| 1 |
<?php |
| 2 |
if (!defined('ABSPATH')) exit; |
| 3 |
|
| 4 |
define ( 'GPTRANSLATE_HDOM_TYPE_ELEMENT', 1 ); |
| 5 |
define ( 'GPTRANSLATE_HDOM_TYPE_COMMENT', 2 ); |
| 6 |
define ( 'GPTRANSLATE_HDOM_TYPE_TEXT', 3 ); |
| 7 |
define ( 'GPTRANSLATE_HDOM_TYPE_ENDTAG', 4 ); |
| 8 |
define ( 'GPTRANSLATE_HDOM_TYPE_ROOT', 5 ); |
| 9 |
define ( 'GPTRANSLATE_HDOM_TYPE_UNKNOWN', 6 ); |
| 10 |
define ( 'GPTRANSLATE_HDOM_QUOTE_DOUBLE', 0 ); |
| 11 |
define ( 'GPTRANSLATE_HDOM_QUOTE_SINGLE', 1 ); |
| 12 |
define ( 'GPTRANSLATE_HDOM_QUOTE_NO', 3 ); |
| 13 |
define ( 'GPTRANSLATE_HDOM_INFO_BEGIN', 0 ); |
| 14 |
define ( 'GPTRANSLATE_HDOM_INFO_END', 1 ); |
| 15 |
define ( 'GPTRANSLATE_HDOM_INFO_QUOTE', 2 ); |
| 16 |
define ( 'GPTRANSLATE_HDOM_INFO_SPACE', 3 ); |
| 17 |
define ( 'GPTRANSLATE_HDOM_INFO_TEXT', 4 ); |
| 18 |
define ( 'GPTRANSLATE_HDOM_INFO_INNER', 5 ); |
| 19 |
define ( 'GPTRANSLATE_HDOM_INFO_OUTER', 6 ); |
| 20 |
define ( 'GPTRANSLATE_HDOM_INFO_ENDSPACE', 7 ); |
| 21 |
define ( 'GPTRANSLATE_DEFAULT_TARGET_CHARSET', 'UTF-8' ); |
| 22 |
define ( 'GPTRANSLATE_DEFAULT_BR_TEXT', "\r\n" ); |
| 23 |
define ( 'GPTRANSLATE_DEFAULT_SPAN_TEXT', " " ); |
| 24 |
define ( 'GPTRANSLATE_MAX_FILE_SIZE', 600000 ); |
| 25 |
// helper functions |
| 26 |
// ----------------------------------------------------------------------------- |
| 27 |
// get html dom from file |
| 28 |
// $maxlen is defined in the code as PHP_STREAM_COPY_ALL which is defined as -1. |
| 29 |
function gptranslate_simplehtmldom_file_get_html($url, $use_include_path = false, $context = null, $offset = -1, $maxLen = -1, $lowercase = true, $forceTagsClosed = true, $target_charset = GPTRANSLATE_DEFAULT_TARGET_CHARSET, $stripRN = true, $defaultBRText = GPTRANSLATE_DEFAULT_BR_TEXT, $defaultSpanText = GPTRANSLATE_DEFAULT_SPAN_TEXT) { |
| 30 |
// We DO force the tags to be terminated. |
| 31 |
$dom = new GPTranslateSimpleHtmlDom ( null, $lowercase, $forceTagsClosed, $target_charset, $stripRN, $defaultBRText, $defaultSpanText ); |
| 32 |
// For sourceforge users: uncomment the next line and comment the retreive_url_contents line 2 lines down if it is not already done. |
| 33 |
$contents = file_get_contents ( $url, $use_include_path, $context, $offset ); |
| 34 |
// Paperg - use our own mechanism for getting the contents as we want to control the timeout. |
| 35 |
// $contents = retrieve_url_contents($url); |
| 36 |
if (empty ( $contents ) || strlen ( $contents ) > GPTRANSLATE_MAX_FILE_SIZE) { |
| 37 |
return false; |
| 38 |
} |
| 39 |
// The second parameter can force the selectors to all be lowercase. |
| 40 |
$dom->load ( $contents, $lowercase, $stripRN ); |
| 41 |
return $dom; |
| 42 |
} |
| 43 |
|
| 44 |
// get html dom from string |
| 45 |
function gptranslate_simplehtmldom_str_get_html($str, $lowercase = true, $forceTagsClosed = true, $target_charset = GPTRANSLATE_DEFAULT_TARGET_CHARSET, $stripRN = true, $defaultBRText = GPTRANSLATE_DEFAULT_BR_TEXT, $defaultSpanText = GPTRANSLATE_DEFAULT_SPAN_TEXT) { |
| 46 |
$dom = new GPTranslateSimpleHtmlDom ( null, $lowercase, $forceTagsClosed, $target_charset, $stripRN, $defaultBRText, $defaultSpanText ); |
| 47 |
if (empty ( $str ) || strlen ( $str ) > GPTRANSLATE_MAX_FILE_SIZE) { |
| 48 |
$dom->clear (); |
| 49 |
return false; |
| 50 |
} |
| 51 |
$dom->load ( $str, $lowercase, $stripRN ); |
| 52 |
return $dom; |
| 53 |
} |
| 54 |
|
| 55 |
/** |
| 56 |
* simple html dom node |
| 57 |
* PaperG - added ability for "find" routine to lowercase the value of the selector. |
| 58 |
* PaperG - added $tag_start to track the start position of the tag in the total byte index |
| 59 |
* |
| 60 |
* @package PlaceLocalInclude |
| 61 |
*/ |
| 62 |
class GPTranslateSimpleHtmlDomNode { |
| 63 |
public $nodetype = GPTRANSLATE_HDOM_TYPE_TEXT; |
| 64 |
public $tag = 'text'; |
| 65 |
public $attr = array (); |
| 66 |
public $children = array (); |
| 67 |
public $nodes = array (); |
| 68 |
public $parent = null; |
| 69 |
// The "info" array - see GPTRANSLATE_HDOM_INFO_... for what each element contains. |
| 70 |
public $_ = array (); |
| 71 |
public $tag_start = 0; |
| 72 |
public $no_remove = null; |
| 73 |
private $dom = null; |
| 74 |
private $gptranslate_debug_object = null; |
| 75 |
function __construct($dom) { |
| 76 |
$this->dom = $dom; |
| 77 |
$dom->nodes [] = $this; |
| 78 |
} |
| 79 |
function __destruct() { |
| 80 |
$this->clear (); |
| 81 |
} |
| 82 |
function __toString() { |
| 83 |
return $this->outertext (); |
| 84 |
} |
| 85 |
|
| 86 |
// clean up memory due to php5 circular references memory leak... |
| 87 |
function clear() { |
| 88 |
$this->dom = null; |
| 89 |
$this->nodes = null; |
| 90 |
$this->parent = null; |
| 91 |
$this->children = null; |
| 92 |
} |
| 93 |
|
| 94 |
// dump node's tree |
| 95 |
function dump($show_attr = true, $deep = 0) { |
| 96 |
$lead = str_repeat ( ' ', $deep ); |
| 97 |
// phpcs:ignore WordPress.Security.EscapeOutput.OutputNotEscaped -- output is safe |
| 98 |
echo $lead . $this->tag; |
| 99 |
if ($show_attr && count ( $this->attr ) > 0) { |
| 100 |
echo '('; |
| 101 |
foreach ( $this->attr as $k => $v ) |
| 102 |
// phpcs:ignore WordPress.Security.EscapeOutput.OutputNotEscaped -- output is safe |
| 103 |
echo "[$k]=>\"" . $this->$k . '", '; |
| 104 |
echo ')'; |
| 105 |
} |
| 106 |
echo "\n"; |
| 107 |
|
| 108 |
if ($this->nodes) { |
| 109 |
foreach ( $this->nodes as $c ) { |
| 110 |
$c->dump ( $show_attr, $deep + 1 ); |
| 111 |
} |
| 112 |
} |
| 113 |
} |
| 114 |
|
| 115 |
// Debugging function to dump a single dom node with a bunch of information about it. |
| 116 |
function dump_node($echo = true) { |
| 117 |
$string = $this->tag; |
| 118 |
if (count ( $this->attr ) > 0) { |
| 119 |
$string .= '('; |
| 120 |
foreach ( $this->attr as $k => $v ) { |
| 121 |
$string .= "[$k]=>\"" . $this->$k . '", '; |
| 122 |
} |
| 123 |
$string .= ')'; |
| 124 |
} |
| 125 |
if (count ( $this->_ ) > 0) { |
| 126 |
$string .= ' $_ ('; |
| 127 |
foreach ( $this->_ as $k => $v ) { |
| 128 |
if (is_array ( $v )) { |
| 129 |
$string .= "[$k]=>("; |
| 130 |
foreach ( $v as $k2 => $v2 ) { |
| 131 |
$string .= "[$k2]=>\"" . $v2 . '", '; |
| 132 |
} |
| 133 |
$string .= ")"; |
| 134 |
} else { |
| 135 |
$string .= "[$k]=>\"" . $v . '", '; |
| 136 |
} |
| 137 |
} |
| 138 |
$string .= ")"; |
| 139 |
} |
| 140 |
|
| 141 |
if (isset ( $this->text )) { |
| 142 |
$string .= " text: (" . $this->text . ")"; |
| 143 |
} |
| 144 |
|
| 145 |
$string .= " GPTRANSLATE_HDOM_INNER_INFO: '"; |
| 146 |
if (isset ( $node->_ [GPTRANSLATE_HDOM_INFO_INNER] )) { |
| 147 |
$string .= $node->_ [GPTRANSLATE_HDOM_INFO_INNER] . "'"; |
| 148 |
} else { |
| 149 |
$string .= ' NULL '; |
| 150 |
} |
| 151 |
|
| 152 |
$string .= " children: " . count ( $this->children ); |
| 153 |
$string .= " nodes: " . count ( $this->nodes ); |
| 154 |
$string .= " tag_start: " . $this->tag_start; |
| 155 |
$string .= "\n"; |
| 156 |
// phpcs:ignore WordPress.Security.EscapeOutput.OutputNotEscaped -- output is safe |
| 157 |
if ($echo) { |
| 158 |
// phpcs:ignore WordPress.Security.EscapeOutput.OutputNotEscaped -- output is safe |
| 159 |
echo $string; |
| 160 |
return; |
| 161 |
} else { |
| 162 |
return $string; |
| 163 |
} |
| 164 |
} |
| 165 |
|
| 166 |
// returns the parent of node |
| 167 |
// If a node is passed in, it will reset the parent of the current node to that one. |
| 168 |
function parent($parent = null) { |
| 169 |
// I am SURE that this doesn't work properly. |
| 170 |
// It fails to unset the current node from it's current parents nodes or children list first. |
| 171 |
if ($parent !== null) { |
| 172 |
$this->parent = $parent; |
| 173 |
$this->parent->nodes [] = $this; |
| 174 |
$this->parent->children [] = $this; |
| 175 |
} |
| 176 |
|
| 177 |
return $this->parent; |
| 178 |
} |
| 179 |
|
| 180 |
// verify that node has children |
| 181 |
function has_child() { |
| 182 |
return ! empty ( $this->children ); |
| 183 |
} |
| 184 |
|
| 185 |
// returns children of node |
| 186 |
function children($idx = -1) { |
| 187 |
if ($idx === - 1) { |
| 188 |
return $this->children; |
| 189 |
} |
| 190 |
if (isset ( $this->children [$idx] )) { |
| 191 |
return $this->children [$idx]; |
| 192 |
} |
| 193 |
return null; |
| 194 |
} |
| 195 |
|
| 196 |
// returns children of node recursively |
| 197 |
function childrenRecursive($element, &$allChildren = array()) { |
| 198 |
if ($element->hasChildNodes()) { |
| 199 |
foreach ($element->childNodes() as $childElement) { |
| 200 |
$allChildren[] = $childElement; |
| 201 |
if($childElement->hasChildNodes()) { |
| 202 |
$this->childrenRecursive($childElement, $allChildren); |
| 203 |
