API tools faq
paste
Advertisement
Guest User

Simple_html_dom_node

a guest
Feb 17th, 2014
115
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
PHP 13.79 KB | None | 0 0
raw download clone embed print report
  1. <?php
  2.  
  3. // simple html dom node
  4. // -----------------------------------------------------------------------------
  5. class Simple_html_dom_node {
  6.     public $nodetype = HDOM_TYPE_TEXT;
  7.     public $tag = 'text';
  8.     public $attr = array();
  9.     public $children = array();
  10.     public $nodes = array();
  11.     public $parent = null;
  12.     public $_ = array();
  13.     private $dom = null;
  14.  
  15.     function __construct($dom=false) {
  16.         if($dom)
  17.         {
  18.             $this->dom = $dom;
  19.             $dom->nodes[] = $this;
  20.         }
  21.     }
  22.  
  23.     function __destruct() {
  24.         $this->clear();
  25.     }
  26.  
  27.     function __toString() {
  28.         return $this->outertext();
  29.     }
  30.  
  31.     // clean up memory due to php5 circular references memory leak...
  32.     function clear() {
  33.         $this->dom = null;
  34.         $this->nodes = null;
  35.         $this->parent = null;
  36.         $this->children = null;
  37.     }
  38.    
  39.     // dump node's tree
  40.     function dump($show_attr=true) {
  41.         dump_html_tree($this, $show_attr);
  42.     }
  43.  
  44.     // returns the parent of node
  45.     function parent() {
  46.         return $this->parent;
  47.     }
  48.  
  49.     // returns children of node
  50.     function children($idx=-1) {
  51.         if ($idx===-1) return $this->children;
  52.         if (isset($this->children[$idx])) return $this->children[$idx];
  53.         return null;
  54.     }
  55.  
  56.     // returns the first child of node
  57.     function first_child() {
  58.         if (count($this->children)>0) return $this->children[0];
  59.         return null;
  60.     }
  61.  
  62.     // returns the last child of node
  63.     function last_child() {
  64.         if (($count=count($this->children))>0) return $this->children[$count-1];
  65.         return null;
  66.     }
  67.  
  68.     // returns the next sibling of node    
  69.     function next_sibling() {
  70.         if ($this->parent===null) return null;
  71.         $idx = 0;
  72.         $count = count($this->parent->children);
  73.         while ($idx<$count && $this!==$this->parent->children[$idx])
  74.             ++$idx;
  75.         if (++$idx>=$count) return null;
  76.         return $this->parent->children[$idx];
  77.     }
  78.  
  79.     // returns the previous sibling of node
  80.     function prev_sibling() {
  81.         if ($this->parent===null) return null;
  82.         $idx = 0;
  83.         $count = count($this->parent->children);
  84.         while ($idx<$count && $this!==$this->parent->children[$idx])
  85.             ++$idx;
  86.         if (--$idx<0) return null;
  87.         return $this->parent->children[$idx];
  88.     }
  89.  
  90.     // get dom node's inner html
  91.     function innertext() {
  92.         if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER];
  93.         if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
  94.  
  95.         $ret = '';
  96.         foreach($this->nodes as $n)
  97.             $ret .= $n->outertext();
  98.         return $ret;
  99.     }
  100.  
  101.     // get dom node's outer text (with tag)
  102.     function outertext() {
  103.         if ($this->tag==='root') return $this->innertext();
  104.  
  105.         // trigger callback
  106.         if ($this->dom->callback!==null)
  107.             call_user_func_array($this->dom->callback, array($this));
  108.  
  109.         if (isset($this->_[HDOM_INFO_OUTER])) return $this->_[HDOM_INFO_OUTER];
  110.         if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
  111.  
  112.         // render begin tag
  113.         $ret = $this->dom->nodes[$this->_[HDOM_INFO_BEGIN]]->makeup();
  114.  
  115.         // render inner text
  116.         if (isset($this->_[HDOM_INFO_INNER]))
  117.             $ret .= $this->_[HDOM_INFO_INNER];
  118.         else {
  119.             foreach($this->nodes as $n)
  120.                 $ret .= $n->outertext();
  121.         }
  122.  
