diff options
Diffstat (limited to 'ext/standard/url_scanner_ex.re')
| -rw-r--r-- | ext/standard/url_scanner_ex.re | 487 |
1 files changed, 0 insertions, 487 deletions
diff --git a/ext/standard/url_scanner_ex.re b/ext/standard/url_scanner_ex.re deleted file mode 100644 index f8208f87a0..0000000000 --- a/ext/standard/url_scanner_ex.re +++ /dev/null @@ -1,487 +0,0 @@ -/* - +----------------------------------------------------------------------+ - | PHP Version 4 | - +----------------------------------------------------------------------+ - | Copyright (c) 1997-2002 The PHP Group | - +----------------------------------------------------------------------+ - | This source file is subject to version 2.02 of the PHP license, | - | that is bundled with this package in the file LICENSE, and is | - | available at through the world-wide-web at | - | http://www.php.net/license/2_02.txt. | - | If you did not receive a copy of the PHP license and are unable to | - | obtain it through the world-wide-web, please send a note to | - | license@php.net so we can mail you a copy immediately. | - +----------------------------------------------------------------------+ - | Author: Sascha Schumann <sascha@schumann.cx> | - +----------------------------------------------------------------------+ -*/ - -#include "php.h" - -#ifdef HAVE_UNISTD_H -#include <unistd.h> -#endif - -#include <stdio.h> -#include <stdlib.h> -#include <string.h> - -#include "php_ini.h" -#include "php_globals.h" -#define STATE_TAG SOME_OTHER_STATE_TAG -#include "basic_functions.h" -#include "url.h" -#undef STATE_TAG - -#define url_scanner url_scanner_ex - -#include "php_smart_str.h" - -static PHP_INI_MH(OnUpdateTags) -{ - url_adapt_state_ex_t *ctx; - char *key; - char *lasts; - char *tmp; - - ctx = &BG(url_adapt_state_ex); - - tmp = estrndup(new_value, new_value_length); - - if (ctx->tags) - zend_hash_destroy(ctx->tags); - else - ctx->tags = malloc(sizeof(HashTable)); - - zend_hash_init(ctx->tags, 0, NULL, NULL, 1); - - for (key = php_strtok_r(tmp, ",", &lasts); - key; - key = php_strtok_r(NULL, ",", &lasts)) { - char *val; - - val = strchr(key, '='); - if (val) { - char *q; - int keylen; - - *val++ = '\0'; - for (q = key; *q; q++) - *q = tolower(*q); - keylen = q - key; - /* key is stored withOUT NUL - val is stored WITH NUL */ - zend_hash_add(ctx->tags, key, keylen, val, strlen(val)+1, NULL); - } - } - - efree(tmp); - - return SUCCESS; -} - -PHP_INI_BEGIN() - STD_PHP_INI_ENTRY("url_rewriter.tags", "a=href,area=href,frame=src,form=fakeentry", PHP_INI_ALL, OnUpdateTags, url_adapt_state_ex, php_basic_globals, basic_globals) -PHP_INI_END() - -/*!re2c -any = [\000-\377]; -N = (any\[<]); -alpha = [a-zA-Z]; -*/ - -#define YYFILL(n) goto done -#define YYCTYPE unsigned char -#define YYCURSOR p -#define YYLIMIT q -#define YYMARKER r - -static inline void append_modified_url(smart_str *url, smart_str *dest, smart_str *url_app, const char *separator) -{ - register const char *p, *q; - const char *bash = NULL; - const char *sep = "?"; - - q = (p = url->c) + url->len; - -scan: -/*!re2c - ":" { smart_str_append(dest, url); return; } - "?" { sep = separator; goto scan; } - "#" { bash = p - 1; goto done; } - (any\[:?