.8.6.1 <CR> or <CR><LF> are both normalized to <LF> */
if (*source_p == LIT_CHAR_CR)
{
has_escape = true;
status_flags = LEXER_LIT_LOCATION_HAS_ESCAPE;
source_p++;
length++;
if (source_p < source_end_p
@@ -1261,7 +1263,7 @@ lexer_parse_string (parser_context_t *context_p, /**< context */
context_p->token.lit_location.char_p = string_start_p;
context_p->token.lit_location.length = (prop_length_t) length;
context_p->token.lit_location.type = LEXER_STRING_LITERAL;
context_p->token.lit_location.has_escape = has_escape;
context_p->token.lit_location.status_flags = (uint8_t) status_flags;
context_p->source_p = source_p + 1;
context_p->line = line;
@@ -1328,7 +1330,7 @@ lexer_parse_number (parser_context_t *context_p) /**< context */
context_p->token.extra_value = LEXER_NUMBER_DECIMAL;
context_p->token.lit_location.char_p = source_p;
context_p->token.lit_location.type = LEXER_NUMBER_LITERAL;
context_p->token.lit_location.has_escape = false;
context_p->token.lit_location.status_flags = LEXER_LIT_LOCATION_IS_ASCII;
if (source_p[0] == LIT_CHAR_0
&& source_p + 1 < source_end_p)
@@ -2240,7 +2242,7 @@ lexer_convert_literal_to_chars (parser_context_t *context_p, /**< context */
{
JERRY_ASSERT (context_p->u.allocated_buffer_p == NULL);
if (!literal_p->has_escape)
if (!(literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
return literal_p->char_p;
}
@@ -2601,6 +2603,11 @@ lexer_construct_literal_object (parser_context_t *context_p, /**< context */
status_flags |= LEXER_FLAG_USED;
}
if (lit_location_p->status_flags & LEXER_LIT_LOCATION_IS_ASCII)
{
literal_p->status_flags |= LEXER_FLAG_ASCII;
}
literal_p->status_flags = status_flags;
context_p->lit_object.literal_p = literal_p;
@@ -3490,7 +3497,7 @@ lexer_compare_identifier_to_string (const lexer_lit_location_t *left_p, /**< lef
return false;
}
if (!left_p->has_escape)
if (!(left_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
return memcmp (left_p->char_p, right_p, size) == 0;
}
@@ -3518,12 +3525,12 @@ lexer_compare_identifiers (parser_context_t *context_p, /**< context */
return false;
}
if (!left_p->has_escape)
if (!(left_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
return lexer_compare_identifier_to_chars (right_p->char_p, left_p->char_p, length);
}
if (!right_p->has_escape)
if (!(right_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
return lexer_compare_identifier_to_chars (left_p->char_p, right_p->char_p, length);
}
@@ -3568,7 +3575,7 @@ lexer_current_is_literal (parser_context_t *context_p, /**< context */
return false;
}
if (!left_ident_p->has_escape && !right_ident_p->has_escape)
if (!((left_ident_p->status_flags | right_ident_p->status_flags) & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
return memcmp (left_ident_p->char_p, right_ident_p->char_p, left_ident_p->length) == 0;
}
@@ -3591,7 +3598,7 @@ lexer_string_is_use_strict (parser_context_t *context_p) /**< context */
&& context_p->token.lit_location.type == LEXER_STRING_LITERAL);
return (context_p->token.lit_location.length == 10
&& !context_p->token.lit_location.has_escape
&& !(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)
&& memcmp (context_p->token.lit_location.char_p, "use strict", 10) == 0);
} /* lexer_string_is_use_strict */
@@ -3649,7 +3656,7 @@ lexer_token_is_let (parser_context_t *context_p) /**< context */
JERRY_ASSERT (context_p->token.type == LEXER_LITERAL);
return (context_p->token.keyword_type == LEXER_KEYW_LET
&& !context_p->token.lit_location.has_escape);
&& !(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE));
} /* lexer_token_is_let */
/**
@@ -3667,7 +3674,7 @@ lexer_token_is_async (parser_context_t *context_p) /**< context */
|| context_p->token.type == LEXER_TEMPLATE_LITERAL);
return (context_p->token.keyword_type == LEXER_KEYW_ASYNC
&& !context_p->token.lit_location.has_escape);
&& !(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE));
} /* lexer_token_is_async */
#endif /* JERRY_ESNEXT */
jerry-core/parser/js/js-lexer.h
+11 -1
View File
@@ -305,6 +305,16 @@ typedef enum
#endif /* JERRY_BUILTIN_BIGINT */
} lexer_number_type_t;
/**
* Lexer literal flags.