} |
| 204 |
} |
| 205 |
} |
| 206 |
return $allChildren; |
| 207 |
} |
| 208 |
|
| 209 |
// returns the first child of node |
| 210 |
function first_child() { |
| 211 |
if (count ( $this->children ) > 0) { |
| 212 |
return $this->children [0]; |
| 213 |
} |
| 214 |
return null; |
| 215 |
} |
| 216 |
|
| 217 |
// returns the last child of node |
| 218 |
function last_child() { |
| 219 |
if (($count = count ( $this->children )) > 0) { |
| 220 |
return $this->children [$count - 1]; |
| 221 |
} |
| 222 |
return null; |
| 223 |
} |
| 224 |
|
| 225 |
// returns the next sibling of node |
| 226 |
function next_sibling() { |
| 227 |
if ($this->parent === null) { |
| 228 |
return null; |
| 229 |
} |
| 230 |
|
| 231 |
$idx = 0; |
| 232 |
$count = count ( $this->parent->children ); |
| 233 |
while ( $idx < $count && $this !== $this->parent->children [$idx] ) { |
| 234 |
++ $idx; |
| 235 |
} |
| 236 |
if (++ $idx >= $count) { |
| 237 |
return null; |
| 238 |
} |
| 239 |
return $this->parent->children [$idx]; |
| 240 |
} |
| 241 |
|
| 242 |
// returns the previous sibling of node |
| 243 |
function prev_sibling() { |
| 244 |
if ($this->parent === null) |
| 245 |
return null; |
| 246 |
$idx = 0; |
| 247 |
$count = count ( $this->parent->children ); |
| 248 |
while ( $idx < $count && $this !== $this->parent->children [$idx] ) |
| 249 |
++ $idx; |
| 250 |
if (-- $idx < 0) |
| 251 |
return null; |
| 252 |
return $this->parent->children [$idx]; |
| 253 |
} |
| 254 |
|
| 255 |
// function to locate a specific ancestor tag in the path to the root. |
| 256 |
function find_ancestor_tag($tag) { |
| 257 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 258 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 259 |
} |
| 260 |
|
| 261 |
// Start by including ourselves in the comparison. |
| 262 |
$returnDom = $this; |
| 263 |
|
| 264 |
while ( ! is_null ( $returnDom ) ) { |
| 265 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 266 |
$this->gptranslate_debug_object->debug_log ( 2, "Current tag is: " . $returnDom->tag ); |
| 267 |
} |
| 268 |
|
| 269 |
if ($returnDom->tag == $tag) { |
| 270 |
break; |
| 271 |
} |
| 272 |
$returnDom = $returnDom->parent; |
| 273 |
} |
| 274 |
return $returnDom; |
| 275 |
} |
| 276 |
|
| 277 |
// get dom node's inner html |
| 278 |
function innertext() { |
| 279 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_INNER] )) |
| 280 |
return $this->_ [GPTRANSLATE_HDOM_INFO_INNER]; |
| 281 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] )) |
| 282 |
return $this->dom->restore_noise ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] ); |
| 283 |
|
| 284 |
$ret = ''; |
| 285 |
foreach ( $this->nodes as $n ) |
| 286 |
$ret .= $n->outertext (); |
| 287 |
return $ret; |
| 288 |
} |
| 289 |
|
| 290 |
// get dom node's outer text (with tag) |
| 291 |
function outertext() { |
| 292 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 293 |
$text = ''; |
| 294 |
if ($this->tag == 'text') { |
| 295 |
if (! empty ( $this->text )) { |
| 296 |
$text = " with text: " . $this->text; |
| 297 |
} |
| 298 |
} |
| 299 |
$this->gptranslate_debug_object->debug_log ( 1, 'Innertext of tag: ' . $this->tag . $text ); |
| 300 |
} |
| 301 |
|
| 302 |
if ($this->tag === 'root') |
| 303 |
return $this->innertext (); |
| 304 |
|
| 305 |
// trigger callback |
| 306 |
if ($this->dom && $this->dom->callback !== null) { |
| 307 |
call_user_func_array ( $this->dom->callback, array ( |
| 308 |
$this |
| 309 |
) ); |
| 310 |
} |
| 311 |
|
| 312 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_OUTER] )) |
| 313 |
return $this->_ [GPTRANSLATE_HDOM_INFO_OUTER]; |
| 314 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] )) |
| 315 |
return $this->dom->restore_noise ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] ); |
| 316 |
|
| 317 |
// render begin tag |
| 318 |
if ($this->dom && $this->dom->nodes [$this->_ [GPTRANSLATE_HDOM_INFO_BEGIN]]) { |
| 319 |
$ret = $this->dom->nodes [$this->_ [GPTRANSLATE_HDOM_INFO_BEGIN]]->makeup (); |
| 320 |
} else { |
| 321 |
$ret = ""; |
| 322 |
} |
| 323 |
|
| 324 |
// render inner text |
| 325 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_INNER] )) { |
| 326 |
// If it's a br tag... don't return the GPTRANSLATE_HDOM_INNER_INFO that we may or may not have added. |
| 327 |
if ($this->tag != "br") { |
| 328 |
$ret .= $this->_ [GPTRANSLATE_HDOM_INFO_INNER]; |
| 329 |
} |
| 330 |
} else { |
| 331 |
if ($this->nodes) { |
| 332 |
foreach ( $this->nodes as $n ) { |
| 333 |
$ret .= $this->convert_text ( $n->outertext () ); |
| 334 |
} |
| 335 |
} |
| 336 |
} |
| 337 |
|
| 338 |
// render end tag |
| 339 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_END] ) && $this->_ [GPTRANSLATE_HDOM_INFO_END] != 0) |
| 340 |
$ret .= '</' . $this->tag . '>'; |
| 341 |
return $ret; |
| 342 |
} |
| 343 |
|
| 344 |
// get dom node's plain text |
| 345 |
function text($script = false) { |
| 346 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_INNER] )) |
| 347 |
return $this->_ [GPTRANSLATE_HDOM_INFO_INNER]; |
| 348 |
switch ($this->nodetype) { |
| 349 |
case GPTRANSLATE_HDOM_TYPE_TEXT : |
| 350 |
return $this->dom->restore_noise ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] ); |
| 351 |
case GPTRANSLATE_HDOM_TYPE_COMMENT : |
| 352 |
return ''; |
| 353 |
case GPTRANSLATE_HDOM_TYPE_UNKNOWN : |
| 354 |
return ''; |
| 355 |
} |
| 356 |
if (strcasecmp ( $this->tag, 'script' ) === 0 && !$script) |
| 357 |
return ''; |
| 358 |
if (strcasecmp ( $this->tag, 'style' ) === 0) |
| 359 |
return ''; |
| 360 |
|
| 361 |
$ret = ''; |
| 362 |
// In rare cases, (always node type 1 or GPTRANSLATE_HDOM_TYPE_ELEMENT - observed for some span tags, and some p tags) $this->nodes is set to NULL. |
| 363 |
// NOTE: This indicates that there is a problem where it's set to NULL without a clear happening. |
| 364 |
// WHY is this happening? |
| 365 |
if (! is_null ( $this->nodes )) { |
| 366 |
foreach ( $this->nodes as $n ) { |
| 367 |
$ret .= $this->convert_text ( $n->text () ); |
| 368 |
} |
| 369 |
|
| 370 |
// If this node is a span... add a space at the end of it so multiple spans don't run into each other. This is plaintext after all. |
| 371 |
if ($this->tag == "span") { |
| 372 |
$ret .= $this->dom->default_span_text; |
| 373 |
} |
| 374 |
} |
| 375 |
return $ret; |
| 376 |
} |
| 377 |
function xmltext() { |
| 378 |
$ret = $this->innertext (); |
| 379 |
$ret = str_ireplace ( '<![CDATA[', '', $ret ); |
| 380 |
$ret = str_replace ( ']]>', '', $ret ); |
| 381 |
return $ret; |
| 382 |
} |
| 383 |
|
| 384 |
// build node's text with tag |
| 385 |
function makeup() { |
| 386 |
// text, comment, unknown |
| 387 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] )) |
| 388 |
return $this->dom->restore_noise ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] ); |
| 389 |
|
| 390 |
$ret = '<' . $this->tag; |
| 391 |
$i = - 1; |
| 392 |
|
| 393 |
foreach ( $this->attr as $key => $val ) { |
| 394 |
++ $i; |
| 395 |
|
| 396 |
// skip removed attribute |
| 397 |
if ($val === null || $val === false) |
| 398 |
continue; |
| 399 |
|
| 400 |
$ret .= $this->_ [GPTRANSLATE_HDOM_INFO_SPACE] [$i] [0]; |
| 401 |
// no value attr: nowrap, checked selected... |
| 402 |
if ($val === true) |
| 403 |
$ret .= $key; |
| 404 |
else { |
| 405 |
switch ($this->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [$i]) { |
| 406 |
case GPTRANSLATE_HDOM_QUOTE_DOUBLE : |
| 407 |
$quote = '"'; |
| 408 |
break; |
| 409 |
case GPTRANSLATE_HDOM_QUOTE_SINGLE : |
| 410 |
$quote = '\''; |
| 411 |
break; |
| 412 |
default : |
| 413 |
$quote = ''; |
| 414 |
} |
| 415 |
$ret .= $key . $this->_ [GPTRANSLATE_HDOM_INFO_SPACE] [$i] [1] . '=' . $this->_ [GPTRANSLATE_HDOM_INFO_SPACE] [$i] [2] . $quote . $val . $quote; |
| 416 |
} |
| 417 |
} |
| 418 |
$ret = $this->dom->restore_noise ( $ret ); |
| 419 |
return $ret . $this->_ [GPTRANSLATE_HDOM_INFO_ENDSPACE] . '>'; |
| 420 |
} |
| 421 |
|
| 422 |
// find elements by css selector |
| 423 |
// PaperG - added ability for find to lowercase the value of the selector. |
| 424 |
function find($selector, $idx = null, $lowercase = false) { |
| 425 |
$selectors = $this->parse_selector ( $selector ); |
| 426 |
if (($count = count ( $selectors )) === 0) |
| 427 |
return array (); |
| 428 |
$found_keys = array (); |
| 429 |
|
| 430 |
// find each selector |
| 431 |
for($c = 0; $c < $count; ++ $c) { |
| 432 |
// The change on the below line was documented on the sourceforge code tracker id 2788009 |
| 433 |
// used to be: if (($levle=count($selectors[0]))===0) return array(); |
| 434 |
if (($levle = count ( $selectors [$c] )) === 0) |
| 435 |
return array (); |
| 436 |
