1 /**
2 * Phoebe DOM Implementation.
3 *
4 * This is a C++ approximation of the W3C DOM model, which follows
5 * fairly closely the specifications in the various .idl files, copies of
6 * which are provided for reference. Most important is this one:
7 *
8 * http://www.w3.org/TR/2004/REC-DOM-Level-3-Core-20040407/idl-definitions.html
9 *
10 * Authors:
11 * Bob Jamison
12 *
13 * Copyright (C) 2005 Bob Jamison
14 *
15 * This library is free software; you can redistribute it and/or
16 * modify it under the terms of the GNU Lesser General Public
17 * License as published by the Free Software Foundation; either
18 * version 2.1 of the License, or (at your option) any later version.
19 *
20 * This library is distributed in the hope that it will be useful,
21 * but WITHOUT ANY WARRANTY; without even the implied warranty of
22 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
23 * Lesser General Public License for more details.
24 *
25 * You should have received a copy of the GNU Lesser General Public
26 * License along with this library; if not, write to the Free Software
27 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
28 */
33 #include "uri.h"
34 #include "charclass.h"
36 #include <stdio.h>
37 #include <stdarg.h>
41 namespace org
42 {
43 namespace w3c
44 {
45 namespace dom
46 {
49 typedef struct
50 {
51 int ival;
52 char *sval;
53 int port;
54 } LookupEntry;
56 LookupEntry schemes[] =
57 {
58 { URI::SCHEME_DATA, "data:", 0 },
59 { URI::SCHEME_HTTP, "http:", 80 },
60 { URI::SCHEME_HTTPS, "https:", 443 },
61 { URI::SCHEME_FTP, "ftp", 12 },
62 { URI::SCHEME_FILE, "file:", 0 },
63 { URI::SCHEME_LDAP, "ldap:", 123 },
64 { URI::SCHEME_MAILTO, "mailto:", 25 },
65 { URI::SCHEME_NEWS, "news:", 117 },
66 { URI::SCHEME_TELNET, "telnet:", 23 },
67 { 0, NULL, 0 }
68 };
72 //#########################################################################
73 //# C O N S T R U C T O R
74 //#########################################################################
76 /**
77 *
78 */
79 URI::URI()
80 {
81 init();
82 }
84 /**
85 *
86 */
87 URI::URI(const DOMString &str)
88 {
89 init();
90 parse(str);
91 }
94 /**
95 *
96 */
97 URI::URI(const char *str)
98 {
99 init();
100 DOMString domStr = str;
101 parse(domStr);
102 }
105 /**
106 *
107 */
108 URI::URI(const URI &other)
109 {
110 init();
111 assign(other);
112 }
115 /**
116 *
117 */
118 URI &URI::operator=(const URI &other)
119 {
120 init();
121 assign(other);
122 return *this;
123 }
126 /**
127 *
128 */
129 URI::~URI()
130 {
131 }
137 /**
138 *
139 */
140 void URI::init()
141 {
142 parsebuf = NULL;
143 parselen = 0;
144 scheme = SCHEME_NONE;
145 schemeStr = "";
146 port = 0;
147 authority = "";
148 path = "";
149 absolute = false;
150 opaque = false;
151 query = "";
152 fragment = "";
153 }
156 /**
157 *
158 */
159 void URI::assign(const URI &other)
160 {
161 scheme = other.scheme;
162 schemeStr = other.schemeStr;
163 authority = other.authority;
164 port = other.port;
165 path = other.path;
166 absolute = other.absolute;
167 opaque = other.opaque;
168 query = other.query;
169 fragment = other.fragment;
170 }
173 //#########################################################################
174 //#A T T R I B U T E S
175 //#########################################################################
177 DOMString URI::toString() const
178 {
179 DOMString str = schemeStr;
180 if (authority.size() > 0)
181 {
182 str.append("//");
183 str.append(authority);
184 }
185 str.append(path);
186 if (query.size() > 0)
187 {
188 str.append("?");
189 str.append(query);
190 }
191 if (fragment.size() > 0)
192 {
193 str.append("#");
194 str.append(fragment);
195 }
196 return str;
197 }
200 int URI::getScheme() const
201 {
202 return scheme;
203 }
205 DOMString URI::getSchemeStr() const
206 {
207 return schemeStr;
208 }
211 DOMString URI::getAuthority() const
212 {
213 DOMString ret = authority;
214 if (portSpecified && port>=0)
215 {
216 char buf[7];
217 snprintf(buf, 6, ":%6d", port);
218 ret.append(buf);
219 }
220 return ret;
221 }
223 DOMString URI::getHost() const
224 {
225 return authority;
226 }
228 int URI::getPort() const
229 {
230 return port;
231 }
234 DOMString URI::getPath() const
