[25344] | 1 | <?php |
---|
| 2 | |
---|
[27403] | 3 | /** |
---|
| 4 | * This class provides functions for converting CSS styles into inline style attributes in your HTML code. |
---|
| 5 | * |
---|
| 6 | * For more information, please see the README.md file. |
---|
| 7 | * |
---|
| 8 | * @author Cameron Brooks |
---|
| 9 | * @author Jaime Prado |
---|
| 10 | */ |
---|
| 11 | class Emogrifier { |
---|
| 12 | /** |
---|
| 13 | * @var string |
---|
| 14 | */ |
---|
| 15 | const ENCODING = 'UTF-8'; |
---|
[25344] | 16 | |
---|
[27403] | 17 | /** |
---|
| 18 | * @var integer |
---|
| 19 | */ |
---|
| 20 | const CACHE_KEY_CSS = 0; |
---|
[25344] | 21 | |
---|
[27403] | 22 | /** |
---|
| 23 | * @var integer |
---|
| 24 | */ |
---|
| 25 | const CACHE_KEY_SELECTOR = 1; |
---|
[25344] | 26 | |
---|
[27403] | 27 | /** |
---|
| 28 | * @var integer |
---|
| 29 | */ |
---|
| 30 | const CACHE_KEY_XPATH = 2; |
---|
| 31 | |
---|
| 32 | /** |
---|
| 33 | * for calculating nth-of-type and nth-child selectors |
---|
| 34 | * |
---|
| 35 | * @var integer |
---|
| 36 | */ |
---|
[25344] | 37 | const INDEX = 0; |
---|
[27403] | 38 | |
---|
| 39 | /** |
---|
| 40 | * for calculating nth-of-type and nth-child selectors |
---|
| 41 | * |
---|
| 42 | * @var integer |
---|
| 43 | */ |
---|
[25344] | 44 | const MULTIPLIER = 1; |
---|
| 45 | |
---|
[27403] | 46 | /** |
---|
| 47 | * @var string |
---|
| 48 | */ |
---|
| 49 | const ID_ATTRIBUTE_MATCHER = '/(\\w+)?\\#([\\w\\-]+)/'; |
---|
| 50 | |
---|
| 51 | /** |
---|
| 52 | * @var string |
---|
| 53 | */ |
---|
| 54 | const CLASS_ATTRIBUTE_MATCHER = '/(\\w+|[\\*\\]])?((\\.[\\w\\-]+)+)/'; |
---|
| 55 | |
---|
| 56 | /** |
---|
| 57 | * @var string |
---|
| 58 | */ |
---|
[25344] | 59 | private $html = ''; |
---|
[27403] | 60 | |
---|
| 61 | /** |
---|
| 62 | * @var string |
---|
| 63 | */ |
---|
[25344] | 64 | private $css = ''; |
---|
| 65 | |
---|
[27403] | 66 | /** |
---|
| 67 | * @var array<string> |
---|
| 68 | */ |
---|
| 69 | private $unprocessableHtmlTags = array('wbr'); |
---|
[25344] | 70 | |
---|
[27403] | 71 | /** |
---|
| 72 | * @var array<array> |
---|
| 73 | */ |
---|
| 74 | private $caches = array( |
---|
| 75 | self::CACHE_KEY_CSS => array(), |
---|
| 76 | self::CACHE_KEY_SELECTOR => array(), |
---|
| 77 | self::CACHE_KEY_XPATH => array(), |
---|
| 78 | ); |
---|
| 79 | |
---|
| 80 | /** |
---|
| 81 | * the visited nodes with the XPath paths as array keys |
---|
| 82 | * |
---|
| 83 | * @var array<\DOMNode> |
---|
| 84 | */ |
---|
| 85 | private $visitedNodes = array(); |
---|
| 86 | |
---|
| 87 | /** |
---|
| 88 | * the styles to apply to the nodes with the XPath paths as array keys for the outer array and the attribute names/values |
---|
| 89 | * as key/value pairs for the inner array |
---|
| 90 | * |
---|
| 91 | * @var array<array><string> |
---|
| 92 | */ |
---|
| 93 | private $styleAttributesForNodes = array(); |
---|
| 94 | |
---|
| 95 | /** |
---|
| 96 | * This attribute applies to the case where you want to preserve your original text encoding. |
---|
| 97 | * |
---|
| 98 | * By default, emogrifier translates your text into HTML entities for two reasons: |
---|
| 99 | * |
---|
| 100 | * 1. Because of client incompatibilities, it is better practice to send out HTML entities rather than unicode over email. |
---|
| 101 | * |
---|
| 102 | * 2. It translates any illegal XML characters that DOMDocument cannot work with. |
---|
| 103 | * |
---|
| 104 | * If you would like to preserve your original encoding, set this attribute to TRUE. |
---|
| 105 | * |
---|
| 106 | * @var boolean |
---|
| 107 | */ |
---|
| 108 | public $preserveEncoding = FALSE; |
---|
| 109 | |
---|
| 110 | /** |
---|
| 111 | * The constructor. |
---|
| 112 | * |
---|
