1 /*
2 +----------------------------------------------------------------------+
3 | PHP Version 7 |
4 +----------------------------------------------------------------------+
5 | Copyright (c) 1997-2017 The PHP Group |
6 +----------------------------------------------------------------------+
7 | This source file is subject to version 3.01 of the PHP license, |
8 | that is bundled with this package in the file LICENSE, and is |
9 | available through the world-wide-web at the following url: |
10 | http://www.php.net/license/3_01.txt |
11 | If you did not receive a copy of the PHP license and are unable to |
12 | obtain it through the world-wide-web, please send a note to |
13 | license@php.net so we can mail you a copy immediately. |
14 +----------------------------------------------------------------------+
15 | Author: Andrei Zmievski <andrei@php.net> |
16 +----------------------------------------------------------------------+
17 */
18
19 /* $Id$ */
20
21 #ifdef HAVE_CONFIG_H
22 #include "config.h"
23 #endif
24
25 #include "php.h"
26 #include "php_ini.h"
27 #include "ext/standard/info.h"
28 #include "php_tokenizer.h"
29
30 #include "zend.h"
31 #include "zend_exceptions.h"
32 #include "zend_language_scanner.h"
33 #include "zend_language_scanner_defs.h"
34 #include <zend_language_parser.h>
35
36 #define zendtext LANG_SCNG(yy_text)
37 #define zendleng LANG_SCNG(yy_leng)
38 #define zendcursor LANG_SCNG(yy_cursor)
39 #define zendlimit LANG_SCNG(yy_limit)
40
41 #define TOKEN_PARSE 1
42
tokenizer_token_get_all_register_constants(INIT_FUNC_ARGS)43 void tokenizer_token_get_all_register_constants(INIT_FUNC_ARGS) {
44 REGISTER_LONG_CONSTANT("TOKEN_PARSE", TOKEN_PARSE, CONST_CS|CONST_PERSISTENT);
45 }
46
47 /* {{{ arginfo */
48 ZEND_BEGIN_ARG_INFO_EX(arginfo_token_get_all, 0, 0, 1)
49 ZEND_ARG_INFO(0, source)
50 ZEND_ARG_INFO(0, flags)
51 ZEND_END_ARG_INFO()
52
53 ZEND_BEGIN_ARG_INFO_EX(arginfo_token_name, 0, 0, 1)
54 ZEND_ARG_INFO(0, token)
55 ZEND_END_ARG_INFO()
56 /* }}} */
57
58 /* {{{ tokenizer_functions[]
59 *
60 * Every user visible function must have an entry in tokenizer_functions[].
61 */
62 const zend_function_entry tokenizer_functions[] = {
63 PHP_FE(token_get_all, arginfo_token_get_all)
64 PHP_FE(token_name, arginfo_token_name)
65 PHP_FE_END
66 };
67 /* }}} */
68
69 /* {{{ tokenizer_module_entry
70 */
71 zend_module_entry tokenizer_module_entry = {
72 STANDARD_MODULE_HEADER,
73 "tokenizer",
74 tokenizer_functions,
75 PHP_MINIT(tokenizer),
76 NULL,
77 NULL,
78 NULL,
79 PHP_MINFO(tokenizer),
80 PHP_TOKENIZER_VERSION,
81 STANDARD_MODULE_PROPERTIES
82 };
83 /* }}} */
84
85 #ifdef COMPILE_DL_TOKENIZER
86 ZEND_GET_MODULE(tokenizer)
87 #endif
88
89 /* {{{ PHP_MINIT_FUNCTION
90 */
PHP_MINIT_FUNCTION(tokenizer)91 PHP_MINIT_FUNCTION(tokenizer)
92 {
93 tokenizer_register_constants(INIT_FUNC_ARGS_PASSTHRU);
94 tokenizer_token_get_all_register_constants(INIT_FUNC_ARGS_PASSTHRU);
95 return SUCCESS;
96 }
97 /* }}} */
98
99 /* {{{ PHP_MINFO_FUNCTION
100 */
PHP_MINFO_FUNCTION(tokenizer)101 PHP_MINFO_FUNCTION(tokenizer)
