php-src/Zend/zend_language_scanner.l

/*
   +----------------------------------------------------------------------+
   | Zend Engine                                                          |
   +----------------------------------------------------------------------+
   | Copyright (c) 1998-2006 Zend Technologies Ltd. (http://www.zend.com) |
   +----------------------------------------------------------------------+
   | This source file is subject to version 2.00 of the Zend license,     |
   | that is bundled with this package in the file LICENSE, and is        |
   | available through the world-wide-web at the following url:           |
   | http://www.zend.com/license/2_00.txt.                                |
   | If you did not receive a copy of the Zend license and are unable to  |
   | obtain it through the world-wide-web, please send a note to          |
   | license@zend.com so we can mail you a copy immediately.              |
   +----------------------------------------------------------------------+
   | Authors: Marcus Boerger <helly@php.net>                              |
   |          Nuno Lopes <nlopess@php.net>                                |
   |          Scott MacVicar <scottmac@php.net>                           |
   | Flex version authors:                                                |
   |          Andi Gutmans <andi@zend.com>                                |
   |          Zeev Suraski <zeev@zend.com>                                |
   +----------------------------------------------------------------------+
*/

/* $Id$ */

#if 0
# define YYDEBUG(s, c) printf("state: %d char: %c\n", s, c)
#else
# define YYDEBUG(s, c)
#endif

#include "zend_language_scanner_defs.h"

#include <errno.h>
#include "zend.h"
#include "zend_alloc.h"
#include <zend_language_parser.h>
#include "zend_compile.h"
#include "zend_language_scanner.h"
#include "zend_highlight.h"
#include "zend_constants.h"
#include "zend_variables.h"
#include "zend_operators.h"
#include "zend_API.h"
#include "zend_strtod.h"
#include "zend_exceptions.h"
#include "tsrm_virtual_cwd.h"
#include "tsrm_config_common.h"

#define YYCTYPE   unsigned char
#define YYFILL(n) { if (YYCURSOR >= YYLIMIT) return 0; }
#define YYCURSOR  SCNG(yy_cursor)
#define YYLIMIT   SCNG(yy_limit)
#define YYMARKER  SCNG(yy_marker)

#define YYGETCONDITION()  SCNG(yy_state)
#define YYSETCONDITION(s) SCNG(yy_state) = s

#define STATE(name)  yyc##name

/* emulate flex constructs */
#define BEGIN(state) YYSETCONDITION(STATE(state))
#define YYSTATE      YYGETCONDITION()
#define yytext       ((char*)SCNG(yy_text))
#define yyleng       SCNG(yy_leng)
#define yyless(x)    YYCURSOR = yytext + x
#define yymore()     goto yymore_restart

#ifdef HAVE_STDARG_H
# include <stdarg.h>
#endif

#ifdef HAVE_UNISTD_H
# include <unistd.h>
#endif

/* Globals Macros */
#define SCNG	LANG_SCNG
#ifdef ZTS
ZEND_API ts_rsrc_id language_scanner_globals_id;
#else
ZEND_API zend_php_scanner_globals language_scanner_globals;
#endif

#define HANDLE_NEWLINES(s, l)													\
do {																			\
	char *p = (s), *boundary = p+(l);											\
																				\
	while (p<boundary) {														\
		if (*p == '\n' || (*p == '\r' && (*(p+1) != '\n'))) {					\
			CG(zend_lineno)++;													\
		}																		\
		p++;																	\
	}																			\
} while (0)

#define HANDLE_NEWLINE(c) \
{ \
	if (c == '\n' || c == '\r') { \
		CG(zend_lineno)++; \
	} \
}

#define ZEND_IS_OCT(c)  ((c)>='0' && (c)<='7')
#define ZEND_IS_HEX(c)  (((c)>='0' && (c)<='9') || ((c)>='a' && (c)<='f') || ((c)>='A' && (c)<='F'))

BEGIN_EXTERN_C()

static void _yy_push_state(int new_state TSRMLS_DC)
{
	zend_stack_push(&SCNG(state_stack), (void *) &YYGETCONDITION(), sizeof(int));
	YYSETCONDITION(new_state);
}

#define yy_push_state(state_and_tsrm) _yy_push_state(yyc##state_and_tsrm)

static void yy_pop_state(TSRMLS_D)
{
	int *stack_state;
	zend_stack_top(&SCNG(state_stack), (void **) &stack_state);
	YYSETCONDITION(*stack_state);
	zend_stack_del_top(&SCNG(state_stack));
}

static void yy_scan_buffer(char *str, unsigned int len TSRMLS_DC)
{
	YYCURSOR       = (YYCTYPE*)str;
	SCNG(yy_start) = YYCURSOR;
	YYLIMIT        = YYCURSOR + len;
}

void startup_scanner(TSRMLS_D)
{
	CG(heredoc) = NULL;
	CG(heredoc_len) = 0;
	CG(doc_comment) = NULL;
	CG(doc_comment_len) = 0;
	zend_llist_init(&SCNG(used_state_stacks), sizeof(zend_stack), (llist_dtor_func_t) zend_stack_destroy, 0);
	zend_stack_init(&SCNG(state_stack));
	zend_llist_add_element(&SCNG(used_state_stacks), &SCNG(state_stack));
}

void shutdown_scanner(TSRMLS_D)
{
	if (CG(heredoc)) {
		efree(CG(heredoc));
		CG(heredoc_len)=0;
	}
	zend_llist_destroy(&SCNG(used_state_stacks));
	RESET_DOC_COMMENT();
}

static int compare_stacks(zend_stack *stack1, zend_stack *stack2)
{
	return (stack1 == stack2);
}

ZEND_API void zend_save_lexical_state(zend_lex_state *lex_state TSRMLS_DC)
{
	lex_state->yy_leng   = SCNG(yy_leng);
	lex_state->yy_start  = SCNG(yy_start);
	lex_state->yy_text   = SCNG(yy_text);
	lex_state->yy_cursor = SCNG(yy_cursor);
	lex_state->yy_marker = SCNG(yy_marker);
	lex_state->yy_limit  = SCNG(yy_limit);

	lex_state->state_stack = SCNG(state_stack);
	zend_stack_init(&SCNG(state_stack));
	zend_llist_add_element(&SCNG(used_state_stacks), &SCNG(state_stack));

	lex_state->in = SCNG(yy_in);
	lex_state->yy_state = YYSTATE;
	lex_state->filename = zend_get_compiled_filename(TSRMLS_C);
	lex_state->lineno = CG(zend_lineno);
}

ZEND_API void zend_restore_lexical_state(zend_lex_state *lex_state TSRMLS_DC)
{
	SCNG(yy_leng)   = lex_state->yy_leng;
	SCNG(yy_start)  = lex_state->yy_start;
	SCNG(yy_text)   = lex_state->yy_text;
	SCNG(yy_cursor) = lex_state->yy_cursor;
	SCNG(yy_marker) = lex_state->yy_marker;
	SCNG(yy_limit)  = lex_state->yy_limit;

	zend_llist_del_element(&SCNG(used_state_stacks), &SCNG(state_stack), (int (*)(void *, void *)) compare_stacks);
	SCNG(state_stack) = lex_state->state_stack;

	SCNG(yy_in) = lex_state->in;
	YYSETCONDITION(lex_state->yy_state);
	CG(zend_lineno) = lex_state->lineno;
	zend_restore_compiled_filename(lex_state->filename TSRMLS_CC);
}

