12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378 |
- /*
- +----------------------------------------------------------------------+
- | Zend Engine |
- +----------------------------------------------------------------------+
- | Copyright (c) 1998-2016 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"
- #ifdef PHP_WIN32
- # include <Winuser.h>
- #endif
- #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 "zend_virtual_cwd.h"
- #include "tsrm_config_common.h"
- #define YYCTYPE unsigned char
- #define YYFILL(n) { if ((YYCURSOR + n) >= (YYLIMIT + ZEND_MMAP_AHEAD)) { 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) do { YYCURSOR = (unsigned char*)yytext + x; \
- yyleng = (unsigned int)x; } while(0)
- #define yymore() goto yymore_restart
- /* perform sanity check. If this message is triggered you should
- increase the ZEND_MMAP_AHEAD value in the zend_streams.h file */
- /*!max:re2c */
- #if ZEND_MMAP_AHEAD < YYMAXFILL
- # error ZEND_MMAP_AHEAD should be greater than or equal to YYMAXFILL
- #endif
- #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)++; \
- } \
- }
- /* To save initial string length after scanning to first variable, CG(doc_comment_len) can be reused */
- #define SET_DOUBLE_QUOTES_SCANNED_LENGTH(len) CG(doc_comment_len) = (len)
- #define GET_DOUBLE_QUOTES_SCANNED_LENGTH() CG(doc_comment_len)
- #define IS_LABEL_START(c) (((c) >= 'a' && (c) <= 'z') || ((c) >= 'A' && (c) <= 'Z') || (c) == '_' || (c) >= 0x7F)
- #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 size_t encoding_filter_script_to_internal(unsigned char **to, size_t *to_length, const unsigned char *from, size_t from_length TSRMLS_DC)
- {
- const zend_encoding *internal_encoding = zend_multibyte_get_internal_encoding(TSRMLS_C);
- assert(internal_encoding && zend_multibyte_check_lexer_compatibility(internal_encoding));
- return zend_multibyte_encoding_converter(to, to_length, from, from_length, internal_encoding, LANG_SCNG(script_encoding) TSRMLS_CC);
- }
- static size_t encoding_filter_script_to_intermediate(unsigned char **to, size_t *to_length, const unsigned char *from, size_t from_length TSRMLS_DC)
- {
- return zend_multibyte_encoding_converter(to, to_length, from, from_length, zend_multibyte_encoding_utf8, LANG_SCNG(script_encoding) TSRMLS_CC);
- }
- static size_t encoding_filter_intermediate_to_script(unsigned char **to, size_t *to_length, const unsigned char *from, size_t from_length TSRMLS_DC)
- {
- return zend_multibyte_encoding_converter(to, to_length, from, from_length,
- LANG_SCNG(script_encoding), zend_multibyte_encoding_utf8 TSRMLS_CC);
- }
- static size_t encoding_filter_intermediate_to_internal(unsigned char **to, size_t *to_length, const unsigned char *from, size_t from_length TSRMLS_DC)
- {
- const zend_encoding *internal_encoding = zend_multibyte_get_internal_encoding(TSRMLS_C);
- assert(internal_encoding && zend_multibyte_check_lexer_compatibility(internal_encoding));
- return zend_multibyte_encoding_converter(to, to_length, from, from_length,
- internal_encoding, zend_multibyte_encoding_utf8 TSRMLS_CC);
- }
- 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;
- YYLIMIT = YYCURSOR + len;
- if (!SCNG(yy_start)) {
- SCNG(yy_start) = YYCURSOR;
- }
- }
- void startup_scanner(TSRMLS_D)
- {
- CG(parse_error) = 0;
- CG(doc_comment) = NULL;
- CG(doc_comment_len) = 0;
- zend_stack_init(&SCNG(state_stack));
- zend_ptr_stack_init(&SCNG(heredoc_label_stack));
- }
- static void heredoc_label_dtor(zend_heredoc_label *heredoc_label) {
- efree(heredoc_label->label);
- }
- void shutdown_scanner(TSRMLS_D)
- {
- CG(parse_error) = 0;
- RESET_DOC_COMMENT();
- zend_stack_destroy(&SCNG(state_stack));
- zend_ptr_stack_clean(&SCNG(heredoc_label_stack), (void (*)(void *)) &heredoc_label_dtor, 1);
- zend_ptr_stack_destroy(&SCNG(heredoc_label_stack));
- }
- 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));
- lex_state->heredoc_label_stack = SCNG(heredoc_label_stack);
- zend_ptr_stack_init(&SCNG(heredoc_label_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);
- lex_state->script_org = SCNG(script_org);
- lex_state->script_org_size = SCNG(script_org_size);
- lex_state->script_filtered = SCNG(script_filtered);
- lex_state->script_filtered_size = SCNG(script_filtered_size);
- lex_state->input_filter = SCNG(input_filter);
- lex_state->output_filter = SCNG(output_filter);
- lex_state->script_encoding = SCNG(script_encoding);
- }
- 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_stack_destroy(&SCNG(state_stack));
- SCNG(state_stack) = lex_state->state_stack;
- zend_ptr_stack_clean(&SCNG(heredoc_label_stack), (void (*)(void *)) &heredoc_label_dtor, 1);
- zend_ptr_stack_destroy(&SCNG(heredoc_label_stack));
- SCNG(heredoc_label_stack) = lex_state->heredoc_label_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);
- if (SCNG(script_filtered)) {
- efree(SCNG(script_filtered));
- SCNG(script_filtered) = NULL;
- }
- SCNG(script_org) = lex_state->script_org;
- SCNG(script_org_size) = lex_state->script_org_size;