#])+ { goto scan; } -*/ -done: - - /* Don't modify URLs of the format "#mark" */ - if (bash && bash - url->c == 0) { - smart_str_append(dest, url); - return; - } - - if (bash) - smart_str_appendl(dest, url->c, bash - url->c); - else - smart_str_append(dest, url); - - smart_str_appends(dest, sep); - smart_str_append(dest, url_app); - - if (bash) - smart_str_appendl(dest, bash, q - bash); -} - - -#undef YYFILL -#undef YYCTYPE -#undef YYCURSOR -#undef YYLIMIT -#undef YYMARKER - -static inline void tag_arg(url_adapt_state_ex_t *ctx, char quotes, char type TSRMLS_DC) -{ - char f = 0; - - if (strncasecmp(ctx->arg.c, ctx->lookup_data, ctx->arg.len) == 0) - f = 1; - - if (quotes) - smart_str_appendc(&ctx->result, type); - if (f) { - append_modified_url(&ctx->val, &ctx->result, &ctx->url_app, PG(arg_separator).output); - } else { - smart_str_append(&ctx->result, &ctx->val); - } - if (quotes) - smart_str_appendc(&ctx->result, type); -} - -enum { - STATE_PLAIN = 0, - STATE_TAG, - STATE_NEXT_ARG, - STATE_ARG, - STATE_BEFORE_VAL, - STATE_VAL -}; - -#define YYFILL(n) goto stop -#define YYCTYPE unsigned char -#define YYCURSOR xp -#define YYLIMIT end -#define YYMARKER q -#define STATE ctx->state - -#define STD_PARA url_adapt_state_ex_t *ctx, char *start, char *YYCURSOR TSRMLS_DC -#define STD_ARGS ctx, start, xp TSRMLS_CC - -static inline void passthru(STD_PARA) -{ - smart_str_appendl(&ctx->result, start, YYCURSOR - start); -} - -static inline void handle_form(STD_PARA) -{ - if (ctx->form_app.len > 0 - && ctx->tag.len == 4 - && strncasecmp(ctx->tag.c, "form", 4) == 0) { - smart_str_append(&ctx->result, &ctx->form_app); - } -} - -/* - * HANDLE_TAG copies the HTML Tag and checks whether we - * have that tag in our table. If we might modify it, - * we continue to scan the tag, otherwise we simply copy the complete - * HTML stuff to the result buffer. - */ - -static inline void handle_tag(STD_PARA) -{ - int ok = 0; - int i; - - ctx->tag.len = 0; - smart_str_appendl(&ctx->tag, start, YYCURSOR - start); - for (i = 0; i < ctx->tag.len; i++) - ctx->tag.c[i] = tolower(ctx->tag.c[i]); - if (zend_hash_find(ctx->tags, ctx->tag.c, ctx->tag.len, (void **) &ctx->lookup_data) == SUCCESS) - ok = 1; - STATE = ok ? STATE_NEXT_ARG : STATE_PLAIN; -} - -static inline void handle_arg(STD_PARA) -{ - ctx->arg.len = 0; - smart_str_appendl(&ctx->arg, start, YYCURSOR - start); -} - -static inline void handle_val(STD_PARA, char quotes, char type) -{ - smart_str_setl(&ctx->val, start + quotes, YYCURSOR - start - quotes * 2); - tag_arg(ctx, quotes, type TSRMLS_CC); -} - -#ifdef SCANNER_DEBUG -#define scdebug(x) printf x -#else -#define scdebug(x) -#endif - -static inline void xx_mainloop(url_adapt_state_ex_t *ctx, const char *newdata, size_t newlen TSRMLS_DC) -{ - char *end, *q; - char *xp; - char *start; - int rest; - - smart_str_appendl(&ctx->buf, newdata, newlen); - - YYCURSOR = ctx->buf.c; - YYLIMIT = ctx->buf.c + ctx->buf.len; - - switch (STATE) { - case STATE_PLAIN: goto state_plain; - case STATE_TAG: goto state_tag; - case STATE_NEXT_ARG: goto state_next_arg; - case STATE_ARG: goto state_arg; - case STATE_BEFORE_VAL: goto state_before_val; - case