235 {
236 return path;
237 }
240 bool URI::isAbsolute() const
241 {
242 return absolute;
243 }
245 bool URI::isOpaque() const
246 {
247 return opaque;
248 }
251 DOMString URI::getQuery() const
252 {
253 return query;
254 }
257 DOMString URI::getFragment() const
258 {
259 return fragment;
260 }
263 URI URI::resolve(const URI &other) const
264 {
265 //### According to w3c, this is handled in 3 cases
267 //## 1
268 if (opaque || other.isAbsolute())
269 return other;
271 //## 2
272 if (other.fragment.size() > 0 &&
273 other.path.size() == 0 &&
274 other.scheme == SCHEME_NONE &&
275 other.authority.size() == 0 &&
276 other.query.size() == 0 )
277 {
278 URI fragUri = *this;
279 fragUri.fragment = other.fragment;
280 return fragUri;
281 }
283 //## 3 http://www.ietf.org/rfc/rfc2396.txt, section 5.2
284 URI newUri;
285 //# 3.1
286 newUri.scheme = scheme;
287 newUri.schemeStr = schemeStr;
288 newUri.query = other.query;
289 newUri.fragment = other.fragment;
290 if (other.authority.size() > 0)
291 {
292 //# 3.2
293 if (absolute || other.absolute)
294 newUri.absolute = true;
295 newUri.authority = other.authority;
296 newUri.port = other.port;//part of authority
297 newUri.path = other.path;
298 }
299 else
300 {
301 //# 3.3
302 if (other.absolute)
303 {
304 newUri.absolute = true;
305 newUri.path = other.path;
306 }
307 else
308 {
309 unsigned int pos = path.rfind('/');
310 if (pos != path.npos)
311 {
312 DOMString tpath = path.substr(0, pos+1);
313 tpath.append(other.path);
314 newUri.path = tpath;
315 }
316 }
317 }
318 newUri.normalize();
319 return newUri;
320 }
323 /**
324 * This follows the Java URI algorithm:
325 * 1. All "." segments are removed.
326 * 2. If a ".." segment is preceded by a non-".." segment
327 * then both of these segments are removed. This step
328 * is repeated until it is no longer applicable.
329 * 3. If the path is relative, and if its first segment
330 * contains a colon character (':'), then a "." segment
331 * is prepended. This prevents a relative URI with a path
332 * such as "a:b/c/d" from later being re-parsed as an
333 * opaque URI with a scheme of "a" and a scheme-specific
334 * part of "b/c/d". (Deviation from RFC 2396)
335 */
336 void URI::normalize()
337 {
338 std::vector<DOMString> segments;
340 //## Collect segments
341 if (path.size()<2)
342 return;
343 bool abs = false;
344 unsigned int pos=0;
345 if (path[0]=='/')
346 {
347 abs = true;
348 pos++;
349 }
350 while (pos < path.size())
351 {
352 unsigned int pos2 = path.find('/', pos);
353 if (pos2==path.npos)
354 {
355 DOMString seg = path.substr(pos);
356 //printf("last segment:%s\n", seg.c_str());
357 segments.push_back(seg);
358 break;
359 }
360 if (pos2>pos)
361 {
362 DOMString seg = path.substr(pos, pos2-pos);
363 //printf("segment:%s\n", seg.c_str());
364 segments.push_back(seg);
365 }
366 pos = pos2;
367 pos++;
368 }
370 //## Clean up (normalize) segments
371 bool edited = false;
372 std::vector<DOMString>::iterator iter;
373 for (iter=segments.begin() ; iter!=segments.end() ; )
374 {
375 DOMString s = *iter;
376 if (s == ".")
377 {
378 iter = segments.erase(iter);
379 edited = true;
380 }
381 else if (s == ".." &&
382 iter != segments.begin() &&
383 *(iter-1) != "..")
384 {
385 iter--; //back up, then erase two entries
386 iter = segments.erase(iter);
387 iter = segments.erase(iter);
388 edited = true;
389 }
390 else
391 iter++;
392 }
394 //## Rebuild path, if necessary
395 if (edited)
396 {
397 path.clear();
398 if (abs)
399 {
400 path.append("/");
401 }
402 std::vector<DOMString>::iterator iter;
403 for (iter=segments.begin() ; iter!=segments.end() ; iter++)
404 {
405 if (iter != segments.begin())
406 path.append("/");
407 path.append(*iter);
408 }
409 }
411 }
415 //#########################################################################
416 //# M E S S A G E S
417 //#########################################################################
419 void URI::error(const char *fmt, ...)
420 {
421 va_list args;
422 fprintf(stderr, "URI error: ");
423 va_start(args, fmt);
424 vfprintf(stderr, fmt, args);
425 va_end(args);
426 fprintf(stderr, "\n");
427 }
429 void URI::trace(const char *fmt, ...)