- SCNG(script_filtered) = lex_state->script_filtered;
- SCNG(script_filtered_size) = lex_state->script_filtered_size;
- SCNG(input_filter) = lex_state->input_filter;
- SCNG(output_filter) = lex_state->output_filter;
- SCNG(script_encoding) = lex_state->script_encoding;
- RESET_DOC_COMMENT();
- }
- 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;
- }
- }
- #define BOM_UTF32_BE "\x00\x00\xfe\xff"
- #define BOM_UTF32_LE "\xff\xfe\x00\x00"
- #define BOM_UTF16_BE "\xfe\xff"
- #define BOM_UTF16_LE "\xff\xfe"
- #define BOM_UTF8 "\xef\xbb\xbf"
- static const zend_encoding *zend_multibyte_detect_utf_encoding(const unsigned char *script, size_t script_size TSRMLS_DC)
- {
- const unsigned char *p;
- int wchar_size = 2;
- int le = 0;
- /* utf-16 or utf-32? */
- p = script;
- while ((p-script) < script_size) {
- p = memchr(p, 0, script_size-(p-script)-2);
- if (!p) {
- break;
- }
- if (*(p+1) == '\0' && *(p+2) == '\0') {
- wchar_size = 4;
- break;
- }
- /* searching for UTF-32 specific byte orders, so this will do */
- p += 4;
- }
- /* BE or LE? */
- p = script;
- while ((p-script) < script_size) {
- if (*p == '\0' && *(p+wchar_size-1) != '\0') {
- /* BE */
- le = 0;
- break;
- } else if (*p != '\0' && *(p+wchar_size-1) == '\0') {
- /* LE* */
- le = 1;
- break;
- }
- p += wchar_size;
- }
- if (wchar_size == 2) {
- return le ? zend_multibyte_encoding_utf16le : zend_multibyte_encoding_utf16be;
- } else {
- return le ? zend_multibyte_encoding_utf32le : zend_multibyte_encoding_utf32be;
- }
- return NULL;
- }
- static const zend_encoding* zend_multibyte_detect_unicode(TSRMLS_D)
- {
- const zend_encoding *script_encoding = NULL;
- int bom_size;
- unsigned char *pos1, *pos2;
- if (LANG_SCNG(script_org_size) < sizeof(BOM_UTF32_LE)-1) {
- return NULL;
- }
- /* check out BOM */
- if (!memcmp(LANG_SCNG(script_org), BOM_UTF32_BE, sizeof(BOM_UTF32_BE)-1)) {
- script_encoding = zend_multibyte_encoding_utf32be;
- bom_size = sizeof(BOM_UTF32_BE)-1;
- } else if (!memcmp(LANG_SCNG(script_org), BOM_UTF32_LE, sizeof(BOM_UTF32_LE)-1)) {
- script_encoding = zend_multibyte_encoding_utf32le;
- bom_size = sizeof(BOM_UTF32_LE)-1;
- } else if (!memcmp(LANG_SCNG(script_org), BOM_UTF16_BE, sizeof(BOM_UTF16_BE)-1)) {
- script_encoding = zend_multibyte_encoding_utf16be;
- bom_size = sizeof(BOM_UTF16_BE)-1;
- } else if (!memcmp(LANG_SCNG(script_org), BOM_UTF16_LE, sizeof(BOM_UTF16_LE)-1)) {
- script_encoding = zend_multibyte_encoding_utf16le;
- bom_size = sizeof(BOM_UTF16_LE)-1;
- } else if (!memcmp(LANG_SCNG(script_org), BOM_UTF8, sizeof(BOM_UTF8)-1)) {
- script_encoding = zend_multibyte_encoding_utf8;
- bom_size = sizeof(BOM_UTF8)-1;
- }
- if (script_encoding) {
- /* remove BOM */
- LANG_SCNG(script_org) += bom_size;
- LANG_SCNG(script_org_size) -= bom_size;
- return script_encoding;
- }
- /* script contains NULL bytes -> auto-detection */
- if ((pos1 = memchr(LANG_SCNG(script_org), 0, LANG_SCNG(script_org_size)))) {
- /* check if the NULL byte is after the __HALT_COMPILER(); */
- pos2 = LANG_SCNG(script_org);
- while (pos1 - pos2 >= sizeof("__HALT_COMPILER();")-1) {
- pos2 = memchr(pos2, '_', pos1 - pos2);
- if (!pos2) break;
- pos2++;
- if (strncasecmp((char*)pos2, "_HALT_COMPILER", sizeof("_HALT_COMPILER")-1) == 0) {
- pos2 += sizeof("_HALT_COMPILER")-1;
- while (*pos2 == ' ' ||
- *pos2 == '\t' ||
- *pos2 == '\r' ||
- *pos2 == '\n') {
- pos2++;
- }
- if (*pos2 == '(') {
- pos2++;
- while (*pos2 == ' ' ||
- *pos2 == '\t' ||
- *pos2 == '\r' ||
- *pos2 == '\n') {
- pos2++;
- }
- if (*pos2 == ')') {
- pos2++;
- while (*pos2 == ' ' ||
- *pos2 == '\t' ||
- *pos2 == '\r' ||
- *pos2 == '\n') {
- pos2++;
- }
- if (*pos2 == ';') {
- return NULL;
- }
- }
- }
- }
- }
- /* make best effort if BOM is missing */
- return zend_multibyte_detect_utf_encoding(LANG_SCNG(script_org), LANG_SCNG(script_org_size) TSRMLS_CC);
- }
- return NULL;
- }
- static const zend_encoding* zend_multibyte_find_script_encoding(TSRMLS_D)
- {
- const zend_encoding *script_encoding;
- if (CG(detect_unicode)) {
- /* check out bom(byte order mark) and see if containing wchars */
- script_encoding = zend_multibyte_detect_unicode(TSRMLS_C);
- if (script_encoding != NULL) {
- /* bom or wchar detection is prior to 'script_encoding' option */
- return script_encoding;
- }
- }
- /* if no script_encoding specified, just leave alone */
- if (!CG(script_encoding_list) || !CG(script_encoding_list_size)) {
- return NULL;
- }
- /* if multiple encodings specified, detect automagically */
- if (CG(script_encoding_list_size) > 1) {
- return zend_multibyte_encoding_detector(LANG_SCNG(script_org), LANG_SCNG(script_org_size), CG(script_encoding_list), CG(script_encoding_list_size) TSRMLS_CC);
- }
- return CG(script_encoding_list)[0];
- }
- ZEND_API int zend_multibyte_set_filter(const zend_encoding *onetime_encoding TSRMLS_DC)
- {
- const zend_encoding *internal_encoding = zend_multibyte_get_internal_encoding(TSRMLS_C);
- const zend_encoding *script_encoding = onetime_encoding ? onetime_encoding: zend_multibyte_find_script_encoding(TSRMLS_C);
