0) {
// There is an author h-card on this page
// Now look for the first h-* object other than an h-card and use that as the object
foreach($mf2['items'] as $item) {
if(!in_array('h-card', $item['type'])) {
if(in_array('h-entry', $item['type']) || in_array('h-cite', $item['type'])) {
return self::parseAsHEntry($mf2, $item, $http);
} elseif(in_array('h-event', $item['type'])) {
return self::parseAsHEvent($mf2, $item, $http);
} elseif(in_array('h-review', $item['type'])) {
return self::parseAsHReview($mf2, $item, $http);
} elseif(in_array('h-recipe', $item['type'])) {
return self::parseAsHRecipe($mf2, $item, $http);
} elseif(in_array('h-product', $item['type'])) {
return self::parseAsHProduct($mf2, $item, $http);
}
}
}
}
}
}
}
// If there was more than one h-entry on the page, treat the whole page as a feed
if(count($mf2['items']) > 1) {
if(count(array_filter($mf2['items'], function($item){
return in_array('h-entry', $item['type']);
})) > 1) {
Parse::debug("mf2:2: Recognized $url as an h-feed because there are more than one object on the page");
return self::parseAsHFeed($mf2, $http);
}
}
// If the first item is an h-feed, parse as a feed
$first = $mf2['items'][0];
if(in_array('h-feed', $first['type'])) {
Parse::debug("mf2:3: Recognized $url as an h-feed because the first item is an h-feed");
return self::parseAsHFeed($mf2, $http);
}
// Fallback case, but hopefully we have found something before this point
foreach($mf2['items'] as $item) {
// Otherwise check for a recognized h-entr* object
if(in_array('h-entry', $item['type']) || in_array('h-cite', $item['type'])) {
Parse::debug("mf2:6: $url is falling back to the first h-entry on the page");
return self::parseAsHEntry($mf2, $item, $http);
} elseif(in_array('h-event', $item['type'])) {
Parse::debug("mf2:6: $url is falling back to the first h-event on the page");
return self::parseAsHEvent($mf2, $item, $http);
} elseif(in_array('h-review', $item['type'])) {
Parse::debug("mf2:6: $url is falling back to the first h-review on the page");
return self::parseAsHReview($mf2, $item, $http);
} elseif(in_array('h-recipe', $item['type'])) {
Parse::debug("mf2:6: $url is falling back to the first h-recipe on the page");
return self::parseAsHReview($mf2, $item, $http);
} elseif(in_array('h-product', $item['type'])) {
Parse::debug("mf2:6: $url is falling back to the first h-product on the page");
return self::parseAsHProduct($mf2, $item, $http);
}
}
Parse::debug("mf2:E: No object at $url was recognized");
return false;
}
private static function collectSingleValues($properties, $urlProperties, $item, &$data) {
foreach($properties as $p) {
if(($v = self::getPlaintext($item, $p)) !== null) {
$data[$p] = $v;
}
}
foreach($urlProperties as $p) {
if(($v = self::getPlaintext($item, $p)) !== null) {
if(self::isURL($v))
$data[$p] = $v;
}
}
}
private static function parseHTMLValue($property, $item) {
if(!array_key_exists($property, $item['properties']))
return null;
$textContent = false;
$htmlContent = false;
$content = $item['properties'][$property][0];
if(is_string($content)) {
$textContent = $content;
} elseif(!is_string($content) && is_array($content) && array_key_exists('value', $content)) {
if(array_key_exists('html', $content)) {
$htmlContent = trim(self::sanitizeHTML($content['html']));
#$textContent = trim(str_replace("
","\r",strip_tags($htmlContent)));
$textContent = trim(str_replace("
","\r",$content['value']));
} else {
$textContent = trim($content['value']);
}
}
$data = [
'text' => $textContent
];
if($htmlContent && $textContent != $htmlContent) {
$data['html'] = $htmlContent;
}
return $data;
}
// Always return arrays, and may contain plaintext content
// Nested objects are added to refs and the URL is used as the value if present
private static function collectArrayValues($properties, $item, &$data, &$refs, &$http) {
foreach($properties as $p) {
if(array_key_exists($p, $item['properties'])) {
foreach($item['properties'][$p] as $v) {
