aboutsummaryrefslogtreecommitdiffhomepage
path: root/inc/simple_html_dom.php
diff options
context:
space:
mode:
Diffstat (limited to 'inc/simple_html_dom.php')
-rw-r--r--inc/simple_html_dom.php1722
1 files changed, 1722 insertions, 0 deletions
diff --git a/inc/simple_html_dom.php b/inc/simple_html_dom.php
new file mode 100644
index 00000000..43b94e57
--- /dev/null
+++ b/inc/simple_html_dom.php
@@ -0,0 +1,1722 @@
1<?php
2/**
3 * Website: http://sourceforge.net/projects/simplehtmldom/
4 * Additional projects that may be used: http://sourceforge.net/projects/debugobject/
5 * Acknowledge: Jose Solorzano (https://sourceforge.net/projects/php-html/)
6 * Contributions by:
7 * Yousuke Kumakura (Attribute filters)
8 * Vadim Voituk (Negative indexes supports of "find" method)
9 * Antcs (Constructor with automatically load contents either text or file/url)
10 *
11 * all affected sections have comments starting with "PaperG"
12 *
13 * Paperg - Added case insensitive testing of the value of the selector.
14 * Paperg - Added tag_start for the starting index of tags - NOTE: This works but not accurately.
15 * This tag_start gets counted AFTER \r\n have been crushed out, and after the remove_noice calls so it will not reflect the REAL position of the tag in the source,
16 * it will almost always be smaller by some amount.
17 * We use this to determine how far into the file the tag in question is. This "percentage will never be accurate as the $dom->size is the "real" number of bytes the dom was created from.
18 * but for most purposes, it's a really good estimation.
19 * Paperg - Added the forceTagsClosed to the dom constructor. Forcing tags closed is great for malformed html, but it CAN lead to parsing errors.
20 * Allow the user to tell us how much they trust the html.
21 * Paperg add the text and plaintext to the selectors for the find syntax. plaintext implies text in the innertext of a node. text implies that the tag is a text node.
22 * This allows for us to find tags based on the text they contain.
23 * Create find_ancestor_tag to see if a tag is - at any level - inside of another specific tag.
24 * Paperg: added parse_charset so that we know about the character set of the source document.
25 * NOTE: If the user's system has a routine called get_last_retrieve_url_contents_content_type availalbe, we will assume it's returning the content-type header from the
26 * last transfer or curl_exec, and we will parse that and use it in preference to any other method of charset detection.
27 *
28 * Found infinite loop in the case of broken html in restore_noise. Rewrote to protect from that.
29 * PaperG (John Schlick) Added get_display_size for "IMG" tags.
30 *
31 * Licensed under The MIT License
32 * Redistributions of files must retain the above copyright notice.
33 *
34 * @author S.C. Chen <me578022@gmail.com>
35 * @author John Schlick
36 * @author Rus Carroll
37 * @version 1.5 ($Rev: 202 $)
38 * @package PlaceLocalInclude
39 * @subpackage simple_html_dom
40 */
41
42/**
43 * All of the Defines for the classes below.
44 * @author S.C. Chen <me578022@gmail.com>
45 */
46define('HDOM_TYPE_ELEMENT', 1);
47define('HDOM_TYPE_COMMENT', 2);
48define('HDOM_TYPE_TEXT', 3);
49define('HDOM_TYPE_ENDTAG', 4);
50define('HDOM_TYPE_ROOT', 5);
51define('HDOM_TYPE_UNKNOWN', 6);
52define('HDOM_QUOTE_DOUBLE', 0);
53define('HDOM_QUOTE_SINGLE', 1);
54define('HDOM_QUOTE_NO', 3);
55define('HDOM_INFO_BEGIN', 0);
56define('HDOM_INFO_END', 1);
57define('HDOM_INFO_QUOTE', 2);
58define('HDOM_INFO_SPACE', 3);
59define('HDOM_INFO_TEXT', 4);
60define('HDOM_INFO_INNER', 5);
61define('HDOM_INFO_OUTER', 6);
62define('HDOM_INFO_ENDSPACE',7);
63define('DEFAULT_TARGET_CHARSET', 'UTF-8');
64define('DEFAULT_BR_TEXT', "\r\n");
65define('DEFAULT_SPAN_TEXT', " ");
66define('MAX_FILE_SIZE', 600000);
67// helper functions
68// -----------------------------------------------------------------------------
69// get html dom from file
70// $maxlen is defined in the code as PHP_STREAM_COPY_ALL which is defined as -1.
71function file_get_html($url, $use_include_path = false, $context=null, $offset = -1, $maxLen=-1, $lowercase = true, $forceTagsClosed=true, $target_charset = DEFAULT_TARGET_CHARSET, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT)
72{
73 // We DO force the tags to be terminated.
74 $dom = new simple_html_dom(null, $lowercase, $forceTagsClosed, $target_charset, $stripRN, $defaultBRText, $defaultSpanText);
75 // For sourceforge users: uncomment the next line and comment the retreive_url_contents line 2 lines down if it is not already done.
76 $contents = file_get_contents($url, $use_include_path, $context, $offset);
77 // Paperg - use our own mechanism for getting the contents as we want to control the timeout.
78 //$contents = retrieve_url_contents($url);
79 if (empty($contents) || strlen($contents) > MAX_FILE_SIZE)
80 {
81 return false;
82 }
83 // The second parameter can force the selectors to all be lowercase.
84 $dom->load($contents, $lowercase, $stripRN);
85 return $dom;
86}
87
88// get html dom from string
89function str_get_html($str, $lowercase=true, $forceTagsClosed=true, $target_charset = DEFAULT_TARGET_CHARSET, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT)
90{
91 $dom = new simple_html_dom(null, $lowercase, $forceTagsClosed, $target_charset, $stripRN, $defaultBRText, $defaultSpanText);
92 if (empty($str) || strlen($str) > MAX_FILE_SIZE)
93 {
94 $dom->clear();
95 return false;
96 }
97 $dom->load($str, $lowercase, $stripRN);
98 return $dom;
99}
100
101// dump html dom tree
102function dump_html_tree($node, $show_attr=true, $deep=0)
103{
104 $node->dump($node);
105}
106
107
108/**
109 * simple html dom node
110 * PaperG - added ability for "find" routine to lowercase the value of the selector.
111 * PaperG - added $tag_start to track the start position of the tag in the total byte index
112 *
113 * @package PlaceLocalInclude
114 */
115class simple_html_dom_node
116{
117 public $nodetype = HDOM_TYPE_TEXT;
118 public $tag = 'text';
119 public $attr = array();
120 public $children = array();
121 public $nodes = array();
122 public $parent = null;
123 // The "info" array - see HDOM_INFO_... for what each element contains.
124 public $_ = array();
125 public $tag_start = 0;
126 private $dom = null;
127
128 function __construct($dom)
129 {
130 $this->dom = $dom;
131 $dom->nodes[] = $this;
132 }
133
134 function __destruct()
135 {
136 $this->clear();
137 }
138
139 function __toString()
140 {
141 return $this->outertext();
142 }
143
144 // clean up memory due to php5 circular references memory leak...
145 function clear()
146 {
147 $this->dom = null;
148 $this->nodes = null;
149 $this->parent = null;
150 $this->children = null;
151 }
152
153 // dump node's tree
154 function dump($show_attr=true, $deep=0)
155 {
156 $lead = str_repeat(' ', $deep);
157
158 echo $lead.$this->tag;
159 if ($show_attr && count($this->attr)>0)
160 {
161 echo '(';
162 foreach ($this->attr as $k=>$v)
163 echo "[$k]=>\"".$this->$k.'", ';
164 echo ')';
165 }
166 echo "\n";
167
168 if ($this->nodes)
169 {
170 foreach ($this->nodes as $c)
171 {
172 $c->dump($show_attr, $deep+1);
173 }
174 }
175 }
176
177
178 // Debugging function to dump a single dom node with a bunch of information about it.