if (! isset ( $this->_ [GPTRANSLATE_HDOM_INFO_BEGIN] )) |
| 437 |
return array (); |
| 438 |
|
| 439 |
$head = array ( |
| 440 |
$this->_ [GPTRANSLATE_HDOM_INFO_BEGIN] => 1 |
| 441 |
); |
| 442 |
|
| 443 |
// handle descendant selectors, no recursive! |
| 444 |
for($l = 0; $l < $levle; ++ $l) { |
| 445 |
$ret = array (); |
| 446 |
foreach ( $head as $k => $v ) { |
| 447 |
$n = ($k === - 1) ? $this->dom->root : $this->dom->nodes [$k]; |
| 448 |
// PaperG - Pass this optional parameter on to the seek function. |
| 449 |
$n->seek ( $selectors [$c] [$l], $ret, $lowercase ); |
| 450 |
} |
| 451 |
$head = $ret; |
| 452 |
} |
| 453 |
|
| 454 |
foreach ( $head as $k => $v ) { |
| 455 |
if (! isset ( $found_keys [$k] )) { |
| 456 |
$found_keys [$k] = 1; |
| 457 |
} |
| 458 |
} |
| 459 |
} |
| 460 |
|
| 461 |
// sort keys |
| 462 |
ksort ( $found_keys ); |
| 463 |
|
| 464 |
$found = array (); |
| 465 |
foreach ( $found_keys as $k => $v ) |
| 466 |
$found [] = $this->dom->nodes [$k]; |
| 467 |
|
| 468 |
// return nth-element or array |
| 469 |
if (is_null ( $idx )) |
| 470 |
return $found; |
| 471 |
else if ($idx < 0) |
| 472 |
$idx = count ( $found ) + $idx; |
| 473 |
return (isset ( $found [$idx] )) ? $found [$idx] : null; |
| 474 |
} |
| 475 |
|
| 476 |
// seek for given conditions |
| 477 |
// PaperG - added parameter to allow for case insensitive testing of the value of a selector. |
| 478 |
protected function seek($selector, &$ret, $lowercase = false) { |
| 479 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 480 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 481 |
} |
| 482 |
|
| 483 |
list ( $tag, $key, $val, $exp, $no_key ) = $selector; |
| 484 |
|
| 485 |
// xpath index |
| 486 |
if ($tag && $key && is_numeric ( $key )) { |
| 487 |
$count = 0; |
| 488 |
foreach ( $this->children as $c ) { |
| 489 |
if ($tag === '*' || $tag === $c->tag) { |
| 490 |
if (++ $count == $key) { |
| 491 |
$ret [$c->_ [GPTRANSLATE_HDOM_INFO_BEGIN]] = 1; |
| 492 |
return; |
| 493 |
} |
| 494 |
} |
| 495 |
} |
| 496 |
return; |
| 497 |
} |
| 498 |
|
| 499 |
$end = (! empty ( $this->_ [GPTRANSLATE_HDOM_INFO_END] )) ? $this->_ [GPTRANSLATE_HDOM_INFO_END] : 0; |
| 500 |
if ($end == 0) { |
| 501 |
$parent = $this->parent; |
| 502 |
while ( ! isset ( $parent->_ [GPTRANSLATE_HDOM_INFO_END] ) && $parent !== null ) { |
| 503 |
$end -= 1; |
| 504 |
$parent = $parent->parent; |
| 505 |
} |
| 506 |
$end += $parent->_ [GPTRANSLATE_HDOM_INFO_END]; |
| 507 |
} |
| 508 |
|
| 509 |
for($i = $this->_ [GPTRANSLATE_HDOM_INFO_BEGIN] + 1; $i < $end; ++ $i) { |
| 510 |
$node = $this->dom->nodes [$i]; |
| 511 |
|
| 512 |
$pass = true; |
| 513 |
|
| 514 |
if ($tag === '*' && ! $key) { |
| 515 |
if (in_array ( $node, $this->children, true )) |
| 516 |
$ret [$i] = 1; |
| 517 |
continue; |
| 518 |
} |
| 519 |
|
| 520 |
// compare tag |
| 521 |
if ($tag && $tag != $node->tag && $tag !== '*') { |
| 522 |
$pass = false; |
| 523 |
} |
| 524 |
// compare key |
| 525 |
if ($pass && $key) { |
| 526 |
if ($no_key) { |
| 527 |
if (isset ( $node->attr [$key] )) |
| 528 |
$pass = false; |
| 529 |
} else { |
| 530 |
if (($key != "plaintext") && ! isset ( $node->attr [$key] )) |
| 531 |
$pass = false; |
| 532 |
} |
| 533 |
} |
| 534 |
// compare value |
| 535 |
if ($pass && $key && $val && $val !== '*') { |
| 536 |
// If they have told us that this is a "plaintext" search then we want the plaintext of the node - right? |
| 537 |
if ($key == "plaintext") { |
| 538 |
// $node->plaintext actually returns $node->text(); |
| 539 |
$nodeKeyValue = $node->text (); |
| 540 |
} else { |
| 541 |
// this is a normal search, we want the value of that attribute of the tag. |
| 542 |
$nodeKeyValue = $node->attr [$key]; |
| 543 |
} |
| 544 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 545 |
$this->gptranslate_debug_object->debug_log ( 2, "testing node: " . $node->tag . " for attribute: " . $key . $exp . $val . " where nodes value is: " . $nodeKeyValue ); |
| 546 |
} |
| 547 |
|
| 548 |
// PaperG - If lowercase is set, do a case insensitive test of the value of the selector. |
| 549 |
if ($lowercase) { |
| 550 |
$check = $this->match ( $exp, strtolower ( $val ), strtolower ( $nodeKeyValue ) ); |
| 551 |
} else { |
| 552 |
$check = $this->match ( $exp, $val, $nodeKeyValue ); |
| 553 |
} |
| 554 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 555 |
$this->gptranslate_debug_object->debug_log ( 2, "after match: " . ($check ? "true" : "false") ); |
| 556 |
} |
| 557 |
|
| 558 |
// handle multiple class |
| 559 |
if (! $check && strcasecmp ( $key, 'class' ) === 0) { |
| 560 |
foreach ( explode ( ' ', $node->attr [$key] ) as $k ) { |
| 561 |
// Without this, there were cases where leading, trailing, or double spaces lead to our comparing blanks - bad form. |
| 562 |
if (! empty ( $k )) { |
| 563 |
if ($lowercase) { |
| 564 |
$check = $this->match ( $exp, strtolower ( $val ), strtolower ( $k ) ); |
| 565 |
} else { |
| 566 |
$check = $this->match ( $exp, $val, $k ); |
| 567 |
} |
| 568 |
if ($check) |
| 569 |
break; |
| 570 |
} |
| 571 |
} |
| 572 |
} |
| 573 |
if (! $check) |
| 574 |
$pass = false; |
| 575 |
} |
| 576 |
if ($pass) |
| 577 |
$ret [$i] = 1; |
| 578 |
unset ( $node ); |
| 579 |
} |
| 580 |
// It's passed by reference so this is actually what this function returns. |
| 581 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 582 |
$this->gptranslate_debug_object->debug_log ( 1, "EXIT - ret: ", $ret ); |
| 583 |
} |
| 584 |
} |
| 585 |
protected function match($exp, $pattern, $value) { |
| 586 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 587 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 588 |
} |
| 589 |
|
| 590 |
switch ($exp) { |
| 591 |
case '=' : |
| 592 |
return ($value === $pattern); |
| 593 |
case '!=' : |
| 594 |
return ($value !== $pattern); |
| 595 |
case '^=' : |
| 596 |
return preg_match ( "/^" . preg_quote ( $pattern, '/' ) . "/", $value ); |
| 597 |
case '$=' : |
| 598 |
return preg_match ( "/" . preg_quote ( $pattern, '/' ) . "$/", $value ); |
| 599 |
case '*=' : |
| 600 |
if ($pattern [0] == '/') { |
| 601 |
return preg_match ( $pattern, $value ); |
| 602 |
} |
| 603 |
return preg_match ( "/" . $pattern . "/i", $value ); |
| 604 |
} |
| 605 |
return false; |
| 606 |
} |
| 607 |
protected function parse_selector($selector_string) { |
| 608 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 609 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 610 |
} |
| 611 |
|
| 612 |
// pattern of CSS selectors, modified from mootools |
| 613 |
// Paperg: Add the colon to the attrbute, so that it properly finds <tag attr:ibute="something" > like google does. |
| 614 |
// Note: if you try to look at this attribute, yo MUST use getAttribute since $dom->x:y will fail the php syntax check. |
| 615 |
// Notice the \[ starting the attbute? and the @? following? This implies that an attribute can begin with an @ sign that is not captured. |
| 616 |
// This implies that an html attribute specifier may start with an @ sign that is NOT captured by the expression. |
| 617 |
// farther study is required to determine of this should be documented or removed. |
| 618 |
// $pattern = "/([\w\-:\*]*)(?:\#([\w\-]+)|\.([\w\-]+))?(?:\[@?(!?[\w\-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; |
| 619 |
$pattern = "/([\w\-:\*]*)(?:\#([\w\-]+)|\.([\w\-]+))?(?:\[@?(!?[\w\-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; |
| 620 |
preg_match_all ( $pattern, trim ( $selector_string ) . ' ', $matches, PREG_SET_ORDER ); |
| 621 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 622 |
$this->gptranslate_debug_object->debug_log ( 2, "Matches Array: ", $matches ); |
| 623 |
} |
| 624 |
|
| 625 |
$selectors = array (); |
| 626 |
$result = array (); |
| 627 |
// print_r($matches); |
| 628 |
|
| 629 |
foreach ( $matches as $m ) { |
| 630 |
$m [0] = trim ( $m [0] ); |
| 631 |
if ($m [0] === '' || $m [0] === '/' || $m [0] === '//') |
| 632 |
continue; |
| 633 |
// for browser generated xpath |
| 634 |
if ($m [1] === 'tbody') |
| 635 |
continue; |
| 636 |
|
| 637 |
list ( $tag, $key, $val, $exp, $no_key ) = array ( |
| 638 |
$m [1], |
| 639 |
'', |
| 640 |
null, |
| 641 |
'=', |
| 642 |
false |
| 643 |
); |
| 644 |
if (! empty ( $m [2] )) { |
| 645 |
$key = 'id'; |
| 646 |
$val = $m [2]; |
| 647 |
} |
| 648 |
if (! empty ( $m [3] )) { |
| 649 |
$key = 'class'; |
| 650 |
$val = $m [3]; |
| 651 |
} |
| 652 |
if (! empty ( $m [4] )) { |
| 653 |
$key = $m [4]; |
| 654 |
} |
| 655 |
if (! empty ( $m [5] )) { |
| 656 |
$exp = $m [5]; |
| 657 |
} |
| 658 |
if (! empty ( $m [6] )) { |
| 659 |
$val = $m [6]; |
| 660 |
} |
| 661 |
|
| 662 |
