Plugin to allow visitor contributions to WordPress posts, wiki style.
Du kannst nicht mehr als 25 Themen auswählen Themen müssen entweder mit einem Buchstaben oder einer Ziffer beginnen. Sie können Bindestriche („-“) enthalten und bis zu 35 Zeichen lang sein.

1742 Zeilen
54KB

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