| 1 |
<?php |
| 2 |
/** |
| 3 |
* Website: https://sourceforge.net/projects/simplehtmldom/ |
| 4 |
* Additional projects that may be used: https://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 |
/** |
| 43 |
* All of the Defines for the classes below. |
| 44 |
* @author S.C. Chen <me578022@gmail.com> |
| 45 |
*/ |
| 46 |
define('HDOM_TYPE_ELEMENT', 1); |
| 47 |
define('HDOM_TYPE_COMMENT', 2); |
| 48 |
define('HDOM_TYPE_TEXT', 3); |
| 49 |
define('HDOM_TYPE_ENDTAG', 4); |
| 50 |
define('HDOM_TYPE_ROOT', 5); |
| 51 |
define('HDOM_TYPE_UNKNOWN', 6); |
| 52 |
define('HDOM_QUOTE_DOUBLE', 0); |
| 53 |
define('HDOM_QUOTE_SINGLE', 1); |
| 54 |
define('HDOM_QUOTE_NO', 3); |
| 55 |
define('HDOM_INFO_BEGIN', 0); |
| 56 |
define('HDOM_INFO_END', 1); |
| 57 |
define('HDOM_INFO_QUOTE', 2); |
| 58 |
define('HDOM_INFO_SPACE', 3); |
| 59 |
define('HDOM_INFO_TEXT', 4); |
| 60 |
define('HDOM_INFO_INNER', 5); |
| 61 |
define('HDOM_INFO_OUTER', 6); |
| 62 |
define('HDOM_INFO_ENDSPACE',7); |
| 63 |
define('DEFAULT_TARGET_CHARSET', 'UTF-8'); |
| 64 |
define('DEFAULT_BR_TEXT', "\r\n"); |
| 65 |
define('DEFAULT_SPAN_TEXT', " "); |
| 66 |
define('MAX_FILE_SIZE', 600000); |
| 67 |
// helper functions |
| 68 |
// ----------------------------------------------------------------------------- |
| 69 |
// get html dom from file |
| 70 |
// $maxlen is defined in the code as PHP_STREAM_COPY_ALL which is defined as -1. |
| 71 |
function file_get_html($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) |
| 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 $dom; |
| 86 |
} |
| 87 |
|
| 88 |
// get html dom from string |
| 89 |
function str_get_html($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 $dom; |
| 99 |
} |
| 100 |
|
| 101 |
// dump html dom tree |
| 102 |
function dump_html_tree($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 |
*/ |
| 115 |
class 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])) |
| 273 |
{ |
| 274 |
return $this->children[$idx]; |
| 275 |
} |
| 276 |
return null; |
| 277 |
} |
| 278 |
|
| 279 |
// returns the first child of node |
| 280 |
function first_child() |
| 281 |
{ |
| 282 |
if (count($this->children)>0) |
| 283 |
{ |
| 284 |
return $this->children[0]; |
| 285 |
} |
| 286 |
return null; |
| 287 |
} |
| 288 |
|
| 289 |
// returns the last child of node |
| 290 |
function last_child() |
| 291 |
{ |
| 292 |
if (($count=count($this->children))>0) |
| 293 |
{ |
| 294 |
return $this->children[$count-1]; |
| 295 |
} |
| 296 |
return null; |
| 297 |
} |
| 298 |
|
| 299 |
// returns the next sibling of node |
| 300 |
function next_sibling() |
| 301 |
{ |
| 302 |
if ($this->parent===null) |
| 303 |
{ |
| 304 |
return null; |
| 305 |
} |
| 306 |
|
| 307 |
$idx = 0; |
| 308 |
$count = count($this->parent->children); |
| 309 |
while ($idx<$count && $this!==$this->parent->children[$idx]) |
| 310 |
{ |
| 311 |
++$idx; |
| 312 |
} |
| 313 |
if (++$idx>=$count) |
| 314 |
{ |
| 315 |
return null; |
| 316 |
} |
| 317 |
return $this->parent->children[$idx]; |
| 318 |
} |
| 319 |
|
| 320 |
// returns the previous sibling of node |
| 321 |
function prev_sibling() |
| 322 |
{ |
| 323 |
if ($this->parent===null) return null; |
| 324 |
$idx = 0; |
| 325 |
$count = count($this->parent->children); |
| 326 |
while ($idx<$count && $this!==$this->parent->children[$idx]) |
| 327 |
++$idx; |
| 328 |
if (--$idx<0) return null; |
| 329 |
return $this->parent->children[$idx]; |
| 330 |
} |
| 331 |
|
| 332 |
// function to locate a specific ancestor tag in the path to the root. |
| 333 |
function find_ancestor_tag($tag) |
| 334 |
{ |
| 335 |
global $debug_object; |
| 336 |
if (is_object($debug_object)) { $debug_object->debug_log_entry(1); } |
| 337 |
|
| 338 |
// Start by including ourselves in the comparison. |
| 339 |
$returnDom = $this; |
| 340 |
|
| 341 |
while (!is_null($returnDom)) |
| 342 |
{ |
| 343 |
if (is_object($debug_object)) { $debug_object->debug_log(2, "Current tag is: " . $returnDom->tag); } |
| 344 |
|
| 345 |
if ($returnDom->tag == $tag) |
| 346 |
{ |
| 347 |
break; |
| 348 |
} |
| 349 |
$returnDom = $returnDom->parent; |
| 350 |
} |
| 351 |
return $returnDom; |
| 352 |
} |
| 353 |
|
| 354 |
// get dom node's inner html |
| 355 |
function innertext() |
| 356 |
{ |
| 357 |
if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER]; |
| 358 |
if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]); |
| 359 |
|
| 360 |
$ret = ''; |
| 361 |
foreach ($this->nodes as $n) |
| 362 |
$ret .= $n->outertext(); |
| 363 |
return $ret; |
| 364 |
} |
| 365 |
|
| 366 |
// get dom node's outer text (with tag) |
| 367 |
function outertext() |
| 368 |
{ |
| 369 |
global $debug_object; |
| 370 |
if (is_object($debug_object)) |
| 371 |
{ |
| 372 |
$text = ''; |
| 373 |
if ($this->tag == 'text') |
| 374 |
{ |
| 375 |
if (!empty($this->text)) |
| 376 |
{ |
| 377 |
$text = " with text: " . $this->text; |
| 378 |
} |
| 379 |
} |
| 380 |
$debug_object->debug_log(1, 'Innertext of tag: ' . $this->tag . $text); |
| 381 |
} |
| 382 |
|
| 383 |
if ($this->tag==='root') return $this->innertext(); |
| 384 |
|
| 385 |
// trigger callback |
| 386 |
if ($this->dom && $this->dom->callback!==null) |
| 387 |
{ |
| 388 |
call_user_func_array($this->dom->callback, array($this)); |
| 389 |
} |
| 390 |
|
| 391 |
if (isset($this->_[HDOM_INFO_OUTER])) return $this->_[HDOM_INFO_OUTER]; |
| 392 |
if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]); |
| 393 |
|
| 394 |
// render begin tag |
| 395 |
if ($this->dom && $this->dom->nodes[$this->_[HDOM_INFO_BEGIN]]) |
| 396 |
{ |
| 397 |
$ret = $this->dom->nodes[$this->_[HDOM_INFO_BEGIN]]->makeup(); |
| 398 |
} else { |
| 399 |
$ret = ""; |
| 400 |
} |
| 401 |
|
| 402 |
// render inner text |
| 403 |
if (isset($this->_[HDOM_INFO_INNER])) |
| 404 |
{ |
| 405 |
// If it's a br tag... don't return the HDOM_INNER_INFO that we may or may not have added. |
| 406 |
if ($this->tag != "br") |
| 407 |
{ |
| 408 |
$ret .= $this->_[HDOM_INFO_INNER]; |
| 409 |
} |
| 410 |
} else { |
| 411 |
if ($this->nodes) |
| 412 |
{ |
| 413 |
foreach ($this->nodes as $n) |
| 414 |
{ |
| 415 |
$ret .= $this->convert_text($n->outertext()); |
| 416 |
} |
| 417 |
} |
| 418 |
} |
| 419 |
|
| 420 |
// render end tag |
| 421 |
if (isset($this->_[HDOM_INFO_END]) && $this->_[HDOM_INFO_END]!=0) |
| 422 |
$ret .= '</'.$this->tag.'>'; |
| 423 |
return $ret; |
| 424 |
} |
| 425 |
|
| 426 |
// get dom node's plain text |
| 427 |
function text() |
| 428 |
{ |
| 429 |
if (isset($this->_[HDOM_INFO_INNER])) return $this->_[HDOM_INFO_INNER]; |
| 430 |
switch ($this->nodetype) |
| 431 |
{ |
| 432 |
case HDOM_TYPE_TEXT: return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]); |
| 433 |
case HDOM_TYPE_COMMENT: return ''; |
| 434 |
case HDOM_TYPE_UNKNOWN: return ''; |
| 435 |
} |
| 436 |
if (strcasecmp($this->tag, 'script')===0) return ''; |
| 437 |
if (strcasecmp($this->tag, 'style')===0) return ''; |
| 438 |
|
| 439 |
$ret = ''; |
| 440 |
// In rare cases, (always node type 1 or HDOM_TYPE_ELEMENT - observed for some span tags, and some p tags) $this->nodes is set to NULL. |
| 441 |
// NOTE: This indicates that there is a problem where it's set to NULL without a clear happening. |