// convert to lowercase |
| 663 |
if ($this->dom->lowercase) { |
| 664 |
$tag = strtolower ( $tag ); |
| 665 |
$key = strtolower ( $key ); |
| 666 |
} |
| 667 |
// elements that do NOT have the specified attribute |
| 668 |
if (isset ( $key [0] ) && $key [0] === '!') { |
| 669 |
$key = substr ( $key, 1 ); |
| 670 |
$no_key = true; |
| 671 |
} |
| 672 |
|
| 673 |
$result [] = array ( |
| 674 |
$tag, |
| 675 |
$key, |
| 676 |
$val, |
| 677 |
$exp, |
| 678 |
$no_key |
| 679 |
); |
| 680 |
if (trim ( $m [7] ) === ',') { |
| 681 |
$selectors [] = $result; |
| 682 |
$result = array (); |
| 683 |
} |
| 684 |
} |
| 685 |
if (count ( $result ) > 0) |
| 686 |
$selectors [] = $result; |
| 687 |
return $selectors; |
| 688 |
} |
| 689 |
function __get($name) { |
| 690 |
if (isset ( $this->attr [$name] )) { |
| 691 |
return $this->convert_text ( $this->attr [$name] ); |
| 692 |
} |
| 693 |
switch ($name) { |
| 694 |
case 'outertext' : |
| 695 |
return $this->outertext (); |
| 696 |
case 'innertext' : |
| 697 |
return $this->innertext (); |
| 698 |
case 'plaintext' : |
| 699 |
return $this->text (); |
| 700 |
case 'xmltext' : |
| 701 |
return $this->xmltext (); |
| 702 |
default : |
| 703 |
return array_key_exists ( $name, $this->attr ); |
| 704 |
} |
| 705 |
} |
| 706 |
function __set($name, $value) { |
| 707 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 708 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 709 |
} |
| 710 |
|
| 711 |
switch ($name) { |
| 712 |
case 'outertext' : |
| 713 |
return $this->_ [GPTRANSLATE_HDOM_INFO_OUTER] = $value; |
| 714 |
case 'innertext' : |
| 715 |
if (isset ( $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] )) |
| 716 |
return $this->_ [GPTRANSLATE_HDOM_INFO_TEXT] = $value; |
| 717 |
return $this->_ [GPTRANSLATE_HDOM_INFO_INNER] = $value; |
| 718 |
} |
| 719 |
if (! isset ( $this->attr [$name] )) { |
| 720 |
$this->_ [GPTRANSLATE_HDOM_INFO_SPACE] [] = array ( |
| 721 |
' ', |
| 722 |
'', |
| 723 |
'' |
| 724 |
); |
| 725 |
$this->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [] = GPTRANSLATE_HDOM_QUOTE_DOUBLE; |
| 726 |
} |
| 727 |
$this->attr [$name] = $value; |
| 728 |
} |
| 729 |
function __isset($name) { |
| 730 |
switch ($name) { |
| 731 |
case 'outertext' : |
| 732 |
return true; |
| 733 |
case 'innertext' : |
| 734 |
return true; |
| 735 |
case 'plaintext' : |
| 736 |
return true; |
| 737 |
} |
| 738 |
// no value attr: nowrap, checked selected... |
| 739 |
return (array_key_exists ( $name, $this->attr )) ? true : isset ( $this->attr [$name] ); |
| 740 |
} |
| 741 |
function __unset($name) { |
| 742 |
if (isset ( $this->attr [$name] )) |
| 743 |
unset ( $this->attr [$name] ); |
| 744 |
} |
| 745 |
|
| 746 |
// PaperG - Function to convert the text from one character set to another if the two sets are not the same. |
| 747 |
function convert_text($text) { |
| 748 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 749 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 750 |
} |
| 751 |
|
| 752 |
$converted_text = $text; |
| 753 |
|
| 754 |
$sourceCharset = ""; |
| 755 |
$targetCharset = ""; |
| 756 |
|
| 757 |
if ($this->dom) { |
| 758 |
$sourceCharset = strtoupper ( $this->dom->_charset ); |
| 759 |
$targetCharset = strtoupper ( $this->dom->_target_charset ); |
| 760 |
} |
| 761 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 762 |
$this->gptranslate_debug_object->debug_log ( 3, "source charset: " . $sourceCharset . " target charaset: " . $targetCharset ); |
| 763 |
} |
| 764 |
|
| 765 |
if (! empty ( $sourceCharset ) && ! empty ( $targetCharset ) && (strcasecmp ( $sourceCharset, $targetCharset ) != 0)) { |
| 766 |
// Check if the reported encoding could have been incorrect and the text is actually already UTF-8 |
| 767 |
if ((strcasecmp ( $targetCharset, 'UTF-8' ) == 0) && ($this->is_utf8 ( $text ))) { |
| 768 |
$converted_text = $text; |
| 769 |
} else { |
| 770 |
$converted_text = iconv ( $sourceCharset, $targetCharset, $text ); |
| 771 |
} |
| 772 |
} |
| 773 |
|
| 774 |
// Lets make sure that we don't have that silly BOM issue with any of the utf-8 text we output. |
| 775 |
if ($targetCharset == 'UTF-8') { |
| 776 |
if (substr ( $converted_text, 0, 3 ) == "\xef\xbb\xbf") { |
| 777 |
$converted_text = substr ( $converted_text, 3 ); |
| 778 |
} |
| 779 |
if (substr ( $converted_text, - 3 ) == "\xef\xbb\xbf") { |
| 780 |
$converted_text = substr ( $converted_text, 0, - 3 ); |
| 781 |
} |
| 782 |
} |
| 783 |
|
| 784 |
return $converted_text; |
| 785 |
} |
| 786 |
|
| 787 |
/** |
| 788 |
* Returns true if $string is valid UTF-8 and false otherwise. |
| 789 |
* |
| 790 |
* @param mixed $str |
| 791 |
* String to be tested |
| 792 |
* @return boolean |
| 793 |
*/ |
| 794 |
static function is_utf8($str) { |
| 795 |
$c = 0; |
| 796 |
$b = 0; |
| 797 |
$bits = 0; |
| 798 |
$len = strlen ( $str ); |
| 799 |
for($i = 0; $i < $len; $i ++) { |
| 800 |
if($str && is_string($str)) { |
| 801 |
$c = ord ( $str [$i] ); |
| 802 |
if ($c > 128) { |
| 803 |
if (($c >= 254)) |
| 804 |
return false; |
| 805 |
elseif ($c >= 252) |
| 806 |
$bits = 6; |
| 807 |
elseif ($c >= 248) |
| 808 |
$bits = 5; |
| 809 |
elseif ($c >= 240) |
| 810 |
$bits = 4; |
| 811 |
elseif ($c >= 224) |
| 812 |
$bits = 3; |
| 813 |
elseif ($c >= 192) |
| 814 |
$bits = 2; |
| 815 |
else |
| 816 |
return false; |
| 817 |
if (($i + $bits) > $len) |
| 818 |
return false; |
| 819 |
while ( $bits > 1 ) { |
| 820 |
$i ++; |
| 821 |
$b = ord ( $str [$i] ); |
| 822 |
if ($b < 128 || $b > 191) |
| 823 |
return false; |
| 824 |
$bits --; |
| 825 |
} |
| 826 |
} |
| 827 |
} |
| 828 |
} |
| 829 |
return true; |
| 830 |
} |
| 831 |
|
| 832 |
/** |
| 833 |
* Function to try a few tricks to determine the displayed size of an img on the page. |
| 834 |
* NOTE: This will ONLY work on an IMG tag. Returns FALSE on all other tag types. |
| 835 |
* |
| 836 |
* @author John Schlick |
| 837 |
* @version April 19 2012 |
| 838 |
* @return array an array containing the 'height' and 'width' of the image on the page or -1 if we can't figure it out. |
| 839 |
*/ |
| 840 |
function get_display_size() { |
| 841 |
$width = - 1; |
| 842 |
$height = - 1; |
| 843 |
|
| 844 |
if ($this->tag !== 'img') { |
| 845 |
return false; |
| 846 |
} |
| 847 |
|
| 848 |
// See if there is aheight or width attribute in the tag itself. |
| 849 |
if (isset ( $this->attr ['width'] )) { |
| 850 |
$width = $this->attr ['width']; |
| 851 |
} |
| 852 |
|
| 853 |
if (isset ( $this->attr ['height'] )) { |
| 854 |
$height = $this->attr ['height']; |
| 855 |
} |
| 856 |
|
| 857 |
// Now look for an inline style. |
| 858 |
if (isset ( $this->attr ['style'] )) { |
| 859 |
// Thanks to user gnarf from stackoverflow for this regular expression. |
| 860 |
$attributes = array (); |
| 861 |
preg_match_all ( "/([\w\-]+)\s*:\s*([^;]+)\s*;?/", $this->attr ['style'], $matches, PREG_SET_ORDER ); |
| 862 |
foreach ( $matches as $match ) { |
| 863 |
$attributes [$match [1]] = $match [2]; |
| 864 |
} |
| 865 |
|
| 866 |
// If there is a width in the style attributes: |
| 867 |
if (isset ( $attributes ['width'] ) && $width == - 1) { |
| 868 |
// check that the last two characters are px (pixels) |
| 869 |
if (strtolower ( substr ( $attributes ['width'], - 2 ) ) == 'px') { |
| 870 |
$proposed_width = substr ( $attributes ['width'], 0, - 2 ); |
| 871 |
// Now make sure that it's an integer and not something stupid. |
| 872 |
if (filter_var ( $proposed_width, FILTER_VALIDATE_INT )) { |
| 873 |
$width = $proposed_width; |
| 874 |
} |
| 875 |
} |
| 876 |
} |
| 877 |
|
| 878 |
// If there is a width in the style attributes: |
| 879 |
if (isset ( $attributes ['height'] ) && $height == - 1) { |
| 880 |
// check that the last two characters are px (pixels) |
| 881 |
if (strtolower ( substr ( $attributes ['height'], - 2 ) ) == 'px') { |
| 882 |
$proposed_height = substr ( $attributes ['height'], 0, - 2 ); |
| 883 |
// Now make sure that it's an integer and not something stupid. |
| 884 |
if (filter_var ( $proposed_height, FILTER_VALIDATE_INT )) { |
| 885 |
$height = $proposed_height; |
| 886 |
} |
| 887 |
} |
| 888 |
} |
| 889 |
} |
| 890 |
|
| 891 |
// Future enhancement: |
| 892 |
// Look in the tag to see if there is a class or id specified that has a height or width attribute to it. |
| 893 |
|
| 894 |
// Far future enhancement |
| 895 |
// 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 |
| 896 |
// Note that in this case, the class or id will have the img subselector for it to apply to the image. |
| 897 |
|
| 898 |
// ridiculously far future development |
| 899 |
// 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. |
| 900 |
