mirrors/php-src
0
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2024-11-24 18:34:21 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fixes to heredoc and cleanup of new re2c scanner. (Patch by Matt Wilmas)

Browse Source
This commit is contained in:
Scott MacVicar 2008-04-09 20:50:58 +00:00
parent 54b70b6015
commit 6f20cea6d8
3 changed files with 4238 additions and 4200 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Zend/zend_globals.h
View File

@ -84,7 +84,7 @@ struct _zend_compiler_globals {
char *compiled_filename;
int zend_lineno;
int comment_start_line;
char *heredoc;
int heredoc_len;

8326
Zend/zend_language_scanner.c
View File

File diff suppressed because it is too large Load Diff

110
Zend/zend_language_scanner.l
View File

@ -598,7 +598,7 @@ yymore_restart:
if (YYCURSOR >= YYLIMIT) {
/* special case */
if (YYSTATE == STATE(ST_COMMENT) || YYSTATE == STATE(ST_DOC_COMMENT)) {
zend_error(E_COMPILE_WARNING,"Unterminated comment starting line %d", CG(comment_start_line));
zend_error(E_COMPILE_WARNING, "Unterminated comment starting line %d", CG(zend_lineno));
}
return 0;
@ -824,13 +824,16 @@ NOWDOC_CHARS ({NEWLINE}*(([^a-zA-Z_\x7f-\xff\n\r][^\n\r]*)|({LABEL}[^a-zA-Z0-9_
return T_OBJECT_OPERATOR;
}
<ST_LOOKING_FOR_PROPERTY>"->" {
return T_OBJECT_OPERATOR;
<ST_IN_SCRIPTING,ST_LOOKING_FOR_PROPERTY>{WHITESPACE}+ {
zendlval->value.str.val = yytext; /* no copying - intentional */
zendlval->value.str.len = yyleng;
zendlval->type = IS_STRING;
HANDLE_NEWLINES(yytext, yyleng);
return T_WHITESPACE;
}
<ST_LOOKING_FOR_PROPERTY>{WHITESPACE}+ {
/* do nothing */
goto restart;
<ST_LOOKING_FOR_PROPERTY>"->" {
return T_OBJECT_OPERATOR;
}
<ST_LOOKING_FOR_PROPERTY>{LABEL} {
@ -1101,7 +1104,6 @@ NOWDOC_CHARS ({NEWLINE}*(([^a-zA-Z_\x7f-\xff\n\r][^\n\r]*)|({LABEL}[^a-zA-Z0-9_
<ST_IN_SCRIPTING>"}" {
RESET_DOC_COMMENT();
/* This is a temporary fix which is dependant on flex and it's implementation */
if (!zend_stack_is_empty(&SCNG(state_stack))) {
yy_pop_state(TSRMLS_C);
}
@ -1430,9 +1432,10 @@ inline_char_handler:
* and "->" will be taken literally
*/
<ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE>"$"{LABEL}"->"[a-zA-Z_\x7f-\xff] {
yyless(yyleng - 3);
yyleng -= 3;
yyless(yyleng);
yy_push_state(ST_LOOKING_FOR_PROPERTY TSRMLS_CC);
zend_copy_value(zendlval, (yytext+1), (yyleng-4));
zend_copy_value(zendlval, (yytext+1), (yyleng-1));
zendlval->type = IS_STRING;
return T_VARIABLE;
}
@ -1440,9 +1443,10 @@ inline_char_handler:
/* A [ always designates a variable offset, regardless of what follows
*/
<ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE>"$"{LABEL}"[" {
yyless(yyleng - 1);
yyleng--;
yyless(yyleng);
yy_push_state(ST_VAR_OFFSET TSRMLS_CC);
zend_copy_value(zendlval, (yytext+1), (yyleng-2));
zend_copy_value(zendlval, (yytext+1), (yyleng-1));
zendlval->type = IS_STRING;
return T_VARIABLE;
}
@ -1466,6 +1470,7 @@ inline_char_handler:
<ST_VAR_OFFSET>[ \n\r\t\\'#] {
/* Invalid rule to return a more explicit parse error with proper line number */
yyless(0);
yyleng = 0;
yy_pop_state(TSRMLS_C);