STATE_VAL: goto state_val; - } - - -state_plain_begin: - STATE = STATE_PLAIN; - -state_plain: - start = YYCURSOR; -/*!re2c - "<" { passthru(STD_ARGS); STATE = STATE_TAG; goto state_tag; } - N+ { passthru(STD_ARGS); goto state_plain; } -*/ - -state_tag: - start = YYCURSOR; -/*!re2c - alpha+ { handle_tag(STD_ARGS); /* Sets STATE */; passthru(STD_ARGS); if (STATE == STATE_PLAIN) goto state_plain; else goto state_next_arg; } - any { passthru(STD_ARGS); goto state_plain_begin; } -*/ - -state_next_arg_begin: - STATE = STATE_NEXT_ARG; - -state_next_arg: - start = YYCURSOR; -/*!re2c - ">" { passthru(STD_ARGS); handle_form(STD_ARGS); goto state_plain_begin; } - [ \v\t\n]+ { passthru(STD_ARGS); goto state_next_arg; } - alpha { --YYCURSOR; STATE = STATE_ARG; goto state_arg; } - any { passthru(STD_ARGS); goto state_plain_begin; } -*/ - -state_arg: - start = YYCURSOR; -/*!re2c - alpha+ { passthru(STD_ARGS); handle_arg(STD_ARGS); STATE = STATE_BEFORE_VAL; goto state_before_val; } - any { passthru(STD_ARGS); STATE = STATE_NEXT_ARG; goto state_next_arg; } -*/ - -state_before_val: - start = YYCURSOR; -/*!re2c - [ ]* "=" [ ]* { passthru(STD_ARGS); STATE = STATE_VAL; goto state_val; } - any { --YYCURSOR; goto state_next_arg_begin; } -*/ - - -state_val: - start = YYCURSOR; -/*!re2c - ["] (any\[">])* ["] { handle_val(STD_ARGS, 1, '"'); goto state_next_arg_begin; } - ['] (any\['>])* ['] { handle_val(STD_ARGS, 1, '\''); goto state_next_arg_begin; } - (any\[ \n>"'])+ { handle_val(STD_ARGS, 0, '"'); goto state_next_arg_begin; } - any { passthru(STD_ARGS); goto state_next_arg_begin; } -*/ - -stop: - rest = YYLIMIT - start; - scdebug(("stopped in state %d at pos %d (%d:%c) %d\n", STATE, YYCURSOR - ctx->buf.c, *YYCURSOR, *YYCURSOR, rest)); - /* XXX: Crash avoidance. Need to work with reporter to figure out what goes wrong */ - if (rest < 0) rest = 0; - - if (rest) memmove(ctx->buf.c, start, rest); - ctx->buf.len = rest; -} - -char *php_url_scanner_adapt_single_url(const char *url, size_t urllen, const char *name, const char *value, size_t *newlen TSRMLS_DC) -{ - smart_str surl = {0}; - smart_str buf = {0}; - smart_str url_app = {0}; - - smart_str_setl(&surl, url, urllen); - - smart_str_appends(&url_app, name); - smart_str_appendc(&url_app, '='); - smart_str_appends(&url_app, value); - - append_modified_url(&surl, &buf, &url_app, PG(arg_separator).output); - - smart_str_0(&buf); - if (newlen) *newlen = buf.len; - - smart_str_free(&url_app); - - return buf.c; -} - - -static char *url_adapt_ext(const char *src, size_t srclen, size_t *newlen, zend_bool do_flush TSRMLS_DC) -{ - url_adapt_state_ex_t *ctx; - char *retval; - - ctx = &BG(url_adapt_state_ex); - - xx_mainloop(ctx, src, srclen TSRMLS_CC); - - *newlen = ctx->result.len; - if (!ctx->result.c) { - smart_str_appendl(&ctx->result, "", 0); - } - smart_str_0(&ctx->result); - if (do_flush) { - smart_str_appendl(&ctx->result, ctx->buf.c, ctx->buf.len); - *newlen += ctx->buf.len; - smart_str_free(&ctx->buf); - } - retval = ctx->result.c; - ctx->result.c = NULL; - ctx->result.len = 0; - return retval; -} - -int php_url_scanner_ex_activate(TSRMLS_D) -{ - url_adapt_state_ex_t *ctx; - - ctx = &BG(url_adapt_state_ex); - - memset(ctx, 