179 function dump_node($echo=true)
180 {
181
182 $string = $this->tag;
183 if (count($this->attr)>0)
184 {
185 $string .= '(';
186 foreach ($this->attr as $k=>$v)
187 {
188 $string .= "[$k]=>\"".$this->$k.'", ';
189 }
190 $string .= ')';
191 }
192 if (count($this->_)>0)
193 {
194 $string .= ' $_ (';
195 foreach ($this->_ as $k=>$v)
196 {
197 if (is_array($v))
198 {
199 $string .= "[$k]=>(";
200 foreach ($v as $k2=>$v2)
201 {
202 $string .= "[$k2]=>\"".$v2.'", ';
203 }
204 $string .= ")";
205 } else {
206 $string .= "[$k]=>\"".$v.'", ';
207 }
208 }
209 $string .= ")";
210 }
211
212 if (isset($this->text))
213 {
214 $string .= " text: (" . $this->text . ")";
215 }
216
217 $string .= " HDOM_INNER_INFO: '";
218 if (isset($node->_[HDOM_INFO_INNER]))
219 {
220 $string .= $node->_[HDOM_INFO_INNER] . "'";
221 }
222 else
223 {
224 $string .= ' NULL ';
225 }
226
227 $string .= " children: " . count($this->children);
228 $string .= " nodes: " . count($this->nodes);
229 $string .= " tag_start: " . $this->tag_start;
230 $string .= "\n";
231
232 if ($echo)
233 {
234 echo $string;
235 return;
236 }
237 else
238 {
239 return $string;
240 }
241 }
242
243 // returns the parent of node
244 // If a node is passed in, it will reset the parent of the current node to that one.
245 function parent($parent=null)
246 {
247 // I am SURE that this doesn't work properly.
248 // It fails to unset the current node from it's current parents nodes or children list first.
249 if ($parent !== null)
250 {
251 $this->parent = $parent;
252 $this->parent->nodes[] = $this;
253 $this->parent->children[] = $this;
254 }
255
256 return $this->parent;
257 }
258
259 // verify that node has children
260 function has_child()
261 {
262 return !empty($this->children);
263 }
264
265 // returns children of node
266 function children($idx=-1)
267 {
268 if ($idx===-1)
269 {
270 return $this->children;
271 }
272 if (isset($this->children[$idx])) return $this->children[$idx];
273 return null;
274 }
275
276 // returns the first child of node
277 function first_child()
278 {
279 if (count($this->children)>0)
280 {
281 return $this->children[0];
282 }
283 return null;
284 }
285
286 // returns the last child of node
287 function last_child()
288 {
289 if (($count=count($this->children))>0)
290 {
291 return $this->children[$count-1];
292 }
293 return null;
294 }
295
296 // returns the next sibling of node
297 function next_sibling()
298 {
299 if ($this->parent===null)
300 {
301 return null;
302 }
303
304 $idx = 0;
305 $count = count($this->parent->children);
306 while ($idx<$count && $this!==$this->parent->children[$idx])
307 {
308 ++$idx;
309 }
310 if (++$idx>=$count)
311 {
312 return null;
313 }
314 return $this->parent->children[$idx];
315 }
316
317 // returns the previous sibling of node
318 function prev_sibling()
319 {
320 if ($this->parent===null) return null;
321 $idx = 0;
322 $count = count($this->parent->children);
323 while ($idx<$count && $this!==$this->parent->children[$idx])
324 ++$idx;
325 if (--$idx<0) return null;
326 return $this->parent->children[$idx];
327 }
328
329 // function to locate a specific ancestor tag in the path to the root.
330 function find_ancestor_tag($tag)
331 {
332 global $debug_object;
333 if (is_object($debug_object)) { $debug_object->debugLogEntry(1); }
334
335 // Start by including ourselves in the comparison.
336 $returnDom = $this;
337
338 while (!is_null($returnDom))
339 {
340 if (is_object($debug_object)) { $debug_object->debugLog(2, "Current tag is: " . $returnDom->tag); }
341
342 if ($returnDom->tag == $tag)
343 {
344 break;
345 }
346 $returnDom = $returnDom->parent;
347 }
348 return $returnDom;
349 }
350
351 // get dom node's inner html
352 function innertext()
353 {
354 if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER];
355 if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
356
357 $ret = '';
358 foreach ($this->nodes as $n)
359 $ret .= $n->outertext();
360 return $ret;
361 }
362
363 // get dom node's outer text (with tag)
364 function outertext()
365 {
366 global $debug_object;
367 if (is_object($debug_object))
368 {
369 $text = '';
370 if ($this->tag == 'text')
371 {
372 if (!empty($this->text))
373 {
374 $text = " with text: " . $this->text;
375 }
376 }
377 $debug_object->debugLog(1, 'Innertext of tag: ' . $this->tag . $text);
378 }
379
380 if ($this->tag==='root') return $this->innertext();
381
382 // trigger callback
383 if ($this->dom && $this->dom->callback!==null)
384 {
385 call_user_func_array($this->dom->callback, array($this));
386 }
387
388 if (isset($this->_[HDOM_INFO_OUTER])) return $this->_[HDOM_INFO_OUTER];
389 if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
390
391 // render begin tag
392 if ($this->dom && $this->dom->nodes[$this->_[HDOM_INFO_BEGIN]])
393 {
394 $ret = $this->dom->nodes[$this->_[HDOM_INFO_BEGIN]]->makeup();
395 } else {
396 $ret = "";
397 }
398
399 // render inner text
400 if (isset($this->_[HDOM_INFO_INNER]))
401 {
402 // If it's a br tag... don't return the HDOM_INNER_INFO that we may or may not have added.
403 if ($this->tag != "br")
404 {
405 $ret .= $this->_[HDOM_INFO_INNER];
406 }
407 } else {
408 if ($this->nodes)
409 {
410 foreach ($this->nodes as $n)
411 {
412 $ret .= $this->convert_text($n->outertext());
413 }
414 }
415 }
416
417 // render end tag
418 if (isset($this->_[HDOM_INFO_END]) && $this->_[HDOM_INFO_END]!=0)
419 $ret .= '</'.$this->tag.'>';
420 return $ret;
421 }
422
423 // get dom node's plain text
424 function text()
425 {
426 if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER];
427 switch ($this->nodetype)
428 {
429 case HDOM_TYPE_TEXT: return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
430 case HDOM_TYPE_COMMENT: return '';
431 case HDOM_TYPE_UNKNOWN: return '';
432 }
433 if (strcasecmp($this->tag, 'script')===0) return '';
434 if (strcasecmp($this->tag, 'style')===0) return '';
435
436 $ret = '';
437 // In rare cases, (always node type 1 or HDOM_TYPE_ELEMENT - observed for some span tags, and some p tags) $this->nodes is set to NULL.
438 // NOTE: This indicates that there is a problem where it's set to NULL without a clear happening.
439 // WHY is this happening?
440 if (!is_null($this->nodes))
441 {
442 foreach ($this->nodes as $n)
443 {
444 $ret .= $this->convert_text($n->text());
445 }
446
447 // If this node is a span... add a space at the end of it so multiple spans don't run into each other. This is plaintext after all.
448 if ($this->tag == "span")
449 {
450 $ret .= $this->dom->default_span_text;
451 }
452
453
454 }
455 return $ret;
456 }
457
458 function xmltext()
459 {
460 $ret = $this->innertext();
461 $ret = str_ireplace('<![CDATA[', '', $ret);
462 $ret = str_replace(']]>', '', $ret);
463 return $ret;
464 }
465
466 // build node's text with tag
467 function makeup()
468 {
469 // text, comment, unknown
470 if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
471
472 $ret = '<'.$this->tag;
473 $i = -1;
474
475 foreach ($this->attr as $key=>$val)
476 {
477 ++$i;
478
479 // skip removed attribute
480 if ($val===null || $val===false)
481 continue;
482
483 $ret .= $this->_[HDOM_INFO_SPACE][$i][0];
484 //no value attr: nowrap, checked selected...