  123.         // render end tag
  124.         if(isset($this->_[HDOM_INFO_END]) && $this->_[HDOM_INFO_END]!=0)
  125.             $ret .= '</'.$this->tag.'>';
  126.         return $ret;
  127.     }
  128.  
  129.     // get dom node's plain text
  130.     function text() {
  131.         if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER];
  132.         switch ($this->nodetype) {
  133.             case HDOM_TYPE_TEXT: return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
  134.             case HDOM_TYPE_COMMENT: return '';
  135.             case HDOM_TYPE_UNKNOWN: return '';
  136.         }
  137.         if (strcasecmp($this->tag, 'script')===0) return '';
  138.         if (strcasecmp($this->tag, 'style')===0) return '';
  139.  
  140.         $ret = '';
  141.         foreach($this->nodes as $n)
  142.             $ret .= $n->text();
  143.         return $ret;
  144.     }
  145.    
  146.     function xmltext() {
  147.         $ret = $this->innertext();
  148.         $ret = str_ireplace('<![CDATA[', '', $ret);
  149.         $ret = str_replace(']]>', '', $ret);
  150.         return $ret;
  151.     }
  152.  
  153.     // build node's text with tag
  154.     function makeup() {
  155.         // text, comment, unknown
  156.         if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]);
  157.  
  158.         $ret = '<'.$this->tag;
  159.         $i = -1;
  160.  
  161.         foreach($this->attr as $key=>$val) {
  162.             ++$i;
  163.  
  164.             // skip removed attribute
  165.             if ($val===null || $val===false)
  166.                 continue;
  167.  
  168.             $ret .= $this->_[HDOM_INFO_SPACE][$i][0];
  169.             //no value attr: nowrap, checked selected...
  170.             if ($val===true)
  171.                 $ret .= $key;
  172.             else {
  173.                 switch($this->_[HDOM_INFO_QUOTE][$i]) {
  174.                     case HDOM_QUOTE_DOUBLE: $quote = '"'; break;
  175.                     case HDOM_QUOTE_SINGLE: $quote = '\''; break;
  176.                     default: $quote = '';
  177.                 }
  178.                 $ret .= $key.$this->_[HDOM_INFO_SPACE][$i][1].'='.$this->_[HDOM_INFO_SPACE][$i][2].$quote.$val.$quote;
  179.             }
  180.         }
  181.         $ret = $this->dom->restore_noise($ret);
  182.         return $ret . $this->_[HDOM_INFO_ENDSPACE] . '>';
  183.     }
  184.  
  185.     // find elements by css selector
  186.     function find($selector, $idx=null) {
  187.         $selectors = $this->parse_selector($selector);
  188.         if (($count=count($selectors))===0) return array();
  189.         $found_keys = array();
  190.  
  191.         // find each selector
  192.         for ($c=0; $c<$count; ++$c) {
  193.             if (($levle=count($selectors[0]))===0) return array();
  194.             if (!isset($this->_[HDOM_INFO_BEGIN])) return array();
  195.  
  196.             $head = array($this->_[HDOM_INFO_BEGIN]=>1);
  197.  
  198.             // handle descendant selectors, no recursive!
  199.             for ($l=0; $l<$levle; ++$l) {
  200.                 $ret = array();
  201.                 foreach($head as $k=>$v) {
  202.                     $n = ($k===-1) ? $this->dom->root : $this->dom->nodes[$k];
  203.                     $n->seek($selectors[$c][$l], $ret);
  204.                 }
  205.                 $head = $ret;
  206.             }
  207.  
  208.             foreach($head as $k=>$v) {
  209.                 if (!isset($found_keys[$k]))
  210.                     $found_keys[$k] = 1;
  211.             }
  212.         }
  213.  
  214.         // sort keys
  215.         ksort($found_keys);
  216.  
  217.         $found = array();
  218.         foreach($found_keys as $k=>$v)
  219.             $found[] = $this->dom->nodes[$k];
  220.  