102 {
103 php_info_print_table_start();
104 php_info_print_table_row(2, "Tokenizer Support", "enabled");
105 php_info_print_table_end();
106 }
107 /* }}} */
108
tokenize(zval * return_value,zend_string * source)109 static zend_bool tokenize(zval *return_value, zend_string *source)
110 {
111 zval source_zval;
112 zend_lex_state original_lex_state;
113 zval token;
114 zval keyword;
115 int token_type;
116 int token_line = 1;
117 int need_tokens = -1; /* for __halt_compiler lexing. -1 = disabled */
118
119 ZVAL_STR_COPY(&source_zval, source);
120 zend_save_lexical_state(&original_lex_state);
121
122 if (zend_prepare_string_for_scanning(&source_zval, "") == FAILURE) {
123 zend_restore_lexical_state(&original_lex_state);
124 return 0;
125 }
126
127 LANG_SCNG(yy_state) = yycINITIAL;
128 array_init(return_value);
129
130 ZVAL_UNDEF(&token);
131 while ((token_type = lex_scan(&token))) {
132 if (token_type == T_CLOSE_TAG && zendtext[zendleng - 1] != '>') {
133 CG(zend_lineno)++;
134 }
135
136 if (token_type >= 256) {
137 array_init(&keyword);
138 add_next_index_long(&keyword, token_type);
139 if (token_type == T_END_HEREDOC) {
140 if (CG(increment_lineno)) {
141 token_line = ++CG(zend_lineno);
142 CG(increment_lineno) = 0;
143 }
144 }
145 add_next_index_stringl(&keyword, (char *)zendtext, zendleng);
146 add_next_index_long(&keyword, token_line);
147 add_next_index_zval(return_value, &keyword);
148 } else {
149 add_next_index_stringl(return_value, (char *)zendtext, zendleng);
150 }
151
152 if (Z_TYPE(token) != IS_UNDEF) {
153 zval_dtor(&token);
154 ZVAL_UNDEF(&token);
155 }
156
157 /* after T_HALT_COMPILER collect the next three non-dropped tokens */
158 if (need_tokens != -1) {
159 if (token_type != T_WHITESPACE && token_type != T_OPEN_TAG
160 && token_type != T_COMMENT && token_type != T_DOC_COMMENT
161 && --need_tokens == 0
162 ) {
163 /* fetch the rest into a T_INLINE_HTML */
164 if (zendcursor != zendlimit) {
165 array_init(&keyword);
166 add_next_index_long(&keyword, T_INLINE_HTML);
167 add_next_index_stringl(&keyword, (char *)zendcursor, zendlimit - zendcursor);
168 add_next_index_long(&keyword, token_line);
169 add_next_index_zval(return_value, &keyword);
170 }
171 break;
172 }
173 } else if (token_type == T_HALT_COMPILER) {
174 need_tokens = 3;
175 }
176
177 token_line = CG(zend_lineno);
178 }
179
180 zval_dtor(&source_zval);
181 zend_restore_lexical_state(&original_lex_state);
182
183 return 1;
184 }
185
186 zval token_stream;
187
on_event(zend_php_scanner_event event,int token,int line)188 void on_event(zend_php_scanner_event event, int token, int line)
189 {
190 zval keyword;
191 HashTable *tokens_ht;
192 zval *token_zv;
193
194 switch (event) {
195 case ON_TOKEN:
196 if (token == END) break;
197 if (token >= 256) {
198 array_init(&keyword);
199 add_next_index_long(&keyword, token);
200 add_next_index_stringl(&keyword, (char *)LANG_SCNG(yy_text), LANG_SCNG(yy_leng));
201 add_next_index_long(&keyword, line);
202 add_next_index_zval(&token_stream, &keyword);
203 } else {