| 113 | * @param string $html the HTML to emogrify, must be UTF-8-encoded |
---|
| 114 | * @param string $css the CSS to merge, must be UTF-8-encoded |
---|
| 115 | */ |
---|
[25344] | 116 | public function __construct($html = '', $css = '') { |
---|
[27403] | 117 | $this->setHtml($html); |
---|
| 118 | $this->setCss($css); |
---|
| 119 | } |
---|
| 120 | |
---|
| 121 | /** |
---|
| 122 | * The destructor. |
---|
| 123 | */ |
---|
| 124 | public function __destruct() { |
---|
| 125 | $this->purgeVisitedNodes(); |
---|
| 126 | } |
---|
| 127 | |
---|
| 128 | /** |
---|
| 129 | * Sets the HTML to emogrify. |
---|
| 130 | * |
---|
| 131 | * @param string $html the HTML to emogrify, must be UTF-8-encoded |
---|
| 132 | * |
---|
| 133 | * @return void |
---|
| 134 | */ |
---|
| 135 | public function setHtml($html = '') { |
---|
[25344] | 136 | $this->html = $html; |
---|
| 137 | } |
---|
| 138 | |
---|
[27403] | 139 | /** |
---|
| 140 | * Sets the CSS to merge with the HTML. |
---|
| 141 | * |
---|
| 142 | * @param string $css the CSS to merge, must be UTF-8-encoded |
---|
| 143 | * |
---|
| 144 | * @return void |
---|
| 145 | */ |
---|
| 146 | public function setCss($css = '') { |
---|
[25344] | 147 | $this->css = $css; |
---|
| 148 | } |
---|
| 149 | |
---|
[27403] | 150 | /** |
---|
| 151 | * Clears all caches. |
---|
| 152 | * |
---|
| 153 | * @return void |
---|
| 154 | */ |
---|
| 155 | private function clearAllCaches() { |
---|
| 156 | $this->clearCache(self::CACHE_KEY_CSS); |
---|
| 157 | $this->clearCache(self::CACHE_KEY_SELECTOR); |
---|
| 158 | $this->clearCache(self::CACHE_KEY_XPATH); |
---|
| 159 | } |
---|
| 160 | |
---|
| 161 | /** |
---|
| 162 | * Clears a single cache by key. |
---|
| 163 | * |
---|
| 164 | * @param integer $key the cache key, must be CACHE_KEY_CSS, CACHE_KEY_SELECTOR or CACHE_KEY_XPATH |
---|
| 165 | * |
---|
| 166 | * @return void |
---|
| 167 | * |
---|
| 168 | * @throws \InvalidArgumentException |
---|
| 169 | */ |
---|
| 170 | private function clearCache($key) { |
---|
| 171 | $allowedCacheKeys = array(self::CACHE_KEY_CSS, self::CACHE_KEY_SELECTOR, self::CACHE_KEY_XPATH); |
---|
| 172 | if (!in_array($key, $allowedCacheKeys, TRUE)) { |
---|
[27404] | 173 | throw new InvalidArgumentException('Invalid cache key: ' . $key, 1391822035); |
---|
[25344] | 174 | } |
---|
[27403] | 175 | |
---|
| 176 | $this->caches[$key] = array(); |
---|
[25344] | 177 | } |
---|
| 178 | |
---|
[27403] | 179 | /** |
---|
| 180 | * Purges the visited nodes. |
---|
| 181 | * |
---|
| 182 | * @return void |
---|
| 183 | */ |
---|
| 184 | private function purgeVisitedNodes() { |
---|
| 185 | $this->visitedNodes = array(); |
---|
| 186 | $this->styleAttributesForNodes = array(); |
---|
[25344] | 187 | } |
---|
| 188 | |
---|
[27403] | 189 | /** |
---|
| 190 | * Marks a tag for removal. |
---|
| 191 | * |
---|
| 192 | * There are some HTML tags that DOMDocument cannot process, and it will throw an error if it encounters them. |
---|
| 193 | * In particular, DOMDocument will complain if you try to use HTML5 tags in an XHTML document. |
---|
| 194 | * |
---|
| 195 | * Note: The tags will not be removed if they have any content. |
---|
| 196 | * |
---|
| 197 | * @param string $tagName the tag name, e.g., "p" |
---|
| 198 | * |
---|
| 199 | * @return void |
---|
| 200 | */ |
---|
| 201 | public function addUnprocessableHtmlTag($tagName) { |
---|
| 202 | $this->unprocessableHtmlTags[] = $tagName; |
---|
| 203 | } |
---|
[25344] | 204 | |
---|
[27403] | 205 | /** |
---|
| 206 | * Drops a tag from the removal list. |
---|
| 207 | * |
---|
| 208 | * @param string $tagName the tag name, e.g., "p" |
---|
| 209 | * |
---|
| 210 | * @return void |
---|
| 211 | */ |
---|
| 212 | public function removeUnprocessableHtmlTag($tagName) { |
---|
