]>
Commit | Line | Data |
---|---|---|
43ae1d95 | 1 | |
2 | /* | |
137a13ea | 3 | * $Id: ESICustomParser.cc,v 1.7 2005/07/03 15:25:08 serassio Exp $ |
43ae1d95 | 4 | * |
5 | * DEBUG: section 86 ESI processing | |
6 | * AUTHOR: Robert Collins | |
7 | * | |
8 | * SQUID Web Proxy Cache http://www.squid-cache.org/ | |
9 | * ---------------------------------------------------------- | |
10 | * | |
11 | * Squid is the result of efforts by numerous individuals from | |
12 | * the Internet community; see the CONTRIBUTORS file for full | |
13 | * details. Many organizations have provided support for Squid's | |
14 | * development; see the SPONSORS file for full details. Squid is | |
15 | * Copyrighted (C) 2001 by the Regents of the University of | |
16 | * California; see the COPYRIGHT file for full details. Squid | |
17 | * incorporates software developed and/or copyrighted by other | |
18 | * sources; see the CREDITS file for full details. | |
19 | * | |
20 | * This program is free software; you can redistribute it and/or modify | |
21 | * it under the terms of the GNU General Public License as published by | |
22 | * the Free Software Foundation; either version 2 of the License, or | |
23 | * (at your option) any later version. | |
24 | * | |
25 | * This program is distributed in the hope that it will be useful, | |
26 | ; but WITHOUT ANY WARRANTY; without even the implied warranty of | |
27 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
28 | * GNU General Public License for more details. | |
29 | * | |
30 | * You should have received a copy of the GNU General Public License | |
31 | * along with this program; if not, write to the Free Software | |
32 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA. | |
33 | * | |
34 | */ | |
35 | ||
36 | #include "squid.h" | |
37 | #include "ESICustomParser.h" | |
38 | #include "Trie.h" | |
924f73bc | 39 | #include "TrieCharTransform.h" |
43ae1d95 | 40 | #include "Array.h" |
41 | ||
42 | Trie *ESICustomParser::SearchTrie=NULL; | |
43ae1d95 | 43 | |
c2d4889f | 44 | RegisterESIParser("custom", ESICustomParser); |
45 | ||
43ae1d95 | 46 | Trie * |
47 | ESICustomParser::GetTrie() | |
48 | { | |
49 | if (SearchTrie) | |
50 | return SearchTrie; | |
51 | ||
924f73bc | 52 | SearchTrie = new Trie(new TrieCaseless); |
43ae1d95 | 53 | |
137a13ea | 54 | static const ESITAG_t ESITAG_value = ESITAG; |
55 | ||
43ae1d95 | 56 | assert (SearchTrie->add |
137a13ea | 57 | ("<esi:",5,(void *)&ESITAG_value)); |
58 | ||
59 | static const ESITAG_t ESIENDTAG_value = ESIENDTAG; | |
43ae1d95 | 60 | |
61 | assert (SearchTrie->add | |
137a13ea | 62 | ("</esi:",6,(void *)&ESIENDTAG_value)); |
63 | ||
64 | static const ESITAG_t ESICOMMENT_value = ESICOMMENT; | |
43ae1d95 | 65 | |
66 | assert (SearchTrie->add | |
137a13ea | 67 | ("<!--",4,(void *)&ESICOMMENT_value)); |
43ae1d95 | 68 | |
69 | return SearchTrie; | |
70 | } | |
71 | ||
43ae1d95 | 72 | ESICustomParser::ESICustomParser(ESIParserClient *aClient) : theClient (aClient) |
73 | {} | |
74 | ||
75 | ESICustomParser::~ESICustomParser() | |
76 | { | |
77 | theClient = NULL; | |
78 | } | |
79 | ||
80 | char const * | |
924f73bc | 81 | ESICustomParser::findTag(char const *buffer, size_t bufferLength) |
43ae1d95 | 82 | { |
83 | size_t myOffset (0); | |