ZEND_API void zend_destroy_file_handle(zend_file_handle *file_handle TSRMLS_DC)
{
	zend_llist_del_element(&CG(open_files), file_handle, (int (*)(void *, void *)) zend_compare_file_handles);
	/* zend_file_handle_dtor() operates on the copy, so we have to NULLify the original here */
	file_handle->opened_path = NULL;
	if (file_handle->free_filename) {
		file_handle->filename = NULL;
	}
}


ZEND_API int open_file_for_scanning(zend_file_handle *file_handle TSRMLS_DC)
{
	char *file_path = NULL, *buf;
	size_t size;

	if (zend_stream_fixup(file_handle, &buf, &size TSRMLS_CC) == FAILURE) {
		return FAILURE;
	}

	zend_llist_add_element(&CG(open_files), file_handle);
	if (file_handle->handle.stream.handle >= (void*)file_handle && file_handle->handle.stream.handle <= (void*)(file_handle+1)) {
		zend_file_handle *fh = (zend_file_handle*)zend_llist_get_last(&CG(open_files));
		size_t diff = (char*)file_handle->handle.stream.handle - (char*)file_handle;
		fh->handle.stream.handle = (void*)(((char*)fh) + diff);
		file_handle->handle.stream.handle = fh->handle.stream.handle;
	}

	/* Reset the scanner for scanning the new file */
	SCNG(yy_in) = file_handle;

	if (size != -1) {
		/* TODO: For MULTIBYTE we need to do some clever re-encoding */
		yy_scan_buffer(buf, size TSRMLS_CC);
	} else {
		zend_error_noreturn(E_COMPILE_ERROR, "zend_stream_mmap() failed");
	}

	BEGIN(INITIAL);

	if (file_handle->opened_path) {
		file_path = file_handle->opened_path;
	} else {
		file_path = file_handle->filename;
	}

	zend_set_compiled_filename(file_path TSRMLS_CC);

	if (CG(start_lineno)) {
		CG(zend_lineno) = CG(start_lineno);
		CG(start_lineno) = 0;
	} else {
		CG(zend_lineno) = 1;
	}

	CG(increment_lineno) = 0;
	return SUCCESS;
}
END_EXTERN_C()


ZEND_API zend_op_array *compile_file(zend_file_handle *file_handle, int type TSRMLS_DC)
{
	zend_lex_state original_lex_state;
	zend_op_array *op_array = (zend_op_array *) emalloc(sizeof(zend_op_array));
	zend_op_array *original_active_op_array = CG(active_op_array);
	zend_op_array *retval=NULL;
	int compiler_result;
	zend_bool compilation_successful=0;
	znode retval_znode;
	zend_bool original_in_compilation = CG(in_compilation);

	retval_znode.op_type = IS_CONST;
	retval_znode.u.constant.type = IS_LONG;
	retval_znode.u.constant.value.lval = 1;
	Z_UNSET_ISREF(retval_znode.u.constant);
	Z_SET_REFCOUNT(retval_znode.u.constant, 1);

	zend_save_lexical_state(&original_lex_state TSRMLS_CC);

	retval = op_array; /* success oriented */

	if (open_file_for_scanning(file_handle TSRMLS_CC)==FAILURE) {
		if (type==ZEND_REQUIRE) {
			zend_message_dispatcher(ZMSG_FAILED_REQUIRE_FOPEN, file_handle->filename);
			zend_bailout();
		} else {
			zend_message_dispatcher(ZMSG_FAILED_INCLUDE_FOPEN, file_handle->filename);
		}
		compilation_successful=0;
	} else {
		init_op_array(op_array, ZEND_USER_FUNCTION, INITIAL_OP_ARRAY_SIZE TSRMLS_CC);
		CG(in_compilation) = 1;
		CG(active_op_array) = op_array;
		compiler_result = zendparse(TSRMLS_C);
		zend_do_return(&retval_znode, 0 TSRMLS_CC);
		CG(in_compilation) = original_in_compilation;
		if (compiler_result==1) { /* parser error */
			zend_bailout();
		}
		compilation_successful=1;
	}

	if (retval) {
		CG(active_op_array) = original_active_op_array;
		if (compilation_successful) {
			pass_two(op_array TSRMLS_CC);
		} else {
			efree(op_array);
			retval = NULL;
		}
	}
	if (compilation_successful) {
		zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
	}
	return retval;
}


zend_op_array *compile_filename(int type, zval *filename TSRMLS_DC)
{
	zend_file_handle file_handle;
	zval tmp;
	zend_op_array *retval;
	char *opened_path = NULL;

	if (filename->type != IS_STRING) {
		tmp = *filename;
		zval_copy_ctor(&tmp);
		convert_to_string(&tmp);
		filename = &tmp;
	}
	file_handle.filename = filename->value.str.val;
	file_handle.free_filename = 0;
	file_handle.type = ZEND_HANDLE_FILENAME;
	file_handle.opened_path = NULL;
	file_handle.handle.fp = NULL;

	retval = zend_compile_file(&file_handle, type TSRMLS_CC);
	if (retval && file_handle.handle.stream.handle) {
		int dummy = 1;

		if (!file_handle.opened_path) {
			file_handle.opened_path = opened_path = estrndup(filename->value.str.val, filename->value.str.len);
		}

		zend_hash_add(&EG(included_files), file_handle.opened_path, strlen(file_handle.opened_path)+1, (void *)&dummy, sizeof(int), NULL);

		if (opened_path) {
			efree(opened_path);
		}
	}
	zend_destroy_file_handle(&file_handle TSRMLS_CC);

	if (filename==&tmp) {
		zval_dtor(&tmp);
	}
	return retval;
}

ZEND_API int zend_prepare_string_for_scanning(zval *str, char *filename TSRMLS_DC)
{
	/* enforce two trailing NULLs for flex... */
	STR_REALLOC(str->value.str.val, str->value.str.len+2);

	str->value.str.val[str->value.str.len+1]=0;

	SCNG(yy_in)=NULL;

	/* TODO: For MULTIBYTE we need to do some clever re-encoding */
	yy_scan_buffer(str->value.str.val, str->value.str.len TSRMLS_CC);

	zend_set_compiled_filename(filename TSRMLS_CC);
	CG(zend_lineno) = 1;
	CG(increment_lineno) = 0;
	return SUCCESS;
}


ZEND_API int zend_get_scanned_file_offset(TSRMLS_D)
{
	return SCNG(yy_cursor) - SCNG(yy_start);
}


zend_op_array *compile_string(zval *source_string, char *filename TSRMLS_DC)
{
	zend_lex_state original_lex_state;
	zend_op_array *op_array = (zend_op_array *) emalloc(sizeof(zend_op_array));
	zend_op_array *original_active_op_array = CG(active_op_array);
	zend_op_array *retval;
	zval tmp;
	int compiler_result;
	zend_bool original_in_compilation = CG(in_compilation);

	if (source_string->value.str.len==0) {
		efree(op_array);
		return NULL;
	}

	CG(in_compilation) = 1;

	tmp = *source_string;
	zval_copy_ctor(&tmp);
	convert_to_string(&tmp);
	source_string = &tmp;

	zend_save_lexical_state(&original_lex_state TSRMLS_CC);
	if (zend_prepare_string_for_scanning(source_string, filename TSRMLS_CC)==FAILURE) {
		efree(op_array);
		retval = NULL;
	} else {
		zend_bool orig_interactive = CG(interactive);