| 442 |
// WHY is this happening? |
| 443 |
if (!is_null($this->nodes)) |
| 444 |
{ |
| 445 |
foreach ($this->nodes as $n) |
| 446 |
{ |
| 447 |
$ret .= $this->convert_text($n->text()); |
| 448 |
} |
| 449 |
|
| 450 |
// 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. |
| 451 |
if ($this->tag == "span") |
| 452 |
{ |
| 453 |
$ret .= $this->dom->default_span_text; |
| 454 |
} |
| 455 |
|
| 456 |
|
| 457 |
} |
| 458 |
return $ret; |
| 459 |
} |
| 460 |
|
| 461 |
function xmltext() |
| 462 |
{ |
| 463 |
$ret = $this->innertext(); |
| 464 |
$ret = str_ireplace('<![CDATA[', '', $ret); |
| 465 |
$ret = str_replace(']]>', '', $ret); |
| 466 |
return $ret; |
| 467 |
} |
| 468 |
|
| 469 |
// build node's text with tag |
| 470 |
function makeup() |
| 471 |
{ |
| 472 |
// text, comment, unknown |
| 473 |
if (isset($this->_[HDOM_INFO_TEXT])) return $this->dom->restore_noise($this->_[HDOM_INFO_TEXT]); |
| 474 |
|
| 475 |
$ret = '<'.$this->tag; |
| 476 |
$i = -1; |
| 477 |
|
| 478 |
foreach ($this->attr as $key=>$val) |
| 479 |
{ |
| 480 |
++$i; |
| 481 |
|
| 482 |
// skip removed attribute |
| 483 |
if ($val===null || $val===false) |
| 484 |
continue; |
| 485 |
|
| 486 |
$ret .= $this->_[HDOM_INFO_SPACE][$i][0]; |
| 487 |
//no value attr: nowrap, checked selected... |
| 488 |
if ($val===true) |
| 489 |
$ret .= $key; |
| 490 |
else { |
| 491 |
switch ($this->_[HDOM_INFO_QUOTE][$i]) |
| 492 |
{ |
| 493 |
case HDOM_QUOTE_DOUBLE: $quote = '"'; break; |
| 494 |
case HDOM_QUOTE_SINGLE: $quote = '\''; break; |
| 495 |
default: $quote = ''; |
| 496 |
} |
| 497 |
$ret .= $key.$this->_[HDOM_INFO_SPACE][$i][1].'='.$this->_[HDOM_INFO_SPACE][$i][2].$quote.$val.$quote; |
| 498 |
} |
| 499 |
} |
| 500 |
$ret = $this->dom->restore_noise($ret); |
| 501 |
return $ret . $this->_[HDOM_INFO_ENDSPACE] . '>'; |
| 502 |
} |
| 503 |
|
| 504 |
// find elements by css selector |
| 505 |
//PaperG - added ability for find to lowercase the value of the selector. |
| 506 |
function find($selector, $idx=null, $lowercase=false) |
| 507 |
{ |
| 508 |
$selectors = $this->parse_selector($selector); |
| 509 |
if (($count=count($selectors))===0) return array(); |
| 510 |
$found_keys = array(); |
| 511 |
|
| 512 |
// find each selector |
| 513 |
for ($c=0; $c<$count; ++$c) |
| 514 |
{ |
| 515 |
// The change on the below line was documented on the sourceforge code tracker id 2788009 |
| 516 |
// used to be: if (($levle=count($selectors[0]))===0) return array(); |
| 517 |
if (($levle=count($selectors[$c]))===0) return array(); |
| 518 |
if (!isset($this->_[HDOM_INFO_BEGIN])) return array(); |
| 519 |
|
| 520 |
$head = array($this->_[HDOM_INFO_BEGIN]=>1); |
| 521 |
|
| 522 |
// handle descendant selectors, no recursive! |
| 523 |
for ($l=0; $l<$levle; ++$l) |
| 524 |
{ |
| 525 |
$ret = array(); |
| 526 |
foreach ($head as $k=>$v) |
| 527 |
{ |
| 528 |
$n = ($k===-1) ? $this->dom->root : $this->dom->nodes[$k]; |
| 529 |
//PaperG - Pass this optional parameter on to the seek function. |
| 530 |
$n->seek($selectors[$c][$l], $ret, $lowercase); |
| 531 |
} |
| 532 |
$head = $ret; |
| 533 |
} |
| 534 |
|
| 535 |
foreach ($head as $k=>$v) |
| 536 |
{ |
| 537 |
if (!isset($found_keys[$k])) |
| 538 |
{ |
| 539 |
$found_keys[$k] = 1; |
| 540 |
} |
| 541 |
} |
| 542 |
} |
| 543 |
|
| 544 |
// sort keys |
| 545 |
ksort($found_keys); |
| 546 |
|
| 547 |
$found = array(); |
| 548 |
foreach ($found_keys as $k=>$v) |
| 549 |
$found[] = $this->dom->nodes[$k]; |
| 550 |
|
| 551 |
// return nth-element or array |
| 552 |
if (is_null($idx)) return $found; |
| 553 |
else if ($idx<0) $idx = count($found) + $idx; |
| 554 |
return (isset($found[$idx])) ? $found[$idx] : null; |
| 555 |
} |
| 556 |
|
| 557 |
// seek for given conditions |
| 558 |
// PaperG - added parameter to allow for case insensitive testing of the value of a selector. |
| 559 |
protected function seek($selector, &$ret, $lowercase=false) |
| 560 |
{ |
| 561 |
global $debug_object; |
| 562 |
if (is_object($debug_object)) { $debug_object->debug_log_entry(1); } |
| 563 |
|
| 564 |
list($tag, $key, $val, $exp, $no_key) = $selector; |
| 565 |
|
| 566 |
// xpath index |
| 567 |
if ($tag && $key && is_numeric($key)) |
| 568 |
{ |
| 569 |
$count = 0; |
| 570 |
foreach ($this->children as $c) |
| 571 |
{ |
| 572 |
if ($tag==='*' || $tag===$c->tag) { |
| 573 |
if (++$count==$key) { |
| 574 |
$ret[$c->_[HDOM_INFO_BEGIN]] = 1; |
| 575 |
return; |
| 576 |
} |
| 577 |
} |
| 578 |
} |
| 579 |
return; |
| 580 |
} |
| 581 |
|
| 582 |
$end = (!empty($this->_[HDOM_INFO_END])) ? $this->_[HDOM_INFO_END] : 0; |
| 583 |
if ($end==0) { |
| 584 |
$parent = $this->parent; |
| 585 |
while (!isset($parent->_[HDOM_INFO_END]) && $parent!==null) { |
| 586 |
$end -= 1; |
| 587 |
$parent = $parent->parent; |
| 588 |
} |
| 589 |
$end += $parent->_[HDOM_INFO_END]; |
| 590 |
} |
| 591 |
|
| 592 |
for ($i=$this->_[HDOM_INFO_BEGIN]+1; $i<$end; ++$i) { |
| 593 |
$node = $this->dom->nodes[$i]; |
| 594 |
|
| 595 |
$pass = true; |
| 596 |
|
| 597 |
if ($tag==='*' && !$key) { |
| 598 |
if (in_array($node, $this->children, true)) |
| 599 |
$ret[$i] = 1; |
| 600 |
continue; |
| 601 |
} |
| 602 |
|
| 603 |
// compare tag |
| 604 |
if ($tag && $tag!=$node->tag && $tag!=='*') {$pass=false;} |
| 605 |
// compare key |
| 606 |
if ($pass && $key) { |
| 607 |
if ($no_key) { |
| 608 |
if (isset($node->attr[$key])) $pass=false; |
| 609 |
} else { |
| 610 |
if (($key != "plaintext") && !isset($node->attr[$key])) $pass=false; |
| 611 |
} |
| 612 |
} |
| 613 |
// compare value |
| 614 |
if ($pass && $key && $val && $val!=='*') { |
| 615 |
// If they have told us that this is a "plaintext" search then we want the plaintext of the node - right? |
| 616 |
if ($key == "plaintext") { |
| 617 |
// $node->plaintext actually returns $node->text(); |
| 618 |
$nodeKeyValue = $node->text(); |
| 619 |
} else { |
| 620 |
// this is a normal search, we want the value of that attribute of the tag. |
| 621 |
$nodeKeyValue = $node->attr[$key]; |
| 622 |
} |
| 623 |
if (is_object($debug_object)) {$debug_object->debug_log(2, "testing node: " . $node->tag . " for attribute: " . $key . $exp . $val . " where nodes value is: " . $nodeKeyValue);} |
| 624 |
|
| 625 |
//PaperG - If lowercase is set, do a case insensitive test of the value of the selector. |
| 626 |
if ($lowercase) { |
| 627 |
$check = $this->match($exp, strtolower($val), strtolower($nodeKeyValue)); |
| 628 |
} else { |
| 629 |
$check = $this->match($exp, $val, $nodeKeyValue); |
| 630 |
} |
| 631 |
if (is_object($debug_object)) {$debug_object->debug_log(2, "after match: " . ($check ? "true" : "false"));} |
| 632 |
|
| 633 |
// handle multiple class |
| 634 |
if (!$check && strcasecmp($key, 'class')===0) { |
| 635 |
foreach (explode(' ',$node->attr[$key]) as $k) { |
| 636 |
// Without this, there were cases where leading, trailing, or double spaces lead to our comparing blanks - bad form. |
| 637 |
if (!empty($k)) { |
| 638 |
if ($lowercase) { |
| 639 |
$check = $this->match($exp, strtolower($val), strtolower($k)); |
| 640 |
} else { |
| 641 |
$check = $this->match($exp, $val, $k); |
| 642 |
} |
| 643 |
if ($check) break; |
| 644 |
} |
| 645 |
} |
| 646 |
} |
| 647 |
if (!$check) $pass = false; |