137a13ea | 84 | ESITAG_t *resulttype = NULL; |
43ae1d95 | 85 | |
924f73bc | 86 | while (myOffset < bufferLength && |
137a13ea | 87 | (resulttype = static_cast<ESITAG_t *>(GetTrie()->findPrefix (buffer + myOffset, bufferLength - myOffset))) |
88 | == NULL) | |
43ae1d95 | 89 | ++myOffset; |
90 | ||
924f73bc | 91 | if (myOffset == bufferLength) |
43ae1d95 | 92 | return NULL; |
93 | ||
137a13ea | 94 | debug (86,9)("ESICustomParser::findTag: found %d\n", *resulttype); |
43ae1d95 | 95 | |
137a13ea | 96 | lastTag = *resulttype; |
43ae1d95 | 97 | |
924f73bc | 98 | return buffer + myOffset; |
43ae1d95 | 99 | } |
100 | ||
101 | bool | |
102 | ESICustomParser::parse(char const *dataToParse, size_t const lengthOfData, bool const endOfStream) | |
103 | { | |
104 | debug (86,9)("ESICustomParser::parse: Appending data to internal buffer\n"); | |
105 | content.append (dataToParse, lengthOfData); | |
106 | ||
107 | if (!endOfStream) { | |
108 | return true; | |
109 | } | |
110 | ||
111 | size_t openESITags (0); | |
112 | char const *currentPos = content.buf(); | |
113 | size_t remainingCount = content.size(); | |
5aeabf95 | 114 | char const *tag = NULL; |
43ae1d95 | 115 | |
116 | while ((tag = findTag(currentPos, remainingCount))) { | |
117 | if (tag - currentPos) | |
118 | theClient->parserDefault (currentPos,tag - currentPos); | |
119 | ||
120 | switch (lastTag) { | |
121 | ||
122 | case ESITAG: { | |
123 | ++openESITags; | |
411c6ea3 | 124 | char *tagEnd = strchr(const_cast<char *>(tag), '>'); |
43ae1d95 | 125 | |
126 | if (!tagEnd) { | |
127 | error = "Could not find end ('>') of tag"; | |
128 | return false; | |
129 | } | |
130 | ||
131 | if (tagEnd - tag > (ssize_t)remainingCount) { | |
132 | error = "Tag ends beyond the parse buffer."; | |
133 | return false; | |
134 | } | |
135 | ||
136 | if (*(tagEnd - 1) == '/') | |
137 | --openESITags; | |
138 | ||
411c6ea3 | 139 | char * endofName = strpbrk(const_cast<char *>(tag), w_space); |
43ae1d95 | 140 | |
141 | if (endofName > tagEnd) | |
142 | endofName = const_cast<char *>(tagEnd); | |
143 | ||
144 | *endofName = '\0'; | |
145 | ||
146 | *tagEnd = '\0'; | |
147 | ||
148 | Vector<char *>attributes; | |
149 | ||
411c6ea3 | 150 | char *attribute = const_cast<char *>(endofName + 1); |
43ae1d95 | 151 | |
152 | while (attribute > tag && attribute < tagEnd) { | |
153 | /* leading spaces */ | |
154 | ||
155 | while (attribute < tagEnd && (xisspace(*attribute) || (*attribute == '/'))) | |
156 | ++attribute; | |
157 | ||
158 | if (! (attribute < tagEnd)) | |
159 | break; | |
160 | ||
161 | /* attribute name */ | |
162 | attributes.push_back(attribute); | |
163 | ||
164 | char *nextSpace = strpbrk(attribute, w_space); | |
165 | ||
166 | char *equals = strchr(attribute, '='); | |
167 | ||
168 | if (!equals) { | |
169 | error = "Missing attribute value."; | |
170 | return false; | |
171 | } | |
172 | ||
173 | if (nextSpace && nextSpace < equals) | |
174 | *nextSpace = '\0'; | |
175 | else | |
176 | *equals = '\0'; | |
177 | ||
178 | ++equals; | |
179 | ||
180 | while (equals < tagEnd && xisspace(*equals)) | |
181 | ++equals; | |
182 | ||
183 | char sep = *equals; | |
184 | ||
185 | if (sep != '\'' && sep != '"') { | |
186 | error = "Unknown identifier ("; | |
187 | error.append (sep); | |
188 | error.append (")"); | |