		CG(interactive) = 0;
		init_op_array(op_array, ZEND_EVAL_CODE, INITIAL_OP_ARRAY_SIZE TSRMLS_CC);
		CG(interactive) = orig_interactive;
		CG(active_op_array) = op_array;
		BEGIN(ST_IN_SCRIPTING);
		compiler_result = zendparse(TSRMLS_C);

		if (compiler_result==1) {
			CG(active_op_array) = original_active_op_array;
			CG(unclean_shutdown)=1;
			retval = NULL;
		} else {
			zend_do_return(NULL, 0 TSRMLS_CC);
			CG(active_op_array) = original_active_op_array;
			pass_two(op_array TSRMLS_CC);
			retval = op_array;
		}
		zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
	}
	zval_dtor(&tmp);
	CG(in_compilation) = original_in_compilation;
	return retval;
}


BEGIN_EXTERN_C()
int highlight_file(char *filename, zend_syntax_highlighter_ini *syntax_highlighter_ini TSRMLS_DC)
{
	zend_lex_state original_lex_state;
	zend_file_handle file_handle;

	file_handle.type = ZEND_HANDLE_FILENAME;
	file_handle.filename = filename;
	file_handle.free_filename = 0;
	file_handle.opened_path = NULL;
	zend_save_lexical_state(&original_lex_state TSRMLS_CC);
	if (open_file_for_scanning(&file_handle TSRMLS_CC)==FAILURE) {
		zend_message_dispatcher(ZMSG_FAILED_HIGHLIGHT_FOPEN, filename);
		return FAILURE;
	}
	zend_highlight(syntax_highlighter_ini TSRMLS_CC);
	zend_destroy_file_handle(&file_handle TSRMLS_CC);
	zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
	return SUCCESS;
}

int highlight_string(zval *str, zend_syntax_highlighter_ini *syntax_highlighter_ini, char *str_name TSRMLS_DC)
{
	zend_lex_state original_lex_state;
	zval tmp = *str;

	str = &tmp;
	zval_copy_ctor(str);
	zend_save_lexical_state(&original_lex_state TSRMLS_CC);
	if (zend_prepare_string_for_scanning(str, str_name TSRMLS_CC)==FAILURE) {
		return FAILURE;
	}
	BEGIN(INITIAL);
	zend_highlight(syntax_highlighter_ini TSRMLS_CC);
	zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
	zval_dtor(str);
	return SUCCESS;
}
END_EXTERN_C()

# define zend_copy_value(zendlval, yytext, yyleng) \
	zendlval->value.str.val = (char *)estrndup(yytext, yyleng); \
	zendlval->value.str.len = yyleng;

static void zend_scan_escape_string(zval *zendlval, char *str, int len, char quote_type TSRMLS_DC)
{
	register char *s, *t;
	char *end;

	ZVAL_STRINGL(zendlval, str, len, 1);

	/* convert escape sequences */
	s = t = zendlval->value.str.val;
	end = s+zendlval->value.str.len;
	while (s<end) {
		if (*s=='\\') {
			s++;
			if (s>=end) {
				continue;
			}
			switch(*s) {
				case 'n':
					*t++ = '\n';
					zendlval->value.str.len--;
					break;
				case 'r':
					*t++ = '\r';
					zendlval->value.str.len--;
					break;
				case 't':
					*t++ = '\t';
					zendlval->value.str.len--;
					break;
				case 'f':
					*t++ = '\f';
					zendlval->value.str.len--;
					break;
				case 'v':
					*t++ = '\v';
					zendlval->value.str.len--;
					break;
				case '"':
				case '`':
					if (*s != quote_type) {
						*t++ = '\\';
						*t++ = *s;
						break;
					}
				case '\\':
				case '$':
					*t++ = *s;
					zendlval->value.str.len--;
					break;
				case 'x':
				case 'X':
					if (ZEND_IS_HEX(*(s+1))) {
						char hex_buf[3] = { 0, 0, 0 };

						zendlval->value.str.len--; /* for the 'x' */

						hex_buf[0] = *(++s);
						zendlval->value.str.len--;
						if (ZEND_IS_HEX(*(s+1))) {
							hex_buf[1] = *(++s);
							zendlval->value.str.len--;
						}
						*t++ = (char) strtol(hex_buf, NULL, 16);
					} else {
						*t++ = '\\';
						*t++ = *s;
					}
					break;
				default:
					/* check for an octal */
					if (ZEND_IS_OCT(*s)) {
						char octal_buf[4] = { 0, 0, 0, 0 };

						octal_buf[0] = *s;
						zendlval->value.str.len--;
						if (ZEND_IS_OCT(*(s+1))) {
							octal_buf[1] = *(++s);
							zendlval->value.str.len--;
							if (ZEND_IS_OCT(*(s+1))) {
								octal_buf[2] = *(++s);
								zendlval->value.str.len--;
							}
						}
						*t++ = (char) strtol(octal_buf, NULL, 8);
					} else {
						*t++ = '\\';
						*t++ = *s;
					}
					break;
			}
		} else {
			*t++ = *s;
		}

		if (*s == '\n' || (*s == '\r' && (*(s+1) != '\n'))) {
			CG(zend_lineno)++;
		}
		s++;
	}
	*t = 0;
}


int lex_scan(zval *zendlval TSRMLS_DC)
{
restart:
	SCNG(yy_text) = YYCURSOR;

yymore_restart:

	/* detect EOF */
	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));
		}

		return 0;
	}

/*!re2c
re2c:yyfill:check = 0;
LNUM	[0-9]+
DNUM	([0-9]*"."[0-9]+)|([0-9]+"."[0-9]*)
EXPONENT_DNUM	(({LNUM}|{DNUM})[eE][+-]?{LNUM})
HNUM	"0x"[0-9a-fA-F]+
LABEL	[a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*
WHITESPACE [ \n\r\t]+
TABS_AND_SPACES [ \t]*
TOKENS [;:,.\[\]()|^&+-/*=%!~$<>?@]
ANY_CHAR [^]
NEWLINE ("\r"|"\n"|"\r\n")

/*
 * LITERAL_DOLLAR matches unescaped $ that aren't followed by a label character
 * or a { and therefore will be taken literally. The case of literal $ before
 * a variable or "${" is handled in a rule for each string type
 */
DOUBLE_QUOTES_LITERAL_DOLLAR ("$"+([^a-zA-Z_\x7f-\xff$"\\{]|("\\"{ANY_CHAR})))
BACKQUOTE_LITERAL_DOLLAR     ("$"+([^a-zA-Z_\x7f-\xff$`\\{]|("\\"{ANY_CHAR})))
HEREDOC_LITERAL_DOLLAR       ("$"+([^a-zA-Z_\x7f-\xff$\n\r\\{]|("\\"[^\n\r])))

/*
 * Usually, HEREDOC_NEWLINE will just function like a simple NEWLINE, but some
 * special cases need to be handled. HEREDOC_CHARS doesn't allow a line to
 * match when { or $, and/or \ is at the end. (("{"*|"$"*)"\\"?) handles that,
 * along with cases where { or $, and/or \ is the ONLY thing on a line
 *
 * The other case is when a line contains a label, followed by ONLY
 * { or $, and/or \  Handled by ({LABEL}";"?((("{"+|"$"+)"\\"?)|"\\"))
 */
HEREDOC_NEWLINE ((({LABEL}";"?((("{"+|"$"+)"\\"?)|"\\"))|(("{"*|"$"*)"\\"?)){NEWLINE})