**/
typedef enum
{
LEXER_LIT_LOCATION_NO_OPTS = 0, /**< no options */
LEXER_LIT_LOCATION_HAS_ESCAPE = (1 << 0), /**< binding has escape */
LEXER_LIT_LOCATION_IS_ASCII = (1 << 1), /**< all characters are ascii characters */
} lexer_lit_location_flags_t;
/**
* Lexer character (string / identifier) literal data.
*/
@@ -313,7 +323,7 @@ typedef struct
const uint8_t *char_p; /**< start of identifier or string token */
prop_length_t length; /**< length or index of a literal */
uint8_t type; /**< type of the current literal */
uint8_t has_escape; /**< has escape sequences */
uint8_t status_flags; /**< any combination of lexer_lit_location_flags_t status bits */
} lexer_lit_location_t;
/**
jerry-core/parser/js/js-parser-expr.c
+2 -2
View File
@@ -1922,7 +1922,7 @@ parser_parse_unary_expression (parser_context_t *context_p, /**< context */
}
#endif /* JERRY_MODULE_SYSTEM */
if (JERRY_UNLIKELY (context_p->token.lit_location.has_escape))
if (JERRY_UNLIKELY (context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
parser_raise_error (context_p, PARSER_ERR_INVALID_KEYWORD);
}
@@ -2281,7 +2281,7 @@ parser_parse_unary_expression (parser_context_t *context_p, /**< context */
JERRY_ASSERT ((context_p->status_flags & PARSER_IS_GENERATOR_FUNCTION)
&& !(context_p->status_flags & PARSER_DISALLOW_AWAIT_YIELD));
if (context_p->token.lit_location.has_escape)
if (context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)
{
parser_raise_error (context_p, PARSER_ERR_INVALID_KEYWORD);
}
jerry-core/parser/js/js-parser-internal.h
+1
View File
@@ -724,6 +724,7 @@ void parser_emit_cbc_forward_branch (parser_context_t *context_p, uint16_t opcod
parser_branch_node_t *parser_emit_cbc_forward_branch_item (parser_context_t *context_p, uint16_t opcode,
parser_branch_node_t *next_p);
void parser_emit_cbc_backward_branch (parser_context_t *context_p, uint16_t opcode, uint32_t offset);
ecma_string_t *parser_new_ecma_string_from_literal (lexer_literal_t *literal_p);
void parser_set_branch_to_current_position (parser_context_t *context_p, parser_branch_t *branch_p);
void parser_set_breaks_to_current_position (parser_context_t *context_p, parser_branch_node_t *current_p);
void parser_set_continues_to_current_position (parser_context_t *context_p, parser_branch_node_t *current_p);
jerry-core/parser/js/js-parser-module.c
+7 -13
View File
@@ -30,7 +30,7 @@
*/
const lexer_lit_location_t lexer_default_literal =
{
(const uint8_t *) "*default*", 9, LEXER_IDENT_LITERAL, false
(const uint8_t *) "*default*", 9, LEXER_IDENT_LITERAL, LEXER_LIT_LOCATION_IS_ASCII
};
/**
@@ -89,9 +89,7 @@ parser_module_append_export_name (parser_context_t *context_p) /**< parser conte
}
context_p->module_identifier_lit_p = context_p->lit_object.literal_p;
ecma_string_t *name_p = ecma_new_ecma_string_from_utf8 (context_p->lit_object.literal_p->u.char_p,
context_p->lit_object.literal_p->prop.length);
ecma_string_t *name_p = parser_new_ecma_string_from_literal (context_p->lit_object.literal_p);
if (parser_module_check_duplicate_export (context_p, name_p))
{
@@ -250,13 +248,11 @@ parser_module_parse_export_clause (parser_context_t *context_p) /**< parser cont
lexer_next_token (context_p);
}
lexer_literal_t *literal_p = PARSER_GET_LITERAL (local_name_index);
local_name_p = ecma_new_ecma_string_from_utf8 (literal_p->u.char_p, literal_p->prop.length);
