| 1 |
<?php |
| 2 |
|
| 3 |
namespace Friends\Mf2; |
| 4 |
|
| 5 |
use DOMDocument; |
| 6 |
use DOMElement; |
| 7 |
use DOMXPath; |
| 8 |
use DOMNode; |
| 9 |
use DOMNodeList; |
| 10 |
use Exception; |
| 11 |
use SplObjectStorage; |
| 12 |
use stdClass; |
| 13 |
|
| 14 |
/** |
| 15 |
* Parse Microformats2 |
| 16 |
* |
| 17 |
* Functional shortcut for the commonest cases of parsing microformats2 from HTML. |
| 18 |
* |
| 19 |
* Example usage: |
| 20 |
* |
| 21 |
* use Mf2; |
| 22 |
* $output = Mf2\parse('<span class="h-card">Barnaby Walters</span>'); |
| 23 |
* echo json_encode($output, JSON_PRETTY_PRINT); |
| 24 |
* |
| 25 |
* Produces: |
| 26 |
* |
| 27 |
* { |
| 28 |
* "items": [ |
| 29 |
* { |
| 30 |
* "type": ["h-card"], |
| 31 |
* "properties": { |
| 32 |
* "name": ["Barnaby Walters"] |
| 33 |
* } |
| 34 |
* } |
| 35 |
* ], |
| 36 |
* "rels": {} |
| 37 |
* } |
| 38 |
* |
| 39 |
* @param string|DOMDocument $input The HTML string or DOMDocument object to parse |
| 40 |
* @param string $url The URL the input document was found at, for relative URL resolution |
| 41 |
* @param bool $convertClassic whether or not to convert classic microformats |
| 42 |
* @return array Canonical MF2 array structure |
| 43 |
*/ |
| 44 |
function parse($input, $url = null, $convertClassic = true) { |
| 45 |
$parser = new Parser($input, $url); |
| 46 |
return $parser->parse($convertClassic); |
| 47 |
} |
| 48 |
|
| 49 |
/** |
| 50 |
* Fetch microformats2 |
| 51 |
* |
| 52 |
* Given a URL, fetches it (following up to 5 redirects) and, if the content-type appears to be HTML, returns the parsed |
| 53 |
* microformats2 array structure. |
| 54 |
* |
| 55 |
* Not that even if the response code was a 4XX or 5XX error, if the content-type is HTML-like then it will be parsed |
| 56 |
* all the same, as there are legitimate cases where error pages might contain useful microformats (for example a deleted |
| 57 |
* h-entry resulting in a 410 Gone page with a stub h-entry explaining the reason for deletion). Look in $curlInfo['http_code'] |
| 58 |
* for the actual value. |
| 59 |
* |
| 60 |
* @param string $url The URL to fetch |
| 61 |
* @param bool $convertClassic (optional, default true) whether or not to convert classic microformats |
| 62 |
* @param &array $curlInfo (optional) the results of curl_getinfo will be placed in this variable for debugging |
| 63 |
* @return array|null canonical microformats2 array structure on success, null on failure |
| 64 |
*/ |
| 65 |
function fetch($url, $convertClassic = true, &$curlInfo=null) { |
| 66 |
$ch = curl_init(); |
| 67 |
curl_setopt($ch, CURLOPT_URL, $url); |
| 68 |
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1); |
| 69 |
curl_setopt($ch, CURLOPT_HEADER, 0); |
| 70 |
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1); |
| 71 |
curl_setopt($ch, CURLOPT_MAXREDIRS, 5); |
| 72 |
curl_setopt($ch, CURLOPT_HTTPHEADER, array( |
| 73 |
'Accept: text/html' |
| 74 |
)); |
| 75 |
$html = curl_exec($ch); |
| 76 |
$info = $curlInfo = curl_getinfo($ch); |
| 77 |
curl_close($ch); |
| 78 |
|
| 79 |
if (strpos(strtolower($info['content_type']), 'html') === false) { |
| 80 |
// The content was not delivered as HTML, do not attempt to parse it. |
| 81 |
return null; |
| 82 |
} |
| 83 |
|
| 84 |
# ensure the final URL is used to resolve relative URLs |
| 85 |
$url = $info['url']; |
| 86 |
|
| 87 |
return parse($html, $url, $convertClassic); |
| 88 |
} |
| 89 |
|
| 90 |
/** |
| 91 |
* Unicode to HTML Entities |
| 92 |
* @param string $input String containing characters to convert into HTML entities |
| 93 |
* @return string |
| 94 |
*/ |
| 95 |
function unicodeToHtmlEntities($input) { |
| 96 |
return mb_convert_encoding($input, 'HTML-ENTITIES', mb_detect_encoding($input)); |
| 97 |
} |
| 98 |
|
| 99 |
/** |
| 100 |
* Collapse Whitespace |
| 101 |
* |
| 102 |
* Collapses any sequences of whitespace within a string into a single space |
| 103 |
* character. |
| 104 |
* |
| 105 |
* @deprecated since v0.2.3 |
| 106 |
* @param string $str |
| 107 |
* @return string |
| 108 |
*/ |
| 109 |
function collapseWhitespace($str) { |
| 110 |
return preg_replace('/[\s|\n]+/', ' ', $str); |
| 111 |
} |
| 112 |
|
| 113 |
function unicodeTrim($str) { |
| 114 |
// this is cheating. TODO: find a better way if this causes any problems |
| 115 |
$str = str_replace(mb_convert_encoding(' ', 'UTF-8', 'HTML-ENTITIES'), ' ', $str); |
| 116 |
$str = preg_replace('/^\s+/', '', $str); |
| 117 |
return preg_replace('/\s+$/', '', $str); |
| 118 |
} |
| 119 |
|
| 120 |
/** |
| 121 |
* Microformat Name From Class string |
| 122 |
* |
| 123 |
* Given the value of @class, get the relevant mf classnames (e.g. h-card, |
| 124 |
* p-name). |
| 125 |
* |
| 126 |
* @param string $class A space delimited list of classnames |
| 127 |
* @param string $prefix The prefix to look for |
| 128 |
* @return string|array The prefixed name of the first microfomats class found or false |
| 129 |
*/ |
| 130 |
function mfNamesFromClass($class, $prefix='h-') { |
| 131 |
$class = str_replace(array(' ', ' ', "\n"), ' ', $class); |
| 132 |
$classes = explode(' ', $class); |
| 133 |
$classes = preg_grep('#^(h|p|u|dt|e)-([a-z0-9]+-)?[a-z]+(-[a-z]+)*$#', $classes); |
| 134 |
$matches = array(); |
| 135 |
|
| 136 |
foreach ($classes as $classname) { |
| 137 |
$compare_classname = ' ' . $classname; |
| 138 |
$compare_prefix = ' ' . $prefix; |
| 139 |
if (strstr($compare_classname, $compare_prefix) !== false && ($compare_classname != $compare_prefix)) { |
| 140 |
$matches[] = ($prefix === 'h-') ? $classname : substr($classname, strlen($prefix)); |
| 141 |
} |
| 142 |
} |
| 143 |
|
| 144 |
return $matches; |
| 145 |
} |
| 146 |
|
| 147 |
/** |
| 148 |
* Registered with the XPath object and used within XPaths for finding root elements. |
| 149 |
* @param string $class |
| 150 |
* @return bool |
| 151 |
*/ |
| 152 |
function classHasMf2RootClassname($class) { |
| 153 |
return count(mfNamesFromClass($class, 'h-')) > 0; |
| 154 |
} |
| 155 |
|
| 156 |
/** |
| 157 |
* Get Nested µf Property Name From Class |
| 158 |
* |
| 159 |
* Returns all the p-, u-, dt- or e- prefixed classnames it finds in a |
| 160 |
* space-separated string. |
| 161 |
* |
| 162 |
* @param string $class |
| 163 |
* @return array |
| 164 |
*/ |
| 165 |
function nestedMfPropertyNamesFromClass($class) { |
| 166 |
$prefixes = array('p-', 'u-', 'dt-', 'e-'); |
| 167 |
$propertyNames = array(); |
| 168 |
|
| 169 |
foreach ($prefixes as $prefix) { |
| 170 |
$classes = mfNamesFromClass($class, $prefix); |
| 171 |
foreach ($classes as $property) { |
| 172 |
$propertyNames[$property][] = $prefix; |
| 173 |
} |
| 174 |
} |
| 175 |
|
| 176 |
foreach ($propertyNames as $property => $prefixes) { |
| 177 |
$propertyNames[$property] = array_unique($prefixes); |
| 178 |
} |
| 179 |
|
| 180 |
return $propertyNames; |
| 181 |
} |
| 182 |
|
| 183 |
/** |
| 184 |
* Wraps mfNamesFromClass to handle an element as input (common) |
| 185 |
* |
| 186 |
* @param DOMElement $e The element to get the classname for |
| 187 |
* @param string $prefix The prefix to look for |
| 188 |
* @return mixed See return value of mf2\Parser::mfNameFromClass() |
| 189 |
*/ |
| 190 |
function mfNamesFromElement(\DOMElement $e, $prefix = 'h-') { |
| 191 |
$class = $e->getAttribute('class'); |
| 192 |
return mfNamesFromClass($class, $prefix); |
| 193 |
} |
| 194 |
|
| 195 |
/** |
| 196 |
* Wraps nestedMfPropertyNamesFromClass to handle an element as input |
| 197 |
*/ |
| 198 |
function nestedMfPropertyNamesFromElement(\DOMElement $e) { |
| 199 |
$class = $e->getAttribute('class'); |
| 200 |
return nestedMfPropertyNamesFromClass($class); |
| 201 |
} |
| 202 |
|
| 203 |
/** |
| 204 |
* Converts various time formats to HH:MM |
| 205 |
* @param string $time The time to convert |
| 206 |
* @return string |
| 207 |
*/ |
| 208 |
function convertTimeFormat($time) { |
| 209 |
$hh = $mm = $ss = ''; |
| 210 |
preg_match('/(\d{1,2}):?(\d{2})?:?(\d{2})?(a\.?m\.?|p\.?m\.?)?/i', $time, $matches); |
| 211 |
|
| 212 |
// If no am/pm is specified: |
| 213 |
if (empty($matches[4])) { |
| 214 |
return $time; |
| 215 |
} else { |
| 216 |
// Otherwise, am/pm is specified. |
| 217 |
$meridiem = strtolower(str_replace('.', '', $matches[4])); |
| 218 |
|
| 219 |
// Hours. |
| 220 |
$hh = $matches[1]; |
| 221 |
|
| 222 |
// Add 12 to hours if pm applies. |
| 223 |
if ($meridiem == 'pm' && ($hh < 12)) { |
| 224 |
$hh += 12; |
| 225 |
} |
| 226 |
|
| 227 |
$hh = str_pad($hh, 2, '0', STR_PAD_LEFT); |
| 228 |
|
| 229 |
// Minutes. |
| 230 |
$mm = (empty($matches[2]) ) ? '00' : $matches[2]; |
| 231 |
|
| 232 |
// Seconds, only if supplied. |
| 233 |
if (!empty($matches[3])) { |
| 234 |
$ss = $matches[3]; |
| 235 |
} |
| 236 |
|
| 237 |
if (empty($ss)) { |
| 238 |
return sprintf('%s:%s', $hh, $mm); |
| 239 |
} |
| 240 |
else { |
| 241 |
return sprintf('%s:%s:%s', $hh, $mm, $ss); |
| 242 |
} |
| 243 |
} |
| 244 |
} |
| 245 |
|
| 246 |
/** |
| 247 |
* Normalize an ordinal date to YYYY-MM-DD |
| 248 |
* This function should only be called after validating the $dtValue |
| 249 |
* matches regex \d{4}-\d{2} |
| 250 |
* @param string $dtValue |
| 251 |
* @return string |
| 252 |
*/ |
| 253 |
function normalizeOrdinalDate($dtValue) { |
| 254 |
list($year, $day) = explode('-', $dtValue, 2); |
| 255 |
$day = intval($day); |
| 256 |
if ($day < 367 && $day > 0) { |
| 257 |
$date = \DateTime::createFromFormat('Y-z', $dtValue); |
| 258 |
$date->modify('-1 day'); # 'z' format is zero-based so need to adjust |
| 259 |
if ($date->format('Y') === $year) { |
| 260 |
return $date->format('Y-m-d'); |
| 261 |
} |
| 262 |
} |
| 263 |
return ''; |
| 264 |
} |
| 265 |
|
| 266 |
/** |
| 267 |
* If a date value has a timezone offset, normalize it. |
| 268 |
* @param string $dtValue |
| 269 |
* @return string isolated, normalized TZ offset for implied TZ for other dt- properties |
| 270 |
*/ |
| 271 |
function normalizeTimezoneOffset(&$dtValue) { |
| 272 |
preg_match('/Z|[+-]\d{1,2}:?(\d{2})?$/i', $dtValue, $matches); |
| 273 |
|
| 274 |
if (empty($matches)) { |
| 275 |
return null; |
| 276 |
} |
| 277 |
|
| 278 |
$timezoneOffset = null; |
| 279 |
|
| 280 |
if ( $matches[0] != 'Z' ) { |
| 281 |
$timezoneString = str_replace(':', '', $matches[0]); |
| 282 |
$plus_minus = substr($timezoneString, 0, 1); |
| 283 |
$timezoneOffset = substr($timezoneString, 1); |
| 284 |
if ( strlen($timezoneOffset) <= 2 ) { |
| 285 |
$timezoneOffset .= '00'; |
| 286 |
} |
| 287 |
$timezoneOffset = str_pad($timezoneOffset, 4, 0, STR_PAD_LEFT); |
| 288 |
$timezoneOffset = $plus_minus . $timezoneOffset; |
| 289 |
$dtValue = preg_replace('/Z?[+-]\d{1,2}:?(\d{2})?$/i', $timezoneOffset, $dtValue); |
| 290 |
} |
| 291 |
|
| 292 |
return $timezoneOffset; |
| 293 |
} |
| 294 |
|
| 295 |
function applySrcsetUrlTransformation($srcset, $transformation) { |
| 296 |
return implode(', ', array_filter(array_map(function ($srcsetPart) use ($transformation) { |
| 297 |
$parts = explode(" \t\n\r\0\x0B", trim($srcsetPart), 2); |
| 298 |
$parts[0] = rtrim($parts[0]); |
| 299 |
|
| 300 |
if (empty($parts[0])) { return false; } |
| 301 |
|
| 302 |
$parts[0] = call_user_func($transformation, $parts[0]); |
| 303 |
|
| 304 |
return $parts[0] . (empty($parts[1]) ? '' : ' ' . $parts[1]); |
| 305 |
}, explode(',', trim($srcset))))); |
| 306 |
} |
| 307 |
|
| 308 |
/** |
| 309 |
* Microformats2 Parser |
| 310 |
* |
| 311 |
* A class which holds state for parsing microformats2 from HTML. |
| 312 |
* |
| 313 |
* Example usage: |
| 314 |
* |
| 315 |
* use Mf2; |
| 316 |
* $parser = new Mf2\Parser('<p class="h-card">Barnaby Walters</p>'); |
| 317 |
* $output = $parser->parse(); |
| 318 |
*/ |
| 319 |