| 648 |
} |
| 649 |
if ($pass) $ret[$i] = 1; |
| 650 |
unset($node); |
| 651 |
} |
| 652 |
// It's passed by reference so this is actually what this function returns. |
| 653 |
if (is_object($debug_object)) {$debug_object->debug_log(1, "EXIT - ret: ", $ret);} |
| 654 |
} |
| 655 |
|
| 656 |
protected function match($exp, $pattern, $value) { |
| 657 |
global $debug_object; |
| 658 |
if (is_object($debug_object)) {$debug_object->debug_log_entry(1);} |
| 659 |
|
| 660 |
switch ($exp) { |
| 661 |
case '=': |
| 662 |
return ($value===$pattern); |
| 663 |
case '!=': |
| 664 |
return ($value!==$pattern); |
| 665 |
case '^=': |
| 666 |
return preg_match("/^".preg_quote($pattern,'/')."/", $value); |
| 667 |
case '$=': |
| 668 |
return preg_match("/".preg_quote($pattern,'/')."$/", $value); |
| 669 |
case '*=': |
| 670 |
if ($pattern[0]=='/') { |
| 671 |
return preg_match($pattern, $value); |
| 672 |
} |
| 673 |
return preg_match("/".$pattern."/i", $value); |
| 674 |
} |
| 675 |
return false; |
| 676 |
} |
| 677 |
|
| 678 |
protected function parse_selector($selector_string) { |
| 679 |
global $debug_object; |
| 680 |
if (is_object($debug_object)) {$debug_object->debug_log_entry(1);} |
| 681 |
|
| 682 |
// pattern of CSS selectors, modified from mootools |
| 683 |
// Paperg: Add the colon to the attrbute, so that it properly finds <tag attr:ibute="something" > like google does. |
| 684 |
// Note: if you try to look at this attribute, yo MUST use getAttribute since $dom->x:y will fail the php syntax check. |
| 685 |
// Notice the \[ starting the attbute? and the @? following? This implies that an attribute can begin with an @ sign that is not captured. |
| 686 |
// This implies that an html attribute specifier may start with an @ sign that is NOT captured by the expression. |
| 687 |
// farther study is required to determine of this should be documented or removed. |
| 688 |
// $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; |
| 689 |
// $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; |
| 690 |
$pattern = "/([\w\-:\*]*)(?:\#([\w\-]+)|\.([\w\-]+))?(?:\[@?(!?[\w\-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; |
| 691 |
preg_match_all($pattern, trim($selector_string).' ', $matches, PREG_SET_ORDER); |
| 692 |
if (is_object($debug_object)) {$debug_object->debug_log(2, "Matches Array: ", $matches);} |
| 693 |
|
| 694 |
$selectors = array(); |
| 695 |
$result = array(); |
| 696 |
//print_r($matches); |
| 697 |
|
| 698 |
foreach ($matches as $m) { |
| 699 |
$m[0] = trim($m[0]); |
| 700 |
if ($m[0]==='' || $m[0]==='/' || $m[0]==='//') continue; |
| 701 |
// for browser generated xpath |
| 702 |
if ($m[1]==='tbody') continue; |
| 703 |
|
| 704 |
list($tag, $key, $val, $exp, $no_key) = array($m[1], null, null, '=', false); |
| 705 |
if (!empty($m[2])) {$key='id'; $val=$m[2];} |
| 706 |
if (!empty($m[3])) {$key='class'; $val=$m[3];} |
| 707 |
if (!empty($m[4])) {$key=$m[4];} |
| 708 |
if (!empty($m[5])) {$exp=$m[5];} |
| 709 |
if (!empty($m[6])) {$val=$m[6];} |
| 710 |
|
| 711 |
// convert to lowercase |
| 712 |
if ($this->dom->lowercase) {$tag=strtolower($tag); $key=strtolower($key);} |
| 713 |
//elements that do NOT have the specified attribute |
| 714 |
if (isset($key[0]) && $key[0]==='!') {$key=substr($key, 1); $no_key=true;} |
| 715 |
|
| 716 |
$result[] = array($tag, $key, $val, $exp, $no_key); |
| 717 |
if (trim($m[7])===',') { |
| 718 |
$selectors[] = $result; |
| 719 |
$result = array(); |
| 720 |
} |
| 721 |
} |
| 722 |
if (count($result)>0) |
| 723 |
$selectors[] = $result; |
| 724 |
return $selectors; |
| 725 |
} |
| 726 |
|
| 727 |
function __get($name) |
| 728 |
{ |
| 729 |
if (isset($this->attr[$name])) |
| 730 |
{ |
| 731 |
return $this->convert_text($this->attr[$name]); |
| 732 |
} |
| 733 |
switch ($name) |
| 734 |
{ |
| 735 |
case 'outertext': return $this->outertext(); |
| 736 |
case 'innertext': return $this->innertext(); |
| 737 |
case 'plaintext': return $this->text(); |
| 738 |
case 'xmltext': return $this->xmltext(); |
| 739 |
default: return array_key_exists($name, $this->attr); |
| 740 |
} |
| 741 |
} |
| 742 |
|
| 743 |
function __set($name, $value) |
| 744 |
{ |
| 745 |
global $debug_object; |
| 746 |
if (is_object($debug_object)) {$debug_object->debug_log_entry(1);} |
| 747 |
|
| 748 |
switch ($name) |
| 749 |
{ |
| 750 |
case 'outertext': return $this->_[HDOM_INFO_OUTER] = $value; |
| 751 |
case 'innertext': |
| 752 |
if (isset($this->_[HDOM_INFO_TEXT])) return $this->_[HDOM_INFO_TEXT] = $value; |
| 753 |
return $this->_[HDOM_INFO_INNER] = $value; |
| 754 |
} |
| 755 |
if (!isset($this->attr[$name])) |
| 756 |
{ |
| 757 |
$this->_[HDOM_INFO_SPACE][] = array(' ', '', ''); |
| 758 |
$this->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE; |
| 759 |
} |
| 760 |
$this->attr[$name] = $value; |
| 761 |
} |
| 762 |
|
| 763 |
function __isset($name) |
| 764 |
{ |
| 765 |
switch ($name) |
| 766 |
{ |
| 767 |
case 'outertext': return true; |
| 768 |
case 'innertext': return true; |
| 769 |
case 'plaintext': return true; |
| 770 |
} |
| 771 |
//no value attr: nowrap, checked selected... |
| 772 |
return (array_key_exists($name, $this->attr)) ? true : isset($this->attr[$name]); |
| 773 |
} |
| 774 |
|
| 775 |
function __unset($name) { |
| 776 |
if (isset($this->attr[$name])) |
| 777 |
unset($this->attr[$name]); |
| 778 |
} |
| 779 |
|
| 780 |
// PaperG - Function to convert the text from one character set to another if the two sets are not the same. |
| 781 |
function convert_text($text) |
| 782 |
{ |
| 783 |
global $debug_object; |
| 784 |
if (is_object($debug_object)) {$debug_object->debug_log_entry(1);} |
| 785 |
|
| 786 |
$converted_text = $text; |
| 787 |
|
| 788 |
$sourceCharset = ""; |
| 789 |
$targetCharset = ""; |
| 790 |
|
| 791 |
if ($this->dom) |
| 792 |
{ |
| 793 |
$sourceCharset = strtoupper($this->dom->_charset); |
| 794 |
$targetCharset = strtoupper($this->dom->_target_charset); |
| 795 |
} |
| 796 |
if (is_object($debug_object)) {$debug_object->debug_log(3, "source charset: " . $sourceCharset . " target charaset: " . $targetCharset);} |
| 797 |
|
| 798 |
if (!empty($sourceCharset) && !empty($targetCharset) && (strcasecmp($sourceCharset, $targetCharset) != 0)) |
| 799 |
{ |
| 800 |
// Check if the reported encoding could have been incorrect and the text is actually already UTF-8 |
| 801 |
if ((strcasecmp($targetCharset, 'UTF-8') == 0) && ($this->is_utf8($text))) |
| 802 |
{ |
| 803 |
$converted_text = $text; |
| 804 |
} |
| 805 |
else |
| 806 |
{ |
| 807 |
$converted_text = iconv($sourceCharset, $targetCharset, $text); |
| 808 |
} |
| 809 |
} |
| 810 |
|
| 811 |
// Lets make sure that we don't have that silly BOM issue with any of the utf-8 text we output. |
| 812 |
if ($targetCharset == 'UTF-8') |
| 813 |
{ |
| 814 |
if (substr($converted_text, 0, 3) == "\xef\xbb\xbf") |
| 815 |
{ |
| 816 |
$converted_text = substr($converted_text, 3); |
| 817 |
} |
| 818 |
if (substr($converted_text, -3) == "\xef\xbb\xbf") |
| 819 |
{ |
| 820 |
$converted_text = substr($converted_text, 0, -3); |
| 821 |
} |
| 822 |
} |
| 823 |
|
| 824 |
return $converted_text; |
| 825 |
} |
| 826 |
|
| 827 |
/** |
| 828 |