  221.         // return nth-element or array
  222.         if (is_null($idx)) return $found;
  223.         else if ($idx<0) $idx = count($found) + $idx;
  224.         return (isset($found[$idx])) ? $found[$idx] : null;
  225.     }
  226.  
  227.     // seek for given conditions
  228.     protected function seek($selector, &$ret) {
  229.         list($tag, $key, $val, $exp, $no_key) = $selector;
  230.  
  231.         // xpath index
  232.         if ($tag && $key && is_numeric($key)) {
  233.             $count = 0;
  234.             foreach ($this->children as $c) {
  235.                 if ($tag==='*' || $tag===$c->tag) {
  236.                     if (++$count==$key) {
  237.                         $ret[$c->_[HDOM_INFO_BEGIN]] = 1;
  238.                         return;
  239.                     }
  240.                 }
  241.             }
  242.             return;
  243.         }
  244.  
  245.         $end = (!empty($this->_[HDOM_INFO_END])) ? $this->_[HDOM_INFO_END] : 0;
  246.         if ($end==0) {
  247.             $parent = $this->parent;
  248.             while (!isset($parent->_[HDOM_INFO_END]) && $parent!==null) {
  249.                 $end -= 1;
  250.                 $parent = $parent->parent;
  251.             }
  252.             $end += $parent->_[HDOM_INFO_END];
  253.         }
  254.  
  255.         for($i=$this->_[HDOM_INFO_BEGIN]+1; $i<$end; ++$i) {
  256.             $node = $this->dom->nodes[$i];
  257.             $pass = true;
  258.  
  259.             if ($tag==='*' && !$key) {
  260.                 if (in_array($node, $this->children, true))
  261.                     $ret[$i] = 1;
  262.                 continue;
  263.             }
  264.  
  265.             // compare tag
  266.             if ($tag && $tag!=$node->tag && $tag!=='*') {$pass=false;}
  267.             // compare key
  268.             if ($pass && $key) {
  269.                 if ($no_key) {
  270.                     if (isset($node->attr[$key])) $pass=false;
  271.                 }
  272.                 else if (!isset($node->attr[$key])) $pass=false;
  273.             }
  274.             // compare value
  275.             if ($pass && $key && $val  && $val!=='*') {
  276.                 $check = $this->match($exp, $val, $node->attr[$key]);
  277.                 // handle multiple class
  278.                 if (!$check && strcasecmp($key, 'class')===0) {
  279.                     foreach(explode(' ',$node->attr[$key]) as $k) {
  280.                         $check = $this->match($exp, $val, $k);
  281.                         if ($check) break;
  282.                     }
  283.                 }
  284.                 if (!$check) $pass = false;
  285.             }
  286.             if ($pass) $ret[$i] = 1;
  287.             unset($node);
  288.         }
  289.     }
  290.  
  291.     protected function match($exp, $pattern, $value) {
  292.         switch ($exp) {
  293.             case '=':
  294.                 return ($value===$pattern);
  295.             case '!=':
  296.                 return ($value!==$pattern);
  297.             case '^=':
  298.                 return preg_match("/^".preg_quote($pattern,'/')."/", $value);
  299.             case '$=':
  300.                 return preg_match("/".preg_quote($pattern,'/')."$/", $value);
  301.             case '*=':
  302.                 if ($pattern[0]=='/')
  303.                     return preg_match($pattern, $value);
  304.                 return preg_match("/".$pattern."/i", $value);
  305.         }
  306.         return false;
  307.     }
  308.  
  309.     protected function parse_selector($selector_string) {
  310.         // pattern of CSS selectors, modified from mootools
  311.         $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is";
  312.         preg_match_all($pattern, trim($selector_string).' ', $matches, PREG_SET_ORDER);
  313.         $selectors = array();
  314.         $result = array();
  315.         //print_r($matches);
  316.  