class Parser { |
| 320 |
/** @var string The baseurl (if any) to use for this parse */ |
| 321 |
public $baseurl; |
| 322 |
|
| 323 |
/** @var DOMXPath object which can be used to query over any fragment*/ |
| 324 |
public $xpath; |
| 325 |
|
| 326 |
/** @var DOMDocument */ |
| 327 |
public $doc; |
| 328 |
|
| 329 |
/** @var SplObjectStorage */ |
| 330 |
protected $parsed; |
| 331 |
|
| 332 |
/** |
| 333 |
* @var bool |
| 334 |
*/ |
| 335 |
public $jsonMode; |
| 336 |
|
| 337 |
/** @var boolean Whether to include experimental language parsing in the result */ |
| 338 |
public $lang = false; |
| 339 |
|
| 340 |
/** @var bool Whether to include alternates object (dropped from spec in favor of rel-urls) */ |
| 341 |
public $enableAlternates = false; |
| 342 |
|
| 343 |
/** |
| 344 |
* Elements upgraded to mf2 during backcompat |
| 345 |
* @var SplObjectStorage |
| 346 |
*/ |
| 347 |
protected $upgraded; |
| 348 |
|
| 349 |
/** |
| 350 |
* Whether to convert classic microformats |
| 351 |
* @var bool |
| 352 |
*/ |
| 353 |
public $convertClassic; |
| 354 |
|
| 355 |
/** |
| 356 |
* Constructor |
| 357 |
* |
| 358 |
* @param DOMDocument|string $input The data to parse. A string of HTML or a DOMDocument |
| 359 |
* @param string $url The URL of the parsed document, for relative URL resolution |
| 360 |
* @param boolean $jsonMode Whether or not to use a stdClass instance for an empty `rels` dictionary. This breaks PHP looping over rels, but allows the output to be correctly serialized as JSON. |
| 361 |
*/ |
| 362 |
public function __construct($input, $url = null, $jsonMode = false) { |
| 363 |
$emptyDocDefault = '<html><body></body></html>'; |
| 364 |
libxml_use_internal_errors(true); |
| 365 |
if (is_string($input)) { |
| 366 |
if (empty($input)) { |
| 367 |
$input = $emptyDocDefault; |
| 368 |
} |
| 369 |
|
| 370 |
if (class_exists('Masterminds\\HTML5')) { |
| 371 |
$doc = new \Masterminds\HTML5(array('disable_html_ns' => true)); |
| 372 |
$doc = $doc->loadHTML($input); |
| 373 |
} else { |
| 374 |
$doc = new DOMDocument(); |
| 375 |
@$doc->loadHTML(unicodeToHtmlEntities($input), \LIBXML_NOWARNING); |
| 376 |
} |
| 377 |
} elseif (is_a($input, 'DOMDocument')) { |
| 378 |
$doc = clone $input; |
| 379 |
} else { |
| 380 |
$doc = new DOMDocument(); |
| 381 |
@$doc->loadHTML($emptyDocDefault); |
| 382 |
} |
| 383 |
|
| 384 |
// Create an XPath object and allow some PHP functions to be used within XPath queries. |
| 385 |
$this->xpath = new DOMXPath($doc); |
| 386 |
$this->xpath->registerNamespace('php', 'http://php.net/xpath'); |
| 387 |
$this->xpath->registerPhpFunctions('\\Friends\\Mf2\\classHasMf2RootClassname'); |
| 388 |
|
| 389 |
$baseurl = $url; |
| 390 |
foreach ($this->xpath->query('//base[@href]') as $base) { |
| 391 |
$baseElementUrl = $base->getAttribute('href'); |
| 392 |
|
| 393 |
if (parse_url($baseElementUrl, PHP_URL_SCHEME) === null) { |
| 394 |
/* The base element URL is relative to the document URL. |
| 395 |
* |
| 396 |
* :/ |
| 397 |
* |
| 398 |
* Perhaps the author was high? */ |
| 399 |
|
| 400 |
$baseurl = resolveUrl($url, $baseElementUrl); |
| 401 |
} else { |
| 402 |
$baseurl = $baseElementUrl; |
| 403 |
} |
| 404 |
break; |
| 405 |
} |
| 406 |
|
| 407 |
// Ignore <template> elements as per the HTML5 spec |
| 408 |
foreach ($this->xpath->query('//template') as $templateEl) { |
| 409 |
$templateEl->parentNode->removeChild($templateEl); |
| 410 |
} |
| 411 |
|
| 412 |
$this->baseurl = $baseurl; |
| 413 |
$this->doc = $doc; |
| 414 |
$this->parsed = new SplObjectStorage(); |
| 415 |
$this->upgraded = new SplObjectStorage(); |
| 416 |
$this->jsonMode = $jsonMode; |
| 417 |
} |
| 418 |
|
| 419 |
private function elementPrefixParsed(\DOMElement $e, $prefix) { |
| 420 |
if (!$this->parsed->offsetExists($e)) |
| 421 |
$this->parsed->offsetSet($e, array()); |
| 422 |
|
| 423 |
$prefixes = $this->parsed[$e]; |
| 424 |
$prefixes[] = $prefix; |
| 425 |
$this->parsed[$e] = $prefixes; |
| 426 |
} |
| 427 |
|
| 428 |
/** |
| 429 |
* Determine if the element has already been parsed |
| 430 |
* @param DOMElement $e |
| 431 |
* @param string $prefix |
| 432 |
* @return bool |
| 433 |
*/ |
| 434 |
private function isElementParsed(\DOMElement $e, $prefix) { |
| 435 |
if (!$this->parsed->offsetExists($e)) { |
| 436 |
return false; |
| 437 |
} |
| 438 |
|
| 439 |
$prefixes = $this->parsed[$e]; |
| 440 |
|
| 441 |
if (!in_array($prefix, $prefixes)) { |
| 442 |
return false; |
| 443 |
} |
| 444 |
|
| 445 |
return true; |
| 446 |
} |
| 447 |
|
| 448 |
/** |
| 449 |
* Determine if the element's specified property has already been upgraded during backcompat |
| 450 |
* @param DOMElement $el |
| 451 |
* @param string $property |
| 452 |
* @return bool |
| 453 |
*/ |
| 454 |
private function isElementUpgraded(\DOMElement $el, $property) { |
| 455 |
if ( $this->upgraded->offsetExists($el) ) { |
| 456 |
if ( in_array($property, $this->upgraded[$el]) ) { |
| 457 |
return true; |
| 458 |
} |
| 459 |
} |
| 460 |
|
| 461 |
return false; |
| 462 |
} |
| 463 |
|
| 464 |
private function resolveChildUrls(DOMElement $el) { |
| 465 |
$hyperlinkChildren = $this->xpath->query('.//*[@src or @href or @data]', $el); |
| 466 |
|
| 467 |
foreach ($hyperlinkChildren as $child) { |
| 468 |
if ($child->hasAttribute('href')) |
| 469 |
$child->setAttribute('href', $this->resolveUrl($child->getAttribute('href'))); |
| 470 |
if ($child->hasAttribute('src')) |
| 471 |
$child->setAttribute('src', $this->resolveUrl($child->getAttribute('src'))); |
| 472 |
if ($child->hasAttribute('srcset')) |
| 473 |
$child->setAttribute('srcset', applySrcsetUrlTransformation($child->getAttribute('href'), array($this, 'resolveUrl'))); |
| 474 |
if ($child->hasAttribute('data')) |
| 475 |
$child->setAttribute('data', $this->resolveUrl($child->getAttribute('data'))); |
| 476 |
} |
| 477 |
} |
| 478 |
|
| 479 |
/** |
| 480 |
* The following two methods implements plain text parsing. |
| 481 |
* @param DOMElement $element |
| 482 |
* @param bool $implied |
| 483 |
* @see https://wiki.zegnat.net/media/textparsing.html |
| 484 |
**/ |
| 485 |
public function textContent(DOMElement $element, $implied=false) |
| 486 |
{ |
| 487 |
return preg_replace( |
| 488 |
'/(^[\t\n\f\r ]+| +(?=\n)|(?<=\n) +| +(?= )|[\t\n\f\r ]+$)/', |
| 489 |
'', |
| 490 |
$this->elementToString($element, $implied) |
| 491 |
); |
| 492 |
} |
| 493 |
private function elementToString(DOMElement $input, $implied=false) |
| 494 |
{ |
| 495 |
$output = ''; |
| 496 |
foreach ($input->childNodes as $child) { |
| 497 |
if ($child->nodeType === XML_TEXT_NODE) { |
| 498 |
$output .= str_replace(array("\t", "\n", "\r") , ' ', $child->textContent); |
| 499 |
} else if ($child->nodeType === XML_ELEMENT_NODE) { |
| 500 |
$tagName = strtoupper($child->tagName); |
| 501 |
if (in_array($tagName, array('SCRIPT', 'STYLE'))) { |
| 502 |
continue; |
| 503 |
} else if ($tagName === 'IMG') { |
| 504 |
if ($child->hasAttribute('alt')) { |
| 505 |
$output .= ' ' . trim($child->getAttribute('alt'), "\t\n\f\r ") . ' '; |
| 506 |
} else if (!$implied && $child->hasAttribute('src')) { |
| 507 |
$output .= ' ' . $this->resolveUrl(trim($child->getAttribute('src'), "\t\n\f\r ")) . ' '; |
| 508 |
} |
| 509 |
} else if ($tagName === 'BR') { |
| 510 |
$output .= "\n"; |
| 511 |
} else if ($tagName === 'P') { |
| 512 |
$output .= "\n" . $this->elementToString($child); |
| 513 |
} else { |
| 514 |
$output .= $this->elementToString($child); |
| 515 |
} |
| 516 |
} |
| 517 |
} |
| 518 |
return $output; |
| 519 |
} |
| 520 |
|
| 521 |
/** |
| 522 |
* Given an img property, parse its value and/or alt text |
| 523 |
* @param DOMElement $el |
| 524 |
* @access public |
| 525 |
* @return string|array |
| 526 |
*/ |
| 527 |
public function parseImg(DOMElement $el) |
| 528 |
{ |
| 529 |
if ($el->hasAttribute('alt')) { |
| 530 |
return [ |
| 531 |
'value' => $this->resolveUrl( $el->getAttribute('src') ), |
| 532 |
'alt' => $el->getAttribute('alt') |
| 533 |
]; |
| 534 |
} |
| 535 |
return $el->getAttribute('src'); |
| 536 |
} |
| 537 |
/** |
| 538 |
* This method parses the language of an element |
| 539 |
* @param DOMElement $el |
| 540 |
* @access public |
| 541 |
* @return string |
| 542 |
*/ |
| 543 |
public function language(DOMElement $el) |
| 544 |
{ |
| 545 |
// element has a lang attribute; use it |
| 546 |
if ($el->hasAttribute('lang')) { |
| 547 |
return unicodeTrim($el->getAttribute('lang')); |
| 548 |
} |
| 549 |
|
| 550 |
if ($el->tagName == 'html') { |
| 551 |
// we're at the <html> element and no lang; check <meta> http-equiv Content-Language |
| 552 |
foreach ( $this->xpath->query('.//meta[@http-equiv]') as $node ) |
| 553 |
{ |
| 554 |
if ($node->hasAttribute('http-equiv') && $node->hasAttribute('content') && strtolower($node->getAttribute('http-equiv')) == 'content-language') { |
| 555 |
return unicodeTrim($node->getAttribute('content')); |
| 556 |
} |
| 557 |
} |
| 558 |
} elseif ($el->parentNode instanceof DOMElement) { |
| 559 |
// check the parent node |
| 560 |
return $this->language($el->parentNode); |
| 561 |
} |
| 562 |
|
| 563 |
return ''; |
| 564 |
} # end method language() |
| 565 |
|
| 566 |
// TODO: figure out if this has problems with sms: and geo: URLs |
| 567 |
public function resolveUrl($url) { |
| 568 |
// If not a string then return. |
| 569 |
if (!is_string($url)){ |
| 570 |
return $url; |
| 571 |
} |
| 572 |
// If the URL is seriously malformed it’s probably beyond the scope of this |
| 573 |
// parser to try to do anything with it. |
| 574 |
if (parse_url($url) === false) { |
| 575 |
return $url; |
| 576 |
} |
| 577 |
|
| 578 |
// per issue #40 valid URLs could have a space on either side |
| 579 |
$url = trim($url); |
| 580 |
|
| 581 |
$scheme = parse_url($url, PHP_URL_SCHEME); |
| 582 |
|
| 583 |
if (empty($scheme) and !empty($this->baseurl)) { |
| 584 |
return resolveUrl($this->baseurl, $url); |
| 585 |
} else { |
| 586 |
return $url; |
| 587 |
} |
| 588 |
} |
| 589 |
|
| 590 |
// Parsing Functions |
| 591 |
|
| 592 |
/** |
| 593 |
* Parse value-class/value-title on an element, joining with $separator if |
| 594 |
* there are multiple. |
| 595 |
* |
| 596 |
* @param \DOMElement $e |
| 597 |
* @param string $separator = '' if multiple value-title elements, join with this string |
| 598 |
* @return string|null the parsed value or null if value-class or -title aren’t in use |
| 599 |
*/ |
| 600 |
public function parseValueClassTitle(\DOMElement $e, $separator = '') { |
| 601 |
$valueClassElements = $this->xpath->query('./*[contains(concat(" ", normalize-space(@class), " "), " value ")]', $e); |
| 602 |
|
| 603 |
if ($valueClassElements->length !== 0) { |
| 604 |
// Process value-class stuff |
| 605 |
$val = ''; |
| 606 |
foreach ($valueClassElements as $el) { |
| 607 |
$val .= $this->textContent($el); |
| 608 |
} |
| 609 |
|
| 610 |
return unicodeTrim($val); |
| 611 |
} |
| 612 |
|
| 613 |
$valueTitleElements = $this->xpath->query('./*[contains(concat(" ", normalize-space(@class), " "), " value-title ")]', $e); |
| 614 |
|
| 615 |
if ($valueTitleElements->length !== 0) { |
| 616 |
// Process value-title stuff |
| 617 |
$val = ''; |
| 618 |
foreach ($valueTitleElements as $el) { |
| 619 |
$val .= $el->getAttribute('title'); |
| 620 |
} |
| 621 |
|
| 622 |
return unicodeTrim($val); |
| 623 |
} |
| 624 |
|
| 625 |
// No value-title or -class in this element |
| 626 |
return null; |
| 627 |
} |
| 628 |
|
| 629 |
/** |
| 630 |
* Given an element with class="p-*", get its value |
| 631 |
* |
| 632 |
* @param DOMElement $p The element to parse |
| 633 |
* @return string The plaintext value of $p, dependant on type |
| 634 |
* @todo Make this adhere to value-class |