|
| 901 |
$result = array ( |
| 902 |
'height' => $height, |
| 903 |
'width' => $width |
| 904 |
); |
| 905 |
return $result; |
| 906 |
} |
| 907 |
|
| 908 |
// camel naming conventions |
| 909 |
function getAllAttributes() { |
| 910 |
return $this->attr; |
| 911 |
} |
| 912 |
function getAttribute($name) { |
| 913 |
return $this->__get ( $name ); |
| 914 |
} |
| 915 |
function setAttribute($name, $value) { |
| 916 |
$this->__set ( $name, $value ); |
| 917 |
} |
| 918 |
function hasAttribute($name) { |
| 919 |
return $this->__isset ( $name ); |
| 920 |
} |
| 921 |
function removeAttribute($name) { |
| 922 |
$this->__set ( $name, null ); |
| 923 |
} |
| 924 |
function getElementById($id) { |
| 925 |
return $this->find ( "#$id", 0 ); |
| 926 |
} |
| 927 |
function getElementsById($id, $idx = null) { |
| 928 |
return $this->find ( "#$id", $idx ); |
| 929 |
} |
| 930 |
function getElementByTagName($name) { |
| 931 |
return $this->find ( $name, 0 ); |
| 932 |
} |
| 933 |
function getElementsByTagName($name, $idx = null) { |
| 934 |
return $this->find ( $name, $idx ); |
| 935 |
} |
| 936 |
function parentNode() { |
| 937 |
return $this->parent (); |
| 938 |
} |
| 939 |
function childNodes($idx = -1) { |
| 940 |
return $this->children ( $idx ); |
| 941 |
} |
| 942 |
function firstChild() { |
| 943 |
return $this->first_child (); |
| 944 |
} |
| 945 |
function lastChild() { |
| 946 |
return $this->last_child (); |
| 947 |
} |
| 948 |
function nextSibling() { |
| 949 |
return $this->next_sibling (); |
| 950 |
} |
| 951 |
function previousSibling() { |
| 952 |
return $this->prev_sibling (); |
| 953 |
} |
| 954 |
function hasChildNodes() { |
| 955 |
return $this->has_child (); |
| 956 |
} |
| 957 |
function nodeName() { |
| 958 |
return $this->tag; |
| 959 |
} |
| 960 |
function appendChild($node) { |
| 961 |
$node->parent ( $this ); |
| 962 |
return $node; |
| 963 |
} |
| 964 |
} |
| 965 |
|
| 966 |
/** |
| 967 |
* simple html dom parser |
| 968 |
* Paperg - in the find routine: allow us to specify that we want case insensitive testing of the value of the selector. |
| 969 |
* Paperg - change $size from protected to public so we can easily access it |
| 970 |
* Paperg - added ForceTagsClosed in the constructor which tells us whether we trust the html or not. Default is to NOT trust it. |
| 971 |
* |
| 972 |
* @package PlaceLocalInclude |
| 973 |
*/ |
| 974 |
class GPTranslateSimpleHtmlDom { |
| 975 |
public $root = null; |
| 976 |
public $nodes = array (); |
| 977 |
public $callback = null; |
| 978 |
public $lowercase = false; |
| 979 |
// Used to keep track of how large the text was when we started. |
| 980 |
public $original_size; |
| 981 |
public $size; |
| 982 |
protected $pos; |
| 983 |
protected $doc; |
| 984 |
protected $char; |
| 985 |
protected $cursor; |
| 986 |
protected $parent; |
| 987 |
protected $noise = array (); |
| 988 |
protected $token_blank = " \t\r\n"; |
| 989 |
protected $token_equal = ' =/>'; |
| 990 |
protected $token_slash = " />\r\n\t"; |
| 991 |
protected $token_attr = ' >'; |
| 992 |
// Note that this is referenced by a child node, and so it needs to be public for that node to see this information. |
| 993 |
public $_charset = ''; |
| 994 |
public $_target_charset = ''; |
| 995 |
protected $default_br_text = ""; |
| 996 |
public $default_span_text = ""; |
| 997 |
protected $gptranslate_debug_object = null; |
| 998 |
|
| 999 |
// use isset instead of in_array, performance boost about 30%... |
| 1000 |
protected $self_closing_tags = array ( |
| 1001 |
'img' => 1, |
| 1002 |
'br' => 1, |
| 1003 |
'input' => 1, |
| 1004 |
'meta' => 1, |
| 1005 |
'link' => 1, |
| 1006 |
'hr' => 1, |
| 1007 |
'base' => 1, |
| 1008 |
'embed' => 1, |
| 1009 |
'spacer' => 1 |
| 1010 |
); |
| 1011 |
protected $block_tags = array ( |
| 1012 |
'root' => 1, |
| 1013 |
'body' => 1, |
| 1014 |
'form' => 1, |
| 1015 |
'div' => 1, |
| 1016 |
'span' => 1, |
| 1017 |
'table' => 1 |
| 1018 |
); |
| 1019 |
// Known sourceforge issue #2977341 |
| 1020 |
// B tags that are not closed cause us to return everything to the end of the document. |
| 1021 |
protected $optional_closing_tags = array ( |
| 1022 |
'tr' => array ( |
| 1023 |
'tr' => 1, |
| 1024 |
'td' => 1, |
| 1025 |
'th' => 1 |
| 1026 |
), |
| 1027 |
'th' => array ( |
| 1028 |
'th' => 1 |
| 1029 |
), |
| 1030 |
'td' => array ( |
| 1031 |
'td' => 1 |
| 1032 |
), |
| 1033 |
'li' => array ( |
| 1034 |
'li' => 1 |
| 1035 |
), |
| 1036 |
'dt' => array ( |
| 1037 |
'dt' => 1, |
| 1038 |
'dd' => 1 |
| 1039 |
), |
| 1040 |
'dd' => array ( |
| 1041 |
'dd' => 1, |
| 1042 |
'dt' => 1 |
| 1043 |
), |
| 1044 |
'dl' => array ( |
| 1045 |
'dd' => 1, |
| 1046 |
'dt' => 1 |
| 1047 |
), |
| 1048 |
'p' => array ( |
| 1049 |
'p' => 1 |
| 1050 |
), |
| 1051 |
'nobr' => array ( |
| 1052 |
'nobr' => 1 |
| 1053 |
), |
| 1054 |
'b' => array ( |
| 1055 |
'b' => 1 |
| 1056 |
), |
| 1057 |
'option' => array ( |
| 1058 |
'option' => 1 |
| 1059 |
) |
| 1060 |
); |
| 1061 |
function __construct($str = null, $lowercase = true, $forceTagsClosed = true, $target_charset = GPTRANSLATE_DEFAULT_TARGET_CHARSET, $stripRN = true, $defaultBRText = GPTRANSLATE_DEFAULT_BR_TEXT, $defaultSpanText = GPTRANSLATE_DEFAULT_SPAN_TEXT) { |
| 1062 |
if ($str) { |
| 1063 |
if (preg_match ( "/^http:\/\//i", $str ) || is_file ( $str )) { |
| 1064 |
$this->load_file ( $str ); |
| 1065 |
} else { |
| 1066 |
$this->load ( $str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText ); |
| 1067 |
} |
| 1068 |
} |
| 1069 |
// 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. |
| 1070 |
if (! $forceTagsClosed) { |
| 1071 |
$this->optional_closing_array = array (); |
| 1072 |
} |
| 1073 |
$this->_target_charset = $target_charset; |
| 1074 |
} |
| 1075 |
function __destruct() { |
| 1076 |
$this->clear (); |
| 1077 |
} |
| 1078 |
|
| 1079 |
// load html from string |
| 1080 |
function load($str, $lowercase = false, $stripRN = false, $defaultBRText = GPTRANSLATE_DEFAULT_BR_TEXT, $defaultSpanText = GPTRANSLATE_DEFAULT_SPAN_TEXT) { |
| 1081 |
|
| 1082 |
// prepare |
| 1083 |
$this->prepare ( $str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText ); |
| 1084 |
// strip out cdata |
| 1085 |
$this->remove_noise ( "'<!\[CDATA\[(.*?)\]\]>'is", true ); |
| 1086 |
// strip out comments |
| 1087 |
$this->remove_noise ( "'<!--(.*?)-->'is" ); |
| 1088 |
// Per sourceforge http://sourceforge.net/tracker/?func=detail&aid=2949097&group_id=218559&atid=1044037 |
| 1089 |
// Script tags removal now preceeds style tag removal. |
| 1090 |
// strip out <script> tags |
| 1091 |
$this->remove_noise ( "'<\s*script[^>]*[^/]>(.*?)<\s*/\s*script\s*>'is" ); |
| 1092 |
$this->remove_noise ( "'<\s*script\s*>(.*?)<\s*/\s*script\s*>'is" ); |
| 1093 |
// strip out <style> tags |
| 1094 |
$this->remove_noise ( "'<\s*style[^>]*[^/]>(.*?)<\s*/\s*style\s*>'is" ); |
| 1095 |
$this->remove_noise ( "'<\s*style\s*>(.*?)<\s*/\s*style\s*>'is" ); |
| 1096 |
// strip out preformatted tags |
| 1097 |
$this->remove_noise ( "'<\s*(?:code)[^>]*>(.*?)<\s*/\s*(?:code)\s*>'is" ); |
| 1098 |
// strip out server side scripts |
| 1099 |
$this->remove_noise ( "'(<\?)(.*?)(\?>)'s", true ); |
| 1100 |
// strip smarty scripts |
| 1101 |
$this->remove_noise ( "'(\{\w)(.*?)(\})'s", true ); |
| 1102 |
|
| 1103 |
// parsing |
| 1104 |
while ( $this->parse () ) |
| 1105 |
; |
| 1106 |
// end |
| 1107 |
$this->root->_ [GPTRANSLATE_HDOM_INFO_END] = $this->cursor; |
| 1108 |
$this->parse_charset (); |
| 1109 |
|
| 1110 |
// make load function chainable |
| 1111 |
return $this; |
| 1112 |
} |
| 1113 |
|
| 1114 |
// load html from file |
| 1115 |
function load_file() { |
| 1116 |
$args = func_get_args (); |
| 1117 |
$this->load ( call_user_func_array ( 'file_get_contents', $args ), true ); |
| 1118 |
// Throw an error if we can't properly load the dom. |
| 1119 |
if (($error = error_get_last ()) !== null) { |
| 1120 |
$this->clear (); |
| 1121 |
return false; |
| 1122 |
} |
| 1123 |
} |
| 1124 |
|
| 1125 |
// set callback function |
| 1126 |
function set_callback($function_name) { |
| 1127 |
$this->callback = $function_name; |
| 1128 |
} |
| 1129 |
|
| 1130 |
// remove callback function |
| 1131 |
function remove_callback() { |
| 1132 |
$this->callback = null; |
| 1133 |
} |
| 1134 |
|
| 1135 |
// save dom as string |
| 1136 |
function save($filepath = '') { |
| 1137 |
$ret = $this->root->innertext (); |
| 1138 |
if ($filepath !== '') |
| 1139 |
file_put_contents ( $filepath, $ret, LOCK_EX ); |
| 1140 |
return $ret; |
| 1141 |
} |
| 1142 |
|
| 1143 |
// find dom node by css selector |
| 1144 |
// Paperg - allow us to specify that we want case insensitive testing of the value of the selector. |