/*
 * This pattern is just used in the next 2 for matching { or literal $, and/or
 * \ escape sequence immediately at the beginning of a line or after a label
 */
HEREDOC_CURLY_OR_ESCAPE_OR_DOLLAR (("{"+[^$\n\r\\{])|("{"*"\\"[^\n\r])|{HEREDOC_LITERAL_DOLLAR})

/*
 * These 2 label-related patterns allow HEREDOC_CHARS to continue "regular"
 * matching after a newline that starts with either a non-label character or a
 * label that isn't followed by a newline. Like HEREDOC_CHARS, they won't match
 * a variable or "{$"  Matching a newline, and possibly label, up TO a variable
 * or "{$", is handled in the heredoc rules
 *
 * The HEREDOC_LABEL_NO_NEWLINE pattern (";"[^$\n\r\\{]) handles cases where ;
 * follows a label. [^a-zA-Z0-9_\x7f-\xff;$\n\r\\{] is needed to prevent a label
 * character or ; from matching on a possible (real) ending label
 */
HEREDOC_NON_LABEL ([^a-zA-Z_\x7f-\xff$\n\r\\{]|{HEREDOC_CURLY_OR_ESCAPE_OR_DOLLAR})
HEREDOC_LABEL_NO_NEWLINE ({LABEL}([^a-zA-Z0-9_\x7f-\xff;$\n\r\\{]|(";"[^$\n\r\\{])|(";"?{HEREDOC_CURLY_OR_ESCAPE_OR_DOLLAR})))

/*
 * CHARS matches everything up to a variable or "{$"
 * {'s are matched as long as they aren't followed by a $
 * The case of { before "{$" is handled in a rule for each string type
 *
 * For heredocs, matching continues across/after newlines if/when it's known
 * that the next line doesn't contain a possible ending label
 */
DOUBLE_QUOTES_CHARS ("{"*([^$"\\{]|("\\"{ANY_CHAR}))|{DOUBLE_QUOTES_LITERAL_DOLLAR})
BACKQUOTE_CHARS     ("{"*([^$`\\{]|("\\"{ANY_CHAR}))|{BACKQUOTE_LITERAL_DOLLAR})
HEREDOC_CHARS       ("{"*([^$\n\r\\{]|("\\"[^\n\r]))|{HEREDOC_LITERAL_DOLLAR}|({HEREDOC_NEWLINE}+({HEREDOC_NON_LABEL}|{HEREDOC_LABEL_NO_NEWLINE})))

NOWDOC_CHARS		({NEWLINE}*(([^a-zA-Z_\x7f-\xff\n\r][^\n\r]*)|({LABEL}[^a-zA-Z0-9_\x7f-\xff;\n\r][^\n\r]*)|({LABEL}[;][^\n\r]+)))

/* compute yyleng before each rule */
<!*> := yyleng = YYCURSOR - SCNG(yy_text);


<ST_IN_SCRIPTING>"exit" {
	return T_EXIT;
}

<ST_IN_SCRIPTING>"die" {
	return T_EXIT;
}

<ST_IN_SCRIPTING>"function" {
	return T_FUNCTION;
}

<ST_IN_SCRIPTING>"const" {
	return T_CONST;
}

<ST_IN_SCRIPTING>"return" {
	return T_RETURN;
}

<ST_IN_SCRIPTING>"try" {
	return T_TRY;
}

<ST_IN_SCRIPTING>"catch" {
	return T_CATCH;
}

<ST_IN_SCRIPTING>"throw" {
	return T_THROW;
}

<ST_IN_SCRIPTING>"if" {
	return T_IF;
}

<ST_IN_SCRIPTING>"elseif" {
	return T_ELSEIF;
}

<ST_IN_SCRIPTING>"endif" {
	return T_ENDIF;
}

<ST_IN_SCRIPTING>"else" {
	return T_ELSE;
}

<ST_IN_SCRIPTING>"while" {
	return T_WHILE;
}

<ST_IN_SCRIPTING>"endwhile" {
	return T_ENDWHILE;
}

<ST_IN_SCRIPTING>"do" {
	return T_DO;
}

<ST_IN_SCRIPTING>"for" {
	return T_FOR;
}

<ST_IN_SCRIPTING>"endfor" {
	return T_ENDFOR;
}

<ST_IN_SCRIPTING>"foreach" {
	return T_FOREACH;
}

<ST_IN_SCRIPTING>"endforeach" {
	return T_ENDFOREACH;
}

<ST_IN_SCRIPTING>"declare" {
	return T_DECLARE;
}

<ST_IN_SCRIPTING>"enddeclare" {
	return T_ENDDECLARE;
}

<ST_IN_SCRIPTING>"instanceof" {
	return T_INSTANCEOF;
}

<ST_IN_SCRIPTING>"as" {
	return T_AS;
}

<ST_IN_SCRIPTING>"switch" {
	return T_SWITCH;
}

<ST_IN_SCRIPTING>"endswitch" {
	return T_ENDSWITCH;
}

<ST_IN_SCRIPTING>"case" {
	return T_CASE;
}

<ST_IN_SCRIPTING>"default" {
	return T_DEFAULT;
}

<ST_IN_SCRIPTING>"break" {
	return T_BREAK;
}

<ST_IN_SCRIPTING>"continue" {
	return T_CONTINUE;
}

<ST_IN_SCRIPTING>"echo" {
	return T_ECHO;
}

<ST_IN_SCRIPTING>"print" {
	return T_PRINT;
}

<ST_IN_SCRIPTING>"class" {
	return T_CLASS;
}

<ST_IN_SCRIPTING>"interface" {
	return T_INTERFACE;
}

<ST_IN_SCRIPTING>"extends" {
	return T_EXTENDS;
}

<ST_IN_SCRIPTING>"implements" {
	return T_IMPLEMENTS;
}

<ST_IN_SCRIPTING>"->" {
	yy_push_state(ST_LOOKING_FOR_PROPERTY TSRMLS_CC);
	return T_OBJECT_OPERATOR;
}

<ST_LOOKING_FOR_PROPERTY>"->" {
	return T_OBJECT_OPERATOR;
}

<ST_LOOKING_FOR_PROPERTY>{WHITESPACE}+ {
	/* do nothing */
	goto restart;
}

<ST_LOOKING_FOR_PROPERTY>{LABEL} {
	yy_pop_state(TSRMLS_C);
	zend_copy_value(zendlval, yytext, yyleng);
	zendlval->type = IS_STRING;
	return T_STRING;
}

<ST_LOOKING_FOR_PROPERTY>{ANY_CHAR} {
	yyless(0);
	yy_pop_state(TSRMLS_C);
	goto restart;
}

<ST_IN_SCRIPTING>"::" {
	return T_PAAMAYIM_NEKUDOTAYIM;
}

<ST_IN_SCRIPTING>"new" {
	return T_NEW;
}

<ST_IN_SCRIPTING>"clone" {
	return T_CLONE;
}

<ST_IN_SCRIPTING>"var" {
	return T_VAR;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}("int"|"integer"){TABS_AND_SPACES}")" {
	return T_INT_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}("real"|"double"|"float"){TABS_AND_SPACES}")" {
	return T_DOUBLE_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}"string"{TABS_AND_SPACES}")" {
	return T_STRING_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}"binary"{TABS_AND_SPACES}")" {
	return T_STRING_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}"array"{TABS_AND_SPACES}")" {
	return T_ARRAY_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}"object"{TABS_AND_SPACES}")" {
	return T_OBJECT_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}("bool"|"boolean"){TABS_AND_SPACES}")" {
	return T_BOOL_CAST;
}