0, ((size_t) &((url_adapt_state_ex_t *)0)->tags)); - - return SUCCESS; -} - -int php_url_scanner_ex_deactivate(TSRMLS_D) -{ - url_adapt_state_ex_t *ctx; - - ctx = &BG(url_adapt_state_ex); - - smart_str_free(&ctx->result); - smart_str_free(&ctx->buf); - smart_str_free(&ctx->tag); - smart_str_free(&ctx->arg); - - return SUCCESS; -} - -static void php_url_scanner_output_handler(char *output, uint output_len, char **handled_output, uint *handled_output_len, int mode TSRMLS_DC) -{ - if (BG(url_adapt_state_ex).url_app.len != 0) { - *handled_output = url_adapt_ext(output, output_len, handled_output_len, (zend_bool) (mode&PHP_OUTPUT_HANDLER_END ? 1 : 0) TSRMLS_CC); - } else { - *handled_output = NULL; - } -} - -int php_url_scanner_add_var(char *name, int name_len, char *value, int value_len, int urlencode TSRMLS_DC) -{ - char *encoded; - int encoded_len; - smart_str val; - - if (! BG(url_adapt_state_ex).active) { - int chunk_size = 4096; /* XXX where should we get chunk_size from? */ - - php_url_scanner_ex_activate(TSRMLS_C); - php_start_ob_buffer(NULL, chunk_size, 1 TSRMLS_CC); - php_ob_set_internal_handler(php_url_scanner_output_handler, chunk_size, "URL-Rewriter", 1 TSRMLS_CC); - BG(url_adapt_state_ex).active = 1; - } - - - if (BG(url_adapt_state_ex).url_app.len != 0) { - smart_str_appends(&BG(url_adapt_state_ex).url_app, PG(arg_separator).output); - } - - if (urlencode) { - encoded = php_url_encode(value, value_len, &encoded_len); - smart_str_setl(&val, encoded, encoded_len); - } else { - smart_str_setl(&val, value, value_len); - } - - smart_str_appendl(&BG(url_adapt_state_ex).url_app, name, name_len); - smart_str_appendc(&BG(url_adapt_state_ex).url_app, '='); - smart_str_append(&BG(url_adapt_state_ex).url_app, &val); - - smart_str_appends(&BG(url_adapt_state_ex).form_app, "<input type=\"hidden\" name=\""); - smart_str_appendl(&BG(url_adapt_state_ex).form_app, name, name_len); - smart_str_appends(&BG(url_adapt_state_ex).form_app, "\" value=\""); - smart_str_append(&BG(url_adapt_state_ex).form_app, &val); - smart_str_appends(&BG(url_adapt_state_ex).form_app, "\" />"); - - if (urlencode) - efree(encoded); - - return SUCCESS; -} - -int php_url_scanner_reset_vars(TSRMLS_D) -{ - BG(url_adapt_state_ex).form_app.len = 0; - BG(url_adapt_state_ex).url_app.len = 0; - - return FAILURE; -} - -PHP_MINIT_FUNCTION(url_scanner) -{ - BG(url_adapt_state_ex).tags = NULL; - - BG(url_adapt_state_ex).form_app.c = BG(url_adapt_state_ex).url_app.c = 0; - BG(url_adapt_state_ex).form_app.len = BG(url_adapt_state_ex).url_app.len = 0; - - REGISTER_INI_ENTRIES(); - return SUCCESS; -} - -PHP_MSHUTDOWN_FUNCTION(url_scanner) -{ - UNREGISTER_INI_ENTRIES(); - zend_hash_destroy(BG(url_adapt_state_ex).tags); - free(BG(url_adapt_state_ex).tags); - - return SUCCESS; -} - -PHP_RINIT_FUNCTION(url_scanner) -{ - BG(url_adapt_state_ex).active = 0; - - return SUCCESS; -} - -PHP_RSHUTDOWN_FUNCTION(url_scanner) -{ - if (BG(url_adapt_state_ex).active) { - php_url_scanner_ex_deactivate(TSRMLS_C); - BG(url_adapt_state_ex).active = 0; - } - - smart_str_free(&BG(url_adapt_state_ex).form_app); - smart_str_free(&BG(url_adapt_state_ex).url_app); - - return SUCCESS; -} |