* Returns true if $string is valid UTF-8 and false otherwise. |
| 829 |
* |
| 830 |
* @param mixed $str String to be tested |
| 831 |
* @return boolean |
| 832 |
*/ |
| 833 |
static function is_utf8($str) |
| 834 |
{ |
| 835 |
$c=0; $b=0; |
| 836 |
$bits=0; |
| 837 |
$len=strlen($str); |
| 838 |
for($i=0; $i<$len; $i++) |
| 839 |
{ |
| 840 |
$c=ord($str[$i]); |
| 841 |
if($c > 128) |
| 842 |
{ |
| 843 |
if(($c >= 254)) return false; |
| 844 |
elseif($c >= 252) $bits=6; |
| 845 |
elseif($c >= 248) $bits=5; |
| 846 |
elseif($c >= 240) $bits=4; |
| 847 |
elseif($c >= 224) $bits=3; |
| 848 |
elseif($c >= 192) $bits=2; |
| 849 |
else return false; |
| 850 |
if(($i+$bits) > $len) return false; |
| 851 |
while($bits > 1) |
| 852 |
{ |
| 853 |
$i++; |
| 854 |
$b=ord($str[$i]); |
| 855 |
if($b < 128 || $b > 191) return false; |
| 856 |
$bits--; |
| 857 |
} |
| 858 |
} |
| 859 |
} |
| 860 |
return true; |
| 861 |
} |
| 862 |
/* |
| 863 |
function is_utf8($string) |
| 864 |
{ |
| 865 |
//this is buggy |
| 866 |
return (utf8_encode(utf8_decode($string)) == $string); |
| 867 |
} |
| 868 |
*/ |
| 869 |
|
| 870 |
/** |
| 871 |
* Function to try a few tricks to determine the displayed size of an img on the page. |
| 872 |
* NOTE: This will ONLY work on an IMG tag. Returns FALSE on all other tag types. |
| 873 |
* |
| 874 |
* @author John Schlick |
| 875 |
* @version April 19 2012 |
| 876 |
* @return array an array containing the 'height' and 'width' of the image on the page or -1 if we can't figure it out. |
| 877 |
*/ |
| 878 |
function get_display_size() |
| 879 |
{ |
| 880 |
global $debug_object; |
| 881 |
|
| 882 |
$width = -1; |
| 883 |
$height = -1; |
| 884 |
|
| 885 |
if ($this->tag !== 'img') |
| 886 |
{ |
| 887 |
return false; |
| 888 |
} |
| 889 |
|
| 890 |
// See if there is aheight or width attribute in the tag itself. |
| 891 |
if (isset($this->attr['width'])) |
| 892 |
{ |
| 893 |
$width = $this->attr['width']; |
| 894 |
} |
| 895 |
|
| 896 |
if (isset($this->attr['height'])) |
| 897 |
{ |
| 898 |
$height = $this->attr['height']; |
| 899 |
} |
| 900 |
|
| 901 |
// Now look for an inline style. |
| 902 |
if (isset($this->attr['style'])) |
| 903 |
{ |
| 904 |
// Thanks to user gnarf from stackoverflow for this regular expression. |
| 905 |
$attributes = array(); |
| 906 |
preg_match_all("/([\w-]+)\s*:\s*([^;]+)\s*;?/", $this->attr['style'], $matches, PREG_SET_ORDER); |
| 907 |
foreach ($matches as $match) { |
| 908 |
$attributes[$match[1]] = $match[2]; |
| 909 |
} |
| 910 |
|
| 911 |
// If there is a width in the style attributes: |
| 912 |
if (isset($attributes['width']) && $width == -1) |
| 913 |
{ |
| 914 |
// check that the last two characters are px (pixels) |
| 915 |
if (strtolower(substr($attributes['width'], -2)) == 'px') |
| 916 |
{ |
| 917 |
$proposed_width = substr($attributes['width'], 0, -2); |
| 918 |
// Now make sure that it's an integer and not something stupid. |
| 919 |
if (filter_var($proposed_width, FILTER_VALIDATE_INT)) |
| 920 |
{ |
| 921 |
$width = $proposed_width; |
| 922 |
} |
| 923 |
} |
| 924 |
} |
| 925 |
|
| 926 |
// If there is a width in the style attributes: |
| 927 |
if (isset($attributes['height']) && $height == -1) |
| 928 |
{ |
| 929 |
// check that the last two characters are px (pixels) |
| 930 |
if (strtolower(substr($attributes['height'], -2)) == 'px') |
| 931 |
{ |
| 932 |
$proposed_height = substr($attributes['height'], 0, -2); |
| 933 |
// Now make sure that it's an integer and not something stupid. |
| 934 |
if (filter_var($proposed_height, FILTER_VALIDATE_INT)) |
| 935 |
{ |
| 936 |
$height = $proposed_height; |
| 937 |
} |
| 938 |
} |
| 939 |
} |
| 940 |
|
| 941 |
} |
| 942 |
|
| 943 |
// Future enhancement: |
| 944 |
// Look in the tag to see if there is a class or id specified that has a height or width attribute to it. |
| 945 |
|
| 946 |
// Far future enhancement |
| 947 |
// 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 |
| 948 |
// Note that in this case, the class or id will have the img subselector for it to apply to the image. |
| 949 |
|
| 950 |
// ridiculously far future development |
| 951 |
// 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. |
| 952 |
|
| 953 |
$result = array('height' => $height, |
| 954 |
'width' => $width); |
| 955 |
return $result; |
| 956 |
} |
| 957 |
|
| 958 |
// camel naming conventions |
| 959 |
function getAllAttributes() {return $this->attr;} |
| 960 |
function getAttribute($name) {return $this->__get($name);} |
| 961 |
function setAttribute($name, $value) {$this->__set($name, $value);} |
| 962 |
function hasAttribute($name) {return $this->__isset($name);} |
| 963 |
function removeAttribute($name) {$this->__set($name, null);} |
| 964 |
function getElementById($id) {return $this->find("#$id", 0);} |
| 965 |
function getElementsById($id, $idx=null) {return $this->find("#$id", $idx);} |
| 966 |
function getElementByTagName($name) {return $this->find($name, 0);} |
| 967 |
function getElementsByTagName($name, $idx=null) {return $this->find($name, $idx);} |
| 968 |
function parentNode() {return $this->parent();} |
| 969 |
function childNodes($idx=-1) {return $this->children($idx);} |
| 970 |
function firstChild() {return $this->first_child();} |
| 971 |
function lastChild() {return $this->last_child();} |
| 972 |
function nextSibling() {return $this->next_sibling();} |
| 973 |
function previousSibling() {return $this->prev_sibling();} |
| 974 |
function hasChildNodes() {return $this->has_child();} |
| 975 |
function nodeName() {return $this->tag;} |
| 976 |
function appendChild($node) {$node->parent($this); return $node;} |
| 977 |
|
| 978 |
} |
| 979 |
|
| 980 |
/** |
| 981 |
* simple html dom parser |
| 982 |
* Paperg - in the find routine: allow us to specify that we want case insensitive testing of the value of the selector. |
| 983 |
* Paperg - change $size from protected to public so we can easily access it |
| 984 |
* Paperg - added ForceTagsClosed in the constructor which tells us whether we trust the html or not. Default is to NOT trust it. |
| 985 |
* |
| 986 |
* @package PlaceLocalInclude |
| 987 |
*/ |
| 988 |
class simple_html_dom |
| 989 |
{ |
| 990 |
public $root = null; |
| 991 |
public $nodes = array(); |
| 992 |
public $callback = null; |
| 993 |
public $lowercase = false; |
| 994 |
// Used to keep track of how large the text was when we started. |
| 995 |
public $original_size; |
| 996 |
public $size; |
| 997 |
protected $pos; |
| 998 |
protected $doc; |
| 999 |
protected $char; |
| 1000 |
protected $cursor; |
| 1001 |
protected $parent; |
| 1002 |
protected $noise = array(); |
| 1003 |
protected $token_blank = " \t\r\n"; |
| 1004 |
protected $token_equal = ' =/>'; |
| 1005 |
protected $token_slash = " />\r\n\t"; |
| 1006 |
protected $token_attr = ' >'; |
| 1007 |
// Note that this is referenced by a child node, and so it needs to be public for that node to see this information. |
| 1008 |
public $_charset = ''; |
| 1009 |
public $_target_charset = ''; |
| 1010 |
protected $default_br_text = ""; |
| 1011 |
public $default_span_text = ""; |
| 1012 |
|
| 1013 |
// use isset instead of in_array, performance boost about 30%... |
| 1014 |