if(is_string($v)) {
if(!array_key_exists($p, $data)) $data[$p] = [];
$data[$p][] = $v;
} elseif(self::isMicroformat($v)) {
if(($u=self::getPlaintext($v, 'url')) && self::isURL($u)) {
if(!array_key_exists($p, $data)) $data[$p] = [];
$data[$p][] = $u;
$ref = self::parse(['items'=>[$v]], $u, $http);
if($ref) {
$refs[$u] = $ref['data'];
}
} else {
if(!array_key_exists($p, $data)) $data[$p] = [];
$data[$p][] = $v['value'];
}
}
}
}
}
}
private static function collectArrayURLValues($properties, $item, &$data, &$refs, &$http) {
foreach($properties as $p) {
if(array_key_exists($p, $item['properties'])) {
foreach($item['properties'][$p] as $v) {
if(is_string($v) && self::isURL($v)) {
if(!array_key_exists($p, $data)) $data[$p] = [];
$data[$p][] = $v;
}
elseif(self::isMicroformat($v) && ($u=self::getPlaintext($v, 'url')) && self::isURL($u)) {
if(!array_key_exists($p, $data)) $data[$p] = [];
$data[$p][] = $u;
// parse the object and put the result in the "refs" object
$ref = self::parse(['items'=>[$v]], $u, $http);
if($ref) {
$refs[$u] = $ref['data'];
}
}
}
}
}
}
private static function determineNameAndContent($item, &$data) {
// Determine if the name is distinct from the content
$name = self::getPlaintext($item, 'name');
$textContent = null;
$htmlContent = null;
$content = self::parseHTMLValue('content', $item);
if($content) {
$htmlContent = array_key_exists('html', $content) ? $content['html'] : null;
$textContent = array_key_exists('text', $content) ? $content['text'] : null;
}
if($content) {
// Trim ellipses from the name
$name = preg_replace('/ ?(\.\.\.|…)$/', '', $name);
// Remove all whitespace when checking equality
$nameCompare = preg_replace('/\s/','',trim($name));
$contentCompare = preg_replace('/\s/','',trim($textContent));
// Check if the name is a prefix of the content
if($contentCompare && $nameCompare && strpos($contentCompare, $nameCompare) === 0) {
$name = null;
}
}
if($name) {
$data['name'] = $name;
}
// If there is content, always return the plaintext content, and return HTML content if it's different
if($content) {
$data['content']['text'] = $content['text'];
if(array_key_exists('html', $content))
$data['content']['html'] = $content['html'];
}
}
private static function parseAsHEntry($mf2, $item, $http) {
$data = [
'type' => 'entry'
];
$refs = [];
// Single plaintext and URL values
self::collectSingleValues(['published','summary','rsvp','swarm-coins'], ['url'], $item, $data);
// These properties are always returned as arrays and may contain plaintext content
self::collectArrayValues(['category','invitee'], $item, $data, $refs, $http);
// These properties are always returned as arrays and always URLs
// If the value is an h-* object with a URL, the URL is used and a "ref" is added as well
self::collectArrayURLValues(['photo','video','audio','syndication','in-reply-to','like-of','repost-of','bookmark-of'], $item, $data, $refs, $http);
self::determineNameAndContent($item, $data);
if($author = self::findAuthor($mf2, $item, $http))
$data['author'] = $author;
$response = [
'data' => $data
];
if(count($refs)) {
$response['refs'] = $refs;
}
return $response;
}
private static function parseAsHReview($mf2, $item, $http) {
$data = [
'type' => 'review'
];
$refs = [];
self::collectSingleValues(['summary','published','rating','best','worst'], ['url'], $item, $data);
// Fallback for Mf1 "description" as content. The PHP parser does not properly map this to "content"
$description = self::parseHTMLValue('description', $item);
if($description) {
$data['content'] = $description;
}
self::collectArrayValues(['category'], $item, $data, $refs, $http);
self::collectArrayURLValues(['item'], $item, $data, $refs, $http);
self::determineNameAndContent($item, $data);
if($author = self::findAuthor($mf2, $item, $http))
$data['author'] = $author;
$response = [
'data' => $data
];
if(count($refs)) {
$response['refs'] = $refs;
}
return $response;
}
private static function parseAsHRecipe($mf2, $item, $http) {