local_name_p = parser_new_ecma_string_from_literal (PARSER_GET_LITERAL (local_name_index));
if (export_name_index != PARSER_MAXIMUM_NUMBER_OF_LITERALS)
{
lexer_literal_t *as_literal_p = PARSER_GET_LITERAL (export_name_index);
export_name_p = ecma_new_ecma_string_from_utf8 (as_literal_p->u.char_p, as_literal_p->prop.length);
export_name_p = parser_new_ecma_string_from_literal (PARSER_GET_LITERAL (export_name_index));
}
else
{
@@ -353,13 +349,11 @@ parser_module_parse_import_clause (parser_context_t *context_p) /**< parser cont
lexer_next_token (context_p);
}
lexer_literal_t *literal_p = PARSER_GET_LITERAL (import_name_index);
import_name_p = ecma_new_ecma_string_from_utf8 (literal_p->u.char_p, literal_p->prop.length);
import_name_p = parser_new_ecma_string_from_literal (PARSER_GET_LITERAL (import_name_index));
if (local_name_index != PARSER_MAXIMUM_NUMBER_OF_LITERALS)
{
lexer_literal_t *as_literal_p = PARSER_GET_LITERAL (local_name_index);
local_name_p = ecma_new_ecma_string_from_utf8 (as_literal_p->u.char_p, as_literal_p->prop.length);
local_name_p = parser_new_ecma_string_from_literal (PARSER_GET_LITERAL (local_name_index));
}
else
{
@@ -460,7 +454,7 @@ parser_module_handle_module_specifier (parser_context_t *context_p, /**< parser
lexer_next_token (context_p);
/* The lexer_next_token may throw an error, so the path is constructed after its call. */
ecma_string_t *path_string_p = ecma_new_ecma_string_from_utf8 (path_p->u.char_p, path_p->prop.length);
ecma_string_t *path_string_p = parser_new_ecma_string_from_literal (path_p);
ecma_module_node_t *node_p = JERRY_CONTEXT (module_current_p)->imports_p;
ecma_module_node_t *last_node_p = NULL;
jerry-core/parser/js/js-parser-statm.c
+7 -10
View File
@@ -1174,7 +1174,7 @@ parser_parse_for_statement_start (parser_context_t *context_p) /**< context */
if (context_p->token.type == LEXER_KEYW_AWAIT)
{
if (JERRY_UNLIKELY (context_p->token.lit_location.has_escape))
if (JERRY_UNLIKELY (context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
parser_raise_error (context_p, PARSER_ERR_INVALID_KEYWORD);
}
@@ -1188,7 +1188,7 @@ parser_parse_for_statement_start (parser_context_t *context_p) /**< context */
#if JERRY_ESNEXT
if (context_p->token.type == LEXER_LITERAL
&& context_p->token.keyword_type == LEXER_KEYW_AWAIT
&& !context_p->token.lit_location.has_escape)
&& !(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
parser_raise_error (context_p, PARSER_ERR_FOR_AWAIT_NO_ASYNC);
}
@@ -2447,8 +2447,7 @@ parser_parse_import_statement (parser_context_t *context_p) /**< parser context
/* Handle ImportedDefaultBinding */
lexer_construct_literal_object (context_p, &context_p->token.lit_location, LEXER_IDENT_LITERAL);
ecma_string_t *local_name_p = ecma_new_ecma_string_from_utf8 (context_p->lit_object.literal_p->u.char_p,
context_p->lit_object.literal_p->prop.length);
ecma_string_t *local_name_p = parser_new_ecma_string_from_literal (context_p->lit_object.literal_p);
if (parser_module_check_duplicate_import (context_p, local_name_p))
{
@@ -2496,8 +2495,7 @@ parser_parse_import_statement (parser_context_t *context_p) /**< parser context
lexer_construct_literal_object (context_p, &context_p->token.lit_location, LEXER_IDENT_LITERAL);
ecma_string_t *local_name_p = ecma_new_ecma_string_from_utf8 (context_p->lit_object.literal_p->u.char_p,
