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

Refactor RegExp builtin (#3136)

Browse Source 
JerryScript-DCO-1.0-Signed-off-by: Dániel Bátyai dbatyai@inf.u-szeged.hu
This commit is contained in:
Dániel Bátyai
2019-10-02 16:55:16 +02:00
committed by Robert Fancsik
parent c3510fc03d
commit c3bb516e4a
20 changed files with 1415 additions and 1337 deletions
Download Patch File Download Diff File Expand all files Collapse all files
jerry-core/parser/regexp/re-parser.c
+16 -274
View File
@@ -40,26 +40,26 @@
* @return true - if lookup number of characters ahead are hex digits
* false - otherwise
*/
static bool
bool
re_hex_lookup (re_parser_ctx_t *parser_ctx_p, /**< RegExp parser context */
uint32_t lookup) /**< size of lookup */
{
bool is_digit = true;
const lit_utf8_byte_t *curr_p = parser_ctx_p->input_curr_p;
for (uint32_t i = 0; is_digit && i < lookup; i++)
if (JERRY_UNLIKELY (curr_p + lookup > parser_ctx_p->input_end_p))
{
if (curr_p < parser_ctx_p->input_end_p)
{
is_digit = lit_char_is_hex_digit (*curr_p++);
}
else
return false;
}
for (uint32_t i = 0; i < lookup; i++)
{
if (!lit_char_is_hex_digit (*curr_p++))
{
return false;
}
}
return is_digit;
return true;
} /* re_hex_lookup */
/**
@@ -86,7 +86,7 @@ re_parse_non_greedy_char (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser cont
*
* @return uint32_t - parsed octal number
*/
static uint32_t
uint32_t
re_parse_octal (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser context */
{
uint32_t number = 0;
@@ -110,7 +110,7 @@ re_parse_octal (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser context */
*
* Returned value must be freed with ecma_free_value
*/
static ecma_value_t
ecma_value_t
re_parse_iterator (re_parser_ctx_t *parser_ctx_p, /**< RegExp parser context */
re_token_t *re_token_p) /**< [out] output token */
{
@@ -253,7 +253,7 @@ static void
re_count_num_of_groups (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser context */
{
int char_class_in = 0;
parser_ctx_p->num_of_groups = 0;
parser_ctx_p->groups_count = 0;
const lit_utf8_byte_t *curr_p = parser_ctx_p->input_start_p;
while (curr_p < parser_ctx_p->input_end_p)
@@ -287,7 +287,7 @@ re_count_num_of_groups (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser contex
&& *curr_p != LIT_CHAR_QUESTION
&& !char_class_in)
{
parser_ctx_p->num_of_groups++;
parser_ctx_p->groups_count++;
}
break;
}
@@ -295,264 +295,6 @@ re_count_num_of_groups (re_parser_ctx_t *parser_ctx_p) /**< RegExp parser contex
}
} /* re_count_num_of_groups */
/**
* Read the input pattern and parse the range of character class
*
* @return empty ecma value - if parsed successfully
* error ecma value - otherwise
*
* Returned value must be freed with ecma_free_value
*/
ecma_value_t
re_parse_char_class (re_parser_ctx_t *parser_ctx_p, /**< number of classes */
re_char_class_callback append_char_class, /**< callback function,
* which adds the char-ranges
* to the bytecode */
void *re_ctx_p, /**< regexp compiler context */
re_token_t *out_token_p) /**< [out] output token */
{
re_token_type_t token_type = ((re_compiler_ctx_t *) re_ctx_p)->current_token.type;
out_token_p->qmax = out_token_p->qmin = 1;
ecma_char_t start = LIT_CHAR_UNDEF;
bool is_range = false;
parser_ctx_p->num_of_classes = 0;
const ecma_char_t prev_char = lit_utf8_peek_prev (parser_ctx_p->input_curr_p);
if (prev_char != LIT_CHAR_LEFT_SQUARE && prev_char != LIT_CHAR_CIRCUMFLEX)
{
lit_utf8_decr (&parser_ctx_p->input_curr_p);
lit_utf8_decr (&parser_ctx_p->input_curr_p);