protected $self_closing_tags = array('img'=>1, 'br'=>1, 'input'=>1, 'meta'=>1, 'link'=>1, 'hr'=>1, 'base'=>1, 'embed'=>1, 'spacer'=>1); |
| 1015 |
protected $block_tags = array('root'=>1, 'body'=>1, 'form'=>1, 'div'=>1, 'span'=>1, 'table'=>1); |
| 1016 |
// Known sourceforge issue #2977341 |
| 1017 |
// B tags that are not closed cause us to return everything to the end of the document. |
| 1018 |
protected $optional_closing_tags = array( |
| 1019 |
'tr'=>array('tr'=>1, 'td'=>1, 'th'=>1), |
| 1020 |
'th'=>array('th'=>1), |
| 1021 |
'td'=>array('td'=>1), |
| 1022 |
'li'=>array('li'=>1), |
| 1023 |
'dt'=>array('dt'=>1, 'dd'=>1), |
| 1024 |
'dd'=>array('dd'=>1, 'dt'=>1), |
| 1025 |
'dl'=>array('dd'=>1, 'dt'=>1), |
| 1026 |
'p'=>array('p'=>1), |
| 1027 |
'nobr'=>array('nobr'=>1), |
| 1028 |
'b'=>array('b'=>1), |
| 1029 |
'option'=>array('option'=>1), |
| 1030 |
); |
| 1031 |
|
| 1032 |
function __construct($str=null, $lowercase=true, $forceTagsClosed=true, $target_charset=DEFAULT_TARGET_CHARSET, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT) |
| 1033 |
{ |
| 1034 |
if ($str) |
| 1035 |
{ |
| 1036 |
if (preg_match("/^https:\/\//i",$str) || is_file($str)) |
| 1037 |
{ |
| 1038 |
$this->load_file($str); |
| 1039 |
} |
| 1040 |
else |
| 1041 |
{ |
| 1042 |
$this->load($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText); |
| 1043 |
} |
| 1044 |
} |
| 1045 |
// Forcing tags to be closed implies that we don't trust the html, but it can lead to parsing errors if we SHOULD trust the html. |
| 1046 |
if (!$forceTagsClosed) { |
| 1047 |
$this->optional_closing_array=array(); |
| 1048 |
} |
| 1049 |
$this->_target_charset = $target_charset; |
| 1050 |
} |
| 1051 |
|
| 1052 |
function __destruct() |
| 1053 |
{ |
| 1054 |
$this->clear(); |
| 1055 |
} |
| 1056 |
|
| 1057 |
// load html from string |
| 1058 |
function load($str, $lowercase=true, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT) |
| 1059 |
{ |
| 1060 |
global $debug_object; |
| 1061 |
|
| 1062 |
// prepare |
| 1063 |
$this->prepare($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText); |
| 1064 |
// strip out cdata |
| 1065 |
$this->remove_noise("'<!\[CDATA\[(.*?)\]\]>'is", true); |
| 1066 |
// strip out comments |
| 1067 |
$this->remove_noise("'<!--(.*?)-->'is"); |
| 1068 |
// Per sourceforge https://sourceforge.net/tracker/?func=detail&aid=2949097&group_id=218559&atid=1044037 |
| 1069 |
// Script tags removal now preceeds style tag removal. |
| 1070 |
// strip out <script> tags |
| 1071 |
$this->remove_noise("'<\s*script[^>]*[^/]>(.*?)<\s*/\s*script\s*>'is"); |
| 1072 |
$this->remove_noise("'<\s*script\s*>(.*?)<\s*/\s*script\s*>'is"); |
| 1073 |
// strip out <style> tags |
| 1074 |
$this->remove_noise("'<\s*style[^>]*[^/]>(.*?)<\s*/\s*style\s*>'is"); |
| 1075 |
$this->remove_noise("'<\s*style\s*>(.*?)<\s*/\s*style\s*>'is"); |
| 1076 |
// strip out preformatted tags |
| 1077 |
$this->remove_noise("'<\s*(?:code)[^>]*>(.*?)<\s*/\s*(?:code)\s*>'is"); |
| 1078 |
// strip out server side scripts |
| 1079 |
$this->remove_noise("'(<\?)(.*?)(\?>)'s", true); |
| 1080 |
// strip smarty scripts |
| 1081 |
$this->remove_noise("'(\{\w)(.*?)(\})'s", true); |
| 1082 |
|
| 1083 |
// parsing |
| 1084 |
while ($this->parse()); |
| 1085 |
// end |
| 1086 |
$this->root->_[HDOM_INFO_END] = $this->cursor; |
| 1087 |
$this->parse_charset(); |
| 1088 |
|
| 1089 |
// make load function chainable |
| 1090 |
return $this; |
| 1091 |
|
| 1092 |
} |
| 1093 |
|
| 1094 |
// load html from file |
| 1095 |
function load_file() |
| 1096 |
{ |
| 1097 |
$args = func_get_args(); |
| 1098 |
$this->load(call_user_func_array('file_get_contents', $args), true); |
| 1099 |
// Throw an error if we can't properly load the dom. |
| 1100 |
if (($error=error_get_last())!==null) { |
| 1101 |
$this->clear(); |
| 1102 |
return false; |
| 1103 |
} |
| 1104 |
} |
| 1105 |
|
| 1106 |
// set callback function |
| 1107 |
function set_callback($function_name) |
| 1108 |
{ |
| 1109 |
$this->callback = $function_name; |
| 1110 |
} |
| 1111 |
|
| 1112 |
// remove callback function |
| 1113 |
function remove_callback() |
| 1114 |
{ |
| 1115 |
$this->callback = null; |
| 1116 |
} |
| 1117 |
|
| 1118 |
// save dom as string |
| 1119 |
function save($filepath='') |
| 1120 |
{ |
| 1121 |
$ret = $this->root->innertext(); |
| 1122 |
if ($filepath!=='') file_put_contents($filepath, $ret, LOCK_EX); |
| 1123 |
return $ret; |
| 1124 |
} |
| 1125 |
|
| 1126 |
// find dom node by css selector |
| 1127 |
// Paperg - allow us to specify that we want case insensitive testing of the value of the selector. |
| 1128 |
function find($selector, $idx=null, $lowercase=false) |
| 1129 |
{ |
| 1130 |
return $this->root->find($selector, $idx, $lowercase); |
| 1131 |
} |
| 1132 |
|
| 1133 |
// clean up memory due to php5 circular references memory leak... |
| 1134 |
function clear() |
| 1135 |
{ |
| 1136 |
foreach ($this->nodes as $n) {$n->clear(); $n = null;} |
| 1137 |
// This add next line is documented in the sourceforge repository. 2977248 as a fix for ongoing memory leaks that occur even with the use of clear. |
| 1138 |
if (isset($this->children)) foreach ($this->children as $n) {$n->clear(); $n = null;} |
| 1139 |
if (isset($this->parent)) {$this->parent->clear(); unset($this->parent);} |
| 1140 |
if (isset($this->root)) {$this->root->clear(); unset($this->root);} |
| 1141 |
unset($this->doc); |
| 1142 |
unset($this->noise); |
| 1143 |
} |
| 1144 |
|
| 1145 |
function dump($show_attr=true) |
| 1146 |
{ |
| 1147 |
$this->root->dump($show_attr); |
| 1148 |
} |
| 1149 |
|
| 1150 |
// prepare HTML data and init everything |
| 1151 |
protected function prepare($str, $lowercase=true, $stripRN=true, $defaultBRText=DEFAULT_BR_TEXT, $defaultSpanText=DEFAULT_SPAN_TEXT) |
| 1152 |
{ |
| 1153 |
$this->clear(); |
| 1154 |
|
| 1155 |
// set the length of content before we do anything to it. |
| 1156 |
$this->size = strlen($str); |
| 1157 |
// Save the original size of the html that we got in. It might be useful to someone. |
| 1158 |
$this->original_size = $this->size; |
| 1159 |
|
| 1160 |
//before we save the string as the doc... strip out the \r \n's if we are told to. |
| 1161 |
if ($stripRN) { |
| 1162 |
$str = str_replace("\r", " ", $str); |
| 1163 |
$str = str_replace("\n", " ", $str); |
| 1164 |
|
| 1165 |
// set the length of content since we have changed it. |
| 1166 |
$this->size = strlen($str); |
| 1167 |
} |
| 1168 |
|
| 1169 |
$this->doc = $str; |
| 1170 |
$this->pos = 0; |
| 1171 |
$this->cursor = 1; |
| 1172 |
$this->noise = array(); |
| 1173 |
$this->nodes = array(); |
| 1174 |
$this->lowercase = $lowercase; |
| 1175 |
$this->default_br_text = $defaultBRText; |
| 1176 |
$this->default_span_text = $defaultSpanText; |
| 1177 |
$this->root = new simple_html_dom_node($this); |
| 1178 |
$this->root->tag = 'root'; |
| 1179 |
$this->root->_[HDOM_INFO_BEGIN] = -1; |
| 1180 |
$this->root->nodetype = HDOM_TYPE_ROOT; |
| 1181 |
$this->parent = $this->root; |
| 1182 |
if ($this->size>0) $this->char = $this->doc[0]; |
| 1183 |
} |
| 1184 |
|
| 1185 |
// parse html content |
| 1186 |
protected function parse() |
| 1187 |
{ |
| 1188 |
if (($s = $this->copy_until_char('<'))==='') |
| 1189 |
{ |
| 1190 |
return $this->read_tag(); |