context_p->lit_object.literal_p->prop.length);
ecma_string_t *local_name_p = parser_new_ecma_string_from_literal (context_p->lit_object.literal_p);
if (parser_module_check_duplicate_import (context_p, local_name_p))
{
@@ -2598,8 +2596,8 @@ parser_parse_export_statement (parser_context_t *context_p) /**< context */
parser_parse_expression_statement (context_p, PARSE_EXPR_NO_COMMA | PARSE_EXPR_HAS_LITERAL);
}
ecma_string_t *name_p = ecma_new_ecma_string_from_utf8 (context_p->module_identifier_lit_p->u.char_p,
context_p->module_identifier_lit_p->prop.length);
ecma_string_t *name_p = parser_new_ecma_string_from_literal (context_p->module_identifier_lit_p);
ecma_string_t *export_name_p = ecma_get_magic_string (LIT_MAGIC_STRING_DEFAULT);
if (parser_module_check_duplicate_export (context_p, export_name_p))
@@ -2637,8 +2635,7 @@ parser_parse_export_statement (parser_context_t *context_p) /**< context */
lexer_construct_literal_object (context_p, &context_p->token.lit_location, LEXER_NEW_IDENT_LITERAL);
lexer_literal_t *literal_p = PARSER_GET_LITERAL (context_p->lit_object.index);
ecma_string_t *export_name_p = ecma_new_ecma_string_from_utf8 (literal_p->u.char_p,
literal_p->prop.length);
ecma_string_t *export_name_p = parser_new_ecma_string_from_literal (literal_p);
if (parser_module_check_duplicate_export (context_p, export_name_p))
{
jerry-core/parser/js/js-parser-tagged-template-literal.c
+10 -4
View File
@@ -43,7 +43,7 @@ parser_tagged_template_literal_append_strings (parser_context_t *context_p, /**<
{
lexer_lit_location_t *lit_loc_p = &context_p->token.lit_location;
if (lit_loc_p->length == 0 && !lit_loc_p->has_escape)
if (lit_loc_p->length == 0 && !(lit_loc_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE))
{
ecma_builtin_helper_def_prop_by_index (template_obj_p,
prop_idx,
@@ -64,10 +64,13 @@ parser_tagged_template_literal_append_strings (parser_context_t *context_p, /**<
LEXER_STRING_NO_OPTS);
ecma_string_t *raw_str_p;
ecma_string_t *cooked_str_p = ecma_new_ecma_string_from_utf8 (source_p, lit_loc_p->length);
ecma_string_t *cooked_str_p = ((lit_loc_p->status_flags & LEXER_FLAG_ASCII)
? ecma_new_ecma_string_from_ascii (source_p, lit_loc_p->length)
: ecma_new_ecma_string_from_utf8 (source_p, lit_loc_p->length));
parser_free_allocated_buffer (context_p);
if (lit_loc_p->has_escape)
if (lit_loc_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)
{
context_p->source_p = context_p->token.lit_location.char_p - 1;
lexer_parse_string (context_p, LEXER_STRING_RAW);
@@ -76,7 +79,10 @@ parser_tagged_template_literal_append_strings (parser_context_t *context_p, /**<
local_byte_array,
LEXER_STRING_RAW);
raw_str_p = ecma_new_ecma_string_from_utf8 (source_p, lit_loc_p->length);
raw_str_p = ((lit_loc_p->status_flags & LEXER_FLAG_ASCII)
? ecma_new_ecma_string_from_ascii (source_p, lit_loc_p->length)
: ecma_new_ecma_string_from_utf8 (source_p, lit_loc_p->length));
parser_free_allocated_buffer (context_p);
}
else
jerry-core/parser/js/js-parser-util.c
+25
View File
@@ -14,6 +14,7 @@
*/
#include "js-parser-internal.h"
#include "ecma-helpers.h"
#if JERRY_PARSER
@@ -650,6 +651,30 @@ parser_emit_cbc_backward_branch (parser_context_t *context_p, /**< context */
#undef PARSER_CHECK_LAST_POSITION
#undef PARSER_APPEND_TO_BYTE_CODE
/**
* Helper function for parser.
*
* @return a new string based on encode.