485 if ($val===true)
486 $ret .= $key;
487 else {
488 switch ($this->_[HDOM_INFO_QUOTE][$i])
489 {
490 case HDOM_QUOTE_DOUBLE: $quote = '"'; break;
491 case HDOM_QUOTE_SINGLE: $quote = '\''; break;
492 default: $quote = '';
493 }
494 $ret .= $key.$this->_[HDOM_INFO_SPACE][$i][1].'='.$this->_[HDOM_INFO_SPACE][$i][2].$quote.$val.$quote;
495 }
496 }
497 $ret = $this->dom->restore_noise($ret);
498 return $ret . $this->_[HDOM_INFO_ENDSPACE] . '>';
499 }
500
501 // find elements by css selector
502 //PaperG - added ability for find to lowercase the value of the selector.
503 function find($selector, $idx=null, $lowercase=false)
504 {
505 $selectors = $this->parse_selector($selector);
506 if (($count=count($selectors))===0) return array();
507 $found_keys = array();
508
509 // find each selector
510 for ($c=0; $c<$count; ++$c)
511 {
512 // The change on the below line was documented on the sourceforge code tracker id 2788009
513 // used to be: if (($levle=count($selectors[0]))===0) return array();
514 if (($levle=count($selectors[$c]))===0) return array();
515 if (!isset($this->_[HDOM_INFO_BEGIN])) return array();
516
517 $head = array($this->_[HDOM_INFO_BEGIN]=>1);
518
519 // handle descendant selectors, no recursive!
520 for ($l=0; $l<$levle; ++$l)
521 {
522 $ret = array();
523 foreach ($head as $k=>$v)
524 {
525 $n = ($k===-1) ? $this->dom->root : $this->dom->nodes[$k];
526 //PaperG - Pass this optional parameter on to the seek function.
527 $n->seek($selectors[$c][$l], $ret, $lowercase);
528 }
529 $head = $ret;
530 }
531
532 foreach ($head as $k=>$v)
533 {
534 if (!isset($found_keys[$k]))
535 $found_keys[$k] = 1;
536 }
537 }
538
539 // sort keys
540 ksort($found_keys);
541
542 $found = array();
543 foreach ($found_keys as $k=>$v)
544 $found[] = $this->dom->nodes[$k];
545
546 // return nth-element or array
547 if (is_null($idx)) return $found;
548 else if ($idx<0) $idx = count($found) + $idx;
549 return (isset($found[$idx])) ? $found[$idx] : null;
550 }
551
552 // seek for given conditions
553 // PaperG - added parameter to allow for case insensitive testing of the value of a selector.
554 protected function seek($selector, &$ret, $lowercase=false)
555 {
556 global $debug_object;
557 if (is_object($debug_object)) { $debug_object->debugLogEntry(1); }
558
559 list($tag, $key, $val, $exp, $no_key) = $selector;
560
561 // xpath index
562 if ($tag && $key && is_numeric($key))
563 {
564 $count = 0;
565 foreach ($this->children as $c)
566 {
567 if ($tag==='*' || $tag===$c->tag) {
568 if (++$count==$key) {
569 $ret[$c->_[HDOM_INFO_BEGIN]] = 1;
570 return;
571 }
572 }
573 }
574 return;
575 }
576
577 $end = (!empty($this->_[HDOM_INFO_END])) ? $this->_[HDOM_INFO_END] : 0;
578 if ($end==0) {
579 $parent = $this->parent;
580 while (!isset($parent->_[HDOM_INFO_END]) && $parent!==null) {
581 $end -= 1;
582 $parent = $parent->parent;
583 }
584 $end += $parent->_[HDOM_INFO_END];
585 }
586
587 for ($i=$this->_[HDOM_INFO_BEGIN]+1; $i<$end; ++$i) {
588 $node = $this->dom->nodes[$i];
589
590 $pass = true;
591
592 if ($tag==='*' && !$key) {
593 if (in_array($node, $this->children, true))
594 $ret[$i] = 1;
595 continue;
596 }
597
598 // compare tag
599 if ($tag && $tag!=$node->tag && $tag!=='*') {$pass=false;}
600 // compare key
601 if ($pass && $key) {
602 if ($no_key) {
603 if (isset($node->attr[$key])) $pass=false;
604 } else {
605 if (($key != "plaintext") && !isset($node->attr[$key])) $pass=false;
606 }
607 }
608 // compare value
609 if ($pass && $key && $val && $val!=='*') {
610 // If they have told us that this is a "plaintext" search then we want the plaintext of the node - right?
611 if ($key == "plaintext") {
612 // $node->plaintext actually returns $node->text();
613 $nodeKeyValue = $node->text();
614 } else {
615 // this is a normal search, we want the value of that attribute of the tag.
616 $nodeKeyValue = $node->attr[$key];
617 }
618 if (is_object($debug_object)) {$debug_object->debugLog(2, "testing node: " . $node->tag . " for attribute: " . $key . $exp . $val . " where nodes value is: " . $nodeKeyValue);}
619
620 //PaperG - If lowercase is set, do a case insensitive test of the value of the selector.
621 if ($lowercase) {
622 $check = $this->match($exp, strtolower($val), strtolower($nodeKeyValue));
623 } else {
624 $check = $this->match($exp, $val, $nodeKeyValue);
625 }
626 if (is_object($debug_object)) {$debug_object->debugLog(2, "after match: " . ($check ? "true" : "false"));}
627
628 // handle multiple class
629 if (!$check && strcasecmp($key, 'class')===0) {
630 foreach (explode(' ',$node->attr[$key]) as $k) {
631 // Without this, there were cases where leading, trailing, or double spaces lead to our comparing blanks - bad form.
632 if (!empty($k)) {
633 if ($lowercase) {
634 $check = $this->match($exp, strtolower($val), strtolower($k));
635 } else {
636 $check = $this->match($exp, $val, $k);
637 }
638 if ($check) break;
639 }
640 }
641 }
642 if (!$check) $pass = false;
643 }
644 if ($pass) $ret[$i] = 1;
645 unset($node);
646 }
647 // It's passed by reference so this is actually what this function returns.
648 if (is_object($debug_object)) {$debug_object->debugLog(1, "EXIT - ret: ", $ret);}
649 }
650
651 protected function match($exp, $pattern, $value) {
652 global $debug_object;
653 if (is_object($debug_object)) {$debug_object->debugLogEntry(1);}
654
655 switch ($exp) {
656 case '=':
657 return ($value===$pattern);
658 case '!=':
659 return ($value!==$pattern);
660 case '^=':
661 return preg_match("/^".preg_quote($pattern,'/')."/", $value);
662 case '$=':
663 return preg_match("/".preg_quote($pattern,'/')."$/", $value);
664 case '*=':
665 if ($pattern[0]=='/') {
666 return preg_match($pattern, $value);
667 }
668 return preg_match("/".$pattern."/i", $value);
669 }
670 return false;
671 }
672
673 protected function parse_selector($selector_string) {
674 global $debug_object;
675 if (is_object($debug_object)) {$debug_object->debugLogEntry(1);}
676
677 // pattern of CSS selectors, modified from mootools
678 // Paperg: Add the colon to the attrbute, so that it properly finds <tag attr:ibute="something" > like google does.
679 // Note: if you try to look at this attribute, yo MUST use getAttribute since $dom->x:y will fail the php syntax check.
680// Notice the \[ starting the attbute? and the @? following? This implies that an attribute can begin with an @ sign that is not captured.