<ST_IN_SCRIPTING>"("{TABS_AND_SPACES}("unset"){TABS_AND_SPACES}")" {
	return T_UNSET_CAST;
}

<ST_IN_SCRIPTING>"eval" {
	return T_EVAL;
}

<ST_IN_SCRIPTING>"include" {
	return T_INCLUDE;
}

<ST_IN_SCRIPTING>"include_once" {
	return T_INCLUDE_ONCE;
}

<ST_IN_SCRIPTING>"require" {
	return T_REQUIRE;
}

<ST_IN_SCRIPTING>"require_once" {
	return T_REQUIRE_ONCE;
}

<ST_IN_SCRIPTING>"namespace" {
	return T_NAMESPACE;
}

<ST_IN_SCRIPTING>"use" {
	return T_USE;
}

<ST_IN_SCRIPTING>"global" {
	return T_GLOBAL;
}

<ST_IN_SCRIPTING>"isset" {
	return T_ISSET;
}

<ST_IN_SCRIPTING>"empty" {
	return T_EMPTY;
}

<ST_IN_SCRIPTING>"__halt_compiler" {
	return T_HALT_COMPILER;
}

<ST_IN_SCRIPTING>"static" {
	return T_STATIC;
}

<ST_IN_SCRIPTING>"abstract" {
	return T_ABSTRACT;
}

<ST_IN_SCRIPTING>"final" {
	return T_FINAL;
}

<ST_IN_SCRIPTING>"private" {
	return T_PRIVATE;
}

<ST_IN_SCRIPTING>"protected" {
	return T_PROTECTED;
}

<ST_IN_SCRIPTING>"public" {
	return T_PUBLIC;
}

<ST_IN_SCRIPTING>"unset" {
	return T_UNSET;
}

<ST_IN_SCRIPTING>"=>" {
	return T_DOUBLE_ARROW;
}

<ST_IN_SCRIPTING>"list" {
	return T_LIST;
}

<ST_IN_SCRIPTING>"array" {
	return T_ARRAY;
}

<ST_IN_SCRIPTING>"++" {
	return T_INC;
}

<ST_IN_SCRIPTING>"--" {
	return T_DEC;
}

<ST_IN_SCRIPTING>"===" {
	return T_IS_IDENTICAL;
}

<ST_IN_SCRIPTING>"!==" {
	return T_IS_NOT_IDENTICAL;
}

<ST_IN_SCRIPTING>"==" {
	return T_IS_EQUAL;
}

<ST_IN_SCRIPTING>"!="|"<>" {
	return T_IS_NOT_EQUAL;
}

<ST_IN_SCRIPTING>"<=" {
	return T_IS_SMALLER_OR_EQUAL;
}

<ST_IN_SCRIPTING>">=" {
	return T_IS_GREATER_OR_EQUAL;
}

<ST_IN_SCRIPTING>"+=" {
	return T_PLUS_EQUAL;
}

<ST_IN_SCRIPTING>"-=" {
	return T_MINUS_EQUAL;
}

<ST_IN_SCRIPTING>"*=" {
	return T_MUL_EQUAL;
}

<ST_IN_SCRIPTING>"/=" {
	return T_DIV_EQUAL;
}

<ST_IN_SCRIPTING>".=" {
	return T_CONCAT_EQUAL;
}

<ST_IN_SCRIPTING>"%=" {
	return T_MOD_EQUAL;
}

<ST_IN_SCRIPTING>"<<=" {
	return T_SL_EQUAL;
}

<ST_IN_SCRIPTING>">>=" {
	return T_SR_EQUAL;
}

<ST_IN_SCRIPTING>"&=" {
	return T_AND_EQUAL;
}

<ST_IN_SCRIPTING>"|=" {
	return T_OR_EQUAL;
}

<ST_IN_SCRIPTING>"^=" {
	return T_XOR_EQUAL;
}

<ST_IN_SCRIPTING>"||" {
	return T_BOOLEAN_OR;
}

<ST_IN_SCRIPTING>"&&" {
	return T_BOOLEAN_AND;
}

<ST_IN_SCRIPTING>"OR" {
	return T_LOGICAL_OR;
}

<ST_IN_SCRIPTING>"AND" {
	return T_LOGICAL_AND;
}

<ST_IN_SCRIPTING>"XOR" {
	return T_LOGICAL_XOR;
}

<ST_IN_SCRIPTING>"<<" {
	return T_SL;
}

<ST_IN_SCRIPTING>">>" {
	return T_SR;
}

<ST_IN_SCRIPTING>{TOKENS} {
	return yytext[0];
}


<ST_IN_SCRIPTING>"{" {
	yy_push_state(ST_IN_SCRIPTING TSRMLS_CC);
	return '{';
}


<ST_DOUBLE_QUOTES,ST_BACKQUOTE,ST_HEREDOC>"${" {
	yy_push_state(ST_LOOKING_FOR_VARNAME TSRMLS_CC);
	return T_DOLLAR_OPEN_CURLY_BRACES;
}


<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);
	}
	return '}';
}


<ST_LOOKING_FOR_VARNAME>{LABEL} {
	zend_copy_value(zendlval, yytext, yyleng);
	zendlval->type = IS_STRING;
	yy_pop_state(TSRMLS_C);
	yy_push_state(ST_IN_SCRIPTING TSRMLS_CC);
	return T_STRING_VARNAME;
}


<ST_LOOKING_FOR_VARNAME>{ANY_CHAR} {
	yyless(0);
	yy_pop_state(TSRMLS_C);
	yy_push_state(ST_IN_SCRIPTING TSRMLS_CC);
	goto restart;
}


<ST_IN_SCRIPTING>{LNUM} {
	if (yyleng < MAX_LENGTH_OF_LONG - 1) { /* Won't overflow */
		zendlval->value.lval = strtol(yytext, NULL, 0);
	} else {
		errno = 0;
		zendlval->value.lval = strtol(yytext, NULL, 0);
		if (errno == ERANGE) { /* Overflow */
			if (yytext[0] == '0') { /* octal overflow */
				zendlval->value.dval = zend_oct_strtod(yytext, NULL);
			} else {
				zendlval->value.dval = zend_strtod(yytext, NULL);
			}
			zendlval->type = IS_DOUBLE;
			return T_DNUMBER;
		}
	}

	zendlval->type = IS_LONG;
	return T_LNUMBER;
}

<ST_IN_SCRIPTING>{HNUM} {
	char *hex = yytext + 2; /* Skip "0x" */
	int len = yyleng - 2;

	/* Skip any leading 0s */
	while (*hex == '0') {
		hex++;
		len--;
	}

	if (len < SIZEOF_LONG * 2 || (len == SIZEOF_LONG * 2 && *hex <= '7')) {
		zendlval->value.lval = strtol(hex, NULL, 16);
		zendlval->type = IS_LONG;
		return T_LNUMBER;
	} else {
		zendlval->value.dval = zend_hex_strtod(hex, NULL);
		zendlval->type = IS_DOUBLE;
		return T_DNUMBER;
	}
}