| 1191 |
} |
| 1192 |
|
| 1193 |
// text |
| 1194 |
$node = new simple_html_dom_node($this); |
| 1195 |
++$this->cursor; |
| 1196 |
$node->_[HDOM_INFO_TEXT] = $s; |
| 1197 |
$this->link_nodes($node, false); |
| 1198 |
return true; |
| 1199 |
} |
| 1200 |
|
| 1201 |
// 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. |
| 1202 |
// 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 |
| 1203 |
// (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. |
| 1204 |
protected function parse_charset() |
| 1205 |
{ |
| 1206 |
global $debug_object; |
| 1207 |
|
| 1208 |
$charset = null; |
| 1209 |
|
| 1210 |
if (function_exists('get_last_retrieve_url_contents_content_type')) |
| 1211 |
{ |
| 1212 |
$contentTypeHeader = get_last_retrieve_url_contents_content_type(); |
| 1213 |
$success = preg_match('/charset=(.+)/', $contentTypeHeader, $matches); |
| 1214 |
if ($success) |
| 1215 |
{ |
| 1216 |
$charset = $matches[1]; |
| 1217 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'header content-type found charset of: ' . $charset);} |
| 1218 |
} |
| 1219 |
|
| 1220 |
} |
| 1221 |
|
| 1222 |
if (empty($charset)) |
| 1223 |
{ |
| 1224 |
$el = $this->root->find('meta[https-equiv=Content-Type]',0, true); |
| 1225 |
if (!empty($el)) |
| 1226 |
{ |
| 1227 |
$fullvalue = $el->content; |
| 1228 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'meta content-type tag found' . $fullvalue);} |
| 1229 |
|
| 1230 |
if (!empty($fullvalue)) |
| 1231 |
{ |
| 1232 |
$success = preg_match('/charset=(.+)/i', $fullvalue, $matches); |
| 1233 |
if ($success) |
| 1234 |
{ |
| 1235 |
$charset = $matches[1]; |
| 1236 |
} |
| 1237 |
else |
| 1238 |
{ |
| 1239 |
// If there is a meta tag, and they don't specify the character set, research says that it's typically ISO-8859-1 |
| 1240 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'meta content-type tag couldn\'t be parsed. using iso-8859 default.');} |
| 1241 |
$charset = 'ISO-8859-1'; |
| 1242 |
} |
| 1243 |
} |
| 1244 |
} |
| 1245 |
} |
| 1246 |
|
| 1247 |
// If we couldn't find a charset above, then lets try to detect one based on the text we got... |
| 1248 |
if (empty($charset)) |
| 1249 |
{ |
| 1250 |
// Use this in case mb_detect_charset isn't installed/loaded on this machine. |
| 1251 |
$charset = false; |
| 1252 |
if (function_exists('mb_detect_encoding')) |
| 1253 |
{ |
| 1254 |
// Have php try to detect the encoding from the text given to us. |
| 1255 |
$charset = mb_detect_encoding($this->root->plaintext . "ascii", $encoding_list = array( "UTF-8", "CP1252" ) ); |
| 1256 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'mb_detect found: ' . $charset);} |
| 1257 |
} |
| 1258 |
|
| 1259 |
// and if this doesn't work... then we need to just wrongheadedly assume it's UTF-8 so that we can move on - cause this will usually give us most of what we need... |
| 1260 |
if ($charset === false) |
| 1261 |
{ |
| 1262 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'since mb_detect failed - using default of utf-8');} |
| 1263 |
$charset = 'UTF-8'; |
| 1264 |
} |
| 1265 |
} |
| 1266 |
|
| 1267 |
// Since CP1252 is a superset, if we get one of it's subsets, we want it instead. |
| 1268 |
if ((strtolower($charset) == strtolower('ISO-8859-1')) || (strtolower($charset) == strtolower('Latin1')) || (strtolower($charset) == strtolower('Latin-1'))) |
| 1269 |
{ |
| 1270 |
if (is_object($debug_object)) {$debug_object->debug_log(2, 'replacing ' . $charset . ' with CP1252 as its a superset');} |
| 1271 |
$charset = 'CP1252'; |
| 1272 |
} |
| 1273 |
|
| 1274 |
if (is_object($debug_object)) {$debug_object->debug_log(1, 'EXIT - ' . $charset);} |
| 1275 |
|
| 1276 |
return $this->_charset = $charset; |
| 1277 |
} |
| 1278 |
|
| 1279 |
// read tag info |
| 1280 |
protected function read_tag() |
| 1281 |
{ |
| 1282 |
if ($this->char!=='<') |
| 1283 |
{ |
| 1284 |
$this->root->_[HDOM_INFO_END] = $this->cursor; |
| 1285 |
return false; |
| 1286 |
} |
| 1287 |
$begin_tag_pos = $this->pos; |
| 1288 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1289 |
|
| 1290 |
// end tag |
| 1291 |
if ($this->char==='/') |
| 1292 |
{ |
| 1293 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1294 |
// This represents the change in the simple_html_dom trunk from revision 180 to 181. |
| 1295 |
// $this->skip($this->token_blank_t); |
| 1296 |
$this->skip($this->token_blank); |
| 1297 |
$tag = $this->copy_until_char('>'); |
| 1298 |
|
| 1299 |
// skip attributes in end tag |
| 1300 |
if (($pos = strpos($tag, ' '))!==false) |
| 1301 |
$tag = substr($tag, 0, $pos); |
| 1302 |
|
| 1303 |
$parent_lower = strtolower($this->parent->tag); |
| 1304 |
$tag_lower = strtolower($tag); |
| 1305 |
|
| 1306 |
if ($parent_lower!==$tag_lower) |
| 1307 |
{ |
| 1308 |
if (isset($this->optional_closing_tags[$parent_lower]) && isset($this->block_tags[$tag_lower])) |
| 1309 |
{ |
| 1310 |
$this->parent->_[HDOM_INFO_END] = 0; |
| 1311 |
$org_parent = $this->parent; |
| 1312 |
|
| 1313 |
while (($this->parent->parent) && strtolower($this->parent->tag)!==$tag_lower) |
| 1314 |
$this->parent = $this->parent->parent; |
| 1315 |
|
| 1316 |
if (strtolower($this->parent->tag)!==$tag_lower) { |
| 1317 |
$this->parent = $org_parent; // restore origonal parent |
| 1318 |
if ($this->parent->parent) $this->parent = $this->parent->parent; |
| 1319 |
$this->parent->_[HDOM_INFO_END] = $this->cursor; |
| 1320 |
return $this->as_text_node($tag); |
| 1321 |
} |
| 1322 |
} |
| 1323 |
else if (($this->parent->parent) && isset($this->block_tags[$tag_lower])) |
| 1324 |
{ |
| 1325 |
$this->parent->_[HDOM_INFO_END] = 0; |
| 1326 |
$org_parent = $this->parent; |
| 1327 |
|
| 1328 |
while (($this->parent->parent) && strtolower($this->parent->tag)!==$tag_lower) |
| 1329 |
$this->parent = $this->parent->parent; |
| 1330 |
|
| 1331 |
if (strtolower($this->parent->tag)!==$tag_lower) |
| 1332 |
{ |
| 1333 |
$this->parent = $org_parent; // restore origonal parent |
| 1334 |
$this->parent->_[HDOM_INFO_END] = $this->cursor; |
| 1335 |
return $this->as_text_node($tag); |
| 1336 |
} |
| 1337 |
} |
| 1338 |
else if (($this->parent->parent) && strtolower($this->parent->parent->tag)===$tag_lower) |
| 1339 |
{ |
| 1340 |
$this->parent->_[HDOM_INFO_END] = 0; |
| 1341 |
$this->parent = $this->parent->parent; |
| 1342 |
} |
| 1343 |
else |
| 1344 |
return $this->as_text_node($tag); |
| 1345 |
} |
| 1346 |
|
| 1347 |
$this->parent->_[HDOM_INFO_END] = $this->cursor; |
| 1348 |
if ($this->parent->parent) $this->parent = $this->parent->parent; |
| 1349 |
|
| 1350 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1351 |
return true; |
| 1352 |
} |
| 1353 |
|
| 1354 |
$node = new simple_html_dom_node($this); |
| 1355 |
$node->_[HDOM_INFO_BEGIN] = $this->cursor; |
| 1356 |
++$this->cursor; |
| 1357 |
$tag = $this->copy_until($this->token_slash); |
| 1358 |
$node->tag_start = $begin_tag_pos; |
| 1359 |
|
| 1360 |
// doctype, cdata & comments... |
| 1361 |
if (isset($tag[0]) && $tag[0]==='!') { |
| 1362 |
$node->_[HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until_char('>'); |
| 1363 |
|
| 1364 |
if (isset($tag[2]) && $tag[1]==='-' && $tag[2]==='-') { |
| 1365 |
$node->nodetype = HDOM_TYPE_COMMENT; |
| 1366 |