- if (!script_encoding) {
- return FAILURE;
- }
- /* judge input/output filter */
- LANG_SCNG(script_encoding) = script_encoding;
- LANG_SCNG(input_filter) = NULL;
- LANG_SCNG(output_filter) = NULL;
- if (!internal_encoding || LANG_SCNG(script_encoding) == internal_encoding) {
- if (!zend_multibyte_check_lexer_compatibility(LANG_SCNG(script_encoding))) {
- /* and if not, work around w/ script_encoding -> utf-8 -> script_encoding conversion */
- LANG_SCNG(input_filter) = encoding_filter_script_to_intermediate;
- LANG_SCNG(output_filter) = encoding_filter_intermediate_to_script;
- } else {
- LANG_SCNG(input_filter) = NULL;
- LANG_SCNG(output_filter) = NULL;
- }
- return SUCCESS;
- }
- if (zend_multibyte_check_lexer_compatibility(internal_encoding)) {
- LANG_SCNG(input_filter) = encoding_filter_script_to_internal;
- LANG_SCNG(output_filter) = NULL;
- } else if (zend_multibyte_check_lexer_compatibility(LANG_SCNG(script_encoding))) {
- LANG_SCNG(input_filter) = NULL;
- LANG_SCNG(output_filter) = encoding_filter_script_to_internal;
- } else {
- /* both script and internal encodings are incompatible w/ flex */
- LANG_SCNG(input_filter) = encoding_filter_script_to_intermediate;
- LANG_SCNG(output_filter) = encoding_filter_intermediate_to_internal;
- }
- return 0;
- }
- ZEND_API int open_file_for_scanning(zend_file_handle *file_handle TSRMLS_DC)
- {
- const char *file_path = NULL;
- char *buf;
- size_t size, offset = 0;
- /* The shebang line was read, get the current position to obtain the buffer start */
- if (CG(start_lineno) == 2 && file_handle->type == ZEND_HANDLE_FP && file_handle->handle.fp) {
- if ((offset = ftell(file_handle->handle.fp)) == -1) {
- offset = 0;
- }
- }
- 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;
- SCNG(yy_start) = NULL;
- if (size != -1) {
- if (CG(multibyte)) {
- SCNG(script_org) = (unsigned char*)buf;
- SCNG(script_org_size) = size;
- SCNG(script_filtered) = NULL;
- zend_multibyte_set_filter(NULL TSRMLS_CC);
- if (SCNG(input_filter)) {
- if ((size_t)-1 == SCNG(input_filter)(&SCNG(script_filtered), &SCNG(script_filtered_size), SCNG(script_org), SCNG(script_org_size) TSRMLS_CC)) {
- zend_error_noreturn(E_COMPILE_ERROR, "Could not convert the script from the detected "
- "encoding \"%s\" to a compatible encoding", zend_multibyte_get_encoding_name(LANG_SCNG(script_encoding)));
- }
- buf = (char*)SCNG(script_filtered);
- size = SCNG(script_filtered_size);
- }
- }
- SCNG(yy_start) = (unsigned char *)buf - offset;
- 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;
- }
- RESET_DOC_COMMENT();
- 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;
- INIT_PZVAL(&retval_znode.u.constant);
- ZVAL_LONG(&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 TSRMLS_CC);
- zend_bailout();
- } else {
- zend_message_dispatcher(ZMSG_FAILED_INCLUDE_FOPEN, file_handle->filename TSRMLS_CC);
- }
- 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;
- zend_stack_push(&CG(context_stack), (void *) &CG(context), sizeof(CG(context)));
- zend_init_compiler_context(TSRMLS_C);
- compiler_result = zendparse(TSRMLS_C);
- zend_do_return(&retval_znode, 0 TSRMLS_CC);
- CG(in_compilation) = original_in_compilation;
- if (compiler_result != 0) { /* 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);
- zend_release_labels(0 TSRMLS_CC);
- } else {
- efree(op_array);
- retval = NULL;
- }
- }
- 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 = Z_STRVAL_P(filename);
- 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(Z_STRVAL_P(filename), Z_STRLEN_P(filename));
- }
- 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)
- {
- char *buf;
- size_t size;
- /* enforce ZEND_MMAP_AHEAD trailing NULLs for flex... */
- Z_STRVAL_P(str) = str_erealloc(Z_STRVAL_P(str), Z_STRLEN_P(str) + ZEND_MMAP_AHEAD);
- memset(Z_STRVAL_P(str) + Z_STRLEN_P(str), 0, ZEND_MMAP_AHEAD);
- SCNG(yy_in) = NULL;
- SCNG(yy_start) = NULL;
- buf = Z_STRVAL_P(str);
- size = Z_STRLEN_P(str);
- if (CG(multibyte)) {
- SCNG(script_org) = (unsigned char*)buf;
- SCNG(script_org_size) = size;
- SCNG(script_filtered) = NULL;
- zend_multibyte_set_filter(zend_multibyte_get_internal_encoding(TSRMLS_C) TSRMLS_CC);
- if (SCNG(input_filter)) {
- if ((size_t)-1 == SCNG(input_filter)(&SCNG(script_filtered), &SCNG(script_filtered_size), SCNG(script_org), SCNG(script_org_size) TSRMLS_CC)) {
- zend_error_noreturn(E_COMPILE_ERROR, "Could not convert the script from the detected "
- "encoding \"%s\" to a compatible encoding", zend_multibyte_get_encoding_name(LANG_SCNG(script_encoding)));
- }
- buf = (char*)SCNG(script_filtered);
- size = SCNG(script_filtered_size);
- }
- }
- yy_scan_buffer(buf, size TSRMLS_CC);
- zend_set_compiled_filename(filename TSRMLS_CC);
- CG(zend_lineno) = 1;
- CG(increment_lineno) = 0;
- RESET_DOC_COMMENT();
- return SUCCESS;
- }
- ZEND_API size_t zend_get_scanned_file_offset(TSRMLS_D)
- {
- size_t offset = SCNG(yy_cursor) - SCNG(yy_start);
- if (SCNG(input_filter)) {
- size_t original_offset = offset, length = 0;
- do {
- unsigned char *p = NULL;