| 635 |
*/ |
| 636 |
public function parseP(\DOMElement $p) { |
| 637 |
$classTitle = $this->parseValueClassTitle($p, ' '); |
| 638 |
|
| 639 |
if ($classTitle !== null) { |
| 640 |
return $classTitle; |
| 641 |
} |
| 642 |
|
| 643 |
$this->resolveChildUrls($p); |
| 644 |
|
| 645 |
if ($p->tagName == 'img' and $p->hasAttribute('alt')) { |
| 646 |
$pValue = $p->getAttribute('alt'); |
| 647 |
} elseif ($p->tagName == 'area' and $p->hasAttribute('alt')) { |
| 648 |
$pValue = $p->getAttribute('alt'); |
| 649 |
} elseif (($p->tagName == 'abbr' or $p->tagName == 'link') and $p->hasAttribute('title')) { |
| 650 |
$pValue = $p->getAttribute('title'); |
| 651 |
} elseif (in_array($p->tagName, array('data', 'input')) and $p->hasAttribute('value')) { |
| 652 |
$pValue = $p->getAttribute('value'); |
| 653 |
} else { |
| 654 |
$pValue = $this->textContent($p); |
| 655 |
} |
| 656 |
|
| 657 |
return $pValue; |
| 658 |
} |
| 659 |
|
| 660 |
/** |
| 661 |
* Given an element with class="u-*", get the value of the URL |
| 662 |
* |
| 663 |
* @param DOMElement $u The element to parse |
| 664 |
* @return string The plaintext value of $u, dependant on type |
| 665 |
* @todo make this adhere to value-class |
| 666 |
*/ |
| 667 |
public function parseU(\DOMElement $u) { |
| 668 |
if (($u->tagName == 'a' or $u->tagName == 'area' or $u->tagName == 'link') and $u->hasAttribute('href')) { |
| 669 |
$uValue = $u->getAttribute('href'); |
| 670 |
} elseif ( $u->tagName == 'img' and $u->hasAttribute('src') ) { |
| 671 |
$uValue = $this->parseImg($u); |
| 672 |
if ( is_array( $uValue ) && isset( $uValue['value'] ) ) { |
| 673 |
$uValue = $uValue['value']; |
| 674 |
} |
| 675 |
} elseif (in_array($u->tagName, array('audio', 'video', 'source', 'iframe')) and $u->hasAttribute('src')) { |
| 676 |
$uValue = $u->getAttribute('src'); |
| 677 |
} elseif ($u->tagName == 'video' and !$u->hasAttribute('src') and $u->hasAttribute('poster')) { |
| 678 |
$uValue = $u->getAttribute('poster'); |
| 679 |
} elseif ($u->tagName == 'object' and $u->hasAttribute('data')) { |
| 680 |
$uValue = $u->getAttribute('data'); |
| 681 |
} elseif (($classTitle = $this->parseValueClassTitle($u)) !== null) { |
| 682 |
$uValue = $classTitle; |
| 683 |
} elseif (($u->tagName == 'abbr' or $u->tagName == 'link') and $u->hasAttribute('title')) { |
| 684 |
$uValue = $u->getAttribute('title'); |
| 685 |
} elseif (in_array($u->tagName, array('data', 'input')) and $u->hasAttribute('value')) { |
| 686 |
$uValue = $u->getAttribute('value'); |
| 687 |
} else { |
| 688 |
$uValue = $this->textContent($u); |
| 689 |
} |
| 690 |
|
| 691 |
return $this->resolveUrl($uValue); |
| 692 |
} |
| 693 |
|
| 694 |
/** |
| 695 |
* Given an element with class="dt-*", get the value of the datetime as a php date object |
| 696 |
* |
| 697 |
* @param DOMElement $dt The element to parse |
| 698 |
* @param array $dates Array of dates processed so far |
| 699 |
* @param string $impliedTimezone |
| 700 |
* @return string The datetime string found |
| 701 |
*/ |
| 702 |
public function parseDT(\DOMElement $dt, &$dates = array(), &$impliedTimezone = null) { |
| 703 |
// Check for value-class pattern |
| 704 |
$valueClassChildren = $this->xpath->query('./*[contains(concat(" ", normalize-space(@class), " "), " value ") or contains(concat(" ", normalize-space(@class), " "), " value-title ")]', $dt); |
| 705 |
$dtValue = false; |
| 706 |
|
| 707 |
if ($valueClassChildren->length > 0) { |
| 708 |
// They’re using value-class |
| 709 |
$dateParts = array(); |
| 710 |
|
| 711 |
foreach ($valueClassChildren as $e) { |
| 712 |
if (strstr(' ' . $e->getAttribute('class') . ' ', ' value-title ')) { |
| 713 |
$title = $e->getAttribute('title'); |
| 714 |
if (!empty($title)) { |
| 715 |
$dateParts[] = $title; |
| 716 |
} |
| 717 |
} |
| 718 |
elseif ($e->tagName == 'img' or $e->tagName == 'area') { |
| 719 |
// Use @alt |
| 720 |
$alt = $e->getAttribute('alt'); |
| 721 |
if (!empty($alt)) { |
| 722 |
$dateParts[] = $alt; |
| 723 |
} |
| 724 |
} |
| 725 |
elseif ($e->tagName == 'data') { |
| 726 |
// Use @value, otherwise innertext |
| 727 |
$value = $e->hasAttribute('value') ? $e->getAttribute('value') : unicodeTrim($e->nodeValue); |
| 728 |
if (!empty($value)) { |
| 729 |
$dateParts[] = $value; |
| 730 |
} |
| 731 |
} |
| 732 |
elseif ($e->tagName == 'abbr') { |
| 733 |
// Use @title, otherwise innertext |
| 734 |
$title = $e->hasAttribute('title') ? $e->getAttribute('title') : unicodeTrim($e->nodeValue); |
| 735 |
if (!empty($title)) { |
| 736 |
$dateParts[] = $title; |
| 737 |
} |
| 738 |
} |
| 739 |
elseif ($e->tagName == 'del' or $e->tagName == 'ins' or $e->tagName == 'time') { |
| 740 |
// Use @datetime if available, otherwise innertext |
| 741 |
$dtAttr = ($e->hasAttribute('datetime')) ? $e->getAttribute('datetime') : unicodeTrim($e->nodeValue); |
| 742 |
if (!empty($dtAttr)) { |
| 743 |
$dateParts[] = $dtAttr; |
| 744 |
} |
| 745 |
} |
| 746 |
else { |
| 747 |
if (!empty($e->nodeValue)) { |
| 748 |
$dateParts[] = unicodeTrim($e->nodeValue); |
| 749 |
} |
| 750 |
} |
| 751 |
} |
| 752 |
|
| 753 |
// Look through dateParts |
| 754 |
$datePart = ''; |
| 755 |
$timePart = ''; |
| 756 |
$timezonePart = ''; |
| 757 |
foreach ($dateParts as $part) { |
| 758 |
// Is this part a full ISO8601 datetime? |
| 759 |
if (preg_match('/^\d{4}-\d{2}-\d{2}[ T]\d{2}:\d{2}(:\d{2})?(Z|[+-]\d{2}:?\d{2})?$/', $part)) { |
| 760 |
// Break completely, we’ve got our value. |
| 761 |
$dtValue = $part; |
| 762 |
break; |
| 763 |
} else { |
| 764 |
// Is the current part a valid time(+TZ?) AND no other time representation has been found? |
| 765 |
if ((preg_match('/^\d{1,2}:\d{2}(:\d{2})?(Z|[+-]\d{1,2}:?\d{2})?$/', $part) or preg_match('/^\d{1,2}(:\d{2})?(:\d{2})?[ap]\.?m\.?$/i', $part)) and empty($timePart)) { |
| 766 |
$timePart = $part; |
| 767 |
|
| 768 |
$timezoneOffset = normalizeTimezoneOffset($timePart); |
| 769 |
if (!$impliedTimezone && $timezoneOffset) { |
| 770 |
$impliedTimezone = $timezoneOffset; |
| 771 |
} |
| 772 |
// Is the current part a valid date AND no other date representation has been found? |
| 773 |
} elseif (preg_match('/^\d{4}-\d{2}-\d{2}$/', $part) and empty($datePart)) { |
| 774 |
$datePart = $part; |
| 775 |
// Is the current part a valid ordinal date AND no other date representation has been found? |
| 776 |
} elseif (preg_match('/^\d{4}-\d{3}$/', $part) and empty($datePart)) { |
| 777 |
$datePart = normalizeOrdinalDate($part); |
| 778 |
// Is the current part a valid timezone offset AND no other timezone part has been found? |
| 779 |
} elseif (preg_match('/^(Z|[+-]\d{1,2}:?(\d{2})?)$/', $part) and empty($timezonePart)) { |
| 780 |
$timezonePart = $part; |
| 781 |
|
| 782 |
$timezoneOffset = normalizeTimezoneOffset($timezonePart); |
| 783 |
if (!$impliedTimezone && $timezoneOffset) { |
| 784 |
$impliedTimezone = $timezoneOffset; |
| 785 |
} |
| 786 |
// Current part already represented by other VCP parts; do nothing with it |
| 787 |
} else { |
| 788 |
continue; |
| 789 |
} |
| 790 |
|
| 791 |
if ( !empty($datePart) && !in_array($datePart, $dates) ) { |
| 792 |
$dates[] = $datePart; |
| 793 |
} |
| 794 |
|
| 795 |
if (!empty($timezonePart) && !empty($timePart)) { |
| 796 |
$timePart .= $timezonePart; |
| 797 |
} |
| 798 |
|
| 799 |
$dtValue = ''; |
| 800 |
|
| 801 |
if ( empty($datePart) && !empty($timePart) ) { |
| 802 |
$timePart = convertTimeFormat($timePart); |
| 803 |
$dtValue = unicodeTrim($timePart); |
| 804 |
} |
| 805 |
else if ( !empty($datePart) && empty($timePart) ) { |
| 806 |
$dtValue = rtrim($datePart, 'T'); |
| 807 |
} |
| 808 |
else { |
| 809 |
$timePart = convertTimeFormat($timePart); |
| 810 |
$dtValue = rtrim($datePart, 'T') . ' ' . unicodeTrim($timePart); |
| 811 |
} |
| 812 |
} |
| 813 |
} |
| 814 |
} else { |
| 815 |
// Not using value-class (phew). |
| 816 |
if ($dt->tagName == 'img' or $dt->tagName == 'area') { |
| 817 |
// Use @alt |
| 818 |
// Is it an entire dt? |
| 819 |
$alt = $dt->getAttribute('alt'); |
| 820 |
if (!empty($alt)) { |
| 821 |
$dtValue = $alt; |
| 822 |
} |
| 823 |
} elseif (in_array($dt->tagName, array('data'))) { |
| 824 |
// Use @value, otherwise innertext |
| 825 |
// Is it an entire dt? |
| 826 |
$value = $dt->getAttribute('value'); |
| 827 |
if (!empty($value)) { |
| 828 |
$dtValue = $value; |
| 829 |
} |
| 830 |
else { |
| 831 |
$dtValue = $this->textContent($dt); |
| 832 |
} |
| 833 |
} elseif ($dt->tagName == 'abbr') { |
| 834 |
// Use @title, otherwise innertext |
| 835 |
// Is it an entire dt? |
| 836 |
$title = $dt->getAttribute('title'); |
| 837 |
if (!empty($title)) { |
| 838 |
$dtValue = $title; |
| 839 |
} |
| 840 |
else { |
| 841 |
$dtValue = $this->textContent($dt); |
| 842 |
} |
| 843 |
} elseif ($dt->tagName == 'del' or $dt->tagName == 'ins' or $dt->tagName == 'time') { |
| 844 |
// Use @datetime if available, otherwise innertext |
| 845 |
// Is it an entire dt? |
| 846 |
$dtAttr = $dt->getAttribute('datetime'); |
| 847 |
if (!empty($dtAttr)) { |
| 848 |
$dtValue = $dtAttr; |
| 849 |
} |
| 850 |
else { |
| 851 |
$dtValue = $this->textContent($dt); |
| 852 |
} |
| 853 |
|
| 854 |
} else { |
| 855 |
$dtValue = $this->textContent($dt); |
| 856 |
} |
| 857 |
|
| 858 |
// if the dtValue is not just YYYY-MM-DD |
| 859 |
if (!preg_match('/^(\d{4}-\d{2}-\d{2})$/', $dtValue)) { |
| 860 |
// no implied timezone set and dtValue has a TZ offset, use un-normalized TZ offset |
| 861 |
preg_match('/Z|[+-]\d{1,2}:?(\d{2})?$/i', $dtValue, $matches); |
| 862 |
if (!$impliedTimezone && !empty($matches[0])) { |
| 863 |
$impliedTimezone = $matches[0]; |
| 864 |
} |
| 865 |
} |
| 866 |
|
| 867 |
$dtValue = unicodeTrim($dtValue); |
| 868 |
|
| 869 |
// Store the date part so that we can use it when assembling the final timestamp if the next one is missing a date part |
| 870 |
if (preg_match('/(\d{4}-\d{2}-\d{2})/', $dtValue, $matches)) { |
| 871 |
$dates[] = $matches[0]; |
| 872 |
} |
| 873 |
} |
| 874 |
|
| 875 |
/** |
| 876 |
* if $dtValue is only a time and there are recently parsed dates, |
| 877 |
* form the full date-time using the most recently parsed dt- value |
| 878 |
*/ |
| 879 |
if ((preg_match('/^\d{1,2}:\d{2}(:\d{2})?(Z|[+-]\d{2}:?\d{2}?)?$/', $dtValue) or preg_match('/^\d{1,2}(:\d{2})?(:\d{2})?[ap]\.?m\.?$/i', $dtValue)) && !empty($dates)) { |
| 880 |
$timezoneOffset = normalizeTimezoneOffset($dtValue); |
| 881 |
if (!$impliedTimezone && $timezoneOffset) { |
| 882 |
$impliedTimezone = $timezoneOffset; |
| 883 |
} |
| 884 |
|
| 885 |
$dtValue = convertTimeFormat($dtValue); |
| 886 |
$dtValue = end($dates) . ' ' . unicodeTrim($dtValue); |
| 887 |
} |
| 888 |
|
| 889 |
return $dtValue; |
| 890 |
} |
| 891 |
|
| 892 |
/** |
| 893 |
* Given the root element of some embedded markup, return a string representing that markup |
| 894 |
* |
| 895 |
* @param DOMElement $e The element to parse |
| 896 |
* @return string $e’s innerHTML |
| 897 |
* |
| 898 |
* @todo need to mark this element as e- parsed so it doesn’t get parsed as it’s parent’s e-* too |
| 899 |
*/ |
| 900 |
public function parseE(\DOMElement $e) { |
| 901 |
$classTitle = $this->parseValueClassTitle($e); |
| 902 |
|
| 903 |
if ($classTitle !== null) |
| 904 |
return $classTitle; |
| 905 |
|
| 906 |
// Expand relative URLs within children of this element |
| 907 |
// TODO: as it is this is not relative to only children, make this .// and rerun tests |
| 908 |
$this->resolveChildUrls($e); |
| 909 |
|
| 910 |
// Temporarily move all descendants into a separate DocumentFragment. |
| 911 |
// This way we can DOMDocument::saveHTML on the entire collection at once. |
| 912 |
// Running DOMDocument::saveHTML per node may add whitespace that isn't in source. |