| 1145 |
function find($selector, $idx = null, $lowercase = false) { |
| 1146 |
return $this->root->find ( $selector, $idx, $lowercase ); |
| 1147 |
} |
| 1148 |
|
| 1149 |
// clean up memory due to php5 circular references memory leak... |
| 1150 |
function clear() { |
| 1151 |
foreach ( $this->nodes as $n ) { |
| 1152 |
$n->clear (); |
| 1153 |
$n = null; |
| 1154 |
} |
| 1155 |
// 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. |
| 1156 |
if (isset ( $this->children )) |
| 1157 |
foreach ( $this->children as $n ) { |
| 1158 |
$n->clear (); |
| 1159 |
$n = null; |
| 1160 |
} |
| 1161 |
if (isset ( $this->parent )) { |
| 1162 |
$this->parent->clear (); |
| 1163 |
unset ( $this->parent ); |
| 1164 |
} |
| 1165 |
if (isset ( $this->root )) { |
| 1166 |
$this->root->clear (); |
| 1167 |
unset ( $this->root ); |
| 1168 |
} |
| 1169 |
unset ( $this->doc ); |
| 1170 |
unset ( $this->noise ); |
| 1171 |
} |
| 1172 |
function dump($show_attr = true) { |
| 1173 |
$this->root->dump ( $show_attr ); |
| 1174 |
} |
| 1175 |
|
| 1176 |
// prepare HTML data and init everything |
| 1177 |
protected function prepare($str, $lowercase = true, $stripRN = true, $defaultBRText = GPTRANSLATE_DEFAULT_BR_TEXT, $defaultSpanText = GPTRANSLATE_DEFAULT_SPAN_TEXT) { |
| 1178 |
$this->clear (); |
| 1179 |
|
| 1180 |
// set the length of content before we do anything to it. |
| 1181 |
$this->size = strlen ( $str ); |
| 1182 |
// Save the original size of the html that we got in. It might be useful to someone. |
| 1183 |
$this->original_size = $this->size; |
| 1184 |
|
| 1185 |
// before we save the string as the doc... strip out the \r \n's if we are told to. |
| 1186 |
if ($stripRN) { |
| 1187 |
$str = str_replace ( "\r", " ", $str ); |
| 1188 |
$str = str_replace ( "\n", " ", $str ); |
| 1189 |
|
| 1190 |
// set the length of content since we have changed it. |
| 1191 |
$this->size = strlen ( $str ); |
| 1192 |
} |
| 1193 |
|
| 1194 |
$this->doc = $str; |
| 1195 |
$this->pos = 0; |
| 1196 |
$this->cursor = 1; |
| 1197 |
$this->noise = array (); |
| 1198 |
$this->nodes = array (); |
| 1199 |
$this->lowercase = $lowercase; |
| 1200 |
$this->default_br_text = $defaultBRText; |
| 1201 |
$this->default_span_text = $defaultSpanText; |
| 1202 |
$this->root = new GPTranslateSimpleHtmlDomNode ( $this ); |
| 1203 |
$this->root->tag = 'root'; |
| 1204 |
$this->root->_ [GPTRANSLATE_HDOM_INFO_BEGIN] = - 1; |
| 1205 |
$this->root->nodetype = GPTRANSLATE_HDOM_TYPE_ROOT; |
| 1206 |
$this->parent = $this->root; |
| 1207 |
if ($this->size > 0) |
| 1208 |
$this->char = $this->doc [0]; |
| 1209 |
} |
| 1210 |
|
| 1211 |
// parse html content |
| 1212 |
protected function parse() { |
| 1213 |
if (($s = $this->copy_until_char ( '<' )) === '') { |
| 1214 |
return $this->read_tag (); |
| 1215 |
} |
| 1216 |
|
| 1217 |
// text |
| 1218 |
$node = new GPTranslateSimpleHtmlDomNode ( $this ); |
| 1219 |
++ $this->cursor; |
| 1220 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = $s; |
| 1221 |
$this->link_nodes ( $node, false ); |
| 1222 |
return true; |
| 1223 |
} |
| 1224 |
|
| 1225 |
// 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. |
| 1226 |
// 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 |
| 1227 |
// (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. |
| 1228 |
protected function parse_charset() { |
| 1229 |
$charset = null; |
| 1230 |
|
| 1231 |
if (function_exists ( 'get_last_retrieve_url_contents_content_type' )) { |
| 1232 |
$contentTypeHeader = get_last_retrieve_url_contents_content_type (); |
| 1233 |
$success = preg_match ( '/charset=(.+)/', $contentTypeHeader, $matches ); |
| 1234 |
if ($success) { |
| 1235 |
$charset = $matches [1]; |
| 1236 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1237 |
$this->gptranslate_debug_object->debug_log ( 2, 'header content-type found charset of: ' . $charset ); |
| 1238 |
} |
| 1239 |
} |
| 1240 |
} |
| 1241 |
|
| 1242 |
if (empty ( $charset )) { |
| 1243 |
$el = $this->root->find ( 'meta[http-equiv=Content-Type]', 0, true ); |
| 1244 |
if (! empty ( $el )) { |
| 1245 |
$fullvalue = $el->content; |
| 1246 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1247 |
$this->gptranslate_debug_object->debug_log ( 2, 'meta content-type tag found' . $fullvalue ); |
| 1248 |
} |
| 1249 |
|
| 1250 |
if (! empty ( $fullvalue )) { |
| 1251 |
$success = preg_match ( '/charset=(.+)/i', $fullvalue, $matches ); |
| 1252 |
if ($success) { |
| 1253 |
$charset = $matches [1]; |
| 1254 |
} else { |
| 1255 |
// If there is a meta tag, and they don't specify the character set, research says that it's typically ISO-8859-1 |
| 1256 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1257 |
$this->gptranslate_debug_object->debug_log ( 2, 'meta content-type tag couldn\'t be parsed. using iso-8859 default.' ); |
| 1258 |
} |
| 1259 |
$charset = 'ISO-8859-1'; |
| 1260 |
} |
| 1261 |
} |
| 1262 |
} |
| 1263 |
} |
| 1264 |
|
| 1265 |
// If we couldn't find a charset above, then lets try to detect one based on the text we got... |
| 1266 |
if (empty ( $charset )) { |
| 1267 |
// Use this in case mb_detect_charset isn't installed/loaded on this machine. |
| 1268 |
$charset = false; |
| 1269 |
if (function_exists ( 'mb_detect_encoding' )) { |
| 1270 |
// Have php try to detect the encoding from the text given to us. |
| 1271 |
$charset = mb_detect_encoding ( $this->root->plaintext . "ascii", $encoding_list = array ( |
| 1272 |
"UTF-8", |
| 1273 |
"CP1252" |
| 1274 |
) ); |
| 1275 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1276 |
$this->gptranslate_debug_object->debug_log ( 2, 'mb_detect found: ' . $charset ); |
| 1277 |
} |
| 1278 |
} |
| 1279 |
|
| 1280 |
// and if this doesn't work... then we need to just wrongheadedly assume it's UTF-8 so that we can move on - cause this will usually give us most of what we need... |
| 1281 |
if ($charset === false) { |
| 1282 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1283 |
$this->gptranslate_debug_object->debug_log ( 2, 'since mb_detect failed - using default of utf-8' ); |
| 1284 |
} |
| 1285 |
$charset = 'UTF-8'; |
| 1286 |
} |
| 1287 |
} |
| 1288 |
|
| 1289 |
// Since CP1252 is a superset, if we get one of it's subsets, we want it instead. |
| 1290 |
if ((strtolower ( $charset ) == strtolower ( 'ISO-8859-1' )) || (strtolower ( $charset ) == strtolower ( 'Latin1' )) || (strtolower ( $charset ) == strtolower ( 'Latin-1' ))) { |
| 1291 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1292 |
$this->gptranslate_debug_object->debug_log ( 2, 'replacing ' . $charset . ' with CP1252 as its a superset' ); |
| 1293 |
} |
| 1294 |
$charset = 'CP1252'; |
| 1295 |
} |
| 1296 |
|
| 1297 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1298 |
$this->gptranslate_debug_object->debug_log ( 1, 'EXIT - ' . $charset ); |
| 1299 |
} |
| 1300 |
|
| 1301 |
return $this->_charset = $charset; |
| 1302 |
} |
| 1303 |
|
| 1304 |
// read tag info |
| 1305 |
protected function read_tag() { |
| 1306 |
if ($this->char !== '<') { |
| 1307 |
$this->root->_ [GPTRANSLATE_HDOM_INFO_END] = $this->cursor; |
| 1308 |
return false; |
| 1309 |
} |
| 1310 |
$begin_tag_pos = $this->pos; |
| 1311 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1312 |
|
| 1313 |
// end tag |
| 1314 |
if ($this->char === '/') { |
| 1315 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1316 |
// This represents the change in the GPTranslateSimpleHtmlDom trunk from revision 180 to 181. |
| 1317 |
// $this->skip($this->token_blank_t); |
| 1318 |
$this->skip ( $this->token_blank ); |
| 1319 |
$tag = $this->copy_until_char ( '>' ); |
| 1320 |
|
| 1321 |
// skip attributes in end tag |
| 1322 |
if (($pos = strpos ( $tag, ' ' )) !== false) |
| 1323 |
$tag = substr ( $tag, 0, $pos ); |
| 1324 |
|
| 1325 |
$parent_lower = strtolower ( $this->parent->tag ); |
| 1326 |
$tag_lower = strtolower ( $tag ); |
| 1327 |
|
| 1328 |
if ($parent_lower !== $tag_lower) { |
| 1329 |
if (isset ( $this->optional_closing_tags [$parent_lower] ) && isset ( $this->block_tags [$tag_lower] )) { |
| 1330 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1331 |
$org_parent = $this->parent; |
| 1332 |
|
| 1333 |
while ( ($this->parent->parent) && strtolower ( $this->parent->tag ) !== $tag_lower ) |
| 1334 |
$this->parent = $this->parent->parent; |
| 1335 |
|
| 1336 |
if (strtolower ( $this->parent->tag ) !== $tag_lower) { |
| 1337 |
$this->parent = $org_parent; // restore origonal parent |
| 1338 |
if ($this->parent->parent) |
| 1339 |
$this->parent = $this->parent->parent; |
| 1340 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = $this->cursor; |
| 1341 |