ZVAL_EMPTY_STRING(zendlval); /* Empty since it won't be used */
return T_ENCAPSED_AND_WHITESPACE;
@ -1478,15 +1483,6 @@ inline_char_handler:
}
<ST_IN_SCRIPTING>{WHITESPACE} {
zendlval->value.str.val = yytext; /* no copying - intentional */
zendlval->value.str.len = yyleng;
zendlval->type = IS_STRING;
HANDLE_NEWLINES(yytext, yyleng);
return T_WHITESPACE;
}
<ST_IN_SCRIPTING>"#"|"//" {
BEGIN(ST_ONE_LINE_COMMENT);
yymore();
@ -1538,14 +1534,12 @@ inline_char_handler:
}
<ST_IN_SCRIPTING>"/**"{WHITESPACE} {
CG(comment_start_line) = CG(zend_lineno);
RESET_DOC_COMMENT();
BEGIN(ST_DOC_COMMENT);
yymore();
}
<ST_IN_SCRIPTING>"/*" {
CG(comment_start_line) = CG(zend_lineno);
BEGIN(ST_COMMENT);
yymore();
}
@ -1591,6 +1585,7 @@ inline_char_handler:
return T_CLOSE_TAG; /* implicit ';' at php-end tag */
} else {
yyless(1);
yyleng = 1;
return yytext[0];
}
}
@ -1659,15 +1654,19 @@ inline_char_handler:
<ST_IN_SCRIPTING>b?"<<<"{TABS_AND_SPACES}({LABEL}|["]{LABEL}["]){NEWLINE} {
char *s;
int bprefix = (yytext[0] != '<') ? 1 : 0;
int quotes = (yytext[bprefix + 3] == '"') ? 2 : 0;
CG(zend_lineno)++;
CG(heredoc_len) = yyleng-bprefix-3-quotes-1-(yytext[yyleng-2]=='\r'?1:0);
s = yytext+bprefix+3+(quotes ? 1 : 0);
CG(heredoc_len) = yyleng-bprefix-3-1-(yytext[yyleng-2]=='\r'?1:0);
s = yytext+bprefix+3;
while ((*s == ' ') || (*s == '\t')) {
s++;
CG(heredoc_len)--;
}
if (*s == '"') {
s++;
CG(heredoc_len) -= 2;
}
CG(heredoc) = estrndup(s, CG(heredoc_len));
BEGIN(ST_START_HEREDOC);
return T_START_HEREDOC;
@ -1694,6 +1693,7 @@ inline_char_handler:
}
yyless(label_len);
yyleng = label_len;
if (label_len==CG(heredoc_len) && !memcmp(yytext, CG(heredoc), label_len)) {
zendlval->value.str.val = CG(heredoc);
@ -1725,16 +1725,22 @@ inline_char_handler:
if (yyleng > CG(heredoc_len) && !memcmp(end - CG(heredoc_len), CG(heredoc), CG(heredoc_len))) {
int len = yyleng - CG(heredoc_len) - 2; /* 2 for newline before and after label */
if (len > 0 && yytext[len - 1] == '\r' && yytext[len] == '\n') {
len--;
/* May have matched fooLABEL; make sure there's a newline before it */
if (yytext[len] != '\n') {
if (yytext[len] != '\r') {
yyless(yyleng - 1);
yymore();
}
} else if (len > 0 && yytext[len - 1] == '\r') {
len--; /* Windows newline */
}
/* Go back before last label char, to match in ST_END_HEREDOC state */
yyless(yyleng - 2);
/* Subtract the remaining label length. yyleng must include newline
/* Subtract the label/newline length. yyleng must include newline
* before label, for zend_highlight/strip, tokenizer, etc. */
yyleng = yyleng - CG(heredoc_len) - 1;
yyleng -= CG(heredoc_len) + 1;
CG(increment_lineno) = 1; /* For newline before label */
BEGIN(ST_END_HEREDOC);
@ -1749,10 +1755,8 @@ inline_char_handler:
}
<ST_END_HEREDOC>{ANY_CHAR} {
zendlval->value.str.val = CG(heredoc);
zendlval->value.str.len = CG(heredoc_len);
SCNG(yy_text) = zendlval->value.str.val;
yyleng = zendlval->value.str.len;
SCNG(yy_text) = Z_STRVAL_P(zendlval) = CG(heredoc);
SCNG(yy_leng) = Z_STRLEN_P(zendlval) = CG(heredoc_len);
CG(heredoc) = NULL;
CG(heredoc_len) = 0;
BEGIN(ST_IN_SCRIPTING);
@ -1760,10 +1764,10 @@ inline_char_handler:
}
<ST_DOUBLE_QUOTES,ST_BACKQUOTE,ST_HEREDOC>"{$" {
/* Will only match when $ follows: "{$" */
<ST_DOUBLE_QUOTES,ST_BACKQUOTE,ST_HEREDOC>"{" {
zendlval->value.lval = (long) '{';
yy_push_state(ST_IN_SCRIPTING TSRMLS_CC);
yyless(1);
return T_CURLY_OPEN;
}
@ -1777,11 +1781,11 @@ inline_char_handler:
* (("{"+|"$"+)["]) handles { or $ at the end of a string
*
* Same for backquotes and heredocs, except the second case doesn't apply to
* heredocs. yyless(yyleng - 1) is used to correct taking one character too many
* heredocs. yyleng--/yyless() is used to correct taking one character too many
*/
<ST_DOUBLE_QUOTES>{DOUBLE_QUOTES_CHARS}*("{"{2,}|"$"{2,}|(("{"+|"$"+)["])) {
yyless(yyleng - 1);
if (yytext[yyleng-1] == '"') --yyleng;
yyleng--;
yyless(yyleng);
zend_scan_escape_string(zendlval, yytext, yyleng, '"' TSRMLS_CC);
return T_ENCAPSED_AND_WHITESPACE;
}
@ -1793,7 +1797,8 @@ inline_char_handler:
}
<ST_BACKQUOTE>{BACKQUOTE_CHARS}*("{"{2,}|"$"{2,}|(("{"+|"$"+)[`])) {
yyless(yyleng - 1);
yyleng--;
yyless(yyleng);
zend_scan_escape_string(zendlval, yytext, yyleng, '`' TSRMLS_CC);
return T_ENCAPSED_AND_WHITESPACE;
}
@ -1812,7 +1817,8 @@ inline_char_handler:
}
<ST_HEREDOC>{HEREDOC_CHARS}*({HEREDOC_NEWLINE}+({LABEL}";"?)?)?("{"{2,}|"$"{2,}) {
yyless(yyleng - 1);
yyleng--;
yyless(yyleng);
zend_scan_escape_string(zendlval, yytext, yyleng, 0 TSRMLS_CC);
return T_ENCAPSED_AND_WHITESPACE;
}
@ -1874,16 +1880,22 @@ inline_char_handler:
if (yyleng > CG(heredoc_len) && !memcmp(end - CG(heredoc_len), CG(heredoc), CG(heredoc_len))) {
int len = yyleng - CG(heredoc_len) - 2; /* 2 for newline before and after label */
if (len > 0 && yytext[len - 1] == '\r' && yytext[len] == '\n') {
len--;
/* May have matched fooLABEL; make sure there's a newline before it */
if (yytext[len] != '\n') {
if (yytext[len] != '\r') {
yyless(yyleng - 1);
yymore();
}
} else if (len > 0 && yytext[len - 1] == '\r') {
len--; /* Windows newline */
}
/* Go back before last label char, to match in ST_END_HEREDOC state */
/* Go back before last label char, to match in ST_END_NOWDOC state */
yyless(yyleng - 2);
/* Subtract the remaining label length. yyleng must include newline
/* Subtract the label/newline length. yyleng must include newline
* before label, for zend_highlight/strip, tokenizer, etc. */
yyleng = yyleng - CG(heredoc_len) - 1;
yyleng -= CG(heredoc_len) + 1;
CG(increment_lineno) = 1; /* For newline before label */
BEGIN(ST_END_NOWDOC);
@ -1894,17 +1906,15 @@ inline_char_handler:
return T_ENCAPSED_AND_WHITESPACE;
} else {
/* Go back to end of label, so the next match works correctly in case of
* a variable or another label at the beginning of the next line */
* another label at the beginning of the next line */
yyless(yyleng - 1);
yymore();
}
}
<ST_END_NOWDOC>{ANY_CHAR} {
Z_STRVAL_P(zendlval) = CG(heredoc);
Z_STRLEN_P(zendlval) = CG(heredoc_len);
SCNG(yy_text) = CG(heredoc);
yyleng = CG(heredoc_len);
SCNG(yy_text) = Z_STRVAL_P(zendlval) = CG(heredoc);
SCNG(yy_leng) = Z_STRLEN_P(zendlval) = CG(heredoc_len);
CG(heredoc) = NULL;
CG(heredoc_len) = 0;
BEGIN(ST_IN_SCRIPTING);