| 913 |
// See https://stackoverflow.com/q/38317903 |
| 914 |
if ($innerNodes = $e->ownerDocument->createDocumentFragment()) { |
| 915 |
while ($e->hasChildNodes()) { |
| 916 |
$innerNodes->appendChild($e->firstChild); |
| 917 |
} |
| 918 |
$html = $e->ownerDocument->saveHtml($innerNodes); |
| 919 |
// Put the nodes back in place. |
| 920 |
if ($innerNodes->hasChildNodes()) { |
| 921 |
$e->appendChild($innerNodes); |
| 922 |
} |
| 923 |
} |
| 924 |
|
| 925 |
$return = array( |
| 926 |
'html' => unicodeTrim($html), |
| 927 |
'value' => $this->textContent($e), |
| 928 |
); |
| 929 |
|
| 930 |
if($this->lang) { |
| 931 |
// Language |
| 932 |
if ( $html_lang = $this->language($e) ) { |
| 933 |
$return['lang'] = $html_lang; |
| 934 |
} |
| 935 |
} |
| 936 |
|
| 937 |
return $return; |
| 938 |
} |
| 939 |
|
| 940 |
private function removeTags(\DOMElement &$e, $tagName) { |
| 941 |
while(($r = $e->getElementsByTagName($tagName)) && $r->length) { |
| 942 |
$r->item(0)->parentNode->removeChild($r->item(0)); |
| 943 |
} |
| 944 |
} |
| 945 |
|
| 946 |
/** |
| 947 |
* Recursively parse microformats |
| 948 |
* |
| 949 |
* @param DOMElement $e The element to parse |
| 950 |
* @param bool $is_backcompat Whether using backcompat parsing or not |
| 951 |
* @param bool $has_nested_mf Whether this microformat has a nested microformat |
| 952 |
* @return array A representation of the values contained within microformat $e |
| 953 |
*/ |
| 954 |
public function parseH(\DOMElement $e, $is_backcompat = false, $has_nested_mf = false) { |
| 955 |
// If it’s already been parsed (e.g. is a child mf), skip |
| 956 |
if ($this->parsed->offsetExists($e)) { |
| 957 |
return null; |
| 958 |
} |
| 959 |
|
| 960 |
// Get current µf name |
| 961 |
$mfTypes = mfNamesFromElement($e, 'h-'); |
| 962 |
|
| 963 |
if (!$mfTypes) { |
| 964 |
return null; |
| 965 |
} |
| 966 |
|
| 967 |
// Initalise var to store the representation in |
| 968 |
$return = array(); |
| 969 |
$children = array(); |
| 970 |
$dates = array(); |
| 971 |
$prefixes = array(); |
| 972 |
$impliedTimezone = null; |
| 973 |
|
| 974 |
if($e->tagName == 'area') { |
| 975 |
$coords = $e->getAttribute('coords'); |
| 976 |
$shape = $e->getAttribute('shape'); |
| 977 |
} |
| 978 |
|
| 979 |
// Handle p-* |
| 980 |
foreach ($this->xpath->query('.//*[contains(concat(" ", normalize-space(@class)) ," p-")]', $e) as $p) { |
| 981 |
// element is already parsed |
| 982 |
if ($this->isElementParsed($p, 'p')) { |
| 983 |
continue; |
| 984 |
// backcompat parsing and element was not upgraded; skip it |
| 985 |
} else if ( $is_backcompat && empty($this->upgraded[$p]) ) { |
| 986 |
$this->elementPrefixParsed($p, 'p'); |
| 987 |
continue; |
| 988 |
} |
| 989 |
|
| 990 |
$prefixes[] = 'p-'; |
| 991 |
$pValue = $this->parseP($p); |
| 992 |
|
| 993 |
// Add the value to the array for it’s p- properties |
| 994 |
foreach (mfNamesFromElement($p, 'p-') as $propName) { |
| 995 |
if (!empty($propName)) { |
| 996 |
$return[$propName][] = $pValue; |
| 997 |
} |
| 998 |
} |
| 999 |
|
| 1000 |
// Make sure this sub-mf won’t get parsed as a top level mf |
| 1001 |
$this->elementPrefixParsed($p, 'p'); |
| 1002 |
} |
| 1003 |
|
| 1004 |
// Handle u-* |
| 1005 |
foreach ($this->xpath->query('.//*[contains(concat(" ", normalize-space(@class))," u-")]', $e) as $u) { |
| 1006 |
// element is already parsed |
| 1007 |
if ($this->isElementParsed($u, 'u')) { |
| 1008 |
continue; |
| 1009 |
// backcompat parsing and element was not upgraded; skip it |
| 1010 |
} else if ( $is_backcompat && empty($this->upgraded[$u]) ) { |
| 1011 |
$this->elementPrefixParsed($u, 'u'); |
| 1012 |
continue; |
| 1013 |
} |
| 1014 |
|
| 1015 |
$prefixes[] = 'u-'; |
| 1016 |
$uValue = $this->parseU($u); |
| 1017 |
|
| 1018 |
// Add the value to the array for it’s property types |
| 1019 |
foreach (mfNamesFromElement($u, 'u-') as $propName) { |
| 1020 |
$return[$propName][] = $uValue; |
| 1021 |
} |
| 1022 |
|
| 1023 |
// Make sure this sub-mf won’t get parsed as a top level mf |
| 1024 |
$this->elementPrefixParsed($u, 'u'); |
| 1025 |
} |
| 1026 |
|
| 1027 |
$temp_dates = array(); |
| 1028 |
|
| 1029 |
// Handle dt-* |
| 1030 |
foreach ($this->xpath->query('.//*[contains(concat(" ", normalize-space(@class)), " dt-")]', $e) as $dt) { |
| 1031 |
// element is already parsed |
| 1032 |
if ($this->isElementParsed($dt, 'dt')) { |
| 1033 |
continue; |
| 1034 |
// backcompat parsing and element was not upgraded; skip it |
| 1035 |
} else if ( $is_backcompat && empty($this->upgraded[$dt]) ) { |
| 1036 |
$this->elementPrefixParsed($dt, 'dt'); |
| 1037 |
continue; |
| 1038 |
} |
| 1039 |
|
| 1040 |
$prefixes[] = 'dt-'; |
| 1041 |
$dtValue = $this->parseDT($dt, $dates, $impliedTimezone); |
| 1042 |
|
| 1043 |
if ($dtValue) { |
| 1044 |
// Add the value to the array for dt- properties |
| 1045 |
foreach (mfNamesFromElement($dt, 'dt-') as $propName) { |
| 1046 |
$temp_dates[$propName][] = $dtValue; |
| 1047 |
} |
| 1048 |
} |
| 1049 |
// Make sure this sub-mf won’t get parsed as a top level mf |
| 1050 |
$this->elementPrefixParsed($dt, 'dt'); |
| 1051 |
} |
| 1052 |
|
| 1053 |
foreach ($temp_dates as $propName => $data) { |
| 1054 |
foreach ( $data as $dtValue ) { |
| 1055 |
// var_dump(preg_match('/[+-]\d{2}(\d{2})?$/i', $dtValue)); |
| 1056 |
if ( $impliedTimezone && preg_match('/(Z|[+-]\d{2}:?(\d{2})?)$/i', $dtValue, $matches) == 0 ) { |
| 1057 |
$dtValue .= $impliedTimezone; |
| 1058 |
} |
| 1059 |
|
| 1060 |
$return[$propName][] = $dtValue; |
| 1061 |
} |
| 1062 |
} |
| 1063 |
|
| 1064 |
// Handle e-* |
| 1065 |
foreach ($this->xpath->query('.//*[contains(concat(" ", normalize-space(@class))," e-")]', $e) as $em) { |
| 1066 |
// element is already parsed |
| 1067 |
if ($this->isElementParsed($em, 'e')) { |
| 1068 |
continue; |
| 1069 |
// backcompat parsing and element was not upgraded; skip it |
| 1070 |
} else if ( $is_backcompat && empty($this->upgraded[$em]) ) { |
| 1071 |
$this->elementPrefixParsed($em, 'e'); |
| 1072 |
continue; |
| 1073 |
} |
| 1074 |
|
| 1075 |
$prefixes[] = 'e-'; |
| 1076 |
$eValue = $this->parseE($em); |
| 1077 |
|
| 1078 |
if ($eValue) { |
| 1079 |
// Add the value to the array for e- properties |
| 1080 |
foreach (mfNamesFromElement($em, 'e-') as $propName) { |
| 1081 |
$return[$propName][] = $eValue; |
| 1082 |
} |
| 1083 |
} |
| 1084 |
// Make sure this sub-mf won’t get parsed as a top level mf |
| 1085 |
$this->elementPrefixParsed($em, 'e'); |
| 1086 |
} |
| 1087 |
|
| 1088 |
// Do we need to imply a name property? |
| 1089 |
// if no explicit "name" property, and no other p-* or e-* properties, and no nested microformats, |
| 1090 |
if (!array_key_exists('name', $return) && !in_array('p-', $prefixes) |
| 1091 |
&& !in_array('e-', $prefixes) && !$has_nested_mf |
| 1092 |
&& !$is_backcompat && empty($this->upgraded[$e])) { |
| 1093 |
$name = false; |
| 1094 |
// img.h-x[alt] or area.h-x[alt] |
| 1095 |
if (($e->tagName === 'img' || $e->tagName === 'area') && $e->hasAttribute('alt')) { |
| 1096 |
$name = $e->getAttribute('alt'); |
| 1097 |
// abbr.h-x[title] |
| 1098 |
} elseif ($e->tagName === 'abbr' && $e->hasAttribute('title')) { |
| 1099 |
$name = $e->getAttribute('title'); |
| 1100 |
} else { |
| 1101 |
$xpaths = array( |
| 1102 |
// .h-x>img:only-child[alt]:not([alt=""]):not[.h-*] |
| 1103 |
'./img[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and @alt and string-length(@alt) != 0]', |
| 1104 |
// .h-x>area:only-child[alt]:not([alt=""]):not[.h-*] |
| 1105 |
'./area[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and @alt and string-length(@alt) != 0]', |
| 1106 |
// .h-x>abbr:only-child[title]:not([title=""]):not[.h-*] |
| 1107 |
'./abbr[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and @title and string-length(@title) != 0]', |
| 1108 |
// .h-x>:only-child:not[.h-*]>img:only-child[alt]:not([alt=""]):not[.h-*] |
| 1109 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(*) = 1]/img[not(contains(concat(" ", @class), " h-")) and @alt and string-length(@alt) != 0]', |
| 1110 |
// .h-x>:only-child:not[.h-*]>area:only-child[alt]:not([alt=""]):not[.h-*] |
| 1111 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(*) = 1]/area[not(contains(concat(" ", @class), " h-")) and @alt and string-length(@alt) != 0]', |
| 1112 |
// .h-x>:only-child:not[.h-*]>abbr:only-child[title]:not([title=""]):not[.h-*] |
| 1113 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(*) = 1]/abbr[not(contains(concat(" ", @class), " h-")) and @title and string-length(@title) != 0]' |
| 1114 |
); |
| 1115 |
foreach ($xpaths as $xpath) { |
| 1116 |
$nameElement = $this->xpath->query($xpath, $e); |
| 1117 |
if ($nameElement !== false && $nameElement->length === 1) { |
| 1118 |
$nameElement = $nameElement->item(0); |
| 1119 |
if ($nameElement->tagName === 'img' || $nameElement->tagName === 'area') { |
| 1120 |
$name = $nameElement->getAttribute('alt'); |
| 1121 |
} else { |
| 1122 |
$name = $nameElement->getAttribute('title'); |
| 1123 |
} |
| 1124 |
break; |
| 1125 |
} |
| 1126 |
} |
| 1127 |
} |
| 1128 |
if ($name === false) { |
| 1129 |
$name = $this->textContent($e, true); |
| 1130 |
} |
| 1131 |
$return['name'][] = unicodeTrim($name); |
| 1132 |
} |
| 1133 |
|
| 1134 |
// Check for u-photo |
| 1135 |
if (!array_key_exists('photo', $return) && !in_array('u-', $prefixes) && !$has_nested_mf && !$is_backcompat) { |
| 1136 |
$photo = $this->parseImpliedPhoto($e); |
| 1137 |
if ($photo !== false) { |
| 1138 |
$return['photo'][] = $photo; |
| 1139 |
} |
| 1140 |
} |
| 1141 |
|
| 1142 |
// Do we need to imply a url property? |
| 1143 |
// if no explicit "url" property, and no other explicit u-* properties, and no nested microformats |
| 1144 |
if (!array_key_exists('url', $return) && !in_array('u-', $prefixes) && !$has_nested_mf && !$is_backcompat) { |
| 1145 |
// a.h-x[href] or area.h-x[href] |
| 1146 |
if (($e->tagName === 'a' || $e->tagName === 'area') && $e->hasAttribute('href')) { |
| 1147 |
$return['url'][] = $this->resolveUrl($e->getAttribute('href')); |
| 1148 |
} else { |
| 1149 |
$xpaths = array( |
| 1150 |
// .h-x>a[href]:only-of-type:not[.h-*] |
| 1151 |
'./a[not(contains(concat(" ", @class), " h-")) and count(../a) = 1 and @href]', |
| 1152 |
// .h-x>area[href]:only-of-type:not[.h-*] |
| 1153 |
'./area[not(contains(concat(" ", @class), " h-")) and count(../area) = 1 and @href]', |
| 1154 |
// .h-x>:only-child:not[.h-*]>a[href]:only-of-type:not[.h-*] |
| 1155 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(a) = 1]/a[not(contains(concat(" ", @class), " h-")) and @href]', |
| 1156 |
// .h-x>:only-child:not[.h-*]>area[href]:only-of-type:not[.h-*] |
| 1157 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(area) = 1]/area[not(contains(concat(" ", @class), " h-")) and @href]' |
| 1158 |
); |
| 1159 |
foreach ($xpaths as $xpath) { |
| 1160 |
$url = $this->xpath->query($xpath, $e); |
| 1161 |
if ($url !== false && $url->length === 1) { |
| 1162 |
$return['url'][] = $this->resolveUrl($url->item(0)->getAttribute('href')); |
| 1163 |
break; |
| 1164 |
} |
| 1165 |
} |
| 1166 |
} |
| 1167 |
} |
| 1168 |
|
| 1169 |
// Make sure things are unique and in alphabetical order |
| 1170 |
$mfTypes = array_unique($mfTypes); |
| 1171 |
sort($mfTypes); |
| 1172 |
|
| 1173 |
// Properties should be an object when JSON serialised |
| 1174 |
if (empty($return) and $this->jsonMode) { |
| 1175 |
$return = new stdClass(); |
| 1176 |
} |
| 1177 |
|
| 1178 |
// Phew. Return the final result. |
| 1179 |
$parsed = array( |
| 1180 |
'type' => $mfTypes, |
| 1181 |
'properties' => $return |
| 1182 |
); |
| 1183 |
|
| 1184 |
if(trim($e->getAttribute('id')) !== '') { |
| 1185 |