$node->tag = 'comment'; |
| 1367 |
} else { |
| 1368 |
$node->nodetype = HDOM_TYPE_UNKNOWN; |
| 1369 |
$node->tag = 'unknown'; |
| 1370 |
} |
| 1371 |
if ($this->char==='>') $node->_[HDOM_INFO_TEXT].='>'; |
| 1372 |
$this->link_nodes($node, true); |
| 1373 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1374 |
return true; |
| 1375 |
} |
| 1376 |
|
| 1377 |
// text |
| 1378 |
if ($pos=strpos($tag, '<')!==false) { |
| 1379 |
$tag = '<' . substr($tag, 0, -1); |
| 1380 |
$node->_[HDOM_INFO_TEXT] = $tag; |
| 1381 |
$this->link_nodes($node, false); |
| 1382 |
$this->char = $this->doc[--$this->pos]; // prev |
| 1383 |
return true; |
| 1384 |
} |
| 1385 |
|
| 1386 |
// /^[\w-:]+$/ |
| 1387 |
|
| 1388 |
if (!preg_match("/^[\w\-:]+$/", $tag)) { |
| 1389 |
$node->_[HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until('<>'); |
| 1390 |
if ($this->char==='<') { |
| 1391 |
$this->link_nodes($node, false); |
| 1392 |
return true; |
| 1393 |
} |
| 1394 |
|
| 1395 |
if ($this->char==='>') $node->_[HDOM_INFO_TEXT].='>'; |
| 1396 |
$this->link_nodes($node, false); |
| 1397 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1398 |
return true; |
| 1399 |
} |
| 1400 |
|
| 1401 |
// begin tag |
| 1402 |
$node->nodetype = HDOM_TYPE_ELEMENT; |
| 1403 |
$tag_lower = strtolower($tag); |
| 1404 |
$node->tag = ($this->lowercase) ? $tag_lower : $tag; |
| 1405 |
|
| 1406 |
// handle optional closing tags |
| 1407 |
if (isset($this->optional_closing_tags[$tag_lower]) ) |
| 1408 |
{ |
| 1409 |
while (isset($this->optional_closing_tags[$tag_lower][strtolower($this->parent->tag)])) |
| 1410 |
{ |
| 1411 |
$this->parent->_[HDOM_INFO_END] = 0; |
| 1412 |
$this->parent = $this->parent->parent; |
| 1413 |
} |
| 1414 |
$node->parent = $this->parent; |
| 1415 |
} |
| 1416 |
|
| 1417 |
$guard = 0; // prevent infinity loop |
| 1418 |
$space = array($this->copy_skip($this->token_blank), '', ''); |
| 1419 |
|
| 1420 |
// attributes |
| 1421 |
do |
| 1422 |
{ |
| 1423 |
if ($this->char!==null && $space[0]==='') |
| 1424 |
{ |
| 1425 |
break; |
| 1426 |
} |
| 1427 |
$name = $this->copy_until($this->token_equal); |
| 1428 |
if ($guard===$this->pos) |
| 1429 |
{ |
| 1430 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1431 |
continue; |
| 1432 |
} |
| 1433 |
$guard = $this->pos; |
| 1434 |
|
| 1435 |
// handle endless '<' |
| 1436 |
if ($this->pos>=$this->size-1 && $this->char!=='>') { |
| 1437 |
$node->nodetype = HDOM_TYPE_TEXT; |
| 1438 |
$node->_[HDOM_INFO_END] = 0; |
| 1439 |
$node->_[HDOM_INFO_TEXT] = '<'.$tag . $space[0] . $name; |
| 1440 |
$node->tag = 'text'; |
| 1441 |
$this->link_nodes($node, false); |
| 1442 |
return true; |
| 1443 |
} |
| 1444 |
|
| 1445 |
// handle mismatch '<' |
| 1446 |
if ($this->doc[$this->pos-1]=='<') { |
| 1447 |
$node->nodetype = HDOM_TYPE_TEXT; |
| 1448 |
$node->tag = 'text'; |
| 1449 |
$node->attr = array(); |
| 1450 |
$node->_[HDOM_INFO_END] = 0; |
| 1451 |
$node->_[HDOM_INFO_TEXT] = substr($this->doc, $begin_tag_pos, $this->pos-$begin_tag_pos-1); |
| 1452 |
$this->pos -= 2; |
| 1453 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1454 |
$this->link_nodes($node, false); |
| 1455 |
return true; |
| 1456 |
} |
| 1457 |
|
| 1458 |
if ($name!=='/' && $name!=='') { |
| 1459 |
$space[1] = $this->copy_skip($this->token_blank); |
| 1460 |
$name = $this->restore_noise($name); |
| 1461 |
if ($this->lowercase) $name = strtolower($name); |
| 1462 |
if ($this->char==='=') { |
| 1463 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1464 |
$this->parse_attr($node, $name, $space); |
| 1465 |
} |
| 1466 |
else { |
| 1467 |
//no value attr: nowrap, checked selected... |
| 1468 |
$node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_NO; |
| 1469 |
$node->attr[$name] = true; |
| 1470 |
if ($this->char!='>') $this->char = $this->doc[--$this->pos]; // prev |
| 1471 |
} |
| 1472 |
$node->_[HDOM_INFO_SPACE][] = $space; |
| 1473 |
$space = array($this->copy_skip($this->token_blank), '', ''); |
| 1474 |
} |
| 1475 |
else |
| 1476 |
break; |
| 1477 |
} while ($this->char!=='>' && $this->char!=='/'); |
| 1478 |
|
| 1479 |
$this->link_nodes($node, true); |
| 1480 |
$node->_[HDOM_INFO_ENDSPACE] = $space[0]; |
| 1481 |
|
| 1482 |
// check self closing |
| 1483 |
if ($this->copy_until_char_escape('>')==='/') |
| 1484 |
{ |
| 1485 |
$node->_[HDOM_INFO_ENDSPACE] .= '/'; |
| 1486 |
$node->_[HDOM_INFO_END] = 0; |
| 1487 |
} |
| 1488 |
else |
| 1489 |
{ |
| 1490 |
// reset parent |
| 1491 |
if (!isset($this->self_closing_tags[strtolower($node->tag)])) $this->parent = $node; |
| 1492 |
} |
| 1493 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1494 |
|
| 1495 |
// If it's a BR tag, we need to set it's text to the default text. |
| 1496 |
// This way when we see it in plaintext, we can generate formatting that the user wants. |
| 1497 |
// since a br tag never has sub nodes, this works well. |
| 1498 |
if ($node->tag == "br") |
| 1499 |
{ |
| 1500 |
$node->_[HDOM_INFO_INNER] = $this->default_br_text; |
| 1501 |
} |
| 1502 |
|
| 1503 |
return true; |
| 1504 |
} |
| 1505 |
|
| 1506 |
// parse attributes |
| 1507 |
protected function parse_attr($node, $name, &$space) |
| 1508 |
{ |
| 1509 |
// Per sourceforge: https://sourceforge.net/tracker/?func=detail&aid=3061408&group_id=218559&atid=1044037 |
| 1510 |
// If the attribute is already defined inside a tag, only pay atetntion to the first one as opposed to the last one. |
| 1511 |
if (isset($node->attr[$name])) |
| 1512 |
{ |
| 1513 |
return; |
| 1514 |
} |
| 1515 |
|
| 1516 |
$space[2] = $this->copy_skip($this->token_blank); |
| 1517 |
switch ($this->char) { |
| 1518 |
case '"': |
| 1519 |
$node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE; |
| 1520 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1521 |
$node->attr[$name] = $this->restore_noise($this->copy_until_char_escape('"')); |
| 1522 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1523 |
break; |
| 1524 |
case '\'': |
| 1525 |
$node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_SINGLE; |
| 1526 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1527 |
$node->attr[$name] = $this->restore_noise($this->copy_until_char_escape('\'')); |
| 1528 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1529 |
break; |
| 1530 |
default: |
| 1531 |
$node->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_NO; |
| 1532 |
$node->attr[$name] = $this->restore_noise($this->copy_until($this->token_attr)); |
| 1533 |
} |
| 1534 |
// PaperG: Attributes should not have \r or \n in them, that counts as html whitespace. |
| 1535 |
$node->attr[$name] = str_replace("\r", "", $node->attr[$name]); |
| 1536 |
$node->attr[$name] = str_replace("\n", "", $node->attr[$name]); |
| 1537 |
// 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. |
| 1538 |
if ($name == "class") { |
| 1539 |
$node->attr[$name] = trim($node->attr[$name]); |
| 1540 |
} |
| 1541 |
} |
| 1542 |
|
| 1543 |
// link node's parent |
| 1544 |
protected function link_nodes(&$node, $is_child) |
| 1545 |
{ |
| 1546 |
$node->parent = $this->parent; |
| 1547 |
$this->parent->nodes[] = $node; |
| 1548 |
if ($is_child) |
| 1549 |
{ |
| 1550 |
$this->parent->children[] = $node; |
| 1551 |
} |
| 1552 |
} |
| 1553 |
|