}
do
{
if (parser_ctx_p->input_curr_p >= parser_ctx_p->input_end_p)
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, end of string"));
}
ecma_char_t ch = lit_utf8_read_next (&parser_ctx_p->input_curr_p);
if (ch == LIT_CHAR_RIGHT_SQUARE)
{
if (start != LIT_CHAR_UNDEF)
{
append_char_class (re_ctx_p, start, start);
}
break;
}
else if (ch == LIT_CHAR_MINUS)
{
if (parser_ctx_p->input_curr_p >= parser_ctx_p->input_end_p)
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, end of string after '-'"));
}
if (start != LIT_CHAR_UNDEF
&& !is_range
&& *parser_ctx_p->input_curr_p != LIT_CHAR_RIGHT_SQUARE)
{
is_range = true;
continue;
}
}
else if (ch == LIT_CHAR_BACKSLASH)
{
if (parser_ctx_p->input_curr_p >= parser_ctx_p->input_end_p)
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, end of string after '\\'"));
}
ch = lit_utf8_read_next (&parser_ctx_p->input_curr_p);
if (ch == LIT_CHAR_LOWERCASE_B)
{
ch = LIT_CHAR_BS;
}
else if (ch == LIT_CHAR_LOWERCASE_F)
{
ch = LIT_CHAR_FF;
}
else if (ch == LIT_CHAR_LOWERCASE_N)
{
ch = LIT_CHAR_LF;
}
else if (ch == LIT_CHAR_LOWERCASE_T)
{
ch = LIT_CHAR_TAB;
}
else if (ch == LIT_CHAR_LOWERCASE_R)
{
ch = LIT_CHAR_CR;
}
else if (ch == LIT_CHAR_LOWERCASE_V)
{
ch = LIT_CHAR_VTAB;
}
else if (ch == LIT_CHAR_LOWERCASE_C)
{
if (parser_ctx_p->input_curr_p < parser_ctx_p->input_end_p)
{
ch = *parser_ctx_p->input_curr_p;
if ((ch >= LIT_CHAR_ASCII_UPPERCASE_LETTERS_BEGIN && ch <= LIT_CHAR_ASCII_UPPERCASE_LETTERS_END)
|| (ch >= LIT_CHAR_ASCII_LOWERCASE_LETTERS_BEGIN && ch <= LIT_CHAR_ASCII_LOWERCASE_LETTERS_END)
|| (ch >= LIT_CHAR_0 && ch <= LIT_CHAR_9))
{
/* See ECMA-262 v5, 15.10.2.10 (Point 3) */
ch = (ch % 32);
parser_ctx_p->input_curr_p++;
}
else
{
ch = LIT_CHAR_LOWERCASE_C;
}
}
}
else if (ch == LIT_CHAR_LOWERCASE_X && re_hex_lookup (parser_ctx_p, 2))
{
ecma_char_t code_unit;
if (!lit_read_code_unit_from_hex (parser_ctx_p->input_curr_p, 2, &code_unit))
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, end of string after '\\x'"));
}
parser_ctx_p->input_curr_p += 2;
if (parser_ctx_p->input_curr_p < parser_ctx_p->input_end_p
&& is_range == false
&& lit_utf8_peek_next (parser_ctx_p->input_curr_p) == LIT_CHAR_MINUS)
{
start = code_unit;
continue;
}
ch = code_unit;
}
else if (ch == LIT_CHAR_LOWERCASE_U && re_hex_lookup (parser_ctx_p, 4))
{
ecma_char_t code_unit;
if (!lit_read_code_unit_from_hex (parser_ctx_p->input_curr_p, 4, &code_unit))
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, end of string after '\\u'"));
}
parser_ctx_p->input_curr_p += 4;
if (parser_ctx_p->input_curr_p < parser_ctx_p->input_end_p
&& is_range == false
&& lit_utf8_peek_next (parser_ctx_p->input_curr_p) == LIT_CHAR_MINUS)
{
start = code_unit;
continue;
}
ch = code_unit;
}
else if (ch == LIT_CHAR_LOWERCASE_D)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_ASCII_DIGITS_BEGIN, LIT_CHAR_ASCII_DIGITS_END);
ch = LIT_CHAR_UNDEF;
}
else if (ch == LIT_CHAR_UPPERCASE_D)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_NULL, LIT_CHAR_ASCII_DIGITS_BEGIN - 1);
append_char_class (re_ctx_p, LIT_CHAR_ASCII_DIGITS_END + 1, LIT_UTF16_CODE_UNIT_MAX);
ch = LIT_CHAR_UNDEF;
}
else if (ch == LIT_CHAR_LOWERCASE_S)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_TAB, LIT_CHAR_CR);
append_char_class (re_ctx_p, LIT_CHAR_SP, LIT_CHAR_SP);
append_char_class (re_ctx_p, LIT_CHAR_NBSP, LIT_CHAR_NBSP);
append_char_class (re_ctx_p, 0x1680UL, 0x1680UL); /* Ogham Space Mark */
append_char_class (re_ctx_p, 0x180EUL, 0x180EUL); /* Mongolian Vowel Separator */