$parsed['id'] = trim($e->getAttribute("id")); |
| 1186 |
} |
| 1187 |
|
| 1188 |
if($this->lang) { |
| 1189 |
// Language |
| 1190 |
if ( $html_lang = $this->language($e) ) { |
| 1191 |
$parsed['lang'] = $html_lang; |
| 1192 |
} |
| 1193 |
} |
| 1194 |
|
| 1195 |
if (!empty($shape)) { |
| 1196 |
$parsed['shape'] = $shape; |
| 1197 |
} |
| 1198 |
|
| 1199 |
if (!empty($coords)) { |
| 1200 |
$parsed['coords'] = $coords; |
| 1201 |
} |
| 1202 |
|
| 1203 |
if (!empty($children)) { |
| 1204 |
$parsed['children'] = array_values(array_filter($children)); |
| 1205 |
} |
| 1206 |
return $parsed; |
| 1207 |
} |
| 1208 |
|
| 1209 |
/** |
| 1210 |
* @see http://microformats.org/wiki/microformats2-parsing#parsing_for_implied_properties |
| 1211 |
*/ |
| 1212 |
public function parseImpliedPhoto(\DOMElement $e) { |
| 1213 |
|
| 1214 |
// img.h-x[src] |
| 1215 |
if ($e->tagName == 'img') { |
| 1216 |
return $this->resolveUrl($this->parseImg($e)); |
| 1217 |
} |
| 1218 |
|
| 1219 |
// object.h-x[data] |
| 1220 |
if ($e->tagName == 'object' && $e->hasAttribute('data')) { |
| 1221 |
return $this->resolveUrl($e->getAttribute('data')); |
| 1222 |
} |
| 1223 |
|
| 1224 |
$xpaths = array( |
| 1225 |
// .h-x>img[src]:only-of-type:not[.h-*] |
| 1226 |
'./img[not(contains(concat(" ", @class), " h-")) and count(../img) = 1 and @src]', |
| 1227 |
// .h-x>object[data]:only-of-type:not[.h-*] |
| 1228 |
'./object[not(contains(concat(" ", @class), " h-")) and count(../object) = 1 and @data]', |
| 1229 |
// .h-x>:only-child:not[.h-*]>img[src]:only-of-type:not[.h-*] |
| 1230 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(img) = 1]/img[not(contains(concat(" ", @class), " h-")) and @src]', |
| 1231 |
// .h-x>:only-child:not[.h-*]>object[data]:only-of-type:not[.h-*] |
| 1232 |
'./*[not(contains(concat(" ", @class), " h-")) and count(../*) = 1 and count(object) = 1]/object[not(contains(concat(" ", @class), " h-")) and @data]', |
| 1233 |
); |
| 1234 |
|
| 1235 |
foreach ($xpaths as $path) { |
| 1236 |
$els = $this->xpath->query($path, $e); |
| 1237 |
|
| 1238 |
if ($els !== false && $els->length === 1) { |
| 1239 |
$el = $els->item(0); |
| 1240 |
if ($el->tagName == 'img') { |
| 1241 |
$return = $this->parseImg($el); |
| 1242 |
return $this->resolveUrl($return); |
| 1243 |
} else if ($el->tagName == 'object') { |
| 1244 |
return $this->resolveUrl($el->getAttribute('data')); |
| 1245 |
} |
| 1246 |
} |
| 1247 |
} |
| 1248 |
|
| 1249 |
// no implied photo |
| 1250 |
return false; |
| 1251 |
} |
| 1252 |
|
| 1253 |
/** |
| 1254 |
* Parse rels and alternates |
| 1255 |
* |
| 1256 |
* Returns [$rels, $rel_urls, $alternates]. |
| 1257 |
* For $rels and $rel_urls, if they are empty and $this->jsonMode = true, they will be returned as stdClass, |
| 1258 |
* optimizing for JSON serialization. Otherwise they will be returned as an empty array. |
| 1259 |
* Note that $alternates is deprecated in the microformats spec in favor of $rel_urls. $alternates only appears |
| 1260 |
* in parsed results if $this->enableAlternates = true. |
| 1261 |
* @return array|stdClass |
| 1262 |
*/ |
| 1263 |
public function parseRelsAndAlternates() { |
| 1264 |
$rels = array(); |
| 1265 |
$rel_urls = array(); |
| 1266 |
$alternates = array(); |
| 1267 |
|
| 1268 |
// Iterate through all a, area and link elements with rel attributes |
| 1269 |
foreach ($this->xpath->query('//a[@rel and @href] | //link[@rel and @href] | //area[@rel and @href]') as $hyperlink) { |
| 1270 |
// Parse the set of rels for the current link |
| 1271 |
$linkRels = array_unique(array_filter(preg_split('/[\t\n\f\r ]/', $hyperlink->getAttribute('rel')))); |
| 1272 |
if (count($linkRels) === 0) { |
| 1273 |
continue; |
| 1274 |
} |
| 1275 |
|
| 1276 |
// Resolve the href |
| 1277 |
$href = $this->resolveUrl($hyperlink->getAttribute('href')); |
| 1278 |
|
| 1279 |
$rel_attributes = array(); |
| 1280 |
|
| 1281 |
if ($hyperlink->hasAttribute('media')) { |
| 1282 |
$rel_attributes['media'] = $hyperlink->getAttribute('media'); |
| 1283 |
} |
| 1284 |
|
| 1285 |
if ($hyperlink->hasAttribute('hreflang')) { |
| 1286 |
$rel_attributes['hreflang'] = $hyperlink->getAttribute('hreflang'); |
| 1287 |
} |
| 1288 |
|
| 1289 |
if ($hyperlink->hasAttribute('title')) { |
| 1290 |
$rel_attributes['title'] = $hyperlink->getAttribute('title'); |
| 1291 |
} |
| 1292 |
|
| 1293 |
if ($hyperlink->hasAttribute('type')) { |
| 1294 |
$rel_attributes['type'] = $hyperlink->getAttribute('type'); |
| 1295 |
} |
| 1296 |
|
| 1297 |
if (strlen($hyperlink->textContent) > 0) { |
| 1298 |
$rel_attributes['text'] = $hyperlink->textContent; |
| 1299 |
} |
| 1300 |
|
| 1301 |
if ($this->enableAlternates) { |
| 1302 |
// If 'alternate' in rels, create 'alternates' structure, append |
| 1303 |
if (in_array('alternate', $linkRels)) { |
| 1304 |
$alternates[] = array_merge( |
| 1305 |
$rel_attributes, |
| 1306 |
array( |
| 1307 |
'url' => $href, |
| 1308 |
'rel' => implode(' ', array_diff($linkRels, array('alternate'))) |
| 1309 |
) |
| 1310 |
); |
| 1311 |
} |
| 1312 |
} |
| 1313 |
|
| 1314 |
foreach ($linkRels as $rel) { |
| 1315 |
if (!array_key_exists($rel, $rels)) { |
| 1316 |
$rels[$rel] = array($href); |
| 1317 |
} elseif (!in_array($href, $rels[$rel])) { |
| 1318 |
$rels[$rel][] = $href; |
| 1319 |
} |
| 1320 |
} |
| 1321 |
|
| 1322 |
if (!array_key_exists($href, $rel_urls)) { |
| 1323 |
$rel_urls[$href] = array('rels' => array()); |
| 1324 |
} |
| 1325 |
|
| 1326 |
// Add the attributes collected only if they were not already set |
| 1327 |
$rel_urls[$href] = array_merge( |
| 1328 |
$rel_attributes, |
| 1329 |
$rel_urls[$href] |
| 1330 |
); |
| 1331 |
|
| 1332 |
// Merge current rels with those already set |
| 1333 |
$rel_urls[$href]['rels'] = array_merge( |
| 1334 |
$rel_urls[$href]['rels'], |
| 1335 |
$linkRels |
| 1336 |
); |
| 1337 |
} |
| 1338 |
|
| 1339 |
// Alphabetically sort the rels arrays after removing duplicates |
| 1340 |
foreach ($rel_urls as $href => $object) { |
| 1341 |
$rel_urls[$href]['rels'] = array_unique($rel_urls[$href]['rels']); |
| 1342 |
sort($rel_urls[$href]['rels']); |
| 1343 |
} |
| 1344 |
|
| 1345 |
if (empty($rels) and $this->jsonMode) { |
| 1346 |
$rels = new stdClass(); |
| 1347 |
} |
| 1348 |
|
| 1349 |
if (empty($rel_urls) and $this->jsonMode) { |
| 1350 |
$rel_urls = new stdClass(); |
| 1351 |
} |
| 1352 |
|
| 1353 |
return array($rels, $rel_urls, $alternates); |
| 1354 |
} |
| 1355 |
|
| 1356 |
/** |
| 1357 |
* Find rel=tag elements that don't have class=category and have an href. |
| 1358 |
* For each element, get the last non-empty URL segment. Append a <data> |
| 1359 |
* element with that value as the category. Uses the mf1 class 'category' |
| 1360 |
* which will then be upgraded to p-category during backcompat. |
| 1361 |
* @param DOMElement $el |
| 1362 |
*/ |
| 1363 |
public function upgradeRelTagToCategory(DOMElement $el) { |
| 1364 |
$rel_tag = $this->xpath->query('.//a[contains(concat(" ",normalize-space(@rel)," ")," tag ") and not(contains(concat(" ", normalize-space(@class), " "), " category ")) and @href]', $el); |
| 1365 |
|
| 1366 |
if ( $rel_tag->length ) { |
| 1367 |
foreach ( $rel_tag as $tempEl ) { |
| 1368 |
$path = trim(parse_url($tempEl->getAttribute('href'), PHP_URL_PATH), ' /'); |
| 1369 |
$segments = explode('/', $path); |
| 1370 |
$value = array_pop($segments); |
| 1371 |
|
| 1372 |
# build the <data> element |
| 1373 |
$dataEl = $tempEl->ownerDocument->createElement('data'); |
| 1374 |
$dataEl->setAttribute('class', 'category'); |
| 1375 |
$dataEl->setAttribute('value', $value); |
| 1376 |
|
| 1377 |
# append as child of input element. this should ensure added element does get parsed inside e-* |
| 1378 |
$el->appendChild($dataEl); |
| 1379 |
} |
| 1380 |
} |
| 1381 |
} |
| 1382 |
|
| 1383 |
/** |
| 1384 |
* Kicks off the parsing routine |
| 1385 |
* @param bool $convertClassic whether to do backcompat parsing on microformats1. Defaults to true. |
| 1386 |
* @param DOMElement $context optionally specify an element from which to parse microformats |
| 1387 |
* @return array An array containing all the microformats found in the current document |
| 1388 |
*/ |
| 1389 |
public function parse($convertClassic = true, ?DOMElement $context = null) { |
| 1390 |
$this->convertClassic = $convertClassic; |
| 1391 |
$mfs = $this->parse_recursive($context); |
| 1392 |
|
| 1393 |
// Parse rels |
| 1394 |
list($rels, $rel_urls, $alternates) = $this->parseRelsAndAlternates(); |
| 1395 |
|
| 1396 |
$top = array( |
| 1397 |
'items' => array_values(array_filter($mfs)), |
| 1398 |
'rels' => $rels, |
| 1399 |
'rel-urls' => $rel_urls, |
| 1400 |
); |
| 1401 |
|
| 1402 |
if ($this->enableAlternates && count($alternates)) { |
| 1403 |
$top['alternates'] = $alternates; |
| 1404 |
} |
| 1405 |
|
| 1406 |
$top['title'] = ''; |
| 1407 |
$x_title = $this->xpath->query( '//title' ); |
| 1408 |
if ( $x_title && $x_title->item( 0 ) ) { |
| 1409 |
$top['title'] = $x_title->item( 0 )->textContent; |
| 1410 |
} |
| 1411 |
|
| 1412 |
return $top; |
| 1413 |
} |
| 1414 |
|
| 1415 |
|
| 1416 |
/** |
| 1417 |
* Parse microformats recursively |
| 1418 |
* Keeps track of whether inside a backcompat root or not |
| 1419 |
* @param DOMElement $context: node to start with |
| 1420 |
* @param int $depth: recursion depth |
| 1421 |
* @return array |
| 1422 |
*/ |
| 1423 |
public function parse_recursive(?DOMElement $context = null, $depth = 0) { |
| 1424 |
if ( $depth > 2 ) { |
| 1425 |
return array(); |
| 1426 |
} |
| 1427 |
|
| 1428 |
$mfs = array(); |
| 1429 |
$mfElements = $this->getRootMF($context); |
| 1430 |
|
| 1431 |
foreach ($mfElements as $node) { |
| 1432 |
$is_backcompat = !$this->hasRootMf2($node); |
| 1433 |
|
| 1434 |
if ($this->convertClassic && $is_backcompat) { |
| 1435 |
$this->backcompat($node); |
| 1436 |
} |
| 1437 |
|
| 1438 |
$recurse = $this->parse_recursive($node, $depth + 1); |
| 1439 |
|
| 1440 |
// set bool flag for nested mf |
| 1441 |
$has_nested_mf = (bool) $recurse; |
| 1442 |
|
| 1443 |
// parse for root mf |
| 1444 |
$result = $this->parseH($node, $is_backcompat, $has_nested_mf); |
| 1445 |
|
| 1446 |
// TODO: Determine if clearing this is required? |
| 1447 |
$this->elementPrefixParsed($node, 'h'); |
| 1448 |
$this->elementPrefixParsed($node, 'p'); |
| 1449 |
$this->elementPrefixParsed($node, 'u'); |
| 1450 |
$this->elementPrefixParsed($node, 'dt'); |
| 1451 |
$this->elementPrefixParsed($node, 'e'); |
| 1452 |
|
| 1453 |
// parseH returned a parsed result |
| 1454 |
if ($result) { |
| 1455 |
|
| 1456 |
// merge recursive results into current results |
| 1457 |
if ($recurse) { |
| 1458 |
$result = array_merge_recursive($result, $recurse); |
| 1459 |
} |
| 1460 |
|
| 1461 |
// currently a nested mf; check if node is an mf property of parent |
| 1462 |
if ($depth > 0) { |
| 1463 |
$temp_properties = nestedMfPropertyNamesFromElement($node); |
| 1464 |
|
| 1465 |
// properties found; set up parsed result in 'properties' |
| 1466 |
if (!empty($temp_properties)) { |
| 1467 |
|
| 1468 |
foreach ($temp_properties as $property => $prefixes) { |
| 1469 |
// Note: handling microformat nesting under multiple conflicting prefixes is not currently specified by the mf2 parsing spec. |
| 1470 |
$prefixSpecificResult = $result; |
| 1471 |
if (in_array('p-', $prefixes)) { |
| 1472 |
$prefixSpecificResult['value'] = (!is_array($prefixSpecificResult['properties']) || empty($prefixSpecificResult['properties']['name'][0])) ? $this->parseP($node) : $prefixSpecificResult['properties']['name'][0]; |
| 1473 |
} elseif (in_array('e-', $prefixes)) { |
| 1474 |
$eParsedResult = $this->parseE($node); |
| 1475 |
$prefixSpecificResult['html'] = $eParsedResult['html']; |