| 213 | $key = array_search($tagName, $this->unprocessableHtmlTags, TRUE); |
---|
| 214 | if ($key !== FALSE) { |
---|
| 215 | unset($this->unprocessableHtmlTags[$key]); |
---|
[25344] | 216 | } |
---|
[27403] | 217 | } |
---|
[25344] | 218 | |
---|
[27403] | 219 | /** |
---|
| 220 | * Applies the CSS you submit to the HTML you submit. |
---|
| 221 | * |
---|
| 222 | * This method places the CSS inline. |
---|
| 223 | * |
---|
| 224 | * @return string |
---|
| 225 | * |
---|
| 226 | * @throws \BadMethodCallException |
---|
| 227 | */ |
---|
| 228 | public function emogrify() { |
---|
| 229 | if ($this->html === '') { |
---|
[27404] | 230 | throw new BadMethodCallException('Please set some HTML first before calling emogrify.', 1390393096); |
---|
[27403] | 231 | } |
---|
[25344] | 232 | |
---|
[27403] | 233 | $xmlDocument = $this->createXmlDocument(); |
---|
[27404] | 234 | $xpath = new DOMXPath($xmlDocument); |
---|
[27403] | 235 | $this->clearAllCaches(); |
---|
[25344] | 236 | |
---|
| 237 | // before be begin processing the CSS file, parse the document and normalize all existing CSS attributes (changes 'DISPLAY: none' to 'display: none'); |
---|
| 238 | // we wouldn't have to do this if DOMXPath supported XPath 2.0. |
---|
| 239 | // also store a reference of nodes with existing inline styles so we don't overwrite them |
---|
[27403] | 240 | $this->purgeVisitedNodes(); |
---|
[25344] | 241 | |
---|
[27403] | 242 | $nodesWithStyleAttributes = $xpath->query('//*[@style]'); |
---|
| 243 | if ($nodesWithStyleAttributes !== FALSE) { |
---|
| 244 | $callback = create_function('$m', 'return strtolower($m[0]);'); |
---|
| 245 | |
---|
| 246 | /** @var $nodeWithStyleAttribute \DOMNode */ |
---|
| 247 | foreach ($nodesWithStyleAttributes as $node) { |
---|
| 248 | $normalizedOriginalStyle = preg_replace_callback( |
---|
| 249 | '/[A-z\\-]+(?=\\:)/S', |
---|
| 250 | $callback, |
---|
| 251 | $node->getAttribute('style') |
---|
| 252 | ); |
---|
| 253 | |
---|
| 254 | // in order to not overwrite existing style attributes in the HTML, we have to save the original HTML styles |
---|
| 255 | $nodePath = $node->getNodePath(); |
---|
| 256 | if (!isset($this->styleAttributesForNodes[$nodePath])) { |
---|
| 257 | $this->styleAttributesForNodes[$nodePath] = $this->parseCssDeclarationBlock($normalizedOriginalStyle); |
---|
| 258 | $this->visitedNodes[$nodePath] = $node; |
---|
| 259 | } |
---|
| 260 | |
---|
| 261 | $node->setAttribute('style', $normalizedOriginalStyle); |
---|
[25344] | 262 | } |
---|
| 263 | } |
---|
| 264 | |
---|
| 265 | // grab any existing style blocks from the html and append them to the existing CSS |
---|
| 266 | // (these blocks should be appended so as to have precedence over conflicting styles in the existing CSS) |
---|
| 267 | $css = $this->css; |
---|
[27403] | 268 | $styleNodes = $xpath->query('//style'); |
---|
| 269 | if ($styleNodes !== FALSE) { |
---|
| 270 | /** @var $styleNode \DOMNode */ |
---|
| 271 | foreach ($styleNodes as $styleNode) { |
---|
| 272 | // append the css |
---|
| 273 | $css .= "\n\n" . $styleNode->nodeValue; |
---|
| 274 | // remove the <style> node |
---|
| 275 | $styleNode->parentNode->removeChild($styleNode); |
---|
[25344] | 276 | } |
---|
| 277 | } |
---|
| 278 | |
---|
| 279 | // filter the CSS |
---|
| 280 | $search = array( |
---|
[27403] | 281 | // get rid of css comment code |
---|
| 282 | '/\\/\\*.*\\*\\//sU', |
---|
| 283 | // strip out any import directives |
---|
| 284 | '/^\\s*@import\\s[^;]+;/misU', |
---|
| 285 | // strip any empty media enclosures |
---|
| 286 | '/^\\s*@media\\s[^{]+{\\s*}/misU', |
---|
| 287 | // strip out all media rules that are not 'screen' or 'all' (these don't apply to email) |
---|
| 288 | '/^\\s*@media\\s+((aural|braille|embossed|handheld|print|projection|speech|tty|tv)\\s*,*\\s*)+{.*}\\s*}/misU', |