| 1554 |
// as a text node |
| 1555 |
protected function as_text_node($tag) |
| 1556 |
{ |
| 1557 |
$node = new simple_html_dom_node($this); |
| 1558 |
++$this->cursor; |
| 1559 |
$node->_[HDOM_INFO_TEXT] = '</' . $tag . '>'; |
| 1560 |
$this->link_nodes($node, false); |
| 1561 |
$this->char = (++$this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1562 |
return true; |
| 1563 |
} |
| 1564 |
|
| 1565 |
protected function skip($chars) |
| 1566 |
{ |
| 1567 |
$this->pos += strspn($this->doc, $chars, $this->pos); |
| 1568 |
$this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1569 |
} |
| 1570 |
|
| 1571 |
protected function copy_skip($chars) |
| 1572 |
{ |
| 1573 |
$pos = $this->pos; |
| 1574 |
$len = strspn($this->doc, $chars, $pos); |
| 1575 |
$this->pos += $len; |
| 1576 |
$this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1577 |
if ($len===0) return ''; |
| 1578 |
return substr($this->doc, $pos, $len); |
| 1579 |
} |
| 1580 |
|
| 1581 |
protected function copy_until($chars) |
| 1582 |
{ |
| 1583 |
$pos = $this->pos; |
| 1584 |
$len = strcspn($this->doc, $chars, $pos); |
| 1585 |
$this->pos += $len; |
| 1586 |
$this->char = ($this->pos<$this->size) ? $this->doc[$this->pos] : null; // next |
| 1587 |
return substr($this->doc, $pos, $len); |
| 1588 |
} |
| 1589 |
|
| 1590 |
protected function copy_until_char($char) |
| 1591 |
{ |
| 1592 |
if ($this->char===null) return ''; |
| 1593 |
|
| 1594 |
if (($pos = strpos($this->doc, $char, $this->pos))===false) { |
| 1595 |
$ret = substr($this->doc, $this->pos, $this->size-$this->pos); |
| 1596 |
$this->char = null; |
| 1597 |
$this->pos = $this->size; |
| 1598 |
return $ret; |
| 1599 |
} |
| 1600 |
|
| 1601 |
if ($pos===$this->pos) return ''; |
| 1602 |
$pos_old = $this->pos; |
| 1603 |
$this->char = $this->doc[$pos]; |
| 1604 |
$this->pos = $pos; |
| 1605 |
return substr($this->doc, $pos_old, $pos-$pos_old); |
| 1606 |
} |
| 1607 |
|
| 1608 |
protected function copy_until_char_escape($char) |
| 1609 |
{ |
| 1610 |
if ($this->char===null) return ''; |
| 1611 |
|
| 1612 |
$start = $this->pos; |
| 1613 |
while (1) |
| 1614 |
{ |
| 1615 |
if (($pos = strpos($this->doc, $char, $start))===false) |
| 1616 |
{ |
| 1617 |
$ret = substr($this->doc, $this->pos, $this->size-$this->pos); |
| 1618 |
$this->char = null; |
| 1619 |
$this->pos = $this->size; |
| 1620 |
return $ret; |
| 1621 |
} |
| 1622 |
|
| 1623 |
if ($pos===$this->pos) return ''; |
| 1624 |
|
| 1625 |
if ($this->doc[$pos-1]==='\\') { |
| 1626 |
$start = $pos+1; |
| 1627 |
continue; |
| 1628 |
} |
| 1629 |
|
| 1630 |
$pos_old = $this->pos; |
| 1631 |
$this->char = $this->doc[$pos]; |
| 1632 |
$this->pos = $pos; |
| 1633 |
return substr($this->doc, $pos_old, $pos-$pos_old); |
| 1634 |
} |
| 1635 |
} |
| 1636 |
|
| 1637 |
// remove noise from html content |
| 1638 |
// save the noise in the $this->noise array. |
| 1639 |
protected function remove_noise($pattern, $remove_tag=false) |
| 1640 |
{ |
| 1641 |
global $debug_object; |
| 1642 |
if (is_object($debug_object)) { $debug_object->debug_log_entry(1); } |
| 1643 |
|
| 1644 |
$count = preg_match_all($pattern, $this->doc, $matches, PREG_SET_ORDER|PREG_OFFSET_CAPTURE); |
| 1645 |
|
| 1646 |
for ($i=$count-1; $i>-1; --$i) |
| 1647 |
{ |
| 1648 |
$key = '___noise___'.sprintf('% 5d', count($this->noise)+1000); |
| 1649 |
if (is_object($debug_object)) { $debug_object->debug_log(2, 'key is: ' . $key); } |
| 1650 |
$idx = ($remove_tag) ? 0 : 1; |
| 1651 |
$this->noise[$key] = $matches[$i][$idx][0]; |
| 1652 |
$this->doc = substr_replace($this->doc, $key, $matches[$i][$idx][1], strlen($matches[$i][$idx][0])); |
| 1653 |
} |
| 1654 |
|
| 1655 |
// reset the length of content |
| 1656 |
$this->size = strlen($this->doc); |
| 1657 |
if ($this->size>0) |
| 1658 |
{ |
| 1659 |
$this->char = $this->doc[0]; |
| 1660 |
} |
| 1661 |
} |
| 1662 |
|
| 1663 |
// restore noise to html content |
| 1664 |
function restore_noise($text) |
| 1665 |
{ |
| 1666 |
global $debug_object; |
| 1667 |
if (is_object($debug_object)) { $debug_object->debug_log_entry(1); } |
| 1668 |
|
| 1669 |
while (($pos=strpos($text, '___noise___'))!==false) |
| 1670 |
{ |
| 1671 |
// Sometimes there is a broken piece of markup, and we don't GET the pos+11 etc... token which indicates a problem outside of us... |
| 1672 |
if (strlen($text) > $pos+15) |
| 1673 |
{ |
| 1674 |
$key = '___noise___'.$text[$pos+11].$text[$pos+12].$text[$pos+13].$text[$pos+14].$text[$pos+15]; |
| 1675 |
if (is_object($debug_object)) { $debug_object->debug_log(2, 'located key of: ' . $key); } |
| 1676 |
|
| 1677 |
if (isset($this->noise[$key])) |
| 1678 |
{ |
| 1679 |
$text = substr($text, 0, $pos).$this->noise[$key].substr($text, $pos+16); |
| 1680 |
} |
| 1681 |
else |
| 1682 |
{ |
| 1683 |
// do this to prevent an infinite loop. |
| 1684 |
$text = substr($text, 0, $pos).'UNDEFINED NOISE FOR KEY: '.$key . substr($text, $pos+16); |
| 1685 |
} |
| 1686 |
} |
| 1687 |
else |
| 1688 |
{ |
| 1689 |
// There is no valid key being given back to us... We must get rid of the ___noise___ or we will have a problem. |
| 1690 |
$text = substr($text, 0, $pos).'NO NUMERIC NOISE KEY' . substr($text, $pos+11); |
| 1691 |
} |
| 1692 |
} |
| 1693 |
return $text; |
| 1694 |
} |
| 1695 |
|
| 1696 |
// Sometimes we NEED one of the noise elements. |
| 1697 |
function search_noise($text) |
| 1698 |
{ |
| 1699 |
global $debug_object; |
| 1700 |
if (is_object($debug_object)) { $debug_object->debug_log_entry(1); } |
| 1701 |
|
| 1702 |
foreach($this->noise as $noiseElement) |
| 1703 |
{ |
| 1704 |
if (strpos($noiseElement, $text)!==false) |
| 1705 |
{ |
| 1706 |
return $noiseElement; |
| 1707 |
} |
| 1708 |
} |
| 1709 |
} |
| 1710 |
function __toString() |
| 1711 |
{ |
| 1712 |
return $this->root->innertext(); |
| 1713 |
} |
| 1714 |
|
| 1715 |
function __get($name) |
| 1716 |
{ |
| 1717 |
switch ($name) |
| 1718 |
{ |
| 1719 |
case 'outertext': |
| 1720 |
return $this->root->innertext(); |
| 1721 |
case 'innertext': |
| 1722 |
return $this->root->innertext(); |
| 1723 |
case 'plaintext': |
| 1724 |
return $this->root->text(); |
| 1725 |
case 'charset': |
| 1726 |
return $this->_charset; |
| 1727 |
case 'target_charset': |
| 1728 |
return $this->_target_charset; |
| 1729 |
} |
| 1730 |
} |
| 1731 |
|
| 1732 |
// camel naming conventions |
| 1733 |
function childNodes($idx=-1) {return $this->root->childNodes($idx);} |
| 1734 |
function firstChild() {return $this->root->first_child();} |
| 1735 |
function lastChild() {return $this->root->last_child();} |
| 1736 |
function createElement($name, $value=null) {return @str_get_html("<$name>$value</$name>")->first_child();} |
| 1737 |
function createTextNode($value) {return @end(str_get_html($value)->nodes);} |
| 1738 |
function getElementById($id) {return $this->find("#$id", 0);} |
| 1739 |
function getElementsById($id, $idx=null) {return $this->find("#$id", $idx);} |
| 1740 |
function getElementByTagName($name) {return $this->find($name, 0);} |
| 1741 |
function getElementsByTagName($name, $idx=-1) {return $this->find($name, $idx);} |
| 1742 |
function loadFile() {$args = func_get_args();$this->load_file($args);} |
| 1743 |
} |
| 1744 |
|
| 1745 |
?> |
| 1746 |
|