| 1476 |
$prefixSpecificResult['value'] = $eParsedResult['value']; |
| 1477 |
} elseif (in_array('u-', $prefixes)) { |
| 1478 |
$prefixSpecificResult['value'] = (!is_array($result['properties']) || empty($result['properties']['url'])) ? $this->parseU($node) : reset($result['properties']['url']); |
| 1479 |
} elseif (in_array('dt-', $prefixes)) { |
| 1480 |
$parsed_property = $this->parseDT($node); |
| 1481 |
$prefixSpecificResult['value'] = ($parsed_property) ? $parsed_property : ''; |
| 1482 |
} |
| 1483 |
$prefixSpecificResult['value'] = is_array($prefixSpecificResult['value']) ? $prefixSpecificResult['value']['value'] : $prefixSpecificResult['value']; |
| 1484 |
|
| 1485 |
$mfs['properties'][$property][] = $prefixSpecificResult; |
| 1486 |
} |
| 1487 |
|
| 1488 |
// otherwise, set up in 'children' |
| 1489 |
} else { |
| 1490 |
$mfs['children'][] = $result; |
| 1491 |
} |
| 1492 |
// otherwise, top-level mf |
| 1493 |
} else { |
| 1494 |
$mfs[] = $result; |
| 1495 |
} |
| 1496 |
} |
| 1497 |
} |
| 1498 |
|
| 1499 |
return $mfs; |
| 1500 |
} |
| 1501 |
|
| 1502 |
|
| 1503 |
/** |
| 1504 |
* Parse From ID |
| 1505 |
* |
| 1506 |
* Given an ID, parse all microformats which are children of the element with |
| 1507 |
* that ID. |
| 1508 |
* |
| 1509 |
* Note that rel values are still document-wide. |
| 1510 |
* |
| 1511 |
* If an element with the ID is not found, an empty skeleton mf2 array structure |
| 1512 |
* will be returned. |
| 1513 |
* |
| 1514 |
* @param string $id |
| 1515 |
* @param bool $htmlSafe = false whether or not to HTML-encode angle brackets in non e-* properties |
| 1516 |
* @return array |
| 1517 |
*/ |
| 1518 |
public function parseFromId($id, $convertClassic=true) { |
| 1519 |
$matches = $this->xpath->query("//*[@id='{$id}']"); |
| 1520 |
|
| 1521 |
if (empty($matches)) |
| 1522 |
return array('items' => array(), 'rels' => array(), 'alternates' => array()); |
| 1523 |
|
| 1524 |
return $this->parse($convertClassic, $matches->item(0)); |
| 1525 |
} |
| 1526 |
|
| 1527 |
/** |
| 1528 |
* Get the root microformat elements |
| 1529 |
* @param DOMElement $context |
| 1530 |
* @return DOMNodeList |
| 1531 |
*/ |
| 1532 |
public function getRootMF(?DOMElement $context = null) { |
| 1533 |
// start with mf2 root class name xpath |
| 1534 |
$xpaths = array( |
| 1535 |
'(php:function("\\Friends\\Mf2\\classHasMf2RootClassname", normalize-space(@class)))' |
| 1536 |
); |
| 1537 |
|
| 1538 |
// add mf1 root class names |
| 1539 |
foreach ( $this->classicRootMap as $old => $new ) { |
| 1540 |
$xpaths[] = '( contains(concat(" ",normalize-space(@class), " "), " ' . $old . ' ") )'; |
| 1541 |
} |
| 1542 |
|
| 1543 |
// final xpath with OR |
| 1544 |
$xpath = '//*[' . implode(' or ', $xpaths) . ']'; |
| 1545 |
|
| 1546 |
$mfElements = (null === $context) |
| 1547 |
? $this->xpath->query($xpath) |
| 1548 |
: $this->xpath->query('.' . $xpath, $context); |
| 1549 |
|
| 1550 |
return $mfElements; |
| 1551 |
} |
| 1552 |
|
| 1553 |
/** |
| 1554 |
* Apply the backcompat algorithm to upgrade mf1 classes to mf2. |
| 1555 |
* This method is called recursively. |
| 1556 |
* @param DOMElement $el |
| 1557 |
* @param string $context |
| 1558 |
* @param bool $isParentMf2 |
| 1559 |
* @see http://microformats.org/wiki/microformats2-parsing#algorithm |
| 1560 |
*/ |
| 1561 |
public function backcompat(DOMElement $el, $context = '', $isParentMf2 = false) { |
| 1562 |
|
| 1563 |
if ( $context ) { |
| 1564 |
$mf1Classes = array($context); |
| 1565 |
} else { |
| 1566 |
$class = str_replace(array("\t", "\n"), ' ', $el->getAttribute('class')); |
| 1567 |
$classes = array_filter(explode(' ', $class)); |
| 1568 |
$mf1Classes = array_intersect($classes, array_keys($this->classicRootMap)); |
| 1569 |
} |
| 1570 |
|
| 1571 |
$elHasMf2 = $this->hasRootMf2($el); |
| 1572 |
|
| 1573 |
foreach ($mf1Classes as $classname) { |
| 1574 |
// special handling for specific properties |
| 1575 |
switch ( $classname ) |
| 1576 |
{ |
| 1577 |
case 'hentry': |
| 1578 |
$this->upgradeRelTagToCategory($el); |
| 1579 |
|
| 1580 |
$rel_bookmark = $this->xpath->query('.//a[contains(concat(" ",normalize-space(@rel)," ")," bookmark ") and @href]', $el); |
| 1581 |
|
| 1582 |
if ( $rel_bookmark->length ) { |
| 1583 |
foreach ( $rel_bookmark as $tempEl ) { |
| 1584 |
$this->addMfClasses($tempEl, 'u-url'); |
| 1585 |
$this->addUpgraded($tempEl, array('bookmark')); |
| 1586 |
} |
| 1587 |
} |
| 1588 |
break; |
| 1589 |
|
| 1590 |
case 'hfeed': |
| 1591 |
$this->upgradeRelTagToCategory($el); |
| 1592 |
break; |
| 1593 |
|
| 1594 |
case 'hproduct': |
| 1595 |
$review_and_hreview_aggregate = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " review ") and contains(concat(" ", normalize-space(@class), " "), " hreview-aggregate ")]', $el); |
| 1596 |
|
| 1597 |
if ( $review_and_hreview_aggregate->length ) { |
| 1598 |
foreach ( $review_and_hreview_aggregate as $tempEl ) { |
| 1599 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1600 |
$this->backcompat($tempEl, 'hreview-aggregate'); |
| 1601 |
$this->addMfClasses($tempEl, 'p-review h-review-aggregate'); |
| 1602 |
$this->addUpgraded($tempEl, array('review hreview-aggregate')); |
| 1603 |
} |
| 1604 |
} |
| 1605 |
} |
| 1606 |
|
| 1607 |
$review_and_hreview = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " review ") and contains(concat(" ", normalize-space(@class), " "), " hreview ")]', $el); |
| 1608 |
|
| 1609 |
if ( $review_and_hreview->length ) { |
| 1610 |
foreach ( $review_and_hreview as $tempEl ) { |
| 1611 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1612 |
$this->backcompat($tempEl, 'hreview'); |
| 1613 |
$this->addMfClasses($tempEl, 'p-review h-review'); |
| 1614 |
$this->addUpgraded($tempEl, array('review hreview')); |
| 1615 |
} |
| 1616 |
} |
| 1617 |
} |
| 1618 |
|
| 1619 |
break; |
| 1620 |
|
| 1621 |
case 'hreview-aggregate': |
| 1622 |
case 'hreview': |
| 1623 |
$item_and_vcard = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " item ") and contains(concat(" ", normalize-space(@class), " "), " vcard ")]', $el); |
| 1624 |
|
| 1625 |
if ( $item_and_vcard->length ) { |
| 1626 |
foreach ( $item_and_vcard as $tempEl ) { |
| 1627 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1628 |
$this->backcompat($tempEl, 'vcard'); |
| 1629 |
$this->addMfClasses($tempEl, 'p-item h-card'); |
| 1630 |
$this->addUpgraded($tempEl, array('item', 'vcard')); |
| 1631 |
} |
| 1632 |
} |
| 1633 |
} |
| 1634 |
|
| 1635 |
$item_and_vevent = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " item ") and contains(concat(" ", normalize-space(@class), " "), " vevent ")]', $el); |
| 1636 |
|
| 1637 |
if ( $item_and_vevent->length ) { |
| 1638 |
foreach ( $item_and_vevent as $tempEl ) { |
| 1639 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1640 |
$this->addMfClasses($tempEl, 'p-item h-event'); |
| 1641 |
$this->backcompat($tempEl, 'vevent'); |
| 1642 |
$this->addUpgraded($tempEl, array('item', 'vevent')); |
| 1643 |
} |
| 1644 |
} |
| 1645 |
} |
| 1646 |
|
| 1647 |
$item_and_hproduct = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " item ") and contains(concat(" ", normalize-space(@class), " "), " hproduct ")]', $el); |
| 1648 |
|
| 1649 |
if ( $item_and_hproduct->length ) { |
| 1650 |
foreach ( $item_and_hproduct as $tempEl ) { |
| 1651 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1652 |
$this->addMfClasses($tempEl, 'p-item h-product'); |
| 1653 |
$this->backcompat($tempEl, 'hproduct'); |
| 1654 |
$this->addUpgraded($tempEl, array('item', 'hproduct')); |
| 1655 |
} |
| 1656 |
} |
| 1657 |
} |
| 1658 |
|
| 1659 |
$rel_self_bookmark = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@rel), " "), " self ") and contains(concat(" ", normalize-space(@rel), " "), " bookmark ")]', $el); |
| 1660 |
|
| 1661 |
if ( $rel_self_bookmark->length ) { |
| 1662 |
foreach ( $rel_self_bookmark as $tempEl ) { |
| 1663 |
$this->addMfClasses($tempEl, 'u-url'); |
| 1664 |
$this->addUpgraded($tempEl, array('self', 'bookmark')); |
| 1665 |
} |
| 1666 |
} |
| 1667 |
|
| 1668 |
$reviewer_nodes = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " reviewer ")]', $el); |
| 1669 |
|
| 1670 |
if ( $reviewer_nodes->length ) { |
| 1671 |
foreach ( $reviewer_nodes as $tempEl ) { |
| 1672 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1673 |
$this->addMfClasses($tempEl, 'p-author h-card'); |
| 1674 |
} |
| 1675 |
} |
| 1676 |
} |
| 1677 |
|
| 1678 |
$this->upgradeRelTagToCategory($el); |
| 1679 |
break; |
| 1680 |
|
| 1681 |
case 'vevent': |
| 1682 |
$location_and_vcard = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " location ") and contains(concat(" ", normalize-space(@class), " "), " vcard ")]', $el); |
| 1683 |
|
| 1684 |
if ( $location_and_vcard->length ) { |
| 1685 |
foreach ( $location_and_vcard as $tempEl ) { |
| 1686 |
if ( !$this->hasRootMf2($tempEl) ) { |
| 1687 |
$this->addMfClasses($tempEl, 'p-location h-card'); |
| 1688 |
$this->backcompat($tempEl, 'vcard'); |
| 1689 |
$this->addUpgraded($tempEl, array('location', 'vcard')); |
| 1690 |
} |
| 1691 |
} |
| 1692 |
} |
| 1693 |
break; |
| 1694 |
} |
| 1695 |
|
| 1696 |
// root class has mf1 properties to be upgraded |
| 1697 |
if ( isset($this->classicPropertyMap[$classname]) ) { |
| 1698 |
// loop through each property of the mf1 root |
| 1699 |
foreach ( $this->classicPropertyMap[$classname] as $property => $data ) { |
| 1700 |
$propertyElements = $this->xpath->query('.//*[contains(concat(" ", normalize-space(@class), " "), " ' . $property . ' ")]', $el); |
| 1701 |
|
| 1702 |
// loop through each element with the property |
| 1703 |
foreach ( $propertyElements as $propertyEl ) { |
| 1704 |
$hasRootMf2 = $this->hasRootMf2($propertyEl); |
| 1705 |
|
| 1706 |
// if the element has not been upgraded and we're not inside an mf2 root, recurse |
| 1707 |
if ( !$this->isElementUpgraded($propertyEl, $property) && !$isParentMf2 ) |
| 1708 |
{ |
| 1709 |
$temp_context = ( isset($data['context']) ) ? $data['context'] : null; |
| 1710 |
$this->backcompat($propertyEl, $temp_context, $hasRootMf2); |
| 1711 |
$this->addMfClasses($propertyEl, $data['replace']); |
| 1712 |
} |
| 1713 |
|
| 1714 |
$this->addUpgraded($propertyEl, $property); |
| 1715 |
} |
| 1716 |
} |
| 1717 |
} |
| 1718 |
|
| 1719 |
if ( empty($context) && isset($this->classicRootMap[$classname]) && !$elHasMf2 ) { |
| 1720 |
$this->addMfClasses($el, $this->classicRootMap[$classname]); |
| 1721 |
} |
| 1722 |
} |
| 1723 |
|
| 1724 |
return; |
| 1725 |
} |
| 1726 |
|
| 1727 |
/** |
| 1728 |
* Add element + property as upgraded during backcompat |
| 1729 |
* @param DOMElement $el |
| 1730 |
* @param string|array $property |
| 1731 |
*/ |
| 1732 |
public function addUpgraded(DOMElement $el, $property) { |
| 1733 |
if ( !is_array($property) ) { |
| 1734 |
$property = array($property); |
| 1735 |
} |
| 1736 |
|
| 1737 |
// add element to list of upgraded elements |
| 1738 |
if ( !$this->upgraded->offsetExists($el) ) { |
| 1739 |
$this->upgraded->offsetSet($el, $property); |
| 1740 |
} else { |
| 1741 |
$this->upgraded[$el] = array_merge($this->upgraded[$el], $property); |
| 1742 |
} |
| 1743 |
} |
| 1744 |
|
| 1745 |
/** |
| 1746 |
* Add the provided classes to an element. |
| 1747 |
* Does not add duplicate if class name already exists. |
| 1748 |
* @param DOMElement $el |
| 1749 |
* @param string $classes |
| 1750 |
*/ |
| 1751 |
public function addMfClasses(DOMElement $el, $classes) { |
| 1752 |
$existingClasses = str_replace(array("\t", "\n"), ' ', $el->getAttribute('class')); |
| 1753 |
$existingClasses = array_filter(explode(' ', $existingClasses)); |
| 1754 |
|
| 1755 |
$addClasses = array_diff(explode(' ', $classes), $existingClasses); |