430 {
431 va_list args;
432 fprintf(stdout, "URI: ");
433 va_start(args, fmt);
434 vfprintf(stdout, fmt, args);
435 va_end(args);
436 fprintf(stdout, "\n");
437 }
441 //#########################################################################
442 //# P A R S I N G
443 //#########################################################################
447 int URI::peek(int p)
448 {
449 if (p<0 || p>=parselen)
450 return -1;
451 return parsebuf[p];
452 }
456 int URI::match(int p0, char *key)
457 {
458 int p = p0;
459 while (p < parselen)
460 {
461 if (*key == '\0')
462 return p;
463 else if (*key != parsebuf[p])
464 break;
465 p++; key++;
466 }
467 return p0;
468 }
470 //#########################################################################
471 //# Parsing is performed according to:
472 //# http://www.gbiv.com/protocols/uri/rfc/rfc3986.html#components
473 //#########################################################################
475 int URI::parseScheme(int p0)
476 {
477 int p = p0;
478 for (LookupEntry *entry = schemes; entry->sval ; entry++)
479 {
480 int p2 = match(p, entry->sval);
481 if (p2 > p)
482 {
483 schemeStr = entry->sval;
484 scheme = entry->ival;
485 port = entry->port;
486 p = p2;
487 return p;
488 }
489 }
491 return p;
492 }
495 int URI::parseHierarchicalPart(int p0)
496 {
497 int p = p0;
498 int ch;
500 //# Authority field (host and port, for example)
501 int p2 = match(p, "//");
502 if (p2 > p)
503 {
504 p = p2;
505 portSpecified = false;
506 DOMString portStr;
507 while (p < parselen)
508 {
509 ch = peek(p);
510 if (ch == '/')
511 break;
512 else if (ch == ':')
513 portSpecified = true;
514 else if (portSpecified)
515 portStr.push_back((XMLCh)ch);
516 else
517 authority.push_back((XMLCh)ch);
518 p++;
519 }
520 if (portStr.size() > 0)
521 {
522 char *pstr = (char *)portStr.c_str();
523 char *endStr;
524 long val = strtol(pstr, &endStr, 10);
525 if (endStr > pstr) //successful parse?
526 port = val;
527 }
528 }
530 //# Are we absolute?
531 ch = peek(p);
532 if (ch == '/')
533 {
534 absolute = true;
535 if (p>p0) //in other words, if '/' is not the first char
536 opaque = true;
537 path.push_back((XMLCh)ch);
538 p++;
539 }
541 while (p < parselen)
542 {
543 ch = peek(p);
544 if (ch == '?' || ch == '#')
545 break;
546 path.push_back((XMLCh)ch);
547 p++;
548 }
550 return p;
551 }
553 int URI::parseQuery(int p0)
554 {
555 int p = p0;
556 int ch = peek(p);
557 if (ch != '?')
558 return p0;
560 p++;
561 while (p < parselen)
562 {
563 ch = peek(p);
564 if (ch == '#')
565 break;
566 query.push_back((XMLCh)ch);
567 p++;
568 }
571 return p;
572 }
574 int URI::parseFragment(int p0)
575 {
577 int p = p0;
578 int ch = peek(p);
579 if (ch != '#')
580 return p0;
582 p++;
583 while (p < parselen)
584 {
585 ch = peek(p);
586 if (ch == '?')
587 break;
588 fragment.push_back((XMLCh)ch);
589 p++;
590 }
593 return p;
594 }
597 int URI::parse(int p0)
598 {
600 int p = p0;
602 int p2 = parseScheme(p);
603 if (p2 < 0)
604 {
605 error("Scheme");
606 return -1;
607 }
608 p = p2;
611 p2 = parseHierarchicalPart(p);
612 if (p2 < 0)
613 {
614 error("Hierarchical part");
615 return -1;
616 }
617 p = p2;
619 p2 = parseQuery(p);
620 if (p2 < 0)
621 {
622 error("Query");
623 return -1;
624 }
625 p = p2;
628 p2 = parseFragment(p);
629 if (p2 < 0)
630 {
631 error("Fragment");
632 return -1;
633 }
634 p = p2;
636 return p;
638 }
642 bool URI::parse(const DOMString &str)
643 {
645 parselen = str.size();
646 DOMString tmp = str;
647 parsebuf = (char *) tmp.c_str();
650 int p = parse(0);
651 normalize();
653 if (p < 0)
654 {
655 error("Syntax error");
656 return false;
657 }
659 //printf("uri:%s\n", toString().c_str());
660 //printf("path:%s\n", path.c_str());
662 return true;
664 }
670 } //namespace dom
671 } //namespace w3c
672 } //namespace org
673 //#########################################################################
674 //# E N D O F F I L E
675 //#########################################################################