681// This implies that an html attribute specifier may start with an @ sign that is NOT captured by the expression.
682// farther study is required to determine of this should be documented or removed.
683// $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is";
684 $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is";
685 preg_match_all($pattern, trim($selector_string).' ', $matches, PREG_SET_ORDER);
686 if (is_object($debug_object)) {$debug_object->debugLog(2, "Matches Array: ", $matches);}
687
688 $selectors = array();
689 $result = array();
690 //print_r($matches);
691
692 foreach ($matches as $m) {
693 $m[0] = trim($m[0]);
694 if ($m[0]==='' || $m[0]==='/' || $m[0]==='//') continue;
695 // for browser generated xpath
696 if ($m[1]==='tbody') continue;
697
698 list($tag, $key, $val, $exp, $no_key) = array($m[1], null, null, '=', false);
699 if (!empty($m[2])) {$key='id'; $val=$m[2];}
700 if (!empty($m[3])) {$key='class'; $val=$m[3];}
701 if (!empty($m[4])) {$key=$m[4];}
702 if (!empty($m[5])) {$exp=$m[5];}
703 if (!empty($m[6])) {$val=$m[6];}
704
705 // convert to lowercase
706 if ($this->dom->lowercase) {$tag=strtolower($tag); $key=strtolower($key);}
707 //elements that do NOT have the specified attribute
708 if (isset($key[0]) && $key[0]==='!') {$key=substr($key, 1); $no_key=true;}
709
710 $result[] = array($tag, $key, $val, $exp, $no_key);
711 if (trim($m[7])===',') {
712 $selectors[] = $result;
713 $result = array();
714 }
715 }
716 if (count($result)>0)
717 $selectors[] = $result;
718 return $selectors;
719 }
720
721 function __get($name) {
722 if (isset($this->attr[$name]))
723 {
724 return $this->convert_text($this->attr[$name]);
725 }
726 switch ($name) {
727 case 'outertext': return $this->outertext();
728 case 'innertext': return $this->innertext();
729 case 'plaintext': return $this->text();
730 case 'xmltext': return $this->xmltext();
731 default: return array_key_exists($name, $this->attr);
732 }
733 }
734
735 function __set($name, $value) {
736 switch ($name) {
737 case 'outertext': return $this->_[HDOM_INFO_OUTER] = $value;
738 case 'innertext':
739 if (isset($this->_[HDOM_INFO_TEXT])) return $this->_[HDOM_INFO_TEXT] = $value;
740 return $this->_[HDOM_INFO_INNER] = $value;
741 }
742 if (!isset($this->attr[$name])) {
743 $this->_[HDOM_INFO_SPACE][] = array(' ', '', '');
744 $this->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE;
745 }
746 $this->attr[$name] = $value;
747 }
748
749 function __isset($name) {
750 switch ($name) {
751 case 'outertext': return true;
752 case 'innertext': return true;
753 case 'plaintext': return true;
754 }
755 //no value attr: nowrap, checked selected...
756 return (array_key_exists($name, $this->attr)) ? true : isset($this->attr[$name]);
757 }
758
759 function __unset($name) {
760 if (isset($this->attr[$name]))
761 unset($this->attr[$name]);
762 }
763
764 // PaperG - Function to convert the text from one character set to another if the two sets are not the same.
765 function convert_text($text)
766 {
767 global $debug_object;
768 if (is_object($debug_object)) {$debug_object->debugLogEntry(1);}
769
770 $converted_text = $text;
771
772 $sourceCharset = "";
773 $targetCharset = "";
774
775 if ($this->dom)
776 {
777 $sourceCharset = strtoupper($this->dom->_charset);
778 $targetCharset = strtoupper($this->dom->_target_charset);
779 }
780 if (is_object($debug_object)) {$debug_object->debugLog(3, "source charset: " . $sourceCharset . " target charaset: " . $targetCharset);}
781
782 if (!empty($sourceCharset) && !empty($targetCharset) && (strcasecmp($sourceCharset, $targetCharset) != 0))
783 {
784 // Check if the reported encoding could have been incorrect and the text is actually already UTF-8
785 if ((strcasecmp($targetCharset, 'UTF-8') == 0) && ($this->is_utf8($text)))
786 {
787 $converted_text = $text;
788 }
789 else
790 {
791 $converted_text = iconv($sourceCharset, $targetCharset, $text);
792 }
793 }
794
795 // Lets make sure that we don't have that silly BOM issue with any of the utf-8 text we output.
796 if ($targetCharset == 'UTF-8')
797 {
798 if (substr($converted_text, 0, 3) == "\xef\xbb\xbf")
799 {
800 $converted_text = substr($converted_text, 3);
801 }
802 if (substr($converted_text, -3) == "\xef\xbb\xbf")
803 {
804 $converted_text = substr($converted_text, 0, -3);
805 }
806 }
807
808 return $converted_text;
809 }
810
811 /**
812 * Returns true if $string is valid UTF-8 and false otherwise.
813 *
814 * @param mixed $str String to be tested
815 * @return boolean
816 */
817 static function is_utf8($str)
818 {
819 $c=0; $b=0;
820 $bits=0;
821 $len=strlen($str);
822 for($i=0; $i<$len; $i++)
823 {
824 $c=ord($str[$i]);
825 if($c > 128)
826 {
827 if(($c >= 254)) return false;
828 elseif($c >= 252) $bits=6;
829 elseif($c >= 248) $bits=5;
830 elseif($c >= 240) $bits=4;
831 elseif($c >= 224) $bits=3;
832 elseif($c >= 192) $bits=2;
833 else return false;
834 if(($i+$bits) > $len) return false;
835 while($bits > 1)
836 {
837 $i++;
838 $b=ord($str[$i]);
839 if($b < 128 || $b > 191) return false;
840 $bits--;
841 }
842 }
843 }
844 return true;
845 }
846 /*
847 function is_utf8($string)
848 {
849 //this is buggy
850 return (utf8_encode(utf8_decode($string)) == $string);
851 }
852 */
853
854 /**
855 * Function to try a few tricks to determine the displayed size of an img on the page.
856 * NOTE: This will ONLY work on an IMG tag. Returns FALSE on all other tag types.
857 *
858 * @author John Schlick
859 * @version April 19 2012
860 * @return array an array containing the 'height' and 'width' of the image on the page or -1 if we can't figure it out.
861 */
862 function get_display_size()
863 {
864 global $debug_object;
865
866 $width = -1;
867 $height = -1;
868
869 if ($this->tag !== 'img')
870 {
871 return false;
872 }
873
874 // See if there is aheight or width attribute in the tag itself.
875 if (isset($this->attr['width']))
876 {
877 $width = $this->attr['width'];
878 }
879
880 if (isset($this->attr['height']))
881 {
882 $height = $this->attr['height'];
883 }
884
885 // Now look for an inline style.
886 if (isset($this->attr['style']))
887 {
888 // Thanks to user gnarf from stackoverflow for this regular expression.
889 $attributes = array();
890 preg_match_all("/([\w-]+)\s*:\s*([^;]+)\s*;?/", $this->attr['style'], $matches, PREG_SET_ORDER);
891 foreach ($matches as $match) {
892 $attributes[$match[1]] = $match[2];
893 }
894
895 // If there is a width in the style attributes:
896 if (isset($attributes['width']) && $width == -1)
897 {
898 // check that the last two characters are px (pixels)
899 if (strtolower(substr($attributes['width'], -2)) == 'px')
900 {
901 $proposed_width = substr($attributes['width'], 0, -2);
902 // Now make sure that it's an integer and not something stupid.