| 1756 |
|
| 1757 |
if ( $addClasses ) { |
| 1758 |
$el->setAttribute('class', $el->getAttribute('class') . ' ' . implode(' ', $addClasses)); |
| 1759 |
} |
| 1760 |
} |
| 1761 |
|
| 1762 |
/** |
| 1763 |
* Check an element for mf2 h-* class, typically to determine if backcompat should be used |
| 1764 |
* @param DOMElement $el |
| 1765 |
*/ |
| 1766 |
public function hasRootMf2(\DOMElement $el) { |
| 1767 |
$class = str_replace(array("\t", "\n"), ' ', $el->getAttribute('class')); |
| 1768 |
|
| 1769 |
// Check for valid mf2 root classnames, not just any classname with a h- prefix. |
| 1770 |
return count(mfNamesFromClass($class, 'h-')) > 0; |
| 1771 |
} |
| 1772 |
|
| 1773 |
/** |
| 1774 |
* Convert Legacy Classnames |
| 1775 |
* |
| 1776 |
* Adds microformats2 classnames into a document containing only legacy |
| 1777 |
* semantic classnames. |
| 1778 |
* |
| 1779 |
* @return Parser $this |
| 1780 |
*/ |
| 1781 |
public function convertLegacy() { |
| 1782 |
$doc = $this->doc; |
| 1783 |
$xp = new DOMXPath($doc); |
| 1784 |
|
| 1785 |
// replace all roots |
| 1786 |
foreach ($this->classicRootMap as $old => $new) { |
| 1787 |
foreach ($xp->query('//*[contains(concat(" ", normalize-space(@class), " "), " ' . $old . ' ") and not(contains(concat(" ", normalize-space(@class), " "), " ' . $new . ' "))]') as $el) { |
| 1788 |
$el->setAttribute('class', $el->getAttribute('class') . ' ' . $new); |
| 1789 |
} |
| 1790 |
} |
| 1791 |
|
| 1792 |
foreach ($this->classicPropertyMap as $oldRoot => $properties) { |
| 1793 |
$newRoot = $this->classicRootMap[$oldRoot]; |
| 1794 |
foreach ($properties as $old => $data) { |
| 1795 |
foreach ($xp->query('//*[contains(concat(" ", normalize-space(@class), " "), " ' . $oldRoot . ' ")]//*[contains(concat(" ", normalize-space(@class), " "), " ' . $old . ' ") and not(contains(concat(" ", normalize-space(@class), " "), " ' . $data['replace'] . ' "))]') as $el) { |
| 1796 |
$el->setAttribute('class', $el->getAttribute('class') . ' ' . $data['replace']); |
| 1797 |
} |
| 1798 |
} |
| 1799 |
} |
| 1800 |
|
| 1801 |
return $this; |
| 1802 |
} |
| 1803 |
|
| 1804 |
/** |
| 1805 |
* XPath Query |
| 1806 |
* |
| 1807 |
* Runs an XPath query over the current document. Works in exactly the same |
| 1808 |
* way as DOMXPath::query. |
| 1809 |
* |
| 1810 |
* @param string $expression |
| 1811 |
* @param DOMNode $context |
| 1812 |
* @return DOMNodeList |
| 1813 |
*/ |
| 1814 |
public function query($expression, $context = null) { |
| 1815 |
return $this->xpath->query($expression, $context); |
| 1816 |
} |
| 1817 |
|
| 1818 |
/** |
| 1819 |
* Classic Root Classname map |
| 1820 |
* @var array |
| 1821 |
*/ |
| 1822 |
public $classicRootMap = array( |
| 1823 |
'vcard' => 'h-card', |
| 1824 |
'hfeed' => 'h-feed', |
| 1825 |
'hentry' => 'h-entry', |
| 1826 |
'hrecipe' => 'h-recipe', |
| 1827 |
'hresume' => 'h-resume', |
| 1828 |
'vevent' => 'h-event', |
| 1829 |
'hreview' => 'h-review', |
| 1830 |
'hreview-aggregate' => 'h-review-aggregate', |
| 1831 |
'hproduct' => 'h-product', |
| 1832 |
'adr' => 'h-adr', |
| 1833 |
'geo' => 'h-geo' |
| 1834 |
); |
| 1835 |
|
| 1836 |
/** |
| 1837 |
* Mapping of mf1 properties to mf2 and the context they're parsed with |
| 1838 |
* @var array |
| 1839 |
*/ |
| 1840 |
public $classicPropertyMap = array( |
| 1841 |
'vcard' => array( |
| 1842 |
'fn' => array( |
| 1843 |
'replace' => 'p-name' |
| 1844 |
), |
| 1845 |
'honorific-prefix' => array( |
| 1846 |
'replace' => 'p-honorific-prefix' |
| 1847 |
), |
| 1848 |
'given-name' => array( |
| 1849 |
'replace' => 'p-given-name' |
| 1850 |
), |
| 1851 |
'additional-name' => array( |
| 1852 |
'replace' => 'p-additional-name' |
| 1853 |
), |
| 1854 |
'family-name' => array( |
| 1855 |
'replace' => 'p-family-name' |
| 1856 |
), |
| 1857 |
'honorific-suffix' => array( |
| 1858 |
'replace' => 'p-honorific-suffix' |
| 1859 |
), |
| 1860 |
'nickname' => array( |
| 1861 |
'replace' => 'p-nickname' |
| 1862 |
), |
| 1863 |
'email' => array( |
| 1864 |
'replace' => 'u-email' |
| 1865 |
), |
| 1866 |
'logo' => array( |
| 1867 |
'replace' => 'u-logo' |
| 1868 |
), |
| 1869 |
'photo' => array( |
| 1870 |
'replace' => 'u-photo' |
| 1871 |
), |
| 1872 |
'url' => array( |
| 1873 |
'replace' => 'u-url' |
| 1874 |
), |
| 1875 |
'uid' => array( |
| 1876 |
'replace' => 'u-uid' |
| 1877 |
), |
| 1878 |
'category' => array( |
| 1879 |
'replace' => 'p-category' |
| 1880 |
), |
| 1881 |
'adr' => array( |
| 1882 |
'replace' => 'p-adr', |
| 1883 |
), |
| 1884 |
'extended-address' => array( |
| 1885 |
'replace' => 'p-extended-address' |
| 1886 |
), |
| 1887 |
'street-address' => array( |
| 1888 |
'replace' => 'p-street-address' |
| 1889 |
), |
| 1890 |
'locality' => array( |
| 1891 |
'replace' => 'p-locality' |
| 1892 |
), |
| 1893 |
'region' => array( |
| 1894 |
'replace' => 'p-region' |
| 1895 |
), |
| 1896 |
'postal-code' => array( |
| 1897 |
'replace' => 'p-postal-code' |
| 1898 |
), |
| 1899 |
'country-name' => array( |
| 1900 |
'replace' => 'p-country-name' |
| 1901 |
), |
| 1902 |
'label' => array( |
| 1903 |
'replace' => 'p-label' |
| 1904 |
), |
| 1905 |
'geo' => array( |
| 1906 |
'replace' => 'p-geo h-geo', |
| 1907 |
'context' => 'geo' |
| 1908 |
), |
| 1909 |
'latitude' => array( |
| 1910 |
'replace' => 'p-latitude' |
| 1911 |
), |
| 1912 |
'longitude' => array( |
| 1913 |
'replace' => 'p-longitude' |
| 1914 |
), |
| 1915 |
'tel' => array( |
| 1916 |
'replace' => 'p-tel' |
| 1917 |
), |
| 1918 |
'note' => array( |
| 1919 |
'replace' => 'p-note' |
| 1920 |
), |
| 1921 |
'bday' => array( |
| 1922 |
'replace' => 'dt-bday' |
| 1923 |
), |
| 1924 |
'key' => array( |
| 1925 |
'replace' => 'u-key' |
| 1926 |
), |
| 1927 |
'org' => array( |
| 1928 |
'replace' => 'p-org' |
| 1929 |
), |
| 1930 |
'organization-name' => array( |
| 1931 |
'replace' => 'p-organization-name' |
| 1932 |
), |
| 1933 |
'organization-unit' => array( |
| 1934 |
'replace' => 'p-organization-unit' |
| 1935 |
), |
| 1936 |
'title' => array( |
| 1937 |
'replace' => 'p-job-title' |
| 1938 |
), |
| 1939 |
'role' => array( |
| 1940 |
'replace' => 'p-role' |
| 1941 |
), |
| 1942 |
'tz' => array( |
| 1943 |
'replace' => 'p-tz' |
| 1944 |
), |
| 1945 |
'rev' => array( |
| 1946 |
'replace' => 'dt-rev' |
| 1947 |
), |
| 1948 |
), |
| 1949 |
'hfeed' => array( |
| 1950 |
'author' => array( |
| 1951 |
'replace' => 'p-author h-card', |
| 1952 |
'context' => 'vcard' |
| 1953 |
), |
| 1954 |
'url' => array( |
| 1955 |
'replace' => 'u-url' |
| 1956 |
), |
| 1957 |
'photo' => array( |
| 1958 |
'replace' => 'u-photo' |
| 1959 |
), |
| 1960 |
'category' => array( |
| 1961 |
'replace' => 'p-category' |
| 1962 |
), |
| 1963 |
), |
| 1964 |
'hentry' => array( |
| 1965 |
'entry-title' => array( |
| 1966 |
'replace' => 'p-name' |
| 1967 |
), |
| 1968 |
'entry-summary' => array( |
| 1969 |
'replace' => 'p-summary' |
| 1970 |
), |
| 1971 |
'entry-content' => array( |
| 1972 |
'replace' => 'e-content' |
| 1973 |
), |
| 1974 |
'published' => array( |
| 1975 |
'replace' => 'dt-published' |
| 1976 |
), |
| 1977 |
'updated' => array( |
| 1978 |
'replace' => 'dt-updated' |
| 1979 |
), |
| 1980 |
'author' => array( |
| 1981 |
'replace' => 'p-author h-card', |
| 1982 |
'context' => 'vcard', |
| 1983 |
), |
| 1984 |
'category' => array( |
| 1985 |
'replace' => 'p-category' |
| 1986 |
), |
| 1987 |
), |
| 1988 |
'hrecipe' => array( |
| 1989 |
'fn' => array( |
| 1990 |
'replace' => 'p-name' |
| 1991 |
), |
| 1992 |
'ingredient' => array( |
| 1993 |
'replace' => 'p-ingredient' |
| 1994 |
/** |
| 1995 |
* TODO: hRecipe 'value' and 'type' child mf not parsing correctly currently. |
| 1996 |
* Per http://microformats.org/wiki/hRecipe#Property_details, they're experimental. |
| 1997 |
*/ |
| 1998 |
), |
| 1999 |
'yield' => array( |
| 2000 |
'replace' => 'p-yield' |
| 2001 |
), |
| 2002 |
'instructions' => array( |
| 2003 |
'replace' => 'e-instructions' |
| 2004 |
), |
| 2005 |
'duration' => array( |
| 2006 |
'replace' => 'dt-duration' |
| 2007 |
), |
| 2008 |
'photo' => array( |
| 2009 |
'replace' => 'u-photo' |
| 2010 |
), |
| 2011 |
'summary' => array( |
| 2012 |
'replace' => 'p-summary' |
| 2013 |
), |
| 2014 |
'author' => array( |
| 2015 |
'replace' => 'p-author h-card', |
| 2016 |
'context' => 'vcard', |
| 2017 |
), |
| 2018 |
'nutrition' => array( |
| 2019 |
'replace' => 'p-nutrition' |
| 2020 |
), |
| 2021 |
'category' => array( |
| 2022 |
'replace' => 'p-category' |
| 2023 |
), |
| 2024 |
), |
| 2025 |
'hresume' => array( |
| 2026 |
'summary' => array( |
| 2027 |
'replace' => 'p-summary' |
| 2028 |
), |
| 2029 |
'contact' => array( |
| 2030 |
'replace' => 'p-contact h-card', |
| 2031 |
'context' => 'vcard', |
| 2032 |
), |
| 2033 |
'education' => array( |
| 2034 |
'replace' => 'p-education h-event', |
| 2035 |
'context' => 'vevent', |
| 2036 |
), |
| 2037 |
'experience' => array( |
| 2038 |
'replace' => 'p-experience h-event', |
| 2039 |
'context' => 'vevent', |
| 2040 |
), |
| 2041 |
'skill' => array( |
| 2042 |
'replace' => 'p-skill' |
| 2043 |
), |
| 2044 |
'affiliation' => array( |
| 2045 |
'replace' => 'p-affiliation h-card', |
| 2046 |
'context' => 'vcard', |
| 2047 |
), |
| 2048 |
), |
| 2049 |
'vevent' => array( |
| 2050 |
'summary' => array( |
| 2051 |
'replace' => 'p-name' |
| 2052 |
), |
| 2053 |
'dtstart' => array( |
| 2054 |
'replace' => 'dt-start' |
| 2055 |
), |
| 2056 |
'dtend' => array( |
| 2057 |
'replace' => 'dt-end' |
| 2058 |
), |
| 2059 |
'duration' => array( |
| 2060 |
'replace' => 'dt-duration' |
| 2061 |
), |
| 2062 |
'description' => array( |
| 2063 |
'replace' => 'p-description' |
| 2064 |
), |
| 2065 |
'url' => array( |
| 2066 |
'replace' => 'u-url' |
| 2067 |
), |
| 2068 |
'category' => array( |
| 2069 |
'replace' => 'p-category' |
| 2070 |
), |
| 2071 |
'location' => array( |
| 2072 |
'replace' => 'p-location', |
| 2073 |
), |
| 2074 |
'geo' => array( |
| 2075 |
'replace' => 'p-location h-geo' |
| 2076 |
), |
| 2077 |
'attendee' => array( |
| 2078 |
'replace' => 'p-attendee h-card', |
| 2079 |
'context' => 'vcard' |
| 2080 |
) |
| 2081 |
), |
| 2082 |
'hreview' => array( |
| 2083 |
'summary' => array( |
| 2084 |
'replace' => 'p-name' |
| 2085 |
), |
| 2086 |
# fn: see item.fn below |
| 2087 |
# photo: see item.photo below |
| 2088 |
# url: see item.url below |
| 2089 |
'item' => array( |
| 2090 |
'replace' => 'p-item h-item', |
| 2091 |
'context' => 'item' |
| 2092 |
), |
| 2093 |
# reviewer: see backcompat() |
| 2094 |
'dtreviewed' => array( |
| 2095 |
'replace' => 'dt-published' |
| 2096 |
), |
| 2097 |
'rating' => array( |
| 2098 |
'replace' => 'p-rating' |
| 2099 |
), |
| 2100 |
'best' => array( |
| 2101 |
'replace' => 'p-best' |
| 2102 |
), |
| 2103 |
'worst' => array( |
| 2104 |
'replace' => 'p-worst' |
| 2105 |
), |
| 2106 |
'description' => array( |
| 2107 |
'replace' => 'e-content' |
| 2108 |
), |
| 2109 |
'category' => array( |
| 2110 |
'replace' => 'p-category' |
| 2111 |
), |
| 2112 |
), |
| 2113 |
'hreview-aggregate' => array( |
| 2114 |
'summary' => array( |
| 2115 |
'replace' => 'p-name' |
| 2116 |
), |
| 2117 |
# fn: see item.fn below |
| 2118 |
# photo: see item.photo below |
| 2119 |
# url: see item.url below |
| 2120 |
'item' => array( |
| 2121 |
'replace' => 'p-item h-item', |
| 2122 |
'context' => 'item' |
| 2123 |
), |
| 2124 |
'rating' => array( |
| 2125 |
'replace' => 'p-rating' |
| 2126 |
), |
| 2127 |
'best' => array( |
| 2128 |
'replace' => 'p-best' |
| 2129 |
), |
| 2130 |
'worst' => array( |
| 2131 |
'replace' => 'p-worst' |
| 2132 |
), |
| 2133 |
'average' => array( |
| 2134 |
'replace' => 'p-average' |
| 2135 |
), |
| 2136 |
'count' => array( |
| 2137 |
'replace' => 'p-count' |