189 | return false; | |
190 | } | |
191 | ||
192 | char *value = equals + 1; | |
193 | char *end = strchr (value, sep); | |
194 | attributes.push_back(value); | |
195 | *end = '\0'; | |
196 | attribute = end + 1; | |
197 | } | |
198 | ||
199 | theClient->start (tag + 1, (const char **)attributes.items, attributes.size() >> 1); | |
200 | /* TODO: attributes */ | |
201 | ||
202 | if (*(tagEnd - 1) == '/') | |
203 | theClient->end (tag + 1); | |
204 | ||
205 | remainingCount -= tagEnd - currentPos + 1; | |
206 | ||
207 | currentPos = tagEnd + 1; | |
208 | } | |
209 | ||
210 | break; | |
211 | ||
212 | case ESIENDTAG: { | |
213 | if (!openESITags) | |
214 | return false; | |
215 | ||
216 | char const *tagEnd = strchr(tag, '>'); | |
217 | ||
218 | if (!tagEnd) | |
219 | return false; | |
220 | ||
221 | if (tagEnd - tag > (ssize_t)remainingCount) | |
222 | return false; | |
223 | ||
411c6ea3 | 224 | char * endofName = strpbrk(const_cast<char *>(tag), w_space); |
43ae1d95 | 225 | |
226 | if (endofName > tagEnd) | |
227 | endofName = const_cast<char *>(tagEnd); | |
228 | ||
229 | *endofName = '\0'; | |
230 | ||
231 | theClient->end (tag + 2); | |
232 | ||
233 | --openESITags; | |
234 | ||
235 | remainingCount -= tagEnd - currentPos + 1; | |
236 | ||
237 | currentPos = tagEnd + 1; | |
238 | } | |
239 | ||
240 | break; | |
241 | ||
242 | case ESICOMMENT: { | |
243 | /* Further optimisation potential: | |
244 | * 1) recognize end comments for esi and don't callback on | |
245 | * comments. | |
246 | * 2) provide the comment length to the caller. | |
247 | */ | |
248 | /* Comments must not be nested, without CDATA | |
249 | * and we don't support CDATA | |
250 | */ | |
411c6ea3 | 251 | char *commentEnd = strstr (const_cast<char *>(tag), "-->"); |
43ae1d95 | 252 | |
253 | if (!commentEnd) { | |
254 | error = "missing end of comment"; | |
255 | return false; | |
256 | } | |
257 | ||
258 | if (commentEnd - tag > (ssize_t)remainingCount) { | |
259 | error = "comment ends beyond parse buffer"; | |
260 | return false; | |
261 | } | |
262 | ||
263 | *commentEnd = '\0'; | |
264 | theClient->parserComment (tag + 4); | |
265 | remainingCount -= commentEnd - currentPos + 3; | |
266 | currentPos = commentEnd + 3; | |
267 | } | |
268 | ||
269 | break; | |
270 | break; | |
271 | ||
272 | default: | |
273 | fatal ("unknown ESI tag type found"); | |
274 | }; | |
275 | ||
276 | /* | |
277 | * Find next esi tag (open or closing) or comment | |
278 | * send tag, or full comment text | |
279 | * rinse | |
280 | */ | |
281 | } | |
282 | ||
283 | if (remainingCount) | |
284 | theClient->parserDefault (currentPos,remainingCount); | |
285 | ||
286 | debug (86,5)("ESICustomParser::parse: Finished parsing, will return %d\n", !openESITags); | |
287 | ||
288 | if (openESITags) | |
289 | error = "ESI Tags still open"; | |
290 | ||
291 | return !openESITags; | |
292 | } | |
293 | ||
137a13ea | 294 | long int |
43ae1d95 | 295 | ESICustomParser::lineNumber() const |
296 | { | |
297 | /* We don't track lines in the body */ | |
298 | return 0; | |
299 | } | |
300 | ||
301 | char const * | |
302 | ESICustomParser::errorString() const | |
303 | { | |
304 | if (error.size()) | |
305 | return error.buf(); | |
306 | else | |
307 | return "Parsing error strings not implemented"; | |
308 | } |