---|
| 289 | // get rid of remaining media type rules |
---|
| 290 | '/^\\s*@media\\s[^{]+{(.*})\\s*}/misU', |
---|
[25344] | 291 | ); |
---|
| 292 | |
---|
| 293 | $replace = array( |
---|
| 294 | '', |
---|
| 295 | '', |
---|
| 296 | '', |
---|
| 297 | '', |
---|
| 298 | '\\1', |
---|
| 299 | ); |
---|
| 300 | |
---|
| 301 | $css = preg_replace($search, $replace, $css); |
---|
| 302 | |
---|
[27403] | 303 | $cssKey = md5($css); |
---|
| 304 | if (!isset($this->caches[self::CACHE_KEY_CSS][$cssKey])) { |
---|
[25344] | 305 | // process the CSS file for selectors and definitions |
---|
[27403] | 306 | preg_match_all('/(?:^|[^{}])\\s*([^{]+){([^}]*)}/mis', $css, $matches, PREG_SET_ORDER); |
---|
[25344] | 307 | |
---|
[27403] | 308 | $allSelectors = array(); |
---|
[25344] | 309 | foreach ($matches as $key => $selectorString) { |
---|
| 310 | // if there is a blank definition, skip |
---|
[27403] | 311 | if (!strlen(trim($selectorString[2]))) { |
---|
| 312 | continue; |
---|
| 313 | } |
---|
[25344] | 314 | |
---|
| 315 | // else split by commas and duplicate attributes so we can sort by selector precedence |
---|
[27403] | 316 | $selectors = explode(',', $selectorString[1]); |
---|
[25344] | 317 | foreach ($selectors as $selector) { |
---|
| 318 | // don't process pseudo-elements and behavioral (dynamic) pseudo-classes; ONLY allow structural pseudo-classes |
---|
[27403] | 319 | if (strpos($selector, ':') !== FALSE && !preg_match('/:\\S+\\-(child|type)\\(/i', $selector)) { |
---|
| 320 | continue; |
---|
| 321 | } |
---|
[25344] | 322 | |
---|
[27403] | 323 | $allSelectors[] = array('selector' => trim($selector), |
---|
| 324 | 'attributes' => trim($selectorString[2]), |
---|
| 325 | // keep track of where it appears in the file, since order is important |
---|
| 326 | 'line' => $key, |
---|
[25344] | 327 | ); |
---|
| 328 | } |
---|
| 329 | } |
---|
| 330 | |
---|
| 331 | // now sort the selectors by precedence |
---|
[27403] | 332 | usort($allSelectors, array($this,'sortBySelectorPrecedence')); |
---|
[25344] | 333 | |
---|
[27403] | 334 | $this->caches[self::CACHE_KEY_CSS][$cssKey] = $allSelectors; |
---|
[25344] | 335 | } |
---|
| 336 | |
---|
[27403] | 337 | foreach ($this->caches[self::CACHE_KEY_CSS][$cssKey] as $value) { |
---|
[25344] | 338 | // query the body for the xpath selector |
---|
[27403] | 339 | $nodesMatchingCssSelectors = $xpath->query($this->translateCssToXpath(trim($value['selector']))); |
---|
[25344] | 340 | |
---|
[27403] | 341 | /** @var $node \DOMNode */ |
---|
| 342 | foreach ($nodesMatchingCssSelectors as $node) { |
---|
[25344] | 343 | // if it has a style attribute, get it, process it, and append (overwrite) new stuff |
---|
| 344 | if ($node->hasAttribute('style')) { |
---|
| 345 | // break it up into an associative array |
---|
[27403] | 346 | $oldStyleDeclarations = $this->parseCssDeclarationBlock($node->getAttribute('style')); |
---|
| 347 | $newStyleDeclarations = $this->parseCssDeclarationBlock($value['attributes']); |
---|
[25344] | 348 | |
---|
| 349 | // new styles overwrite the old styles (not technically accurate, but close enough) |
---|
[27403] | 350 | $combinedArray = array_merge($oldStyleDeclarations, $newStyleDeclarations); |
---|
[25344] | 351 | $style = ''; |
---|
[27403] | 352 | foreach ($combinedArray as $attributeName => $attributeValue) { |
---|
| 353 | $style .= (strtolower($attributeName) . ':' . $attributeValue . ';'); |
---|
| 354 | } |
---|
[25344] | 355 | } else { |
---|
| 356 | // otherwise create a new style |
---|
| 357 | $style = trim($value['attributes']); |
---|
| 358 | } |
---|
| 359 | $node->setAttribute('style', $style); |
---|
| 360 | } |
---|
| 361 | } |
---|
| 362 | |
---|