*/
ecma_string_t *
parser_new_ecma_string_from_literal (lexer_literal_t *literal_p) /**< literal */
{
JERRY_ASSERT (literal_p != NULL);
ecma_string_t *new_string = NULL;
if (literal_p->status_flags & LEXER_FLAG_ASCII)
{
new_string = ecma_new_ecma_string_from_ascii (literal_p->u.char_p, literal_p->prop.length);
}
else
{
new_string = ecma_new_ecma_string_from_utf8 (literal_p->u.char_p, literal_p->prop.length);
}
return new_string;
} /* parser_new_ecma_string_from_literal */
/**
* Set a branch to the current byte code position
*/
jerry-core/parser/js/js-parser.c
+7 -3
View File
@@ -130,7 +130,8 @@ parser_compute_indicies (parser_context_t *context_p, /**< context */
if (char_p != NULL)
{
literal_p->u.value = ecma_find_or_create_literal_string (char_p,
literal_p->prop.length);
literal_p->prop.length,
(literal_p->status_flags & LEXER_FLAG_ASCII) != 0);
if (!(literal_p->status_flags & LEXER_FLAG_SOURCE_PTR))
{
@@ -1274,7 +1275,8 @@ parser_post_processing (parser_context_t *context_p) /**< context */
uint32_t source_data = literal_p->u.source_data;
const uint8_t *char_p = context_p->source_end_p - (source_data & 0xfffff);
ecma_value_t lit_value = ecma_find_or_create_literal_string (char_p,
source_data >> 20);
source_data >> 20,
(literal_p->status_flags & LEXER_FLAG_ASCII) != 0);
literal_pool_p[literal_p->prop.index] = lit_value;
}
}
@@ -2896,7 +2898,9 @@ parser_compiled_code_set_function_name (parser_context_t *context_p, /**< contex
memcpy (name_buffer_p + 4, name_lit_p->u.char_p, name_lit_p->prop.length);
}
*func_name_start_p = ecma_find_or_create_literal_string (name_buffer_p, name_length);
*func_name_start_p = ecma_find_or_create_literal_string (name_buffer_p,
name_length,
(status_flags & LEXER_FLAG_ASCII) != 0);
if (name_buffer_p != name_lit_p->u.char_p)
{
jerry-core/parser/js/js-scanner-util.c
+20 -16
View File
@@ -406,9 +406,9 @@ scanner_scope_find_lexical_declaration (parser_context_t *context_p, /**< contex
return false;
}
if (JERRY_LIKELY (!literal_p->has_escape))
if (JERRY_LIKELY (!(literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
name_p = ecma_new_ecma_string_from_utf8 (literal_p->char_p, literal_p->length);
name_p = parser_new_ecma_string_from_literal ((lexer_literal_t *) literal_p);
}
else
{
@@ -416,7 +416,8 @@ scanner_scope_find_lexical_declaration (parser_context_t *context_p, /**< contex
lexer_convert_ident_to_cesu8 (destination_p, literal_p->char_p, literal_p->length);
name_p = ecma_new_ecma_string_from_utf8 (destination_p, literal_p->length);
name_p = parser_new_ecma_string_from_literal ((lexer_literal_t *) literal_p);
scanner_free (destination_p, literal_p->length);
}
@@ -1114,7 +1115,7 @@ scanner_pop_literal_pool (parser_context_t *context_p, /**< context */
}
#endif /* JERRY_ESNEXT */
if (literal_p->has_escape)
if (literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)
{
type |= SCANNER_STREAM_HAS_ESCAPE;
}
@@ -1384,13 +1385,13 @@ scanner_add_custom_literal (parser_context_t *context_p, /**< context */
const uint8_t *char_p = literal_location_p->char_p;
prop_length_t length = literal_location_p->length;
if (JERRY_LIKELY (!literal_location_p->has_escape))
if (JERRY_LIKELY (!(literal_location_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
while ((literal_p = (lexer_lit_location_t *) parser_list_iterator_next (&literal_iterator)) != NULL)
{
if (literal_p->length == length)
{
if (JERRY_LIKELY (!literal_p->has_escape))
if (JERRY_LIKELY (!(literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
if (memcmp (literal_p->char_p, char_p, length) == 0)
{
@@ -1401,7 +1402,7 @@ scanner_add_custom_literal (parser_context_t *context_p, /**< context */
{
/* The non-escaped version is preferred. */
literal_p->char_p = char_p;
literal_p->has_escape = 0;
literal_p->status_flags = LEXER_LIT_LOCATION_NO_OPTS;
return literal_p;
}
}
@@ -1494,13 +1495,13 @@ scanner_append_argument (parser_context_t *context_p, /**< context */
JERRY_ASSERT (SCANNER_LITERAL_POOL_MAY_HAVE_ARGUMENTS (literal_pool_p->status_flags));
if (JERRY_LIKELY (!context_p->token.lit_location.has_escape))