- if ((size_t)-1 == SCNG(input_filter)(&p, &length, SCNG(script_org), offset TSRMLS_CC)) {
- return (size_t)-1;
- }
- efree(p);
- if (length > original_offset) {
- offset--;
- } else if (length < original_offset) {
- offset++;
- }
- } while (original_offset != length);
- }
- return offset;
- }
- 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 (Z_STRLEN_P(source_string)==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;
- zend_stack_push(&CG(context_stack), (void *) &CG(context), sizeof(CG(context)));
- zend_init_compiler_context(TSRMLS_C);
- BEGIN(ST_IN_SCRIPTING);
- compiler_result = zendparse(TSRMLS_C);
- if (SCNG(script_filtered)) {
- efree(SCNG(script_filtered));
- SCNG(script_filtered) = NULL;
- }
- if (compiler_result != 0) {
- CG(active_op_array) = original_active_op_array;
- CG(unclean_shutdown)=1;
- destroy_op_array(op_array TSRMLS_CC);
- efree(op_array);
- retval = NULL;
- } else {
- zend_do_return(NULL, 0 TSRMLS_CC);
- CG(active_op_array) = original_active_op_array;
- pass_two(op_array TSRMLS_CC);
- zend_release_labels(0 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 TSRMLS_CC);
- zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
- return FAILURE;
- }
- zend_highlight(syntax_highlighter_ini TSRMLS_CC);
- if (SCNG(script_filtered)) {
- efree(SCNG(script_filtered));
- SCNG(script_filtered) = NULL;
- }
- 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) {
- zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
- return FAILURE;
- }
- BEGIN(INITIAL);
- zend_highlight(syntax_highlighter_ini TSRMLS_CC);
- if (SCNG(script_filtered)) {
- efree(SCNG(script_filtered));
- SCNG(script_filtered) = NULL;
- }
- zend_restore_lexical_state(&original_lex_state TSRMLS_CC);
- zval_dtor(str);
- return SUCCESS;
- }
- ZEND_API void zend_multibyte_yyinput_again(zend_encoding_filter old_input_filter, const zend_encoding *old_encoding TSRMLS_DC)
- {
- size_t length;
- unsigned char *new_yy_start;
- /* convert and set */
- if (!SCNG(input_filter)) {
- if (SCNG(script_filtered)) {
- efree(SCNG(script_filtered));
- SCNG(script_filtered) = NULL;
- }
- SCNG(script_filtered_size) = 0;
- length = SCNG(script_org_size);
- new_yy_start = SCNG(script_org);
- } else {
- if ((size_t)-1 == SCNG(input_filter)(&new_yy_start, &length, SCNG(script_org), SCNG(script_org_size) TSRMLS_CC)) {
- zend_error_noreturn(E_COMPILE_ERROR, "Could not convert the script from the detected "
- "encoding \"%s\" to a compatible encoding", zend_multibyte_get_encoding_name(LANG_SCNG(script_encoding)));
- }
- SCNG(script_filtered) = new_yy_start;
- SCNG(script_filtered_size) = length;
- }
- SCNG(yy_cursor) = new_yy_start + (SCNG(yy_cursor) - SCNG(yy_start));
- SCNG(yy_marker) = new_yy_start + (SCNG(yy_marker) - SCNG(yy_start));
- SCNG(yy_text) = new_yy_start + (SCNG(yy_text) - SCNG(yy_start));
- SCNG(yy_limit) = new_yy_start + (SCNG(yy_limit) - SCNG(yy_start));
- SCNG(yy_start) = new_yy_start;
- }
- # define zend_copy_value(zendlval, yytext, yyleng) \
- if (SCNG(output_filter)) { \
- size_t sz = 0; \
- SCNG(output_filter)((unsigned char **)&Z_STRVAL_P(zendlval), &sz, (unsigned char *)yytext, (size_t)yyleng TSRMLS_CC); \
- Z_STRLEN_P(zendlval) = sz; \
- } else { \
- Z_STRVAL_P(zendlval) = (char *) estrndup(yytext, yyleng); \
- Z_STRLEN_P(zendlval) = 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 = Z_STRVAL_P(zendlval);
- end = s+Z_STRLEN_P(zendlval);
- while (s<end) {
- if (*s=='\\') {
- s++;
- if (s >= end) {
- *t++ = '\\';
- break;
- }
- switch(*s) {
- case 'n':
- *t++ = '\n';
- Z_STRLEN_P(zendlval)--;
- break;
- case 'r':
- *t++ = '\r';
- Z_STRLEN_P(zendlval)--;
- break;
- case 't':
- *t++ = '\t';
- Z_STRLEN_P(zendlval)--;
- break;
- case 'f':
- *t++ = '\f';
- Z_STRLEN_P(zendlval)--;
- break;
- case 'v':
- *t++ = '\v';
- Z_STRLEN_P(zendlval)--;
- break;
- case 'e':
- #ifdef PHP_WIN32
- *t++ = VK_ESCAPE;
- #else
- *t++ = '\e';
- #endif
- Z_STRLEN_P(zendlval)--;
- break;
- case '"':
- case '`':
- if (*s != quote_type) {
- *t++ = '\\';
- *t++ = *s;
- break;
- }
- case '\\':
- case '$':
- *t++ = *s;
- Z_STRLEN_P(zendlval)--;
- break;
- case 'x':
- case 'X':
- if (ZEND_IS_HEX(*(s+1))) {
- char hex_buf[3] = { 0, 0, 0 };
- Z_STRLEN_P(zendlval)--; /* for the 'x' */
- hex_buf[0] = *(++s);
- Z_STRLEN_P(zendlval)--;
- if (ZEND_IS_HEX(*(s+1))) {
- hex_buf[1] = *(++s);
- Z_STRLEN_P(zendlval)--;
- }
- *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;
- Z_STRLEN_P(zendlval)--;
- if (ZEND_IS_OCT(*(s+1))) {
- octal_buf[1] = *(++s);
- Z_STRLEN_P(zendlval)--;
- if (ZEND_IS_OCT(*(s+1))) {
- octal_buf[2] = *(++s);
- Z_STRLEN_P(zendlval)--;
- }
- }
- *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;
- if (SCNG(output_filter)) {
- size_t sz = 0;
- s = Z_STRVAL_P(zendlval);
- SCNG(output_filter)((unsigned char **)&Z_STRVAL_P(zendlval), &sz, (unsigned char *)s, (size_t)Z_STRLEN_P(zendlval) TSRMLS_CC);
- Z_STRLEN_P(zendlval) = sz;
- efree(s);
- }
- }
- int lex_scan(zval *zendlval TSRMLS_DC)
- {
- restart:
- SCNG(yy_text) = YYCURSOR;
- yymore_restart:
- /*!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]+
- BNUM "0b"[01]+
- 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")
- /* 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>"yield" {
- return T_YIELD;
- }
- <ST_IN_SCRIPTING>"try" {
- return T_TRY;
- }
- <ST_IN_SCRIPTING>"catch" {
- return T_CATCH;
- }
- <ST_IN_SCRIPTING>"finally" {
- return T_FINALLY;
- }
- <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>"goto" {
- return T_GOTO;
- }
- <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>"trait" {
- return T_TRAIT;
- }
- <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_IN_SCRIPTING,ST_LOOKING_FOR_PROPERTY>{WHITESPACE}+ {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- HANDLE_NEWLINES(yytext, yyleng);
- return T_WHITESPACE;
- }
- <ST_LOOKING_FOR_PROPERTY>"->" {
- return T_OBJECT_OPERATOR;
- }
- <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>"\\" {
- return T_NS_SEPARATOR;
- }
- <ST_IN_SCRIPTING>"..." {
- return T_ELLIPSIS;
- }
- <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"|"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>"insteadof" {
- return T_INSTEADOF;
- }
- <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>"callable" {
- return T_CALLABLE;
- }
- <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_POW;
- }
- <ST_IN_SCRIPTING>"*\*=" {
- return T_POW_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();
- if (!zend_stack_is_empty(&SCNG(state_stack))) {
- yy_pop_state(TSRMLS_C);
- }
- return '}';
- }
- <ST_LOOKING_FOR_VARNAME>{LABEL}[[}] {
- yyless(yyleng - 1);
- 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>{BNUM} {
- char *bin = yytext + 2; /* Skip "0b" */
- int len = yyleng - 2;
- /* Skip any leading 0s */
- while (*bin == '0') {
- ++bin;
- --len;
- }
- if (len < SIZEOF_LONG * 8) {
- if (len == 0) {
- Z_LVAL_P(zendlval) = 0;
- } else {
- Z_LVAL_P(zendlval) = strtol(bin, NULL, 2);
- }
- zendlval->type = IS_LONG;
- return T_LNUMBER;
- } else {
- ZVAL_DOUBLE(zendlval, zend_bin_strtod(bin, NULL));
- return T_DNUMBER;
- }
- }
- <ST_IN_SCRIPTING>{LNUM} {
- if (yyleng < MAX_LENGTH_OF_LONG - 1) { /* Won't overflow */
- Z_LVAL_P(zendlval) = strtol(yytext, NULL, 0);
- } else {
- errno = 0;
- Z_LVAL_P(zendlval) = strtol(yytext, NULL, 0);
- if (errno == ERANGE) { /* Overflow */
- if (yytext[0] == '0') { /* octal overflow */
- Z_DVAL_P(zendlval) = zend_oct_strtod(yytext, NULL);
- } else {
- Z_DVAL_P(zendlval) = 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')) {
- if (len == 0) {
- Z_LVAL_P(zendlval) = 0;
- } else {
- Z_LVAL_P(zendlval) = strtol(hex, NULL, 16);
- }
- zendlval->type = IS_LONG;
- return T_LNUMBER;
- } else {
- ZVAL_DOUBLE(zendlval, zend_hex_strtod(hex, NULL));
- 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)) {
- ZVAL_LONG(zendlval, strtol(yytext, NULL, 10));
- } else {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 1);
- }
- return T_NUM_STRING;
- }
- <ST_VAR_OFFSET>{LNUM}|{HNUM}|{BNUM} { /* Offset must be treated as a string */
- ZVAL_STRINGL(zendlval, yytext, yyleng, 1);
- return T_NUM_STRING;
- }
- <ST_IN_SCRIPTING>{DNUM}|{EXPONENT_DNUM} {
- ZVAL_DOUBLE(zendlval, zend_strtod(yytext, NULL));
- return T_DNUMBER;
- }
- <ST_IN_SCRIPTING>"__CLASS__" {
- zend_class_entry *ce = CG(active_class_entry);
- if (ce && ZEND_ACC_TRAIT == (ce->ce_flags & ZEND_ACC_TRAIT)) {
- /* We create a special __CLASS__ constant that is going to be resolved
- at run-time */
- Z_STRLEN_P(zendlval) = sizeof("__CLASS__")-1;
- Z_STRVAL_P(zendlval) = estrndup("__CLASS__", Z_STRLEN_P(zendlval));
- zendlval->type = IS_CONSTANT;
- } else {
- if (ce && ce->name) {
- ZVAL_STRINGL(zendlval, ce->name, ce->name_length, 1);
- } else {
- ZVAL_EMPTY_STRING(zendlval);
- }
- }
- return T_CLASS_C;
- }
- <ST_IN_SCRIPTING>"__TRAIT__" {
- zend_class_entry *ce = CG(active_class_entry);
- if (ce && ce->name && ZEND_ACC_TRAIT == (ce->ce_flags & ZEND_ACC_TRAIT)) {
- ZVAL_STRINGL(zendlval, ce->name, ce->name_length, 1);
- } else {
- ZVAL_EMPTY_STRING(zendlval);
- }
- return T_TRAIT_C;
- }
- <ST_IN_SCRIPTING>"__FUNCTION__" {
- zend_op_array *op_array = CG(active_op_array);
- if (op_array && op_array->function_name) {
- ZVAL_STRING(zendlval, op_array->function_name, 1);
- } else {
- ZVAL_EMPTY_STRING(zendlval);
- }
- return T_FUNC_C;
- }
- <ST_IN_SCRIPTING>"__METHOD__" {
- const char *class_name = CG(active_class_entry) ? CG(active_class_entry)->name : NULL;
- const char *func_name = CG(active_op_array)? CG(active_op_array)->function_name : NULL;
- Z_STRLEN_P(zendlval) = zend_spprintf(&Z_STRVAL_P(zendlval), 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__" {
- ZVAL_LONG(zendlval, CG(zend_lineno));
- return T_LINE;
- }
- <ST_IN_SCRIPTING>"__FILE__" {
- char *filename = zend_get_compiled_filename(TSRMLS_C);
- if (!filename) {
- filename = "";
- }
- ZVAL_STRING(zendlval, filename, 1);
- 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
- }
- ZVAL_STRING(zendlval, dirname, 0);
- 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}*">" {