| 2138 |
), |
| 2139 |
'votes' => array( |
| 2140 |
'replace' => 'p-votes' |
| 2141 |
), |
| 2142 |
), |
| 2143 |
'hproduct' => array( |
| 2144 |
'fn' => array( |
| 2145 |
'replace' => 'p-name', |
| 2146 |
), |
| 2147 |
'photo' => array( |
| 2148 |
'replace' => 'u-photo', |
| 2149 |
), |
| 2150 |
'brand' => array( |
| 2151 |
'replace' => 'p-brand', |
| 2152 |
), |
| 2153 |
'category' => array( |
| 2154 |
'replace' => 'p-category', |
| 2155 |
), |
| 2156 |
'description' => array( |
| 2157 |
'replace' => 'p-description', |
| 2158 |
), |
| 2159 |
'identifier' => array( |
| 2160 |
'replace' => 'u-identifier', |
| 2161 |
), |
| 2162 |
'url' => array( |
| 2163 |
'replace' => 'u-url', |
| 2164 |
), |
| 2165 |
// review is handled in the special processing section to allow for 'review hreview-aggregate' |
| 2166 |
'price' => array( |
| 2167 |
'replace' => 'p-price' |
| 2168 |
), |
| 2169 |
), |
| 2170 |
'item' => array( |
| 2171 |
'fn' => array( |
| 2172 |
'replace' => 'p-name' |
| 2173 |
), |
| 2174 |
'url' => array( |
| 2175 |
'replace' => 'u-url' |
| 2176 |
), |
| 2177 |
'photo' => array( |
| 2178 |
'replace' => 'u-photo' |
| 2179 |
), |
| 2180 |
), |
| 2181 |
'adr' => array( |
| 2182 |
'post-office-box' => array( |
| 2183 |
'replace' => 'p-post-office-box' |
| 2184 |
), |
| 2185 |
'extended-address' => array( |
| 2186 |
'replace' => 'p-extended-address' |
| 2187 |
), |
| 2188 |
'street-address' => array( |
| 2189 |
'replace' => 'p-street-address' |
| 2190 |
), |
| 2191 |
'locality' => array( |
| 2192 |
'replace' => 'p-locality' |
| 2193 |
), |
| 2194 |
'region' => array( |
| 2195 |
'replace' => 'p-region' |
| 2196 |
), |
| 2197 |
'postal-code' => array( |
| 2198 |
'replace' => 'p-postal-code' |
| 2199 |
), |
| 2200 |
'country-name' => array( |
| 2201 |
'replace' => 'p-country-name' |
| 2202 |
), |
| 2203 |
), |
| 2204 |
'geo' => array( |
| 2205 |
'latitude' => array( |
| 2206 |
'replace' => 'p-latitude' |
| 2207 |
), |
| 2208 |
'longitude' => array( |
| 2209 |
'replace' => 'p-longitude' |
| 2210 |
), |
| 2211 |
), |
| 2212 |
); |
| 2213 |
} |
| 2214 |
|
| 2215 |
function parseUriToComponents($uri) { |
| 2216 |
$result = array( |
| 2217 |
'scheme' => null, |
| 2218 |
'authority' => null, |
| 2219 |
'path' => null, |
| 2220 |
'query' => null, |
| 2221 |
'fragment' => null |
| 2222 |
); |
| 2223 |
|
| 2224 |
if ( ! is_string( $uri ) ) { |
| 2225 |
return $result; |
| 2226 |
} |
| 2227 |
|
| 2228 |
$u = parse_url($uri); |
| 2229 |
if ( ! $u ) { |
| 2230 |
return $result; |
| 2231 |
} |
| 2232 |
|
| 2233 |
if(array_key_exists('scheme', $u)) |
| 2234 |
$result['scheme'] = $u['scheme']; |
| 2235 |
|
| 2236 |
if(array_key_exists('host', $u)) { |
| 2237 |
if(array_key_exists('user', $u)) |
| 2238 |
$result['authority'] = $u['user']; |
| 2239 |
if(array_key_exists('pass', $u)) |
| 2240 |
$result['authority'] .= ':' . $u['pass']; |
| 2241 |
if(array_key_exists('user', $u) || array_key_exists('pass', $u)) |
| 2242 |
$result['authority'] .= '@'; |
| 2243 |
$result['authority'] .= $u['host']; |
| 2244 |
if(array_key_exists('port', $u)) |
| 2245 |
$result['authority'] .= ':' . $u['port']; |
| 2246 |
} |
| 2247 |
|
| 2248 |
if(array_key_exists('path', $u)) |
| 2249 |
$result['path'] = $u['path']; |
| 2250 |
|
| 2251 |
if(array_key_exists('query', $u)) |
| 2252 |
$result['query'] = $u['query']; |
| 2253 |
|
| 2254 |
if(array_key_exists('fragment', $u)) |
| 2255 |
$result['fragment'] = $u['fragment']; |
| 2256 |
|
| 2257 |
return $result; |
| 2258 |
} |
| 2259 |
|
| 2260 |
function resolveUrl($baseURI, $referenceURI) { |
| 2261 |
$target = array( |
| 2262 |
'scheme' => null, |
| 2263 |
'authority' => null, |
| 2264 |
'path' => null, |
| 2265 |
'query' => null, |
| 2266 |
'fragment' => null |
| 2267 |
); |
| 2268 |
|
| 2269 |
# 5.2.1 Pre-parse the Base URI |
| 2270 |
# The base URI (Base) is established according to the procedure of |
| 2271 |
# Section 5.1 and parsed into the five main components described in |
| 2272 |
# Section 3 |
| 2273 |
$base = parseUriToComponents($baseURI); |
| 2274 |
|
| 2275 |
# If base path is blank (http://example.com) then set it to / |
| 2276 |
# (I can't tell if this is actually in the RFC or not, but seems like it makes sense) |
| 2277 |
if($base['path'] == null) |
| 2278 |
$base['path'] = '/'; |
| 2279 |
|
| 2280 |
# 5.2.2. Transform References |
| 2281 |
|
| 2282 |
# The URI reference is parsed into the five URI components |
| 2283 |
# (R.scheme, R.authority, R.path, R.query, R.fragment) = parse(R); |
| 2284 |
$reference = parseUriToComponents($referenceURI); |
| 2285 |
|
| 2286 |
# A non-strict parser may ignore a scheme in the reference |
| 2287 |
# if it is identical to the base URI's scheme. |
| 2288 |
# TODO |
| 2289 |
|
| 2290 |
if($reference['scheme']) { |
| 2291 |
$target['scheme'] = $reference['scheme']; |
| 2292 |
$target['authority'] = $reference['authority']; |
| 2293 |
$target['path'] = removeDotSegments($reference['path']); |
| 2294 |
$target['query'] = $reference['query']; |
| 2295 |
} else { |
| 2296 |
if($reference['authority']) { |
| 2297 |
$target['authority'] = $reference['authority']; |
| 2298 |
$target['path'] = removeDotSegments($reference['path']); |
| 2299 |
$target['query'] = $reference['query']; |
| 2300 |
} else { |
| 2301 |
if($reference['path'] == '') { |
| 2302 |
$target['path'] = $base['path']; |
| 2303 |
if($reference['query']) { |
| 2304 |
$target['query'] = $reference['query']; |
| 2305 |
} else { |
| 2306 |
$target['query'] = $base['query']; |
| 2307 |
} |
| 2308 |
} else { |
| 2309 |
if(substr($reference['path'], 0, 1) == '/') { |
| 2310 |
$target['path'] = removeDotSegments($reference['path']); |
| 2311 |
} else { |
| 2312 |
$target['path'] = mergePaths($base, $reference); |
| 2313 |
$target['path'] = removeDotSegments($target['path']); |
| 2314 |
} |
| 2315 |
$target['query'] = $reference['query']; |
| 2316 |
} |
| 2317 |
$target['authority'] = $base['authority']; |
| 2318 |
} |
| 2319 |
$target['scheme'] = $base['scheme']; |
| 2320 |
} |
| 2321 |
$target['fragment'] = $reference['fragment']; |
| 2322 |
|
| 2323 |
# 5.3 Component Recomposition |
| 2324 |
$result = ''; |
| 2325 |
if($target['scheme']) { |
| 2326 |
$result .= $target['scheme'] . ':'; |
| 2327 |
} |
| 2328 |
if($target['authority']) { |
| 2329 |
$result .= '//' . $target['authority']; |
| 2330 |
} |
| 2331 |
$result .= $target['path']; |
| 2332 |
if($target['query']) { |
| 2333 |
$result .= '?' . $target['query']; |
| 2334 |
} |
| 2335 |
if($target['fragment']) { |
| 2336 |
$result .= '#' . $target['fragment']; |
| 2337 |
} elseif($referenceURI == '#') { |
| 2338 |
$result .= '#'; |
| 2339 |
} |
| 2340 |
return $result; |
| 2341 |
} |
| 2342 |
|
| 2343 |
# 5.2.3 Merge Paths |
| 2344 |
function mergePaths($base, $reference) { |
| 2345 |
# If the base URI has a defined authority component and an empty |
| 2346 |
# path, |
| 2347 |
if($base['authority'] && $base['path'] == null) { |
| 2348 |
# then return a string consisting of "/" concatenated with the |
| 2349 |
# reference's path; otherwise, |
| 2350 |
$merged = '/' . $reference['path']; |
| 2351 |
} else { |
| 2352 |
if(($pos=strrpos($base['path'], '/')) !== false) { |
| 2353 |
# return a string consisting of the reference's path component |
| 2354 |
# appended to all but the last segment of the base URI's path (i.e., |
| 2355 |
# excluding any characters after the right-most "/" in the base URI |
| 2356 |
# path, |
| 2357 |
$merged = substr($base['path'], 0, $pos + 1) . $reference['path']; |
| 2358 |
} else { |
| 2359 |
# or excluding the entire base URI path if it does not contain |
| 2360 |
# any "/" characters). |
| 2361 |
$merged = $base['path']; |
| 2362 |
} |
| 2363 |
} |
| 2364 |
return $merged; |
| 2365 |
} |
| 2366 |
|
| 2367 |
# 5.2.4.A Remove leading ../ or ./ |
| 2368 |
function removeLeadingDotSlash(&$input) { |
| 2369 |
if(substr($input, 0, 3) == '../') { |
| 2370 |
$input = substr($input, 3); |
| 2371 |
} elseif(substr($input, 0, 2) == './') { |
| 2372 |
$input = substr($input, 2); |
| 2373 |
} |
| 2374 |
} |
| 2375 |
|
| 2376 |
# 5.2.4.B Replace leading /. with / |
| 2377 |
function removeLeadingSlashDot(&$input) { |
| 2378 |
if(substr($input, 0, 3) == '/./') { |
| 2379 |
$input = '/' . substr($input, 3); |
| 2380 |
} else { |
| 2381 |
$input = '/' . substr($input, 2); |
| 2382 |
} |
| 2383 |
} |
| 2384 |
|
| 2385 |
# 5.2.4.C Given leading /../ remove component from output buffer |
| 2386 |
function removeOneDirLevel(&$input, &$output) { |
| 2387 |
if(substr($input, 0, 4) == '/../') { |
| 2388 |
$input = '/' . substr($input, 4); |
| 2389 |
} else { |
| 2390 |
$input = '/' . substr($input, 3); |
| 2391 |
} |
| 2392 |
$output = substr($output, 0, strrpos($output, '/')); |
| 2393 |
} |
| 2394 |
|
| 2395 |
# 5.2.4.D Remove . and .. if it's the only thing in the input |
| 2396 |
function removeLoneDotDot(&$input) { |
| 2397 |
if($input == '.') { |
| 2398 |
$input = substr($input, 1); |
| 2399 |
} else { |
| 2400 |
$input = substr($input, 2); |
| 2401 |
} |
| 2402 |
} |
| 2403 |
|
| 2404 |
# 5.2.4.E Move one segment from input to output |
| 2405 |
function moveOneSegmentFromInput(&$input, &$output) { |
| 2406 |
if(substr($input, 0, 1) != '/') { |
| 2407 |
$pos = strpos($input, '/'); |
| 2408 |
} else { |
| 2409 |
$pos = strpos($input, '/', 1); |
| 2410 |
} |
| 2411 |
|
| 2412 |
if($pos === false) { |
| 2413 |
$output .= $input; |
| 2414 |
$input = ''; |
| 2415 |
} else { |
| 2416 |
$output .= substr($input, 0, $pos); |
| 2417 |
$input = substr($input, $pos); |
| 2418 |
} |
| 2419 |
} |
| 2420 |
|
| 2421 |
# 5.2.4 Remove Dot Segments |
| 2422 |
function removeDotSegments($path) { |
| 2423 |
# 1. The input buffer is initialized with the now-appended path |
| 2424 |
# components and the output buffer is initialized to the empty |
| 2425 |
# string. |
| 2426 |
$input = $path; |
| 2427 |
$output = ''; |
| 2428 |
|
| 2429 |
$step = 0; |
| 2430 |
|
| 2431 |
# 2. While the input buffer is not empty, loop as follows: |
| 2432 |
while($input) { |
| 2433 |
$step++; |
| 2434 |
|
| 2435 |
if(substr($input, 0, 3) == '../' || substr($input, 0, 2) == './') { |
| 2436 |
# A. If the input buffer begins with a prefix of "../" or "./", |
| 2437 |
# then remove that prefix from the input buffer; otherwise, |
| 2438 |
removeLeadingDotSlash($input); |
| 2439 |
} elseif(substr($input, 0, 3) == '/./' || $input == '/.') { |
| 2440 |
# B. if the input buffer begins with a prefix of "/./" or "/.", |
| 2441 |
# where "." is a complete path segment, then replace that |
| 2442 |
# prefix with "/" in the input buffer; otherwise, |
| 2443 |
removeLeadingSlashDot($input); |
| 2444 |
} elseif(substr($input, 0, 4) == '/../' || $input == '/..') { |
| 2445 |
# C. if the input buffer begins with a prefix of "/../" or "/..", |
| 2446 |
# where ".." is a complete path segment, then replace that |
| 2447 |
# prefix with "/" in the input buffer and remove the last |
| 2448 |
# segment and its preceding "/" (if any) from the output |
| 2449 |
# buffer; otherwise, |
| 2450 |
removeOneDirLevel($input, $output); |
| 2451 |
} elseif($input == '.' || $input == '..') { |
| 2452 |
# D. if the input buffer consists only of "." or "..", then remove |
| 2453 |
# that from the input buffer; otherwise, |
| 2454 |
removeLoneDotDot($input); |
| 2455 |
} else { |
| 2456 |
# E. move the first path segment in the input buffer to the end of |
| 2457 |
# the output buffer and any subsequent characters up to, but not including, |
| 2458 |
# the next "/" character or the end of the input buffer |
| 2459 |
moveOneSegmentFromInput($input, $output); |
| 2460 |
} |
| 2461 |
} |
| 2462 |
|
| 2463 |
return $output; |
| 2464 |
} |
| 2465 |
|