  317.         foreach ($matches as $m) {
  318.             $m[0] = trim($m[0]);
  319.             if ($m[0]==='' || $m[0]==='/' || $m[0]==='//') continue;
  320.             // for borwser grnreated xpath
  321.             if ($m[1]==='tbody') continue;
  322.  
  323.             list($tag, $key, $val, $exp, $no_key) = array($m[1], null, null, '=', false);
  324.             if(!empty($m[2])) {$key='id'; $val=$m[2];}
  325.             if(!empty($m[3])) {$key='class'; $val=$m[3];}
  326.             if(!empty($m[4])) {$key=$m[4];}
  327.             if(!empty($m[5])) {$exp=$m[5];}
  328.             if(!empty($m[6])) {$val=$m[6];}
  329.  
  330.             // convert to lowercase
  331.             if ($this->dom->lowercase) {$tag=strtolower($tag); $key=strtolower($key);}
  332.             //elements that do NOT have the specified attribute
  333.             if (isset($key[0]) && $key[0]==='!') {$key=substr($key, 1); $no_key=true;}
  334.  
  335.             $result[] = array($tag, $key, $val, $exp, $no_key);
  336.             if (trim($m[7])===',') {
  337.                 $selectors[] = $result;
  338.                 $result = array();
  339.             }
  340.         }
  341.         if (count($result)>0)
  342.             $selectors[] = $result;
  343.         return $selectors;
  344.     }
  345.  
  346.     function __get($name) {
  347.         if (isset($this->attr[$name])) return $this->attr[$name];
  348.         switch($name) {
  349.             case 'outertext': return $this->outertext();
  350.             case 'innertext': return $this->innertext();
  351.             case 'plaintext': return $this->text();
  352.             case 'xmltext': return $this->xmltext();
  353.             default: return array_key_exists($name, $this->attr);
  354.         }
  355.     }
  356.  
  357.     function __set($name, $value) {
  358.         switch($name) {
  359.             case 'outertext': return $this->_[HDOM_INFO_OUTER] = $value;
  360.             case 'innertext':
  361.                 if (isset($this->_[HDOM_INFO_TEXT])) return $this->_[HDOM_INFO_TEXT] = $value;
  362.                 return $this->_[HDOM_INFO_INNER] = $value;
  363.         }
  364.         if (!isset($this->attr[$name])) {
  365.             $this->_[HDOM_INFO_SPACE][] = array(' ', '', '');
  366.             $this->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE;
  367.         }
  368.         $this->attr[$name] = $value;
  369.     }
  370.  
  371.     function __isset($name) {
  372.         switch($name) {
  373.             case 'outertext': return true;
  374.             case 'innertext': return true;
  375.             case 'plaintext': return true;
  376.         }
  377.         //no value attr: nowrap, checked selected...
  378.         return (array_key_exists($name, $this->attr)) ? true : isset($this->attr[$name]);
  379.     }
  380.  
  381.     function __unset($name) {
  382.         if (isset($this->attr[$name]))
  383.             unset($this->attr[$name]);
  384.     }
  385.  
  386.     // camel naming conventions
  387.     function getAllAttributes() {return $this->attr;}
  388.     function getAttribute($name) {return $this->__get($name);}
  389.     function setAttribute($name, $value) {$this->__set($name, $value);}
  390.     function hasAttribute($name) {return $this->__isset($name);}
  391.     function removeAttribute($name) {$this->__set($name, null);}
  392.     function getElementById($id) {return $this->find("#$id", 0);}
  393.     function getElementsById($id, $idx=null) {return $this->find("#$id", $idx);}
  394.     function getElementByTagName($name) {return $this->find($name, 0);}
  395.     function getElementsByTagName($name, $idx=null) {return $this->find($name, $idx);}
  396.     function parentNode() {return $this->parent();}
  397.     function childNodes($idx=-1) {return $this->children($idx);}
  398.     function firstChild() {return $this->first_child();}
  399.     function lastChild() {return $this->last_child();}
  400.     function nextSibling() {return $this->next_sibling();}
  401.     function previousSibling() {return $this->prev_sibling();}
  402. }
  403.  
  404. ?>
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!