- YYCTYPE *bracket = (YYCTYPE*)zend_memrchr(yytext, '<', yyleng - (sizeof("script language=php>") - 1));
- if (bracket != SCNG(yy_text)) {
- /* Handle previously scanned HTML, as possible <script> tags found are assumed to not be PHP's */
- YYCURSOR = bracket;
- goto inline_html;
- }
- HANDLE_NEWLINES(yytext, yyleng);
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG;
- }
- <INITIAL>"<%=" {
- if (CG(asp_tags)) {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG_WITH_ECHO;
- } else {
- goto inline_char_handler;
- }
- }
- <INITIAL>"<?=" {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG_WITH_ECHO;
- }
- <INITIAL>"<%" {
- if (CG(asp_tags)) {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG;
- } else {
- goto inline_char_handler;
- }
- }
- <INITIAL>"<?php"([ \t]|{NEWLINE}) {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- HANDLE_NEWLINE(yytext[yyleng-1]);
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG;
- }
- <INITIAL>"<?" {
- if (CG(short_tags)) {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(ST_IN_SCRIPTING);
- return T_OPEN_TAG;
- } else {
- goto inline_char_handler;
- }
- }
- <INITIAL>{ANY_CHAR} {
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- inline_char_handler:
- while (1) {
- YYCTYPE *ptr = memchr(YYCURSOR, '<', YYLIMIT - YYCURSOR);
- YYCURSOR = ptr ? ptr + 1 : YYLIMIT;
- if (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR) {
- case '?':
- if (CG(short_tags) || !strncasecmp((char*)YYCURSOR + 1, "php", 3) || (*(YYCURSOR + 1) == '=')) { /* Assume [ \t\n\r] follows "php" */
- break;
- }
- continue;
- case '%':
- if (CG(asp_tags)) {
- break;
- }
- continue;
- case 's':
- case 'S':
- /* Probably NOT an opening PHP <script> tag, so don't end the HTML chunk yet
- * If it is, the PHP <script> tag rule checks for any HTML scanned before it */
- YYCURSOR--;
- yymore();
- default:
- continue;
- }
- YYCURSOR--;
- }
- break;
- }
- inline_html:
- yyleng = YYCURSOR - SCNG(yy_text);
- if (SCNG(output_filter)) {
- int readsize;
- size_t sz = 0;
- readsize = SCNG(output_filter)((unsigned char **)&Z_STRVAL_P(zendlval), &sz, (unsigned char *)yytext, (size_t)yyleng TSRMLS_CC);
- Z_STRLEN_P(zendlval) = sz;
- if (readsize < yyleng) {
- yyless(readsize);
- }
- } else {
- Z_STRVAL_P(zendlval) = (char *) estrndup(yytext, yyleng);
- Z_STRLEN_P(zendlval) = 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>"$"{LABEL}"->"[a-zA-Z_\x7f-\xff] {
- yyless(yyleng - 3);
- yy_push_state(ST_LOOKING_FOR_PROPERTY TSRMLS_CC);
- zend_copy_value(zendlval, (yytext+1), (yyleng-1));
- zendlval->type = IS_STRING;
- return T_VARIABLE;
- }
- /* A [ always designates a variable offset, regardless of what follows
- */
- <ST_DOUBLE_QUOTES,ST_HEREDOC,ST_BACKQUOTE>"$"{LABEL}"[" {
- yyless(yyleng - 1);
- yy_push_state(ST_VAR_OFFSET TSRMLS_CC);
- zend_copy_value(zendlval, (yytext+1), (yyleng-1));
- 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);
- 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>"#"|"//" {
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '\r':
- if (*YYCURSOR == '\n') {
- YYCURSOR++;
- }
- /* fall through */
- case '\n':
- CG(zend_lineno)++;
- break;
- case '%':
- if (!CG(asp_tags)) {
- continue;
- }
- /* fall through */
- case '?':
- if (*YYCURSOR == '>') {
- YYCURSOR--;
- break;
- }
- /* fall through */
- default:
- continue;
- }
- break;
- }
- yyleng = YYCURSOR - SCNG(yy_text);
- return T_COMMENT;
- }
- <ST_IN_SCRIPTING>"/*"|"/**"{WHITESPACE} {
- int doc_com;
- if (yyleng > 2) {
- doc_com = 1;
- RESET_DOC_COMMENT();
- } else {
- doc_com = 0;
- }
- while (YYCURSOR < YYLIMIT) {
- if (*YYCURSOR++ == '*' && *YYCURSOR == '/') {
- break;
- }
- }
- if (YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- } else {
- zend_error(E_COMPILE_WARNING, "Unterminated comment starting line %d", CG(zend_lineno));
- }
- yyleng = YYCURSOR - SCNG(yy_text);
- HANDLE_NEWLINES(yytext, yyleng);
- if (doc_com) {
- CG(doc_comment) = estrndup(yytext, yyleng);
- CG(doc_comment_len) = yyleng;
- return T_DOC_COMMENT;
- }
- return T_COMMENT;
- }
- <ST_IN_SCRIPTING>("?>"|"</script"{WHITESPACE}*">"){NEWLINE}? {
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- BEGIN(INITIAL);
- return T_CLOSE_TAG; /* implicit ';' at php-end tag */
- }
- <ST_IN_SCRIPTING>"%>"{NEWLINE}? {
- if (CG(asp_tags)) {
- BEGIN(INITIAL);
- ZVAL_STRINGL(zendlval, yytext, yyleng, 0); /* no copying - intentional */
- return T_CLOSE_TAG; /* implicit ';' at php-end tag */
- } else {
- yyless(1);
- return yytext[0];
- }
- }
- <ST_IN_SCRIPTING>b?['] {
- register char *s, *t;
- char *end;
- int bprefix = (yytext[0] != '\'') ? 1 : 0;
- while (1) {
- if (YYCURSOR < YYLIMIT) {
- if (*YYCURSOR == '\'') {
- YYCURSOR++;
- yyleng = YYCURSOR - SCNG(yy_text);
- break;
- } else if (*YYCURSOR++ == '\\' && YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- } else {
- yyleng = YYLIMIT - SCNG(yy_text);
- /* Unclosed single quotes; treat similar to double quotes, but without a separate token
- * for ' (unrecognized by parser), instead of old flex fallback to "Unexpected character..."