<ST_VAR_OFFSET>[0]|([1-9][0-9]*) { /* Offset could be treated as a long */
	if (yyleng < MAX_LENGTH_OF_LONG - 1 || (yyleng == MAX_LENGTH_OF_LONG - 1 && strcmp(yytext, long_min_digits) < 0)) {
		zendlval->value.lval = strtol(yytext, NULL, 10);
		zendlval->type = IS_LONG;
	} else {
		zendlval->value.str.val = (char *)estrndup(yytext, yyleng);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
	}
	return T_NUM_STRING;
}

<ST_VAR_OFFSET>{LNUM}|{HNUM} { /* Offset must be treated as a string */
	zendlval->value.str.val = (char *)estrndup(yytext, yyleng);
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	return T_NUM_STRING;
}

<ST_IN_SCRIPTING>{DNUM}|{EXPONENT_DNUM} {
	zendlval->value.dval = zend_strtod(yytext, NULL);
	zendlval->type = IS_DOUBLE;
	return T_DNUMBER;
}

<ST_IN_SCRIPTING>'__CLASS__' {
	char *class_name = NULL;

	if (CG(active_class_entry)) {
		class_name = CG(active_class_entry)->name;
	}

	if (!class_name) {
		class_name = "";
	}
	zendlval->value.str.len = strlen(class_name);
	zendlval->value.str.val = estrndup(class_name, zendlval->value.str.len);
	zendlval->type = IS_STRING;
	return T_CLASS_C;
}

<ST_IN_SCRIPTING>'__FUNCTION__' {
	char *func_name = NULL;

	if (CG(active_op_array)) {
		func_name = CG(active_op_array)->function_name;
	}

	if (!func_name) {
		func_name = "";
	}
	zendlval->value.str.len = strlen(func_name);
	zendlval->value.str.val = estrndup(func_name, zendlval->value.str.len);
	zendlval->type = IS_STRING;
	return T_FUNC_C;
}

<ST_IN_SCRIPTING>'__METHOD__' {
	char *class_name = CG(active_class_entry) ? CG(active_class_entry)->name : NULL;
	char *func_name = CG(active_op_array)? CG(active_op_array)->function_name : NULL;
	size_t len = 0;

	if (class_name) {
		len += strlen(class_name) + 2;
	}
	if (func_name) {
		len += strlen(func_name);
	}

	zendlval->value.str.len = zend_spprintf(&zendlval->value.str.val, 0, "%s%s%s", 
		class_name ? class_name : "",
		class_name && func_name ? "::" : "",
		func_name ? func_name : ""
		);
	zendlval->type = IS_STRING;
	return T_METHOD_C;
}

<ST_IN_SCRIPTING>'__LINE__' {
	zendlval->value.lval = CG(zend_lineno);
	zendlval->type = IS_LONG;
	return T_LINE;
}

<ST_IN_SCRIPTING>'__FILE__' {
	char *filename = zend_get_compiled_filename(TSRMLS_C);

	if (!filename) {
		filename = "";
	}
	zendlval->value.str.len = strlen(filename);
	zendlval->value.str.val = estrndup(filename, zendlval->value.str.len);
	zendlval->type = IS_STRING;
	return T_FILE;
}

<ST_IN_SCRIPTING>'__DIR__' {
	char *filename = zend_get_compiled_filename(TSRMLS_C);
	const size_t filename_len = strlen(filename);
	char *dirname;

	if (!filename) {
		filename = "";
	}

	dirname = estrndup(filename, filename_len);
	zend_dirname(dirname, filename_len);

	if (strcmp(dirname, ".") == 0) {
		dirname = erealloc(dirname, MAXPATHLEN);
#if HAVE_GETCWD
		VCWD_GETCWD(dirname, MAXPATHLEN);
#elif HAVE_GETWD
		VCWD_GETWD(dirname);
#endif
	}

	zendlval->value.str.len = strlen(dirname);
	zendlval->value.str.val = dirname;
	zendlval->type = IS_STRING;
	return T_DIR;
}

<ST_IN_SCRIPTING>'__NAMESPACE__' {
	if (CG(current_namespace)) {
		*zendlval = *CG(current_namespace);
		zval_copy_ctor(zendlval);
	} else {
		ZVAL_EMPTY_STRING(zendlval);
	}
	return T_NS_C;
}

<INITIAL>"<script"{WHITESPACE}+"language"{WHITESPACE}*"="{WHITESPACE}*("php"|"\"php\""|"'php'"){WHITESPACE}*">" {
	HANDLE_NEWLINES(yytext, yyleng);
	zendlval->value.str.val = yytext; /* no copying - intentional */
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	BEGIN(ST_IN_SCRIPTING);
	return T_OPEN_TAG;
}


<INITIAL>"<%=" {
	if (CG(asp_tags)) {
		zendlval->value.str.val = yytext; /* no copying - intentional */
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		BEGIN(ST_IN_SCRIPTING);
		return T_OPEN_TAG_WITH_ECHO;
	} else {
		zendlval->value.str.val = (char *) estrndup(yytext, yyleng);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		return T_INLINE_HTML;
	}
}


<INITIAL>"<?=" {
	if (CG(short_tags)) {
		zendlval->value.str.val = yytext; /* no copying - intentional */
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		BEGIN(ST_IN_SCRIPTING);
		return T_OPEN_TAG_WITH_ECHO;
	} else {
		zendlval->value.str.val = (char *) estrndup(yytext, yyleng);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		return T_INLINE_HTML;
	}
}


<INITIAL>"<%" {
	if (CG(asp_tags)) {
		zendlval->value.str.val = yytext; /* no copying - intentional */
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		BEGIN(ST_IN_SCRIPTING);
		return T_OPEN_TAG;
	} else {
		zendlval->value.str.val = (char *) estrndup(yytext, yyleng);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		return T_INLINE_HTML;
	}
}


<INITIAL>"<?php"([ \t]|{NEWLINE}) {
	zendlval->value.str.val = yytext; /* no copying - intentional */
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	HANDLE_NEWLINE(yytext[yyleng-1]);
	BEGIN(ST_IN_SCRIPTING);
	return T_OPEN_TAG;
}


<INITIAL>"<?" {
	HANDLE_NEWLINES(yytext, yyleng);
	if (CG(short_tags)) {
		zendlval->value.str.val = yytext; /* no copying - intentional */
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		BEGIN(ST_IN_SCRIPTING);
		return T_OPEN_TAG;
	} else {
		zendlval->value.str.val = (char *) estrndup(yytext, yyleng);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		return T_INLINE_HTML;
	}
}

<INITIAL>"#".+ {NEWLINE} {
	if ((YYCTYPE*)yytext == SCNG(yy_start)) {
		/* ignore first line when it's started with a # */
		goto restart;
	} else {
		goto inline_char_handler;
	}
}

<INITIAL>{ANY_CHAR} {

	while (1) {
		YYCTYPE *ptr = memchr(YYCURSOR, '<', YYLIMIT - YYCURSOR);

		if (ptr == NULL) {
			YYCURSOR = YYLIMIT;
			yyleng   = YYCURSOR - SCNG(yy_text);
			break;

		} else {
			YYCURSOR = ptr + 1;