| 363 | // now iterate through the nodes that contained inline styles in the original HTML |
---|
[27403] | 364 | foreach ($this->styleAttributesForNodes as $nodePath => $styleAttributesForNode) { |
---|
| 365 | $node = $this->visitedNodes[$nodePath]; |
---|
| 366 | $currentStyleAttributes = $this->parseCssDeclarationBlock($node->getAttribute('style')); |
---|
[25344] | 367 | |
---|
[27403] | 368 | $combinedArray = array_merge($currentStyleAttributes, $styleAttributesForNode); |
---|
[25344] | 369 | $style = ''; |
---|
[27403] | 370 | foreach ($combinedArray as $attributeName => $attributeValue) { |
---|
| 371 | $style .= (strtolower($attributeName) . ':' . $attributeValue . ';'); |
---|
| 372 | } |
---|
[25344] | 373 | |
---|
| 374 | $node->setAttribute('style', $style); |
---|
| 375 | } |
---|
| 376 | |
---|
| 377 | // This removes styles from your email that contain display:none. |
---|
| 378 | // We need to look for display:none, but we need to do a case-insensitive search. Since DOMDocument only supports XPath 1.0, |
---|
| 379 | // lower-case() isn't available to us. We've thus far only set attributes to lowercase, not attribute values. Consequently, we need |
---|
| 380 | // to translate() the letters that would be in 'NONE' ("NOE") to lowercase. |
---|
[27403] | 381 | $nodesWithStyleDisplayNone = $xpath->query('//*[contains(translate(translate(@style," ",""),"NOE","noe"),"display:none")]'); |
---|
[25344] | 382 | // The checks on parentNode and is_callable below ensure that if we've deleted the parent node, |
---|
| 383 | // we don't try to call removeChild on a nonexistent child node |
---|
[27403] | 384 | if ($nodesWithStyleDisplayNone->length > 0) { |
---|
| 385 | /** @var $node \DOMNode */ |
---|
| 386 | foreach ($nodesWithStyleDisplayNone as $node) { |
---|
| 387 | if ($node->parentNode && is_callable(array($node->parentNode,'removeChild'))) { |
---|
| 388 | $node->parentNode->removeChild($node); |
---|
| 389 | } |
---|
| 390 | } |
---|
| 391 | } |
---|
[25344] | 392 | |
---|
| 393 | if ($this->preserveEncoding) { |
---|
[27403] | 394 | return mb_convert_encoding($xmlDocument->saveHTML(), self::ENCODING, 'HTML-ENTITIES'); |
---|
[25344] | 395 | } else { |
---|
[27403] | 396 | return $xmlDocument->saveHTML(); |
---|
[25344] | 397 | } |
---|
| 398 | } |
---|
| 399 | |
---|
[27403] | 400 | /** |
---|
| 401 | * Creates a DOMDocument instance with the current HTML. |
---|
| 402 | * |
---|
| 403 | * @return \DOMDocument |
---|
| 404 | */ |
---|
| 405 | private function createXmlDocument() { |
---|
[27404] | 406 | $xmlDocument = new DOMDocument; |
---|
[27403] | 407 | $xmlDocument->encoding = self::ENCODING; |
---|
| 408 | $xmlDocument->strictErrorChecking = FALSE; |
---|
| 409 | $xmlDocument->formatOutput = TRUE; |
---|
| 410 | $libxmlState = libxml_use_internal_errors(TRUE); |
---|
| 411 | $xmlDocument->loadHTML($this->getUnifiedHtml()); |
---|
| 412 | libxml_clear_errors(); |
---|
| 413 | libxml_use_internal_errors($libxmlState); |
---|
| 414 | $xmlDocument->normalizeDocument(); |
---|
[25344] | 415 | |
---|
[27403] | 416 | return $xmlDocument; |
---|
[25344] | 417 | } |
---|
| 418 | |
---|
[27403] | 419 | /** |
---|
| 420 | * Returns the HTML with the non-ASCII characters converts into HTML entities and the unprocessable HTML tags removed. |
---|
| 421 | * |
---|
| 422 | * @return string the unified HTML |
---|
| 423 | * |
---|
| 424 | * @throws \BadMethodCallException |
---|
| 425 | */ |
---|
| 426 | private function getUnifiedHtml() { |
---|
| 427 | if (!empty($this->unprocessableHtmlTags)) { |
---|
| 428 | $unprocessableHtmlTags = implode('|', $this->unprocessableHtmlTags); |
---|
| 429 | $bodyWithoutUnprocessableTags = preg_replace('/<\\/?(' . $unprocessableHtmlTags . ')[^>]*>/i', '', $this->html); |
---|
| 430 | } else { |
---|