- * rule, which continued in ST_IN_SCRIPTING state after the quote */
- return T_ENCAPSED_AND_WHITESPACE;
- }
- }
- ZVAL_STRINGL(zendlval, yytext+bprefix+1, yyleng-bprefix-2, 1);
- /* convert escape sequences */
- s = t = Z_STRVAL_P(zendlval);
- end = s+Z_STRLEN_P(zendlval);
- while (s<end) {
- if (*s=='\\') {
- s++;
- switch(*s) {
- case '\\':
- case '\'':
- *t++ = *s;
- Z_STRLEN_P(zendlval)--;
- break;
- default:
- *t++ = '\\';
- *t++ = *s;
- break;
- }
- } else {
- *t++ = *s;
- }
- if (*s == '\n' || (*s == '\r' && (*(s+1) != '\n'))) {
- CG(zend_lineno)++;
- }
- s++;
- }
- *t = 0;
- if (SCNG(output_filter)) {
- size_t sz = 0;
- s = Z_STRVAL_P(zendlval);
- SCNG(output_filter)((unsigned char **)&Z_STRVAL_P(zendlval), &sz, (unsigned char *)s, (size_t)Z_STRLEN_P(zendlval) TSRMLS_CC);
- Z_STRLEN_P(zendlval) = sz;
- efree(s);
- }
- return T_CONSTANT_ENCAPSED_STRING;
- }
- <ST_IN_SCRIPTING>b?["] {
- int bprefix = (yytext[0] != '"') ? 1 : 0;
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '"':
- yyleng = YYCURSOR - SCNG(yy_text);
- zend_scan_escape_string(zendlval, yytext+bprefix+1, yyleng-bprefix-2, '"' TSRMLS_CC);
- return T_CONSTANT_ENCAPSED_STRING;
- case '$':
- if (IS_LABEL_START(*YYCURSOR) || *YYCURSOR == '{') {
- break;
- }
- continue;
- case '{':
- if (*YYCURSOR == '$') {
- break;
- }
- continue;
- case '\\':
- if (YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- /* fall through */
- default:
- continue;
- }
- YYCURSOR--;
- break;
- }
- /* Remember how much was scanned to save rescanning */
- SET_DOUBLE_QUOTES_SCANNED_LENGTH(YYCURSOR - SCNG(yy_text) - yyleng);
- YYCURSOR = SCNG(yy_text) + yyleng;
- BEGIN(ST_DOUBLE_QUOTES);
- return '"';
- }
- <ST_IN_SCRIPTING>b?"<<<"{TABS_AND_SPACES}({LABEL}|([']{LABEL}['])|(["]{LABEL}["])){NEWLINE} {
- char *s;
- int bprefix = (yytext[0] != '<') ? 1 : 0;
- zend_heredoc_label *heredoc_label = emalloc(sizeof(zend_heredoc_label));
- CG(zend_lineno)++;
- heredoc_label->length = yyleng-bprefix-3-1-(yytext[yyleng-2]=='\r'?1:0);
- s = yytext+bprefix+3;
- while ((*s == ' ') || (*s == '\t')) {
- s++;
- heredoc_label->length--;
- }
- if (*s == '\'') {
- s++;
- heredoc_label->length -= 2;
- BEGIN(ST_NOWDOC);
- } else {
- if (*s == '"') {
- s++;
- heredoc_label->length -= 2;
- }
- BEGIN(ST_HEREDOC);
- }
- heredoc_label->label = estrndup(s, heredoc_label->length);
- /* Check for ending label on the next line */
- if (heredoc_label->length < YYLIMIT - YYCURSOR && !memcmp(YYCURSOR, s, heredoc_label->length)) {
- YYCTYPE *end = YYCURSOR + heredoc_label->length;
- if (*end == ';') {
- end++;
- }
- if (*end == '\n' || *end == '\r') {
- BEGIN(ST_END_HEREDOC);
- }
- }
- zend_ptr_stack_push(&SCNG(heredoc_label_stack), (void *) heredoc_label);
- return T_START_HEREDOC;
- }
- <ST_IN_SCRIPTING>[`] {
- BEGIN(ST_BACKQUOTE);
- return '`';
- }
- <ST_END_HEREDOC>{ANY_CHAR} {
- zend_heredoc_label *heredoc_label = zend_ptr_stack_pop(&SCNG(heredoc_label_stack));
- YYCURSOR += heredoc_label->length - 1;
- yyleng = heredoc_label->length;
- heredoc_label_dtor(heredoc_label);
- efree(heredoc_label);
- BEGIN(ST_IN_SCRIPTING);
- return T_END_HEREDOC;
- }
- <ST_DOUBLE_QUOTES,ST_BACKQUOTE,ST_HEREDOC>"{$" {
- Z_LVAL_P(zendlval) = (long) '{';
- yy_push_state(ST_IN_SCRIPTING TSRMLS_CC);
- yyless(1);
- return T_CURLY_OPEN;
- }
- <ST_DOUBLE_QUOTES>["] {
- BEGIN(ST_IN_SCRIPTING);
- return '"';
- }
- <ST_BACKQUOTE>[`] {
- BEGIN(ST_IN_SCRIPTING);
- return '`';
- }
- <ST_DOUBLE_QUOTES>{ANY_CHAR} {
- if (GET_DOUBLE_QUOTES_SCANNED_LENGTH()) {
- YYCURSOR += GET_DOUBLE_QUOTES_SCANNED_LENGTH() - 1;
- SET_DOUBLE_QUOTES_SCANNED_LENGTH(0);
- goto double_quotes_scan_done;
- }
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- if (yytext[0] == '\\' && YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '"':
- break;
- case '$':
- if (IS_LABEL_START(*YYCURSOR) || *YYCURSOR == '{') {
- break;
- }
- continue;
- case '{':
- if (*YYCURSOR == '$') {
- break;
- }
- continue;
- case '\\':
- if (YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- /* fall through */