			/* if it can be an opening tag, stop */
			if (ptr < YYLIMIT && (*YYCURSOR == '?' || *YYCURSOR == '%')) {
				--YYCURSOR;
				yyleng = YYCURSOR - SCNG(yy_text);
				break;
			}
		}
	}

inline_char_handler:

	zendlval->value.str.val = (char *) estrndup(yytext, yyleng);
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	HANDLE_NEWLINES(yytext, yyleng);
	return T_INLINE_HTML;
}


/* Make sure a label character follows "->", otherwise there is no property
 * and "->" will be taken literally
 */
<ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE>"$"{LABEL}"->"[a-zA-Z_\x7f-\xff] {
	yyless(yyleng - 3);
	yy_push_state(ST_LOOKING_FOR_PROPERTY TSRMLS_CC);
	zend_copy_value(zendlval, (yytext+1), (yyleng-4));
	zendlval->type = IS_STRING;
	return T_VARIABLE;
}

/* A [ always designates a variable offset, regardless of what follows
 */
<ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE>"$"{LABEL}"[" {
	yyless(yyleng - 1);
	yy_push_state(ST_VAR_OFFSET TSRMLS_CC);
	zend_copy_value(zendlval, (yytext+1), (yyleng-2));
	zendlval->type = IS_STRING;
	return T_VARIABLE;
}

<ST_IN_SCRIPTING,ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE,ST_VAR_OFFSET>"$"{LABEL} {
	zend_copy_value(zendlval, (yytext+1), (yyleng-1));
	zendlval->type = IS_STRING;
	return T_VARIABLE;
}

<ST_VAR_OFFSET>"]" {
	yy_pop_state(TSRMLS_C);
	return ']';
}

<ST_VAR_OFFSET>{TOKENS}|[{}"`] {
	/* Only '[' can be valid, but returning other tokens will allow a more explicit parse error */
	return yytext[0];
}

<ST_VAR_OFFSET>[ \n\r\t\\'#] {
	/* Invalid rule to return a more explicit parse error with proper line number */
	yyless(0);
	yy_pop_state(TSRMLS_C);
	ZVAL_EMPTY_STRING(zendlval); /* Empty since it won't be used */
	return T_ENCAPSED_AND_WHITESPACE;
}

<ST_IN_SCRIPTING,ST_VAR_OFFSET>{LABEL} {
	zend_copy_value(zendlval, yytext, yyleng);
	zendlval->type = IS_STRING;
	return T_STRING;
}


<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();
}

<ST_ONE_LINE_COMMENT>"?"|"%"|">" {
	yymore();
}

<ST_ONE_LINE_COMMENT>[^\n\r?%>]*{ANY_CHAR} {
	switch (yytext[yyleng-1]) {
		case '?': case '%': case '>':
			yyless(yyleng-1);
			yymore();
			break;
		case '\n':
			CG(zend_lineno)++;
			/* intentional fall through */
		default:
			zendlval->value.str.val = yytext; /* no copying - intentional */
			zendlval->value.str.len = yyleng;
			zendlval->type = IS_STRING;
			BEGIN(ST_IN_SCRIPTING);
			return T_COMMENT;
	}
}

<ST_ONE_LINE_COMMENT>{NEWLINE} {
	zendlval->value.str.val = yytext; /* no copying - intentional */
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	BEGIN(ST_IN_SCRIPTING);
	CG(zend_lineno)++;
	return T_COMMENT;
}

<ST_ONE_LINE_COMMENT>"?>"|"%>" {
    if (CG(asp_tags) || yytext[yyleng-2] != '%') { /* asp comment? */
		zendlval->value.str.val = yytext; /* no copying - intentional */
		zendlval->value.str.len = yyleng-2;
		zendlval->type = IS_STRING;
		yyleng -= 2;
		yyless(yyleng);
		BEGIN(ST_IN_SCRIPTING);
		return T_COMMENT;
	} else {
		yymore();
	}
}

<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();
}


<ST_COMMENT,ST_DOC_COMMENT>[^*]+ {
	yymore();
}

<ST_DOC_COMMENT>"*/" {
	CG(doc_comment) = estrndup(yytext, yyleng);
	CG(doc_comment_len) = yyleng;
	HANDLE_NEWLINES(yytext, yyleng);
	BEGIN(ST_IN_SCRIPTING);
	return T_DOC_COMMENT;
}

<ST_COMMENT>"*/" {
	HANDLE_NEWLINES(yytext, yyleng);
	BEGIN(ST_IN_SCRIPTING);
	return T_COMMENT;
}

<ST_COMMENT,ST_DOC_COMMENT>"*" {
	yymore();
}

<ST_IN_SCRIPTING>("?>"|"</script"{WHITESPACE}*">"){NEWLINE}? {
	zendlval->value.str.val = yytext; /* no copying - intentional */
	zendlval->value.str.len = yyleng;
	zendlval->type = IS_STRING;
	BEGIN(INITIAL);
	return T_CLOSE_TAG;  /* implicit ';' at php-end tag */
}


<ST_IN_SCRIPTING>"%>"{NEWLINE}? {
	if (CG(asp_tags)) {
		BEGIN(INITIAL);
		zendlval->value.str.len = yyleng;
		zendlval->type = IS_STRING;
		zendlval->value.str.val = yytext; /* no copying - intentional */
		return T_CLOSE_TAG;  /* implicit ';' at php-end tag */
	} else {
		yyless(1);
		return yytext[0];
	}
}


/* ("{"*|"$"*) handles { or $ at the end of a string (or the entire contents)
 */
<ST_IN_SCRIPTING>(b?["]{DOUBLE_QUOTES_CHARS}*("{"*|"$"*)["]) {
	int bprefix = (yytext[0] != '"') ? 1 : 0;

	zend_scan_escape_string(zendlval, yytext+bprefix+1, yyleng-bprefix-2, '"' TSRMLS_CC);
	return T_CONSTANT_ENCAPSED_STRING;
}


<ST_IN_SCRIPTING>(b?[']([^'\\]|("\\"{ANY_CHAR}))*[']) {
	register char *s, *t;
	char *end;
	int bprefix = (yytext[0] != '\'') ? 1 : 0;

	zendlval->value.str.val = estrndup(yytext+bprefix+1, yyleng-bprefix-2);
	zendlval->value.str.len = yyleng-bprefix-2;
	zendlval->type = IS_STRING;

	/* convert escape sequences */
	s = t = zendlval->value.str.val;
	end = s+zendlval->value.str.len;
	while (s<end) {
		if (*s=='\\') {
			s++;
			if (s>=end) {
				continue;
			}
			switch(*s) {
				case '\\':
				case '\'':
					*t++ = *s;
					zendlval->value.str.len--;
					break;
				default:
					*t++ = '\\';
					*t++ = *s;
					break;
			}
		} else {
			*t++ = *s;
		}

		if (*s == '\n' || (*s == '\r' && (*(s+1) != '\n'))) {
			CG(zend_lineno)++;
		}
		s++;
	}
	*t = 0;

	return T_CONSTANT_ENCAPSED_STRING;
}


<ST_IN_SCRIPTING>b?["] {
	BEGIN(ST_DOUBLE_QUOTES);
	return '"';
}


<ST_IN_SCRIPTING>b?"<<<"{TABS_AND_SPACES}{LABEL}{NEWLINE} {
	char *s;
	int bprefix = (yytext[0] != '<') ? 1 : 0;