903 if (filter_var($proposed_width, FILTER_VALIDATE_INT))
904 {
905 $width = $proposed_width;
906 }
907 }
908 }
909
910 // If there is a width in the style attributes:
911 if (isset($attributes['height']) && $height == -1)
912 {
913 // check that the last two characters are px (pixels)
914 if (strtolower(substr($attributes['height'], -2)) == 'px')
915 {
916 $proposed_height = substr($attributes['height'], 0, -2);
917 // Now make sure that it's an integer and not something stupid.
918 if (filter_var($proposed_height, FILTER_VALIDATE_INT))
919 {
920 $height = $proposed_height;
921 }
922 }
923 }
924
925 }
926
927 // Future enhancement:
928 // Look in the tag to see if there is a class or id specified that has a height or width attribute to it.
929
930 // Far future enhancement
931 // Look at all the parent tags of this image to see if they specify a class or id that has an img selector that specifies a height or width
932 // Note that in this case, the class or id will have the img subselector for it to apply to the image.
933
934 // ridiculously far future development
935 // If the class or id is specified in a SEPARATE css file thats not on the page, go get it and do what we were just doing for the ones on the page.
936
937 $result = array('height' => $height,
938 'width' => $width);
939 return $result;
940 }
941
942 // camel naming conventions
943 function getAllAttributes() {return $this->attr;}
944 function getAttribute($name) {return $this->__get($name);}
945 function setAttribute($name, $value) {$this->__set($name, $value);}
946 function hasAttribute($name) {return $this->__isset($name);}
947 function removeAttribute($name) {$this->__set($name, null);}
948 function getElementById($id) {return $this->find("#$id", 0);}
949 function getElementsById($id, $idx=null) {return $this->find("#$id", $idx);}
950 function getElementByTagName($name) {return $this->find($name, 0);}
951 function getElementsByTagName($name, $idx=null) {return $this->find($name, $idx);}
952 function parentNode() {return $this->parent();}
953 function childNodes($idx=-1) {return $this->children($idx);}
954 function firstChild() {return $this->first_child();}
955 function lastChild() {return $this->last_child();}
956 function nextSibling() {return $this->next_sibling();}
957 function previousSibling() {return $this->prev_sibling();}
958 function hasChildNodes() {return $this->has_child();}
959 function nodeName() {return $this->tag;}
960 function appendChild($node) {$node->parent($this); return $node;}
961
962}
963
964/**
965 * simple html dom parser
966 * Paperg - in the find routine: allow us to specify that we want case insensitive testing of the value of the selector.
967 * Paperg - change $size from protected to public so we can easily access it
968 * Paperg - added ForceTagsClosed in the constructor which tells us whether we trust the html or not. Default is to NOT trust it.
969 *
970 * @package PlaceLocalInclude
971 */
972class simple_html_dom
973{
974 public $root = null;
975 public $nodes = array();
976 public $callback = null;
977 public $lowercase = false;
978 // Used to keep track of how large the text was when we started.
979 public $original_size;
980 public $size;
981 protected $pos;
982 protected $doc;
983 protected $char;
984 protected $cursor;
985 protected $parent;
986 protected $noise = array();
987 protected $token_blank = " \t\r\n";
988 protected $token_equal = ' =/>';
989 protected $token_slash = " />\r\n\t";
990 protected $token_attr = ' >';
991 // Note that this is referenced by a child node, and so it needs to be public for that node to see this information.
992 public $_charset = '';
993 public $_target_charset = '';
994 protected $default_br_text = "";
995 public $default_span_text = "";
996
997 // use isset instead of in_array, performance boost about 30%...
998 protected $self_closing_tags = array('img'=>1, 'br'=>1, 'input'=>1, 'meta'=>1, 'link'=>1, 'hr'=>1, 'base'=>1, 'embed'=>1, 'spacer'=>1);
999 protected $block_tags = array('root'=>1, 'body'=>1, 'form'=>1, 'div'=>1, 'span'=>1, 'table'=>1);
1000 // Known sourceforge issue #2977341
1001 // B tags that are not closed cause us to return everything to the end of the document.
1002 protected $optional_closing_tags = array(
1003 'tr'=>array('tr'=>1, 'td'=>1, 'th'=>1),
1004 'th'=>array('th'=>1),
1005 'td'=>array('td'=>1),
1006 'li'=>array('li'=>1),
1007 'dt'=>array('dt'=>1, 'dd'=>1),
1008 'dd'=>array('dd'=>1, 'dt'=>1),
1009 'dl'=>array('dd'=>1, 'dt'=>1),
1010 'p'=>array('p'=>1),
1011 'nobr'=>array('nobr'=>1),
1012 'b'=>array('b'=>1),
1013 'option'=>array('option'=>1),
1014 );
1015
1016 function __construct($str=null, $lowercase=true, $forceTagsClosed=true, $target_charset=DEFAULT_TARGET_CHARSET, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT)
1017 {
1018 if ($str)
1019 {
1020 if (preg_match("/^http:\/\//i",$str) || is_file($str))
1021 {
1022 $this->load_file($str);
1023 }
1024 else
1025 {
1026 $this->load($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText);
1027 }
1028 }
1029 // Forcing tags to be closed implies that we don't trust the html, but it can lead to parsing errors if we SHOULD trust the html.
1030 if (!$forceTagsClosed) {
1031 $this->optional_closing_array=array();
1032 }
1033 $this->_target_charset = $target_charset;
1034 }
1035
1036 function __destruct()
1037 {
1038 $this->clear();
1039 }
1040
1041 // load html from string
1042 function load($str, $lowercase=true, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT)
1043 {
1044 global $debug_object;
1045
1046 // prepare
1047 $this->prepare($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText);
1048 // strip out comments
1049 $this->remove_noise("'<!--(.*?)-->'is");
1050 // strip out cdata
1051 $this->remove_noise("'<!\[CDATA\[(.*?)\]\]>'is", true);
1052 // Per sourceforge http://sourceforge.net/tracker/?func=detail&aid=2949097&group_id=218559&atid=1044037
1053 // Script tags removal now preceeds style tag removal.
1054 // strip out <script> tags
1055 $this->remove_noise("'<\s*script[^>]*[^/]>(.*?)<\s*/\s*script\s*>'is");
1056 $this->remove_noise("'<\s*script\s*>(.*?)<\s*/\s*script\s*>'is");
1057 // strip out <style> tags
1058 $this->remove_noise("'<\s*style[^>]*[^/]>(.*?)<\s*/\s*style\s*>'is");
1059 $this->remove_noise("'<\s*style\s*>(.*?)<\s*/\s*style\s*>'is");
1060 // strip out preformatted tags
1061 $this->remove_noise("'<\s*(?:code)[^>]*>(.*?)<\s*/\s*(?:code)\s*>'is");
1062 // strip out server side scripts
1063 $this->remove_noise("'(<\?)(.*?)(\?>)'s", true);
1064 // strip smarty scripts
1065 $this->remove_noise("'(\{\w)(.*?)(\})'s", true);
1066
1067 // parsing
1068 while ($this->parse());
1069 // end
1070 $this->root->_[HDOM_INFO_END] = $this->cursor;
1071 $this->parse_charset();
1072
1073 // make load function chainable
1074 return $this;
1075
1076 }
1077
1078 // load html from file
1079 function load_file()
1080 {
1081 $args = func_get_args();
1082 $this->load(call_user_func_array('file_get_contents', $args), true);
1083 // Throw an error if we can't properly load the dom.
1084 if (($error=error_get_last())!==null) {
1085 $this->clear();
1086 return false;
1087 }
1088 }
1089
1090 // set callback function
1091 function set_callback($function_name)
1092 {
1093 $this->callback = $function_name;
1094 }
1095
1096 // remove callback function
1097 function remove_callback()
1098 {
1099 $this->callback = null;
1100 }
1101
1102 // save dom as string
1103 function save($filepath='')
1104 {
1105 $ret = $this->root->innertext();
1106 if ($filepath!=='') file_put_contents($filepath, $ret, LOCK_EX);
1107 return $ret;
1108 }
1109
1110 // find dom node by css selector
1111 // Paperg - allow us to specify that we want case insensitive testing of the value of the selector.