- default:
- continue;
- }
- YYCURSOR--;
- break;
- }
- double_quotes_scan_done:
- yyleng = YYCURSOR - SCNG(yy_text);
- zend_scan_escape_string(zendlval, yytext, yyleng, '"' TSRMLS_CC);
- return T_ENCAPSED_AND_WHITESPACE;
- }
- <ST_BACKQUOTE>{ANY_CHAR} {
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- if (yytext[0] == '\\' && YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '`':
- break;
- case '$':
- if (IS_LABEL_START(*YYCURSOR) || *YYCURSOR == '{') {
- break;
- }
- continue;
- case '{':
- if (*YYCURSOR == '$') {
- break;
- }
- continue;
- case '\\':
- if (YYCURSOR < YYLIMIT) {
- YYCURSOR++;
- }
- /* fall through */
- default:
- continue;
- }
- YYCURSOR--;
- break;
- }
- yyleng = YYCURSOR - SCNG(yy_text);
- zend_scan_escape_string(zendlval, yytext, yyleng, '`' TSRMLS_CC);
- return T_ENCAPSED_AND_WHITESPACE;
- }
- <ST_HEREDOC>{ANY_CHAR} {
- int newline = 0;
- zend_heredoc_label *heredoc_label = zend_ptr_stack_top(&SCNG(heredoc_label_stack));
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- YYCURSOR--;
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '\r':
- if (*YYCURSOR == '\n') {
- YYCURSOR++;
- }
- /* fall through */
- case '\n':
- /* Check for ending label on the next line */
- if (IS_LABEL_START(*YYCURSOR) && heredoc_label->length < YYLIMIT - YYCURSOR && !memcmp(YYCURSOR, heredoc_label->label, heredoc_label->length)) {
- YYCTYPE *end = YYCURSOR + heredoc_label->length;
- if (*end == ';') {
- end++;
- }
- if (*end == '\n' || *end == '\r') {
- /* newline before label will be subtracted from returned text, but
- * yyleng/yytext will include it, for zend_highlight/strip, tokenizer, etc. */
- if (YYCURSOR[-2] == '\r' && YYCURSOR[-1] == '\n') {
- newline = 2; /* Windows newline */
- } else {
- newline = 1;
- }
- CG(increment_lineno) = 1; /* For newline before label */
- BEGIN(ST_END_HEREDOC);
- goto heredoc_scan_done;
- }
- }
- continue;
- case '$':
- if (IS_LABEL_START(*YYCURSOR) || *YYCURSOR == '{') {
- break;
- }
- continue;
- case '{':
- if (*YYCURSOR == '$') {
- break;
- }
- continue;
- case '\\':
- if (YYCURSOR < YYLIMIT && *YYCURSOR != '\n' && *YYCURSOR != '\r') {
- YYCURSOR++;
- }
- /* fall through */
- default:
- continue;
- }
- YYCURSOR--;
- break;
- }
- heredoc_scan_done:
- yyleng = YYCURSOR - SCNG(yy_text);
- zend_scan_escape_string(zendlval, yytext, yyleng - newline, 0 TSRMLS_CC);
- return T_ENCAPSED_AND_WHITESPACE;
- }
- <ST_NOWDOC>{ANY_CHAR} {
- int newline = 0;
- zend_heredoc_label *heredoc_label = zend_ptr_stack_top(&SCNG(heredoc_label_stack));
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- YYCURSOR--;
- while (YYCURSOR < YYLIMIT) {
- switch (*YYCURSOR++) {
- case '\r':
- if (*YYCURSOR == '\n') {
- YYCURSOR++;
- }
- /* fall through */
- case '\n':
- /* Check for ending label on the next line */
- if (IS_LABEL_START(*YYCURSOR) && heredoc_label->length < YYLIMIT - YYCURSOR && !memcmp(YYCURSOR, heredoc_label->label, heredoc_label->length)) {
- YYCTYPE *end = YYCURSOR + heredoc_label->length;
- if (*end == ';') {
- end++;
- }
- if (*end == '\n' || *end == '\r') {
- /* newline before label will be subtracted from returned text, but
- * yyleng/yytext will include it, for zend_highlight/strip, tokenizer, etc. */
- if (YYCURSOR[-2] == '\r' && YYCURSOR[-1] == '\n') {
- newline = 2; /* Windows newline */
- } else {
- newline = 1;
- }
- CG(increment_lineno) = 1; /* For newline before label */
- BEGIN(ST_END_HEREDOC);
- goto nowdoc_scan_done;
- }
- }
- /* fall through */
- default:
- continue;
- }
- }
- nowdoc_scan_done:
- yyleng = YYCURSOR - SCNG(yy_text);
- zend_copy_value(zendlval, yytext, yyleng - newline);
- zendlval->type = IS_STRING;
- HANDLE_NEWLINES(yytext, yyleng - newline);
- return T_ENCAPSED_AND_WHITESPACE;
- }
- <ST_IN_SCRIPTING,ST_VAR_OFFSET>{ANY_CHAR} {
- if (YYCURSOR > YYLIMIT) {
- return 0;
- }
- zend_error(E_COMPILE_WARNING,"Unexpected character in input: '%c' (ASCII=%d) state=%d", yytext[0], yytext[0], YYSTATE);
- goto restart;
- }
- */
- }
|