	CG(zend_lineno)++;
	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)--;
	}
	CG(heredoc) = estrndup(s, CG(heredoc_len));
	BEGIN(ST_START_HEREDOC);
	return T_START_HEREDOC;
}


<ST_IN_SCRIPTING>[`] {
	BEGIN(ST_BACKQUOTE);
	return '`';
}


<ST_START_HEREDOC>{ANY_CHAR} {
	yyless(0);
	BEGIN(ST_HEREDOC);
	goto restart;
}

<ST_START_HEREDOC>{LABEL}";"?[\n\r] {
	int label_len = yyleng - 1;

	if (yytext[label_len-1]==';') {
		label_len--;
	}

	yyless(label_len);

	if (label_len==CG(heredoc_len) && !memcmp(yytext, CG(heredoc), label_len)) {
		zendlval->value.str.val = CG(heredoc);
		zendlval->value.str.len = label_len;
		CG(heredoc)=NULL;
		CG(heredoc_len)=0;
		BEGIN(ST_IN_SCRIPTING);
		return T_END_HEREDOC;
	} else {
		BEGIN(ST_HEREDOC);
		yymore();
	}
}

/* Match everything up to and including a possible ending label, so if the label
 * doesn't match, it's kept with the rest of the string
 *
 * {HEREDOC_NEWLINE}+ handles the case of more than one newline sequence that
 * couldn't be matched with HEREDOC_CHARS, because of the following label
 */
<ST_HEREDOC>{HEREDOC_CHARS}*{HEREDOC_NEWLINE}+{LABEL}";"?[\n\r] {
	char *end = yytext + yyleng - 1;

	if (end[-1] == ';') {
		end--;
		yyleng--;
	}

	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--;
		}

		/* 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
		 * before label, for zend_highlight/strip, tokenizer, etc. */
		yyleng = yyleng - CG(heredoc_len) - 1;

		CG(increment_lineno) = 1; /* For newline before label */
		BEGIN(ST_END_HEREDOC);
		zend_scan_escape_string(zendlval, yytext, len, 0 TSRMLS_CC);
		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 */
		yyless(yyleng - 1);
		yymore();
	}
}

<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;
	CG(heredoc) = NULL;
	CG(heredoc_len) = 0;
	BEGIN(ST_IN_SCRIPTING);
	return T_END_HEREDOC;
}


<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;
}


<ST_DOUBLE_QUOTES>{DOUBLE_QUOTES_CHARS}+ {
	zend_scan_escape_string(zendlval, yytext, yyleng, '"' TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}

/* "{"{2,}|"$"{2,} handles { before "{$" or literal $ before a variable or "${"
 * (("{"+|"$"+)["]) 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
 */
<ST_DOUBLE_QUOTES>{DOUBLE_QUOTES_CHARS}*("{"{2,}|"$"{2,}|(("{"+|"$"+)["])) {
	yyless(yyleng - 1);
	if (yytext[yyleng-1] == '"') --yyleng;
	zend_scan_escape_string(zendlval, yytext, yyleng, '"' TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}


<ST_BACKQUOTE>{BACKQUOTE_CHARS}+ {
	zend_scan_escape_string(zendlval, yytext, yyleng, '`' TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}

<ST_BACKQUOTE>{BACKQUOTE_CHARS}*("{"{2,}|"$"{2,}|(("{"+|"$"+)[`])) {
	yyless(yyleng - 1);
	zend_scan_escape_string(zendlval, yytext, yyleng, '`' TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}


/* ({HEREDOC_NEWLINE}+({LABEL}";"?)?)? handles the possible case of newline
 * sequences, possibly followed by a label, that couldn't be matched with
 * HEREDOC_CHARS because of a following variable or "{$"
 *
 * This doesn't affect real ending labels, as they are followed by a newline,
 * which will result in a longer match for the correct rule if present
 */
<ST_HEREDOC>{HEREDOC_CHARS}*({HEREDOC_NEWLINE}+({LABEL}";"?)?)? {
	zend_scan_escape_string(zendlval, yytext, yyleng, 0 TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}

<ST_HEREDOC>{HEREDOC_CHARS}*({HEREDOC_NEWLINE}+({LABEL}";"?)?)?("{"{2,}|"$"{2,}) {
	yyless(yyleng - 1);
	zend_scan_escape_string(zendlval, yytext, yyleng, 0 TSRMLS_CC);
	return T_ENCAPSED_AND_WHITESPACE;
}


/* BEGIN nowdoc */

<ST_IN_SCRIPTING>b?"<<<"{TABS_AND_SPACES}[']{LABEL}[']{NEWLINE} {
	int bprefix = (yytext[0] != '<') ? 1 : 0;
	char *s;
	CG(zend_lineno)++;
	/* 3 is <<<, 2 is quotes, 1 is newline */
	CG(heredoc_len) = yyleng-bprefix-3-2-1-(yytext[yyleng-2]=='\r'?1:0);
	s = yytext+bprefix+3;
	while ((*s == ' ') || (*s == '\t')) {
		s++;
		CG(heredoc_len)--;
	}
	s++; /* first quote */
	CG(heredoc) = estrndup(s, CG(heredoc_len));
	BEGIN(ST_START_NOWDOC);
	return T_START_NOWDOC;
}

<ST_START_NOWDOC>{ANY_CHAR} {
	yyless(0);
	BEGIN(ST_NOWDOC);
	goto restart;
}

<ST_START_NOWDOC>{LABEL}";"?[\r\n] {
	int label_len = yyleng - 1;

	if (yytext[label_len-1]==';') {
		label_len--;
	}

	if (label_len==CG(heredoc_len) && !memcmp(yytext, CG(heredoc), label_len)) {
		yyless(label_len-1);
		yyleng = 0;
		BEGIN(ST_END_NOWDOC);
		ZVAL_EMPTY_STRING(zendlval);
		return T_ENCAPSED_AND_WHITESPACE;
	} else {
		BEGIN(ST_NOWDOC);
		yyless(label_len);
		yymore();
	}
}

<ST_NOWDOC>{NOWDOC_CHARS}*{NEWLINE}+{LABEL}";"?[\n\r] {
	char *end = yytext + yyleng - 1;

	if (end[-1] == ';') {
		end--;
		yyleng--;
	}

	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--;
		}

		/* 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
		 * before label, for zend_highlight/strip, tokenizer, etc. */
		yyleng = yyleng - CG(heredoc_len) - 1;

		CG(increment_lineno) = 1; /* For newline before label */
		BEGIN(ST_END_NOWDOC);

		HANDLE_NEWLINES(yytext, len);
		zend_copy_value(zendlval, yytext, len);
		zendlval->type = IS_STRING;
		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 */
		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);
	CG(heredoc) = NULL;
	CG(heredoc_len) = 0;
	BEGIN(ST_IN_SCRIPTING);
	return T_END_NOWDOC;
}

/* END nowdoc */

<ST_DOUBLE_QUOTES>["] {
	BEGIN(ST_IN_SCRIPTING);
	return '"';
}


<ST_BACKQUOTE>[`] {
	BEGIN(ST_IN_SCRIPTING);
	return '`';
}


<ST_IN_SCRIPTING,ST_VAR_OFFSET>{ANY_CHAR} {
	zend_error(E_COMPILE_WARNING,"Unexpected character in input:  '%c' (ASCII=%d) state=%d", yytext[0], yytext[0], YYSTATE);
	goto restart;
}

*/
}