1112 function find($selector, $idx=null, $lowercase=false)
1113 {
1114 return $this->root->find($selector, $idx, $lowercase);
1115 }
1116
1117 // clean up memory due to php5 circular references memory leak...
1118 function clear()
1119 {
1120 foreach ($this->nodes as $n) {$n->clear(); $n = null;}
1121 // This add next line is documented in the sourceforge repository. 2977248 as a fix for ongoing memory leaks that occur even with the use of clear.
1122 if (isset($this->children)) foreach ($this->children as $n) {$n->clear(); $n = null;}
1123 if (isset($this->parent)) {$this->parent->clear(); unset($this->parent);}
1124 if (isset($this->root)) {$this->root->clear(); unset($this->root);}
1125 unset($this->doc);
1126 unset($this->noise);
1127 }
1128
1129 function dump($show_attr=true)
1130 {
1131 $this->root->dump($show_attr);
1132 }
1133
1134 // prepare HTML data and init everything
1135 protected function prepare($str, $lowercase=true, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT)
1136 {
1137 $this->clear();
1138
1139 // set the length of content before we do anything to it.
1140 $this->size = strlen($str);
1141 // Save the original size of the html that we got in. It might be useful to someone.
1142 $this->original_size = $this->size;
1143
1144 //before we save the string as the doc... strip out the \r \n's if we are told to.
1145 if ($stripRN) {
1146 $str = str_replace("\r", " ", $str);
1147 $str = str_replace("\n", " ", $str);
1148
1149 // set the length of content since we have changed it.
1150 $this->size = strlen($str);
1151 }
1152
1153 $this->doc = $str;
1154 $this->pos = 0;
1155 $this->cursor = 1;
1156 $this->noise = array();
1157 $this->nodes = array();
1158 $this->lowercase = $lowercase;
1159 $this->default_br_text = $defaultBRText;
1160 $this->default_span_text = $defaultSpanText;
1161 $this->root = new simple_html_dom_node($this);
1162 $this->root->tag = 'root';
1163 $this->root->_[HDOM_INFO_BEGIN] = -1;
1164 $this->root->nodetype = HDOM_TYPE_ROOT;
1165 $this->parent = $this->root;
1166 if ($this->size>0) $this->char = $this->doc[0];
1167 }
1168
1169 // parse html content
1170 protected function parse()
1171 {
1172 if (($s = $this->copy_until_char('<'))==='')
1173 {
1174 return $this->read_tag();
1175 }
1176
1177 // text
1178 $node = new simple_html_dom_node($this);
1179 ++$this->cursor;
1180 $node->_[HDOM_INFO_TEXT] = $s;
1181 $this->link_nodes($node, false);
1182 return true;
1183 }
1184
1185 // PAPERG - dkchou - added this to try to identify the character set of the page we have just parsed so we know better how to spit it out later.
1186 // NOTE: IF you provide a routine called get_last_retrieve_url_contents_content_type which returns the CURLINFO_CONTENT_TYPE from the last curl_exec
1187 // (or the content_type header from the last transfer), we will parse THAT, and if a charset is specified, we will use it over any other mechanism.
1188 protected function parse_charset()
1189 {
1190 global $debug_object;
1191
1192 $charset = null;
1193
1194 if (function_exists('get_last_retrieve_url_contents_content_type'))
1195 {
1196 $contentTypeHeader = get_last_retrieve_url_contents_content_type();
1197 $success = preg_match('/charset=(.+)/', $contentTypeHeader, $matches);
1198 if ($success)
1199 {
1200 $charset = $matches[1];
1201 if (is_object($debug_object)) {$debug_object->debugLog(2, 'header content-type found charset of: ' . $charset);}
1202 }
1203
1204 }
1205
1206 if (empty($charset))
1207 {
1208 $el = $this->root->find('meta[http-equiv=Content-Type]',0);
1209 if (!empty($el))
1210 {
1211 $fullvalue = $el->content;
1212 if (is_object($debug_object)) {$debug_object->debugLog(2, 'meta content-type tag found' . $fullvalue);}
1213
1214 if (!empty($fullvalue))
1215 {
1216 $success = preg_match('/charset=(.+)/', $fullvalue, $matches);
1217 if ($success)
1218 {
1219 $charset = $matches[1];
1220 }
1221 else
1222 {
1223 // If there is a meta tag, and they don't specify the character set, research says that it's typically ISO-8859-1
1224 if (is_object($debug_object)) {$debug_object->debugLog(2, 'meta content-type tag couldn\'t be parsed. using iso-8859 default.');}
1225 $charset = 'ISO-8859-1';
1226 }
1227 }
1228 }
1229 }
1230
1231 // If we couldn't find a charset above, then lets try to detect one based on the text we got...
1232 if (empty($charset))
1233 {
1234 // Have php try to detect the encoding from the text given to us.
1235 $charset = mb_detect_encoding($this->root->plaintext . "ascii", $encoding_list = array( "UTF-8", "CP1252" ) );
1236 if (is_object($debug_object)) {$debug_object->debugLog(2, 'mb_detect found: ' . $charset);}
1237
1238 // and if this doesn't work... then we need to just wrongheadedly assume it's UTF-8 so that we can move on - cause this will usually give us most of what we need...
1239 if ($charset === false)
1240 {
1241 if (is_object($debug_object)) {$debug_object->debugLog(2, 'since mb_detect failed - using default of utf-8');}
1242 $charset = 'UTF-8';
1243 }
1244 }
1245
1246 // Since CP1252 is a superset, if we get one of it's subsets, we want it instead.
1247 if ((strtolower($charset) == strtolower('ISO-8859-1')) || (strtolower($charset) == strtolower('Latin1')) || (strtolower($charset) == strtolower('Latin-1')))
1248 {
1249 if (is_object($debug_object)) {$debug_object->debugLog(2, 'replacing ' . $charset . ' with CP1252 as its a superset');}
1250 $charset = 'CP1252';
1251 }
1252
1253 if (is_object($debug_object)) {$debug_object->debugLog(1, 'EXIT - ' . $charset);}
1254
1255 return $this->_charset = $charset;
1256 }
1257
1258 // read tag info
1259 protected function read_tag()
1260 {
1261 if ($this->char!=='<')
1262 {
1263 $this->root->_[HDOM_INFO_END] = $this->cursor;
1264 return false;
1265 }
1266 $begin_tag_pos = $this->pos;
1267 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1268
1269 // end tag
1270 if ($this->char==='/')
1271 {
1272 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1273 // This represents the change in the simple_html_dom trunk from revision 180 to 181.