| 431 | $bodyWithoutUnprocessableTags = $this->html; |
---|
| 432 | } |
---|
| 433 | |
---|
| 434 | return mb_convert_encoding($bodyWithoutUnprocessableTags, 'HTML-ENTITIES', self::ENCODING); |
---|
| 435 | } |
---|
| 436 | |
---|
| 437 | /** |
---|
| 438 | * @param array $a |
---|
| 439 | * @param array $b |
---|
| 440 | * |
---|
| 441 | * @return integer |
---|
| 442 | */ |
---|
| 443 | private function sortBySelectorPrecedence(array $a, array $b) { |
---|
| 444 | $precedenceA = $this->getCssSelectorPrecedence($a['selector']); |
---|
| 445 | $precedenceB = $this->getCssSelectorPrecedence($b['selector']); |
---|
| 446 | |
---|
| 447 | // We want these sorted in ascending order so selectors with lesser precedence get processed first and |
---|
| 448 | // selectors with greater precedence get sorted last. |
---|
| 449 | // The parenthesis around the -1 are necessary to avoid a PHP_CodeSniffer warning about missing spaces around |
---|
| 450 | // arithmetic operators. |
---|
| 451 | // @see http://forge.typo3.org/issues/55605 |
---|
| 452 | $precedenceForEquals = ($a['line'] < $b['line'] ? (-1) : 1); |
---|
| 453 | $precedenceForNotEquals = ($precedenceA < $precedenceB ? (-1) : 1); |
---|
| 454 | return ($precedenceA === $precedenceB) ? $precedenceForEquals : $precedenceForNotEquals; |
---|
| 455 | } |
---|
| 456 | |
---|
| 457 | /** |
---|
| 458 | * @param string $selector |
---|
| 459 | * |
---|
| 460 | * @return integer |
---|
| 461 | */ |
---|
| 462 | private function getCssSelectorPrecedence($selector) { |
---|
| 463 | $selectorKey = md5($selector); |
---|
| 464 | if (!isset($this->caches[self::CACHE_KEY_SELECTOR][$selectorKey])) { |
---|
[25344] | 465 | $precedence = 0; |
---|
| 466 | $value = 100; |
---|
[27403] | 467 | // ids: worth 100, classes: worth 10, elements: worth 1 |
---|
| 468 | $search = array('\\#','\\.',''); |
---|
[25344] | 469 | |
---|
| 470 | foreach ($search as $s) { |
---|
[27403] | 471 | if (trim($selector == '')) { |
---|
| 472 | break; |
---|
| 473 | } |
---|
| 474 | $number = 0; |
---|
| 475 | $selector = preg_replace('/' . $s . '\\w+/', '', $selector, -1, $number); |
---|
| 476 | $precedence += ($value * $number); |
---|
[25344] | 477 | $value /= 10; |
---|
| 478 | } |
---|
[27403] | 479 | $this->caches[self::CACHE_KEY_SELECTOR][$selectorKey] = $precedence; |
---|
[25344] | 480 | } |
---|
| 481 | |
---|
[27403] | 482 | return $this->caches[self::CACHE_KEY_SELECTOR][$selectorKey]; |
---|
[25344] | 483 | } |
---|
| 484 | |
---|
[27403] | 485 | /** |
---|
| 486 | * Right now, we support all CSS 1 selectors and most CSS2/3 selectors. |
---|
| 487 | * |
---|
| 488 | * @see http://plasmasturm.org/log/444/ |
---|
| 489 | * |
---|
| 490 | * @param string $cssSelector |
---|
| 491 | * |
---|
| 492 | * @return string |
---|
| 493 | */ |
---|
| 494 | private function translateCssToXpath($cssSelector) { |
---|
| 495 | $cssSelector = trim($cssSelector); |
---|
| 496 | $xpathKey = md5($cssSelector); |
---|
| 497 | if (!isset($this->caches[self::CACHE_KEY_XPATH][$xpathKey])) { |
---|
[25344] | 498 | // returns an Xpath selector |
---|
| 499 | $search = array( |
---|
[27403] | 500 | // Matches any element that is a child of parent. |
---|
| 501 | '/\\s+>\\s+/', |
---|
| 502 | // Matches any element that is an adjacent sibling. |
---|
| 503 | '/\\s+\\+\\s+/', |
---|
| 504 | // Matches any element that is a descendant of an parent element element. |
---|
| 505 | '/\\s+/', |
---|
| 506 | // first-child pseudo-selector |
---|
| 507 | '/([^\\/]+):first-child/i', |
---|
| 508 | // last-child pseudo-selector |
---|
| 509 | '/([^\\/]+):last-child/i', |
---|
| 510 | // Matches element with attribute |
---|
| 511 | '/(\\w)\\[(\\w+)\\]/', |
---|
| 512 | // Matches element with EXACT attribute |
---|