if (JERRY_LIKELY (!(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
while ((literal_p = (lexer_lit_location_t *) parser_list_iterator_next (&literal_iterator)) != NULL)
{
if (literal_p->length == length)
{
if (JERRY_LIKELY (!literal_p->has_escape))
if (JERRY_LIKELY (!(literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
if (memcmp (literal_p->char_p, char_p, length) == 0)
{
@@ -1603,7 +1604,7 @@ scanner_detect_invalid_var (parser_context_t *context_p, /**< context */
parser_list_iterator_init (&literal_pool_p->literal_pool, &literal_iterator);
lexer_lit_location_t *literal_p;
if (JERRY_LIKELY (!context_p->token.lit_location.has_escape))
if (JERRY_LIKELY (!(context_p->token.lit_location.status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
while ((literal_p = (lexer_lit_location_t *) parser_list_iterator_next (&literal_iterator)) != NULL)
{
@@ -1614,7 +1615,7 @@ scanner_detect_invalid_var (parser_context_t *context_p, /**< context */
&& (literal_p->type & SCANNER_LITERAL_IS_LOCAL) != SCANNER_LITERAL_IS_LOCAL
&& literal_p->length == length)
{
if (JERRY_LIKELY (!literal_p->has_escape))
if (JERRY_LIKELY (!(literal_p->status_flags & LEXER_LIT_LOCATION_HAS_ESCAPE)))
{
if (memcmp (literal_p->char_p, char_p, length) == 0)
{
@@ -1862,7 +1863,7 @@ scanner_append_hole (parser_context_t *context_p, scanner_context_t *scanner_con
literal_p->char_p = NULL;
literal_p->length = 0;
literal_p->type = SCANNER_LITERAL_IS_ARG;
literal_p->has_escape = 0;
literal_p->status_flags = LEXER_LIT_LOCATION_NO_OPTS;
} /* scanner_append_hole */
#endif /* JERRY_ESNEXT */
@@ -2203,7 +2204,7 @@ scanner_try_scan_new_target (parser_context_t *context_p) /**< parser/scanner co
*/
const lexer_lit_location_t lexer_arguments_literal =
{
(const uint8_t *) "arguments", 9, LEXER_IDENT_LITERAL, false
(const uint8_t *) "arguments", 9, LEXER_IDENT_LITERAL, LEXER_LIT_LOCATION_IS_ASCII
};
/**
@@ -2279,7 +2280,8 @@ scanner_check_variables (parser_context_t *context_p) /**< context */
literal.length = data_p[1];
literal.type = LEXER_IDENT_LITERAL;
literal.has_escape = (data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? 1 : 0;
literal.status_flags = ((data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? LEXER_LIT_LOCATION_HAS_ESCAPE
: LEXER_LIT_LOCATION_NO_OPTS);
lexer_construct_literal_object (context_p, &literal, LEXER_NEW_IDENT_LITERAL);
literal.char_p += data_p[1];
@@ -2524,7 +2526,8 @@ scanner_create_variables (parser_context_t *context_p, /**< context */
{
literal.length = data_p[1];
literal.type = LEXER_IDENT_LITERAL;
literal.has_escape = (data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? 1 : 0;
literal.status_flags = ((data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? LEXER_LIT_LOCATION_HAS_ESCAPE
: LEXER_LIT_LOCATION_NO_OPTS);
/* Literal must be exists. */
lexer_construct_literal_object (context_p, &literal, LEXER_IDENT_LITERAL);
@@ -2549,7 +2552,8 @@ scanner_create_variables (parser_context_t *context_p, /**< context */
literal.length = data_p[1];
literal.type = LEXER_IDENT_LITERAL;
literal.has_escape = (data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? 1 : 0;
literal.status_flags = ((data_p[0] & SCANNER_STREAM_HAS_ESCAPE) ? LEXER_LIT_LOCATION_HAS_ESCAPE
: LEXER_LIT_LOCATION_NO_OPTS);
lexer_construct_literal_object (context_p, &literal, LEXER_NEW_IDENT_LITERAL);
literal.char_p += data_p[1];
tests/unit-core/test-literal-storage.c
+6 -6
View File
@@ -78,7 +78,7 @@ main (void)
{
lengths[j] = (lit_utf8_size_t) (rand () % max_characters_in_string + 1);
generate_string (strings[j], lengths[j]);
ecma_find_or_create_literal_string (strings[j], lengths[j]);
ecma_find_or_create_literal_string (strings[j], lengths[j], false);
strings[j][lengths[j]] = '\0';
ptrs[j] = strings[j];
TEST_ASSERT (ptrs[j]);
@@ -89,7 +89,7 @@ main (void)
ptrs[j] = lit_get_magic_string_utf8 (msi);
TEST_ASSERT (ptrs[j]);
lengths[j] = (lit_utf8_size_t) lit_zt_utf8_string_size (ptrs[j]);
ecma_find_or_create_literal_string (ptrs[j], lengths[j]);
ecma_find_or_create_literal_string (ptrs[j], lengths[j], false);