1274 // $this->skip($this->token_blank_t);
1275 $this->skip($this->token_blank);
1276 $tag = $this->copy_until_char('>');
1277
1278 // skip attributes in end tag
1279 if (($pos = strpos($tag, ' '))!==false)
1280 $tag = substr($tag, 0, $pos);
1281
1282 $parent_lower = strtolower($this->parent->tag);
1283 $tag_lower = strtolower($tag);
1284
1285 if ($parent_lower!==$tag_lower)
1286 {
1287 if (isset($this->optional_closing_tags[$parent_lower]) && isset($this->block_tags[$tag_lower]))
1288 {
1289 $this->parent->_[HDOM_INFO_END] = 0;
1290 $org_parent = $this->parent;
1291
1292 while (($this->parent->parent) && strtolower($this->parent->tag)!==$tag_lower)
1293 $this->parent = $this->parent->parent;
1294
1295 if (strtolower($this->parent->tag)!==$tag_lower) {
1296 $this->parent = $org_parent; // restore origonal parent
1297 if ($this->parent->parent) $this->parent = $this->parent->parent;
1298 $this->parent->_[HDOM_INFO_END] = $this->cursor;
1299 return $this->as_text_node($tag);
1300 }
1301 }
1302 else if (($this->parent->parent) && isset($this->block_tags[$tag_lower]))
1303 {
1304 $this->parent->_[HDOM_INFO_END] = 0;
1305 $org_parent = $this->parent;
1306
1307 while (($this->parent->parent) && strtolower($this->parent->tag)!==$tag_lower)
1308 $this->parent = $this->parent->parent;
1309
1310 if (strtolower($this->parent->tag)!==$tag_lower)
1311 {
1312 $this->parent = $org_parent; // restore origonal parent
1313 $this->parent->_[HDOM_INFO_END] = $this->cursor;
1314 return $this->as_text_node($tag);
1315 }
1316 }
1317 else if (($this->parent->parent) && strtolower($this->parent->parent->tag)===$tag_lower)
1318 {
1319 $this->parent->_[HDOM_INFO_END] = 0;
1320 $this->parent = $this->parent->parent;
1321 }
1322 else
1323 return $this->as_text_node($tag);
1324 }
1325
1326 $this->parent->_[HDOM_INFO_END] = $this->cursor;
1327 if ($this->parent->parent) $this->parent = $this->parent->parent;
1328
1329 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1330 return true;
1331 }
1332
1333 $node = new simple_html_dom_node($this);
1334 $node->_[HDOM_INFO_BEGIN] = $this->cursor;
1335 ++$this->cursor;
1336 $tag = $this->copy_until($this->token_slash);
1337 $node->tag_start = $begin_tag_pos;
1338
1339 // doctype, cdata & comments...
1340 if (isset($tag[0]) && $tag[0]==='!') {
1341 $node->_[HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until_char('>');
1342
1343 if (isset($tag[2]) && $tag[1]==='-' && $tag[2]==='-') {
1344 $node->nodetype = HDOM_TYPE_COMMENT;
1345 $node->tag = 'comment';
1346 } else {
1347 $node->nodetype = HDOM_TYPE_UNKNOWN;
1348 $node->tag = 'unknown';
1349 }
1350 if ($this->char==='>') $node->_[HDOM_INFO_TEXT].='>';
1351 $this->link_nodes($node, true);
1352 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1353 return true;
1354 }
1355
1356 // text
1357 if ($pos=strpos($tag, '<')!==false) {
1358 $tag = '<' . substr($tag, 0, -1);
1359 $node->_[HDOM_INFO_TEXT] = $tag;
1360 $this->link_nodes($node, false);
1361 $this->char = $this->doc[--$this->pos]; // prev
1362 return true;
1363 }
1364
1365 if (!preg_match("/^[\w-:]+$/", $tag)) {
1366 $node->_[HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until('<>');
1367 if ($this->char==='<') {
1368 $this->link_nodes($node, false);
1369 return true;
1370 }
1371
1372 if ($this->char==='>') $node->_[HDOM_INFO_TEXT].='>';
1373 $this->link_nodes($node, false);
1374 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1375 return true;
1376 }
1377
1378 // begin tag
1379 $node->nodetype = HDOM_TYPE_ELEMENT;
1380 $tag_lower = strtolower($tag);
1381 $node->tag = ($this->lowercase) ? $tag_lower : $tag;
1382
1383 // handle optional closing tags
1384 if (isset($this->optional_closing_tags[$tag_lower]) )
1385 {
1386 while (isset($this->optional_closing_tags[$tag_lower][strtolower($this->parent->tag)]))
1387 {
1388 $this->parent->_[HDOM_INFO_END] = 0;
1389 $this->parent = $this->parent->parent;
1390 }
1391 $node->parent = $this->parent;
1392 }
1393
1394 $guard = 0; // prevent infinity loop
1395 $space = array($this->copy_skip($this->token_blank), '', '');
1396
1397 // attributes
1398 do
1399 {
1400 if ($this->char!==null && $space[0]==='')
1401 {
1402 break;
1403 }
1404 $name = $this->copy_until($this->token_equal);
1405 if ($guard===$this->pos)
1406 {
1407 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1408 continue;
1409 }
1410 $guard = $this->pos;
1411
1412 // handle endless '<'
1413 if ($this->pos>=$this->size-1 && $this->char!=='>') {
1414 $node->nodetype = HDOM_TYPE_TEXT;
1415 $node->_[HDOM_INFO_END] = 0;
1416 $node->_[HDOM_INFO_TEXT] = '<'.$tag . $space[0] . $name;
1417 $node->tag = 'text';
1418 $this->link_nodes($node, false);
1419 return true;
1420 }
1421
1422 // handle mismatch '<'
1423 if ($this->doc[$this->pos-1]=='<') {
1424 $node->nodetype = HDOM_TYPE_TEXT;
1425 $node->tag = 'text';
1426 $node->attr = array();
1427 $node->_[HDOM_INFO_END] = 0;
1428 $node->_[HDOM_INFO_TEXT] = substr($this->doc, $begin_tag_pos, $this->pos-$begin_tag_pos-1);
1429 $this->pos -= 2;
1430 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1431 $this->link_nodes($node, false);
1432 return true;
1433 }
1434
1435 if ($name!=='/' && $name!=='') {
1436 $space[1] = $this->copy_skip($this->token_blank);
1437 $name = $this->restore_noise($name);
1438 if ($this->lowercase) $name = strtolower($name);
1439 if ($this->char==='=') {
1440 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1441 $this->parse_attr($node, $name, $space);
1442 }
1443 else {
1444 //no value attr: nowrap, checked selected...
1445 $node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_NO;
1446 $node->attr[$name] = true;
1447 if ($this->char!='>') $this->char = $this->doc[--$this->pos]; // prev
1448 }
1449 $node->_[HDOM_INFO_SPACE][] = $space;
1450 $space = array($this->copy_skip($this->token_blank), '', '');
1451 }
1452 else
1453 break;
1454 } while ($this->char!=='>' && $this->char!=='/');
1455
1456 $this->link_nodes($node, true);
1457 $node->_[HDOM_INFO_ENDSPACE] = $space[0];
1458
1459 // check self closing
1460 if ($this->copy_until_char_escape('>')==='/')
1461 {
1462 $node->_[HDOM_INFO_ENDSPACE] .= '/';
1463 $node->_[HDOM_INFO_END] = 0;
1464 }
1465 else
1466 {
1467 // reset parent
1468 if (!isset($this->self_closing_tags[strtolower($node->tag)])) $this->parent = $node;
1469 }
1470 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1471
1472 // If it's a BR tag, we need to set it's text to the default text.
1473 // This way when we see it in plaintext, we can generate formatting that the user wants.
1474 // since a br tag never has sub nodes, this works well.
1475 if ($node->tag == "br")
1476 {
1477 $node->_[HDOM_INFO_INNER] = $this->default_br_text;
1478 }
1479
1480 return true;
1481 }
1482
1483 // parse attributes
1484 protected function parse_attr($node, $name, &$space)
1485 {
1486 // Per sourceforge: http://sourceforge.net/tracker/?func=detail&aid=3061408&group_id=218559&atid=1044037
1487 // If the attribute is already defined inside a tag, only pay atetntion to the first one as opposed to the last one.
1488 if (isset($node->attr[$name]))
1489 {
1490 return;
1491 }
1492
1493 $space[2] = $this->copy_skip($this->token_blank);
1494 switch ($this->char) {
1495 case '"':
1496 $node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE;
1497 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1498 $node->attr[$name] = $this->restore_noise($this->copy_until_char_escape('"'));
1499 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1500 break;
1501 case '\'':
1502 $node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_SINGLE;
1503 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1504 $node->attr[$name] = $this->restore_noise($this->copy_until_char_escape('\''));
1505 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1506 break;
1507 default:
1508 $node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_NO;
1509 $node->attr[$name] = $this->restore_noise($this->copy_until($this->token_attr));
1510 }
1511 // PaperG: Attributes should not have \r or \n in them, that counts as html whitespace.