append_char_class (re_ctx_p, 0x2000UL, 0x200AUL); /* En Quad - Hair Space */
append_char_class (re_ctx_p, LIT_CHAR_LS, LIT_CHAR_PS);
append_char_class (re_ctx_p, 0x202FUL, 0x202FUL); /* Narrow No-Break Space */
append_char_class (re_ctx_p, 0x205FUL, 0x205FUL); /* Medium Mathematical Space */
append_char_class (re_ctx_p, 0x3000UL, 0x3000UL); /* Ideographic Space */
append_char_class (re_ctx_p, LIT_CHAR_BOM, LIT_CHAR_BOM);
ch = LIT_CHAR_UNDEF;
}
else if (ch == LIT_CHAR_UPPERCASE_S)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_NULL, LIT_CHAR_TAB - 1);
append_char_class (re_ctx_p, LIT_CHAR_CR + 1, LIT_CHAR_SP - 1);
append_char_class (re_ctx_p, LIT_CHAR_SP + 1, LIT_CHAR_NBSP - 1);
append_char_class (re_ctx_p, LIT_CHAR_NBSP + 1, 0x167FUL);
append_char_class (re_ctx_p, 0x1681UL, 0x180DUL);
append_char_class (re_ctx_p, 0x180FUL, 0x1FFFUL);
append_char_class (re_ctx_p, 0x200BUL, LIT_CHAR_LS - 1);
append_char_class (re_ctx_p, LIT_CHAR_PS + 1, 0x202EUL);
append_char_class (re_ctx_p, 0x2030UL, 0x205EUL);
append_char_class (re_ctx_p, 0x2060UL, 0x2FFFUL);
append_char_class (re_ctx_p, 0x3001UL, LIT_CHAR_BOM - 1);
append_char_class (re_ctx_p, LIT_CHAR_BOM + 1, LIT_UTF16_CODE_UNIT_MAX);
ch = LIT_CHAR_UNDEF;
}
else if (ch == LIT_CHAR_LOWERCASE_W)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_0, LIT_CHAR_9);
append_char_class (re_ctx_p, LIT_CHAR_UPPERCASE_A, LIT_CHAR_UPPERCASE_Z);
append_char_class (re_ctx_p, LIT_CHAR_UNDERSCORE, LIT_CHAR_UNDERSCORE);
append_char_class (re_ctx_p, LIT_CHAR_LOWERCASE_A, LIT_CHAR_LOWERCASE_Z);
ch = LIT_CHAR_UNDEF;
}
else if (ch == LIT_CHAR_UPPERCASE_W)
{
/* See ECMA-262 v5, 15.10.2.12 */
append_char_class (re_ctx_p, LIT_CHAR_NULL, LIT_CHAR_0 - 1);
append_char_class (re_ctx_p, LIT_CHAR_9 + 1, LIT_CHAR_UPPERCASE_A - 1);
append_char_class (re_ctx_p, LIT_CHAR_UPPERCASE_Z + 1, LIT_CHAR_UNDERSCORE - 1);
append_char_class (re_ctx_p, LIT_CHAR_UNDERSCORE + 1, LIT_CHAR_LOWERCASE_A - 1);
append_char_class (re_ctx_p, LIT_CHAR_LOWERCASE_Z + 1, LIT_UTF16_CODE_UNIT_MAX);
ch = LIT_CHAR_UNDEF;
}
else if (lit_char_is_octal_digit ((ecma_char_t) ch)
&& ch != LIT_CHAR_0)
{
lit_utf8_decr (&parser_ctx_p->input_curr_p);
ch = (ecma_char_t) re_parse_octal (parser_ctx_p);
}
} /* ch == LIT_CHAR_BACKSLASH */
if (start != LIT_CHAR_UNDEF)
{
if (is_range)
{
if (start > ch)
{
return ecma_raise_syntax_error (ECMA_ERR_MSG ("invalid character class, wrong order"));
}
else
{
append_char_class (re_ctx_p, start, ch);
start = LIT_CHAR_UNDEF;
is_range = false;
}
}
else
{
append_char_class (re_ctx_p, start, start);
start = ch;
}
}
else
{
start = ch;
}
}
while (token_type == RE_TOK_START_CHAR_CLASS || token_type == RE_TOK_START_INV_CHAR_CLASS);
return re_parse_iterator (parser_ctx_p, out_token_p);
} /* re_parse_char_class */
/**
* Read the input pattern and parse the next token for the RegExp compiler
*
@@ -730,12 +472,12 @@ re_parse_next_token (re_parser_ctx_t *parser_ctx_p, /**< RegExp parser context *
}
else
{
if (parser_ctx_p->num_of_groups == -1)
if (parser_ctx_p->groups_count == -1)
{
re_count_num_of_groups (parser_ctx_p);
}
if (parser_ctx_p->num_of_groups)
if (parser_ctx_p->groups_count)
{
parser_ctx_p->input_curr_p--;
uint32_t number = 0;
@@ -765,7 +507,7 @@ re_parse_next_token (re_parser_ctx_t *parser_ctx_p, /**< RegExp parser context *
}
while (true);
if ((int) number <= parser_ctx_p->num_of_groups)
if ((int) number <= parser_ctx_p->groups_count)
{
out_token_p->type = RE_TOK_BACKREFERENCE;
}