}
else
{
@@ -100,7 +100,7 @@ main (void)
}
/* Add empty string. */
ecma_find_or_create_literal_string (NULL, 0);
ecma_find_or_create_literal_string (NULL, 0, false);
for (uint32_t j = 0; j < test_sub_iters; j++)
{
@@ -108,8 +108,8 @@ main (void)
ecma_value_t lit2;
if (ptrs[j])
{
lit1 = ecma_find_or_create_literal_string (ptrs[j], lengths[j]);
lit2 = ecma_find_or_create_literal_string (ptrs[j], lengths[j]);
lit1 = ecma_find_or_create_literal_string (ptrs[j], lengths[j], false);
lit2 = ecma_find_or_create_literal_string (ptrs[j], lengths[j], false);
TEST_ASSERT (ecma_is_value_string (lit1));
TEST_ASSERT (ecma_is_value_string (lit2));
TEST_ASSERT (lit1 == lit2);
@@ -125,7 +125,7 @@ main (void)
}
/* Check empty string exists. */
TEST_ASSERT (ecma_find_or_create_literal_string (NULL, 0) != JMEM_CP_NULL);
TEST_ASSERT (ecma_find_or_create_literal_string (NULL, 0, false) != JMEM_CP_NULL);
}
ecma_finalize_lit_storage ();
tests/unit-core/test-stringbuilder.c
+8 -8
View File
@@ -35,7 +35,7 @@ main (void)
ecma_stringbuilder_append_raw (&builder, string_data, sizeof (string_data) - 1);
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (string_data, sizeof (string_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (string_data, sizeof (string_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);
@@ -57,7 +57,7 @@ main (void)
ecma_stringbuilder_append_char (&builder, LIT_CHAR_LOWERCASE_A);
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (string_data, sizeof (string_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (string_data, sizeof (string_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);
@@ -65,7 +65,7 @@ main (void)
{
static const lit_utf8_byte_t string_data[] = "A simple string";
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (string_data, sizeof (string_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (string_data, sizeof (string_data) - 1);
ecma_stringbuilder_t builder = ecma_stringbuilder_create ();
ecma_stringbuilder_append (&builder, str_p);
@@ -94,7 +94,7 @@ main (void)
ecma_stringbuilder_append_char (&builder, LIT_CHAR_LOWERCASE_C);
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (string_data, sizeof (string_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (string_data, sizeof (string_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);
@@ -125,7 +125,7 @@ main (void)
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
static const lit_utf8_byte_t expected_data[] = "1abc234string";
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (expected_data, sizeof (expected_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (expected_data, sizeof (expected_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);
@@ -177,14 +177,14 @@ main (void)
ecma_stringbuilder_append_char (&builder, LIT_CHAR_1);
ecma_stringbuilder_append_raw (&builder, string_data, sizeof (string_data) - 1);
ecma_string_t *another_string = ecma_new_ecma_string_from_utf8 (string_data, sizeof (string_data) - 1);
ecma_string_t *another_string = ecma_new_ecma_string_from_ascii (string_data, sizeof (string_data) - 1);
ecma_stringbuilder_append (&builder, uint_str_p);
ecma_stringbuilder_append_magic (&builder, LIT_MAGIC_STRING_STRING);
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
static const lit_utf8_byte_t expected_data[] = "1abc234string";
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (expected_data, sizeof (expected_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (expected_data, sizeof (expected_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);
@@ -201,7 +201,7 @@ main (void)
ecma_string_t *result_p = ecma_stringbuilder_finalize (&builder);
static const lit_utf8_byte_t expected_data[] = "234abcstring";
ecma_string_t *str_p = ecma_new_ecma_string_from_utf8 (expected_data, sizeof (expected_data) - 1);
ecma_string_t *str_p = ecma_new_ecma_string_from_ascii (expected_data, sizeof (expected_data) - 1);
TEST_ASSERT (ecma_compare_ecma_strings (result_p, str_p));
ecma_deref_ecma_string (result_p);
ecma_deref_ecma_string (str_p);