1512 $node->attr[$name] = str_replace("\r", "", $node->attr[$name]);
1513 $node->attr[$name] = str_replace("\n", "", $node->attr[$name]);
1514 // PaperG: If this is a "class" selector, lets get rid of the preceeding and trailing space since some people leave it in the multi class case.
1515 if ($name == "class") {
1516 $node->attr[$name] = trim($node->attr[$name]);
1517 }
1518 }
1519
1520 // link node's parent
1521 protected function link_nodes(&$node, $is_child)
1522 {
1523 $node->parent = $this->parent;
1524 $this->parent->nodes[] = $node;
1525 if ($is_child)
1526 {
1527 $this->parent->children[] = $node;
1528 }
1529 }
1530
1531 // as a text node
1532 protected function as_text_node($tag)
1533 {
1534 $node = new simple_html_dom_node($this);
1535 ++$this->cursor;
1536 $node->_[HDOM_INFO_TEXT] = '</' . $tag . '>';
1537 $this->link_nodes($node, false);
1538 $this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1539 return true;
1540 }
1541
1542 protected function skip($chars)
1543 {
1544 $this->pos += strspn($this->doc, $chars, $this->pos);
1545 $this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1546 }
1547
1548 protected function copy_skip($chars)
1549 {
1550 $pos = $this->pos;
1551 $len = strspn($this->doc, $chars, $pos);
1552 $this->pos += $len;
1553 $this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1554 if ($len===0) return '';
1555 return substr($this->doc, $pos, $len);
1556 }
1557
1558 protected function copy_until($chars)
1559 {
1560 $pos = $this->pos;
1561 $len = strcspn($this->doc, $chars, $pos);
1562 $this->pos += $len;
1563 $this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next
1564 return substr($this->doc, $pos, $len);
1565 }
1566
1567 protected function copy_until_char($char)
1568 {
1569 if ($this->char===null) return '';
1570
1571 if (($pos = strpos($this->doc, $char, $this->pos))===false) {
1572 $ret = substr($this->doc, $this->pos, $this->size-$this->pos);
1573 $this->char = null;
1574 $this->pos = $this->size;
1575 return $ret;
1576 }
1577
1578 if ($pos===$this->pos) return '';
1579 $pos_old = $this->pos;
1580 $this->char = $this->doc[$pos];
1581 $this->pos = $pos;
1582 return substr($this->doc, $pos_old, $pos-$pos_old);
1583 }
1584
1585 protected function copy_until_char_escape($char)
1586 {
1587 if ($this->char===null) return '';
1588
1589 $start = $this->pos;
1590 while (1)
1591 {
1592 if (($pos = strpos($this->doc, $char, $start))===false)
1593 {
1594 $ret = substr($this->doc, $this->pos, $this->size-$this->pos);
1595 $this->char = null;
1596 $this->pos = $this->size;
1597 return $ret;
1598 }
1599
1600 if ($pos===$this->pos) return '';
1601
1602 if ($this->doc[$pos-1]==='\\') {
1603 $start = $pos+1;
1604 continue;
1605 }
1606
1607 $pos_old = $this->pos;
1608 $this->char = $this->doc[$pos];
1609 $this->pos = $pos;
1610 return substr($this->doc, $pos_old, $pos-$pos_old);
1611 }
1612 }
1613
1614 // remove noise from html content
1615 // save the noise in the $this->noise array.
1616 protected function remove_noise($pattern, $remove_tag=false)
1617 {
1618 global $debug_object;
1619 if (is_object($debug_object)) { $debug_object->debugLogEntry(1); }
1620
1621 $count = preg_match_all($pattern, $this->doc, $matches, PREG_SET_ORDER|PREG_OFFSET_CAPTURE);
1622
1623 for ($i=$count-1; $i>-1; --$i)
1624 {
1625 $key = '___noise___'.sprintf('% 5d', count($this->noise)+1000);
1626 if (is_object($debug_object)) { $debug_object->debugLog(2, 'key is: ' . $key); }
1627 $idx = ($remove_tag) ? 0 : 1;
1628 $this->noise[$key] = $matches[$i][$idx][0];
1629 $this->doc = substr_replace($this->doc, $key, $matches[$i][$idx][1], strlen($matches[$i][$idx][0]));
1630 }
1631
1632 // reset the length of content
1633 $this->size = strlen($this->doc);
1634 if ($this->size>0)
1635 {
1636 $this->char = $this->doc[0];
1637 }
1638 }
1639
1640 // restore noise to html content
1641 function restore_noise($text)
1642 {
1643 global $debug_object;
1644 if (is_object($debug_object)) { $debug_object->debugLogEntry(1); }
1645
1646 while (($pos=strpos($text, '___noise___'))!==false)
1647 {
1648 // Sometimes there is a broken piece of markup, and we don't GET the pos+11 etc... token which indicates a problem outside of us...
1649 if (strlen($text) > $pos+15)
1650 {
1651 $key = '___noise___'.$text[$pos+11].$text[$pos+12].$text[$pos+13].$text[$pos+14].$text[$pos+15];
1652 if (is_object($debug_object)) { $debug_object->debugLog(2, 'located key of: ' . $key); }
1653
1654 if (isset($this->noise[$key]))
1655 {
1656 $text = substr($text, 0, $pos).$this->noise[$key].substr($text, $pos+16);
1657 }
1658 else
1659 {
1660 // do this to prevent an infinite loop.
1661 $text = substr($text, 0, $pos).'UNDEFINED NOISE FOR KEY: '.$key . substr($text, $pos+16);
1662 }
1663 }
1664 else
1665 {
1666 // There is no valid key being given back to us... We must get rid of the ___noise___ or we will have a problem.
1667 $text = substr($text, 0, $pos).'NO NUMERIC NOISE KEY' . substr($text, $pos+11);
1668 }
1669 }
1670 return $text;
1671 }
1672
1673 // Sometimes we NEED one of the noise elements.
1674 function search_noise($text)
1675 {
1676 global $debug_object;
1677 if (is_object($debug_object)) { $debug_object->debugLogEntry(1); }
1678
1679 foreach($this->noise as $noiseElement)
1680 {
1681 if (strpos($noiseElement, $text)!==false)
1682 {
1683 return $noiseElement;
1684 }
1685 }
1686 }
1687 function __toString()
1688 {
1689 return $this->root->innertext();
1690 }
1691
1692 function __get($name)
1693 {
1694 switch ($name)
1695 {
1696 case 'outertext':
1697 return $this->root->innertext();
1698 case 'innertext':
1699 return $this->root->innertext();
1700 case 'plaintext':
1701 return $this->root->text();
1702 case 'charset':
1703 return $this->_charset;
1704 case 'target_charset':
1705 return $this->_target_charset;
1706 }
1707 }
1708
1709 // camel naming conventions
1710 function childNodes($idx=-1) {return $this->root->childNodes($idx);}
1711 function firstChild() {return $this->root->first_child();}
1712 function lastChild() {return $this->root->last_child();}
1713 function createElement($name, $value=null) {return @str_get_html("<$name>$value</$name>")->first_child();}
1714 function createTextNode($value) {return @end(str_get_html($value)->nodes);}
1715 function getElementById($id) {return $this->find("#$id", 0);}
1716 function getElementsById($id, $idx=null) {return $this->find("#$id", $idx);}
1717 function getElementByTagName($name) {return $this->find($name, 0);}
1718 function getElementsByTagName($name, $idx=-1) {return $this->find($name, $idx);}
1719 function loadFile() {$args = func_get_args();$this->load_file($args);}
1720}
1721
1722?> \ No newline at end of file