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