return $this->as_text_node ( $tag ); |
| 1342 |
} |
| 1343 |
} else if (($this->parent->parent) && isset ( $this->block_tags [$tag_lower] )) { |
| 1344 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1345 |
$org_parent = $this->parent; |
| 1346 |
|
| 1347 |
while ( ($this->parent->parent) && strtolower ( $this->parent->tag ) !== $tag_lower ) |
| 1348 |
$this->parent = $this->parent->parent; |
| 1349 |
|
| 1350 |
if (strtolower ( $this->parent->tag ) !== $tag_lower) { |
| 1351 |
$this->parent = $org_parent; // restore origonal parent |
| 1352 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = $this->cursor; |
| 1353 |
return $this->as_text_node ( $tag ); |
| 1354 |
} |
| 1355 |
} else if (($this->parent->parent) && strtolower ( $this->parent->parent->tag ) === $tag_lower) { |
| 1356 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1357 |
$this->parent = $this->parent->parent; |
| 1358 |
} else |
| 1359 |
return $this->as_text_node ( $tag ); |
| 1360 |
} |
| 1361 |
|
| 1362 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = $this->cursor; |
| 1363 |
if ($this->parent->parent) |
| 1364 |
$this->parent = $this->parent->parent; |
| 1365 |
|
| 1366 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1367 |
return true; |
| 1368 |
} |
| 1369 |
|
| 1370 |
$node = new GPTranslateSimpleHtmlDomNode ( $this ); |
| 1371 |
$node->_ [GPTRANSLATE_HDOM_INFO_BEGIN] = $this->cursor; |
| 1372 |
++ $this->cursor; |
| 1373 |
$tag = $this->copy_until ( $this->token_slash ); |
| 1374 |
$node->tag_start = $begin_tag_pos; |
| 1375 |
|
| 1376 |
// doctype, cdata & comments... |
| 1377 |
if (isset ( $tag [0] ) && $tag [0] === '!') { |
| 1378 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until_char ( '>' ); |
| 1379 |
|
| 1380 |
if (isset ( $tag [2] ) && $tag [1] === '-' && $tag [2] === '-') { |
| 1381 |
$node->nodetype = GPTRANSLATE_HDOM_TYPE_COMMENT; |
| 1382 |
$node->tag = 'comment'; |
| 1383 |
} else { |
| 1384 |
$node->nodetype = GPTRANSLATE_HDOM_TYPE_UNKNOWN; |
| 1385 |
$node->tag = 'unknown'; |
| 1386 |
} |
| 1387 |
if ($this->char === '>') |
| 1388 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] .= '>'; |
| 1389 |
$this->link_nodes ( $node, true ); |
| 1390 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1391 |
return true; |
| 1392 |
} |
| 1393 |
|
| 1394 |
// text |
| 1395 |
if ($pos = strpos ( $tag, '<' ) !== false) { |
| 1396 |
$tag = '<' . substr ( $tag, 0, - 1 ); |
| 1397 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = $tag; |
| 1398 |
$this->link_nodes ( $node, false ); |
| 1399 |
$this->char = $this->doc [-- $this->pos]; // prev |
| 1400 |
return true; |
| 1401 |
} |
| 1402 |
|
| 1403 |
if (! preg_match ( "/^[\w\-:]+$/", $tag )) { |
| 1404 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = '<' . $tag . $this->copy_until ( '<>' ); |
| 1405 |
if ($this->char === '<') { |
| 1406 |
$this->link_nodes ( $node, false ); |
| 1407 |
return true; |
| 1408 |
} |
| 1409 |
|
| 1410 |
if ($this->char === '>') |
| 1411 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] .= '>'; |
| 1412 |
$this->link_nodes ( $node, false ); |
| 1413 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1414 |
return true; |
| 1415 |
} |
| 1416 |
|
| 1417 |
// begin tag |
| 1418 |
$node->nodetype = GPTRANSLATE_HDOM_TYPE_ELEMENT; |
| 1419 |
$tag_lower = strtolower ( $tag ); |
| 1420 |
$node->tag = ($this->lowercase) ? $tag_lower : $tag; |
| 1421 |
|
| 1422 |
// handle optional closing tags |
| 1423 |
if (isset ( $this->optional_closing_tags [$tag_lower] )) { |
| 1424 |
while ( isset ( $this->optional_closing_tags [$tag_lower] [strtolower ( $this->parent->tag )] ) ) { |
| 1425 |
$this->parent->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1426 |
$this->parent = $this->parent->parent; |
| 1427 |
} |
| 1428 |
$node->parent = $this->parent; |
| 1429 |
} |
| 1430 |
|
| 1431 |
$guard = 0; // prevent infinity loop |
| 1432 |
$space = array ( |
| 1433 |
$this->copy_skip ( $this->token_blank ), |
| 1434 |
'', |
| 1435 |
'' |
| 1436 |
); |
| 1437 |
|
| 1438 |
// attributes |
| 1439 |
do { |
| 1440 |
if ($this->char !== null && $space [0] === '') { |
| 1441 |
break; |
| 1442 |
} |
| 1443 |
$name = $this->copy_until ( $this->token_equal ); |
| 1444 |
if ($guard === $this->pos) { |
| 1445 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1446 |
continue; |
| 1447 |
} |
| 1448 |
$guard = $this->pos; |
| 1449 |
|
| 1450 |
// handle endless '<' |
| 1451 |
if ($this->pos >= $this->size - 1 && $this->char !== '>') { |
| 1452 |
$node->nodetype = GPTRANSLATE_HDOM_TYPE_TEXT; |
| 1453 |
$node->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1454 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = '<' . $tag . $space [0] . $name; |
| 1455 |
$node->tag = 'text'; |
| 1456 |
$this->link_nodes ( $node, false ); |
| 1457 |
return true; |
| 1458 |
} |
| 1459 |
|
| 1460 |
// handle mismatch '<' |
| 1461 |
if ($this->doc [$this->pos - 1] == '<') { |
| 1462 |
$node->nodetype = GPTRANSLATE_HDOM_TYPE_TEXT; |
| 1463 |
$node->tag = 'text'; |
| 1464 |
$node->attr = array (); |
| 1465 |
$node->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1466 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = substr ( $this->doc, $begin_tag_pos, $this->pos - $begin_tag_pos - 1 ); |
| 1467 |
$this->pos -= 2; |
| 1468 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1469 |
$this->link_nodes ( $node, false ); |
| 1470 |
return true; |
| 1471 |
} |
| 1472 |
|
| 1473 |
if ($name !== '/' && $name !== '') { |
| 1474 |
$space [1] = $this->copy_skip ( $this->token_blank ); |
| 1475 |
$name = $this->restore_noise ( $name ); |
| 1476 |
if ($this->lowercase) |
| 1477 |
$name = strtolower ( $name ); |
| 1478 |
if ($this->char === '=') { |
| 1479 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1480 |
$this->parse_attr ( $node, $name, $space ); |
| 1481 |
} else { |
| 1482 |
// no value attr: nowrap, checked selected... |
| 1483 |
$node->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [] = GPTRANSLATE_HDOM_QUOTE_NO; |
| 1484 |
$node->attr [$name] = true; |
| 1485 |
if ($this->char != '>') |
| 1486 |
$this->char = $this->doc [-- $this->pos]; // prev |
| 1487 |
} |
| 1488 |
$node->_ [GPTRANSLATE_HDOM_INFO_SPACE] [] = $space; |
| 1489 |
$space = array ( |
| 1490 |
$this->copy_skip ( $this->token_blank ), |
| 1491 |
'', |
| 1492 |
'' |
| 1493 |
); |
| 1494 |
} else |
| 1495 |
break; |
| 1496 |
} while ( $this->char !== '>' && $this->char !== '/' ); |
| 1497 |
|
| 1498 |
$this->link_nodes ( $node, true ); |
| 1499 |
$node->_ [GPTRANSLATE_HDOM_INFO_ENDSPACE] = $space [0]; |
| 1500 |
|
| 1501 |
// check self closing |
| 1502 |
if ($this->copy_until_char_escape ( '>' ) === '/') { |
| 1503 |
$node->_ [GPTRANSLATE_HDOM_INFO_ENDSPACE] .= '/'; |
| 1504 |
$node->_ [GPTRANSLATE_HDOM_INFO_END] = 0; |
| 1505 |
} else { |
| 1506 |
// reset parent |
| 1507 |
if (! isset ( $this->self_closing_tags [strtolower ( $node->tag )] )) |
| 1508 |
$this->parent = $node; |
| 1509 |
} |
| 1510 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1511 |
|
| 1512 |
// If it's a BR tag, we need to set it's text to the default text. |
| 1513 |
// This way when we see it in plaintext, we can generate formatting that the user wants. |
| 1514 |
// since a br tag never has sub nodes, this works well. |
| 1515 |
if ($node->tag == "br") { |
| 1516 |
$node->_ [GPTRANSLATE_HDOM_INFO_INNER] = $this->default_br_text; |
| 1517 |
} |
| 1518 |
|
| 1519 |
return true; |
| 1520 |
} |
| 1521 |
|
| 1522 |
// parse attributes |
| 1523 |
protected function parse_attr($node, $name, &$space) { |
| 1524 |
// Per sourceforge: http://sourceforge.net/tracker/?func=detail&aid=3061408&group_id=218559&atid=1044037 |
| 1525 |
// If the attribute is already defined inside a tag, only pay atetntion to the first one as opposed to the last one. |
| 1526 |
if (isset ( $node->attr [$name] )) { |
| 1527 |
return; |
| 1528 |
} |
| 1529 |
|
| 1530 |
$space [2] = $this->copy_skip ( $this->token_blank ); |
| 1531 |
switch ($this->char) { |
| 1532 |
case '"' : |
| 1533 |
$node->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [] = GPTRANSLATE_HDOM_QUOTE_DOUBLE; |
| 1534 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1535 |
$node->attr [$name] = $this->restore_noise ( $this->copy_until_char_escape ( '"' ) ); |
| 1536 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1537 |
break; |
| 1538 |
case '\'' : |
| 1539 |
$node->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [] = GPTRANSLATE_HDOM_QUOTE_SINGLE; |
| 1540 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1541 |
$node->attr [$name] = $this->restore_noise ( $this->copy_until_char_escape ( '\'' ) ); |