204 add_next_index_stringl(&token_stream, (char *)LANG_SCNG(yy_text), LANG_SCNG(yy_leng));
205 }
206 break;
207 case ON_FEEDBACK:
208 tokens_ht = Z_ARRVAL(token_stream);
209 token_zv = zend_hash_index_find(tokens_ht, zend_hash_num_elements(tokens_ht) - 1);
210 if (token_zv && Z_TYPE_P(token_zv) == IS_ARRAY) {
211 ZVAL_LONG(zend_hash_index_find(Z_ARRVAL_P(token_zv), 0), token);
212 }
213 break;
214 case ON_STOP:
215 if (LANG_SCNG(yy_cursor) != LANG_SCNG(yy_limit)) {
216 array_init(&keyword);
217 add_next_index_long(&keyword, T_INLINE_HTML);
218 add_next_index_stringl(&keyword,
219 (char *)LANG_SCNG(yy_cursor), LANG_SCNG(yy_limit) - LANG_SCNG(yy_cursor));
220 add_next_index_long(&keyword, CG(zend_lineno));
221 add_next_index_zval(&token_stream, &keyword);
222 }
223 break;
224 }
225 }
226
tokenize_parse(zval * return_value,zend_string * source)227 static zend_bool tokenize_parse(zval *return_value, zend_string *source)
228 {
229 zval source_zval;
230 zend_lex_state original_lex_state;
231 zend_bool original_in_compilation;
232 zend_bool success;
233
234 ZVAL_STR_COPY(&source_zval, source);
235
236 original_in_compilation = CG(in_compilation);
237 CG(in_compilation) = 1;
238 zend_save_lexical_state(&original_lex_state);
239
240 if ((success = (zend_prepare_string_for_scanning(&source_zval, "") == SUCCESS))) {
241 CG(ast) = NULL;
242 CG(ast_arena) = zend_arena_create(1024 * 32);
243 LANG_SCNG(yy_state) = yycINITIAL;
244 LANG_SCNG(on_event) = on_event;
245
246 array_init(&token_stream);
247 if((success = (zendparse() == SUCCESS))) {
248 ZVAL_COPY_VALUE(return_value, &token_stream);
249 } else {
250 zval_ptr_dtor(&token_stream);
251 }
252
253 zend_ast_destroy(CG(ast));
254 zend_arena_destroy(CG(ast_arena));
255 }
256
257 /* restore compiler and scanner global states */
258 zend_restore_lexical_state(&original_lex_state);
259 CG(in_compilation) = original_in_compilation;
260
261 zval_dtor(&source_zval);
262
263 return success;
264 }
265
266 /* }}} */
267
268 /* {{{ proto array token_get_all(string source [, int flags])
269 */
PHP_FUNCTION(token_get_all)270 PHP_FUNCTION(token_get_all)
271 {
272 zend_string *source;
273 zend_long flags = 0;
274 zend_bool success;
275
276 if (zend_parse_parameters(ZEND_NUM_ARGS(), "S|l", &source, &flags) == FAILURE) {
277 return;
278 }
279
280 if (flags & TOKEN_PARSE) {
281 success = tokenize_parse(return_value, source);
282 } else {
283 success = tokenize(return_value, source);
284 /* Normal token_get_all() should not throw. */
285 zend_clear_exception();
286 }
287
288 if (!success) RETURN_FALSE;
289 }
290 /* }}} */
291
292 /* {{{ proto string token_name(int type)
293 */
PHP_FUNCTION(token_name)294 PHP_FUNCTION(token_name)
295 {
296 zend_long type;
297
298 if (zend_parse_parameters(ZEND_NUM_ARGS(), "l", &type) == FAILURE) {
299 return;
300 }
301
302 RETVAL_STRING(get_token_type_name(type));
303 }
304 /* }}} */
305
306 /*
307 * Local variables:
308 * tab-width: 4
309 * c-basic-offset: 4
310 * End:
311 * vim600: noet sw=4 ts=4 fdm=marker
312 * vim<600: noet sw=4 ts=4
313 */
314