| 513 | '/(\\w)\\[(\\w+)\\=[\'"]?(\\w+)[\'"]?\\]/', |
---|
[25344] | 514 | ); |
---|
| 515 | $replace = array( |
---|
[27403] | 516 | '/', |
---|
| 517 | '/following-sibling::*[1]/self::', |
---|
| 518 | '//', |
---|
| 519 | '*[1]/self::\\1', |
---|
| 520 | '*[last()]/self::\\1', |
---|
| 521 | '\\1[@\\2]', |
---|
| 522 | '\\1[@\\2="\\3"]', |
---|
[25344] | 523 | ); |
---|
| 524 | |
---|
[27403] | 525 | $cssSelector = '//' . preg_replace($search, $replace, $cssSelector); |
---|
[25344] | 526 | |
---|
[27403] | 527 | $cssSelector = preg_replace_callback(self::ID_ATTRIBUTE_MATCHER, array($this, 'matchIdAttributes'), $cssSelector); |
---|
| 528 | $cssSelector = preg_replace_callback(self::CLASS_ATTRIBUTE_MATCHER, array($this, 'matchClassAttributes'), $cssSelector); |
---|
[26972] | 529 | |
---|
[27403] | 530 | // Advanced selectors are going to require a bit more advanced emogrification. |
---|
| 531 | // When we required PHP 5.3, we could do this with closures. |
---|
| 532 | $cssSelector = preg_replace_callback( |
---|
| 533 | '/([^\\/]+):nth-child\\(\s*(odd|even|[+\-]?\\d|[+\\-]?\\d?n(\\s*[+\\-]\\s*\\d)?)\\s*\\)/i', |
---|
| 534 | array($this, 'translateNthChild'), $cssSelector |
---|
| 535 | ); |
---|
| 536 | $cssSelector = preg_replace_callback( |
---|
| 537 | '/([^\\/]+):nth-of-type\\(\s*(odd|even|[+\-]?\\d|[+\\-]?\\d?n(\\s*[+\\-]\\s*\\d)?)\\s*\\)/i', |
---|
| 538 | array($this, 'translateNthOfType'), $cssSelector |
---|
| 539 | ); |
---|
[25344] | 540 | |
---|
[27403] | 541 | $this->caches[self::CACHE_KEY_SELECTOR][$xpathKey] = $cssSelector; |
---|
[25344] | 542 | } |
---|
[27403] | 543 | return $this->caches[self::CACHE_KEY_SELECTOR][$xpathKey]; |
---|
[25344] | 544 | } |
---|
| 545 | |
---|
[27403] | 546 | /** |
---|
| 547 | * @param array $match |
---|
| 548 | * |
---|
| 549 | * @return string |
---|
| 550 | */ |
---|
| 551 | private function matchIdAttributes(array $match) { |
---|
| 552 | return (strlen($match[1]) ? $match[1] : '*') . '[@id="' . $match[2] . '"]'; |
---|
[26972] | 553 | } |
---|
| 554 | |
---|
[27403] | 555 | /** |
---|
| 556 | * @param array $match |
---|
| 557 | * |
---|
| 558 | * @return string |
---|
| 559 | */ |
---|
| 560 | private function matchClassAttributes(array $match) { |
---|
| 561 | return (strlen($match[1]) ? $match[1] : '*') . '[contains(concat(" ",@class," "),concat(" ","' . |
---|
| 562 | implode( |
---|
| 563 | '"," "))][contains(concat(" ",@class," "),concat(" ","', |
---|
| 564 | explode('.', substr($match[2], 1)) |
---|
| 565 | ) . '"," "))]'; |
---|
[26972] | 566 | } |
---|
| 567 | |
---|
[27403] | 568 | /** |
---|
| 569 | * @param array $match |
---|
| 570 | * |
---|
| 571 | * @return string |
---|
| 572 | */ |
---|
| 573 | private function translateNthChild(array $match) { |
---|
[25344] | 574 | $result = $this->parseNth($match); |
---|
| 575 | |
---|
| 576 | if (isset($result[self::MULTIPLIER])) { |
---|
| 577 | if ($result[self::MULTIPLIER] < 0) { |
---|
| 578 | $result[self::MULTIPLIER] = abs($result[self::MULTIPLIER]); |
---|
[27403] | 579 | return sprintf('*[(last() - position()) mod %u = %u]/self::%s', $result[self::MULTIPLIER], $result[self::INDEX], $match[1]); |
---|
[25344] | 580 | } else { |
---|
[27403] | 581 | return sprintf('*[position() mod %u = %u]/self::%s', $result[self::MULTIPLIER], $result[self::INDEX], $match[1]); |
---|
[25344] | 582 | } |
---|
| 583 | } else { |
---|
[27403] | 584 | return sprintf('*[%u]/self::%s', $result[self::INDEX], $match[1]); |
---|
[25344] | 585 | } |
---|
| 586 | } |
---|
| 587 | |
---|
[27403] | 588 | /** |
---|
| 589 | * @param array $match |
---|
| 590 | * |
---|
| 591 | * @return string |
---|
| 592 | */ |
---|
| 593 | private function translateNthOfType(array $match) { |
---|
[25344] | 594 | $result = $this->parseNth($match); |
---|