| 1542 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1543 |
break; |
| 1544 |
default : |
| 1545 |
$node->_ [GPTRANSLATE_HDOM_INFO_QUOTE] [] = GPTRANSLATE_HDOM_QUOTE_NO; |
| 1546 |
$node->attr [$name] = $this->restore_noise ( $this->copy_until ( $this->token_attr ) ); |
| 1547 |
} |
| 1548 |
// PaperG: Attributes should not have \r or \n in them, that counts as html whitespace. |
| 1549 |
$node->attr [$name] = str_replace ( "\r", "", $node->attr [$name] ); |
| 1550 |
$node->attr [$name] = str_replace ( "\n", "", $node->attr [$name] ); |
| 1551 |
// 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. |
| 1552 |
if ($name == "class") { |
| 1553 |
$node->attr [$name] = trim ( $node->attr [$name] ); |
| 1554 |
} |
| 1555 |
} |
| 1556 |
|
| 1557 |
// link node's parent |
| 1558 |
protected function link_nodes(&$node, $is_child) { |
| 1559 |
$node->parent = $this->parent; |
| 1560 |
$this->parent->nodes [] = $node; |
| 1561 |
if ($is_child) { |
| 1562 |
$this->parent->children [] = $node; |
| 1563 |
} |
| 1564 |
} |
| 1565 |
|
| 1566 |
// as a text node |
| 1567 |
protected function as_text_node($tag) { |
| 1568 |
$node = new GPTranslateSimpleHtmlDomNode ( $this ); |
| 1569 |
++ $this->cursor; |
| 1570 |
$node->_ [GPTRANSLATE_HDOM_INFO_TEXT] = '</' . $tag . '>'; |
| 1571 |
$this->link_nodes ( $node, false ); |
| 1572 |
$this->char = (++ $this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1573 |
return true; |
| 1574 |
} |
| 1575 |
protected function skip($chars) { |
| 1576 |
$this->pos += strspn ( $this->doc, $chars, $this->pos ); |
| 1577 |
$this->char = ($this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1578 |
} |
| 1579 |
protected function copy_skip($chars) { |
| 1580 |
$pos = $this->pos; |
| 1581 |
$len = strspn ( $this->doc, $chars, $pos ); |
| 1582 |
$this->pos += $len; |
| 1583 |
$this->char = ($this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1584 |
if ($len === 0) |
| 1585 |
return ''; |
| 1586 |
return substr ( $this->doc, $pos, $len ); |
| 1587 |
} |
| 1588 |
protected function copy_until($chars) { |
| 1589 |
$pos = $this->pos; |
| 1590 |
$len = strcspn ( $this->doc, $chars, $pos ); |
| 1591 |
$this->pos += $len; |
| 1592 |
$this->char = ($this->pos < $this->size) ? $this->doc [$this->pos] : null; // next |
| 1593 |
return substr ( $this->doc, $pos, $len ); |
| 1594 |
} |
| 1595 |
protected function copy_until_char($char) { |
| 1596 |
if ($this->char === null) |
| 1597 |
return ''; |
| 1598 |
|
| 1599 |
if (($pos = strpos ( $this->doc, $char, $this->pos )) === false) { |
| 1600 |
$ret = substr ( $this->doc, $this->pos, $this->size - $this->pos ); |
| 1601 |
$this->char = null; |
| 1602 |
$this->pos = $this->size; |
| 1603 |
return $ret; |
| 1604 |
} |
| 1605 |
|
| 1606 |
if ($pos === $this->pos) |
| 1607 |
return ''; |
| 1608 |
$pos_old = $this->pos; |
| 1609 |
$this->char = $this->doc [$pos]; |
| 1610 |
$this->pos = $pos; |
| 1611 |
return substr ( $this->doc, $pos_old, $pos - $pos_old ); |
| 1612 |
} |
| 1613 |
protected function copy_until_char_escape($char) { |
| 1614 |
if ($this->char === null) |
| 1615 |
return ''; |
| 1616 |
|
| 1617 |
$start = $this->pos; |
| 1618 |
while ( 1 ) { |
| 1619 |
if (($pos = strpos ( $this->doc, $char, $start )) === false) { |
| 1620 |
$ret = substr ( $this->doc, $this->pos, $this->size - $this->pos ); |
| 1621 |
$this->char = null; |
| 1622 |
$this->pos = $this->size; |
| 1623 |
return $ret; |
| 1624 |
} |
| 1625 |
|
| 1626 |
if ($pos === $this->pos) |
| 1627 |
return ''; |
| 1628 |
|
| 1629 |
if ($this->doc [$pos - 1] === '\\') { |
| 1630 |
$start = $pos + 1; |
| 1631 |
continue; |
| 1632 |
} |
| 1633 |
|
| 1634 |
$pos_old = $this->pos; |
| 1635 |
$this->char = $this->doc [$pos]; |
| 1636 |
$this->pos = $pos; |
| 1637 |
return substr ( $this->doc, $pos_old, $pos - $pos_old ); |
| 1638 |
} |
| 1639 |
} |
| 1640 |
|
| 1641 |
// remove noise from html content |
| 1642 |
// save the noise in the $this->noise array. |
| 1643 |
protected function remove_noise($pattern, $remove_tag = false) { |
| 1644 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1645 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 1646 |
} |
| 1647 |
|
| 1648 |
$count = preg_match_all ( $pattern, $this->doc, $matches, PREG_SET_ORDER | PREG_OFFSET_CAPTURE ); |
| 1649 |
|
| 1650 |
for($i = $count - 1; $i > - 1; -- $i) { |
| 1651 |
$key = '___noise___' . sprintf ( '% 5d', count ( $this->noise ) + 1000 ); |
| 1652 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1653 |
$this->gptranslate_debug_object->debug_log ( 2, 'key is: ' . $key ); |
| 1654 |
} |
| 1655 |
$idx = ($remove_tag) ? 0 : 1; |
| 1656 |
$this->noise [$key] = $matches [$i] [$idx] [0]; |
| 1657 |
$this->doc = substr_replace ( $this->doc, $key, $matches [$i] [$idx] [1], strlen ( $matches [$i] [$idx] [0] ) ); |
| 1658 |
} |
| 1659 |
|
| 1660 |
// reset the length of content |
| 1661 |
$this->size = strlen ( $this->doc ); |
| 1662 |
if ($this->size > 0) { |
| 1663 |
$this->char = $this->doc [0]; |
| 1664 |
} |
| 1665 |
} |
| 1666 |
|
| 1667 |
// restore noise to html content |
| 1668 |
function restore_noise($text) { |
| 1669 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1670 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 1671 |
} |
| 1672 |
|
| 1673 |
while ( ($pos = strpos ( $text, '___noise___' )) !== false ) { |
| 1674 |
// 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... |
| 1675 |
if (strlen ( $text ) > $pos + 15) { |
| 1676 |
$key = '___noise___' . $text [$pos + 11] . $text [$pos + 12] . $text [$pos + 13] . $text [$pos + 14] . $text [$pos + 15]; |
| 1677 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1678 |
$this->gptranslate_debug_object->debug_log ( 2, 'located key of: ' . $key ); |
| 1679 |
} |
| 1680 |
|
| 1681 |
if (isset ( $this->noise [$key] )) { |
| 1682 |
$text = substr ( $text, 0, $pos ) . $this->noise [$key] . substr ( $text, $pos + 16 ); |
| 1683 |
} else { |
| 1684 |
// do this to prevent an infinite loop. |
| 1685 |
$text = substr ( $text, 0, $pos ) . 'UNDEFINED NOISE FOR KEY: ' . $key . substr ( $text, $pos + 16 ); |
| 1686 |
} |
| 1687 |
} else { |
| 1688 |
// There is no valid key being given back to us... We must get rid of the ___noise___ or we will have a problem. |
| 1689 |
$text = substr ( $text, 0, $pos ) . 'NO NUMERIC NOISE KEY' . substr ( $text, $pos + 11 ); |
| 1690 |
} |
| 1691 |
} |
| 1692 |
return $text; |
| 1693 |
} |
| 1694 |
|
| 1695 |
// Sometimes we NEED one of the noise elements. |
| 1696 |
function search_noise($text) { |
| 1697 |
if (is_object ( $this->gptranslate_debug_object )) { |
| 1698 |
$this->gptranslate_debug_object->debug_log_entry ( 1 ); |
| 1699 |
} |
| 1700 |
|
| 1701 |
foreach ( $this->noise as $noiseElement ) { |
| 1702 |
if (strpos ( $noiseElement, $text ) !== false) { |
| 1703 |
return $noiseElement; |
| 1704 |
} |
| 1705 |
} |
| 1706 |
} |
| 1707 |
function __toString() { |
| 1708 |
return $this->root->innertext (); |
| 1709 |
} |
| 1710 |
function __get($name) { |
| 1711 |
switch ($name) { |
| 1712 |
case 'outertext' : |
| 1713 |
return $this->root->innertext (); |
| 1714 |
case 'innertext' : |
| 1715 |
return $this->root->innertext (); |
| 1716 |
case 'plaintext' : |
| 1717 |
return $this->root->text (); |
| 1718 |
case 'charset' : |
| 1719 |
return $this->_charset; |
| 1720 |
case 'target_charset' : |
| 1721 |
return $this->_target_charset; |
| 1722 |
} |
| 1723 |
} |
| 1724 |
|
| 1725 |
// camel naming conventions |
| 1726 |
function childNodes($idx = -1) { |
| 1727 |
return $this->root->childNodes ( $idx ); |
| 1728 |
} |
| 1729 |
function firstChild() { |
| 1730 |
return $this->root->first_child (); |
| 1731 |
} |
| 1732 |
function lastChild() { |
| 1733 |
return $this->root->last_child (); |
| 1734 |
} |
| 1735 |
function createElement($name, $value = null) { |
| 1736 |
return @gptranslate_simplehtmldom_str_get_html ( "<$name>$value</$name>" )->first_child (); |
| 1737 |
} |
| 1738 |
function createTextNode($value) { |
| 1739 |
return @end ( gptranslate_simplehtmldom_str_get_html ( $value )->nodes ); |
| 1740 |
} |
| 1741 |
function getElementById($id) { |
| 1742 |
return $this->find ( "#$id", 0 ); |
| 1743 |
} |
| 1744 |
function getElementsById($id, $idx = null) { |
| 1745 |
return $this->find ( "#$id", $idx ); |
| 1746 |
} |
| 1747 |
function getElementByTagName($name) { |
| 1748 |
return $this->find ( $name, 0 ); |
| 1749 |
} |
| 1750 |
function getElementsByTagName($name, $idx = -1) { |
| 1751 |
return $this->find ( $name, $idx ); |
| 1752 |
} |
| 1753 |
function loadFile() { |
| 1754 |
$args = func_get_args (); |
| 1755 |
$this->load_file ( $args ); |
| 1756 |
} |
| 1757 |
} |