| 595 | |
---|
| 596 | if (isset($result[self::MULTIPLIER])) { |
---|
| 597 | if ($result[self::MULTIPLIER] < 0) { |
---|
| 598 | $result[self::MULTIPLIER] = abs($result[self::MULTIPLIER]); |
---|
[27403] | 599 | return sprintf('%s[(last() - position()) mod %u = %u]', $match[1], $result[self::MULTIPLIER], $result[self::INDEX]); |
---|
[25344] | 600 | } else { |
---|
[27403] | 601 | return sprintf('%s[position() mod %u = %u]', $match[1], $result[self::MULTIPLIER], $result[self::INDEX]); |
---|
[25344] | 602 | } |
---|
| 603 | } else { |
---|
[27403] | 604 | return sprintf('%s[%u]', $match[1], $result[self::INDEX]); |
---|
[25344] | 605 | } |
---|
| 606 | } |
---|
| 607 | |
---|
[27403] | 608 | /** |
---|
| 609 | * @param array $match |
---|
| 610 | * |
---|
| 611 | * @return array |
---|
| 612 | */ |
---|
| 613 | private function parseNth(array $match) { |
---|
[25344] | 614 | if (in_array(strtolower($match[2]), array('even','odd'))) { |
---|
| 615 | $index = strtolower($match[2]) == 'even' ? 0 : 1; |
---|
| 616 | return array(self::MULTIPLIER => 2, self::INDEX => $index); |
---|
[27403] | 617 | } elseif (stripos($match[2], 'n') === FALSE) { |
---|
| 618 | // if there is a multiplier |
---|
[25344] | 619 | $index = intval(str_replace(' ', '', $match[2])); |
---|
| 620 | return array(self::INDEX => $index); |
---|
| 621 | } else { |
---|
| 622 | if (isset($match[3])) { |
---|
[27403] | 623 | $multipleTerm = str_replace($match[3], '', $match[2]); |
---|
[25344] | 624 | $index = intval(str_replace(' ', '', $match[3])); |
---|
| 625 | } else { |
---|
[27403] | 626 | $multipleTerm = $match[2]; |
---|
[25344] | 627 | $index = 0; |
---|
| 628 | } |
---|
| 629 | |
---|
[27403] | 630 | $multiplier = str_ireplace('n', '', $multipleTerm); |
---|
[25344] | 631 | |
---|
[27403] | 632 | if (!strlen($multiplier)) { |
---|
| 633 | $multiplier = 1; |
---|
| 634 | } elseif ($multiplier == 0) { |
---|
| 635 | return array(self::INDEX => $index); |
---|
| 636 | } else { |
---|
| 637 | $multiplier = intval($multiplier); |
---|
| 638 | } |
---|
[25344] | 639 | |
---|
[27403] | 640 | while ($index < 0) { |
---|
| 641 | $index += abs($multiplier); |
---|
| 642 | } |
---|
[25344] | 643 | |
---|
| 644 | return array(self::MULTIPLIER => $multiplier, self::INDEX => $index); |
---|
| 645 | } |
---|
| 646 | } |
---|
| 647 | |
---|
[27403] | 648 | /** |
---|
| 649 | * Parses a CSS declaration block into property name/value pairs. |
---|
| 650 | * |
---|
| 651 | * Example: |
---|
| 652 | * |
---|
| 653 | * The declaration block |
---|
| 654 | * |
---|
| 655 | * "color: #000; font-weight: bold;" |
---|
| 656 | * |
---|
| 657 | * will be parsed into the following array: |
---|
| 658 | * |
---|
| 659 | * "color" => "#000" |
---|
| 660 | * "font-weight" => "bold" |
---|
| 661 | * |
---|
| 662 | * @param string $cssDeclarationBlock the CSS declaration block without the curly braces, may be empty |
---|
| 663 | * |
---|
| 664 | * @return array the CSS declarations with the property names as array keys and the property values as array values |
---|
| 665 | */ |
---|
| 666 | private function parseCssDeclarationBlock($cssDeclarationBlock) { |
---|
| 667 | $properties = array(); |
---|
| 668 | |
---|
| 669 | $declarations = explode(';', $cssDeclarationBlock); |
---|
| 670 | foreach ($declarations as $declaration) { |
---|
| 671 | $matches = array(); |
---|
| 672 | if (!preg_match('/ *([a-z\-]+) *: *([^;]+) */', $declaration, $matches)) { |
---|
| 673 | continue; |
---|
| 674 | } |
---|
| 675 | $propertyName = $matches[1]; |
---|
| 676 | $propertyValue = $matches[2]; |
---|
| 677 | $properties[$propertyName] = $propertyValue; |
---|
[25344] | 678 | } |
---|
[27403] | 679 | |
---|
| 680 | return $properties; |
---|
[25344] | 681 | } |
---|
[27403] | 682 | } |
---|