$data = [
'type' => 'recipe'
];
$refs = [];
self::collectSingleValues(['name','summary','published','duration','yield','nutrition'], ['url'], $item, $data);
$instructions = self::parseHTMLValue('instructions', $item);
if($instructions) {
$data['instructions'] = $instructions;
}
self::collectArrayValues(['category','ingredient'], $item, $data, $refs, $http);
self::collectArrayURLValues(['photo'], $item, $data, $refs, $http);
if($author = self::findAuthor($mf2, $item, $http))
$data['author'] = $author;
$response = [
'data' => $data
];
if(count($refs)) {
$response['refs'] = $refs;
}
return $response;
}
private static function parseAsHProduct($mf2, $item, $http) {
$data = [
'type' => 'product'
];
self::collectSingleValues(['name','identifier','price'], ['url'], $item, $data);
$description = self::parseHTMLValue('description', $item);
if($description) {
$data['description'] = $description;
}
self::collectArrayValues(['category','brand'], $item, $data, $refs, $http);
self::collectArrayURLValues(['photo','video','audio'], $item, $data, $refs, $http);
$response = [
'data' => $data
];
if(count($refs)) {
$response['refs'] = $refs;
}
return $response;
}
private static function parseAsHEvent($mf2, $item, $http) {
$data = [
'type' => 'event'
];
$refs = [];
// Single plaintext and URL values
self::collectSingleValues(['name','summary','published','start','end','duration'], ['url'], $item, $data);
// These properties are always returned as arrays and may contain plaintext content
self::collectArrayValues(['category','location','attendee'], $item, $data, $refs, $http);
// These properties are always returned as arrays and always URLs
// If the value is an h-* object with a URL, the URL is used and a "ref" is added as well
self::collectArrayURLValues(['photo','video','audio','syndication'], $item, $data, $refs, $http);
// If there is a description, always return the plaintext description, and return HTML description if it's different
$textDescription = null;
$htmlDescription = null;
if(array_key_exists('description', $item['properties'])) {
$description = $item['properties']['description'][0];
if(is_string($description)) {
$textDescription = $description;
} elseif(!is_string($description) && is_array($description) && array_key_exists('value', $description)) {
if(array_key_exists('html', $description)) {
$htmlDescription = trim(self::sanitizeHTML($description['html']));
$textDescription = trim(str_replace("
","\r",strip_tags($htmlDescription)));
$textDescription = trim(str_replace("
","\r",$description['value']));
} else {
$textDescription = trim($description['value']);
}
}
}
if($textDescription) {
$data['description'] = [
'text' => $textDescription
];
if($htmlDescription && $textDescription != $htmlDescription) {
$data['description']['html'] = $htmlDescription;
}
}
$response = [
'data' => $data
];
if(count($refs)) {
$response['refs'] = $refs;
}
return $response;
}
private static function parseAsHFeed($mf2, $http) {
$data = [
'type' => 'feed',
'author' => [
'type' => 'card',
'name' => null,
'url' => null,
'photo' => null
],
'todo' => 'Not yet implemented. Please see https://github.com/aaronpk/XRay/issues/1'
];
return [
'data' => $data,
'entries' => []
];
}
private static function parseAsHCard($item, $http, $authorURL=false) {
$data = [
'type' => 'card',
'name' => null,
'url' => null,
'photo' => null
];
$properties = ['url','name','photo'];
foreach($properties as $p) {
if($p == 'url' && $authorURL) {
// If there is a matching author URL, use that one
$found = false;
foreach($item['properties']['url'] as $url) {
if(self::isURL($url)) {
$url = self::normalize_url($url);
if($url == $authorURL) {
$data['url'] = $url;
$found = true;
}
}
}
if(!$found && self::isURL($item['properties']['url'][0])) {
$data['url'] = $item['properties']['url'][0];
}
} else if(($v = self::getPlaintext($item, $p)) !== null) {
// Make sure the URL property is actually a URL
if($p == 'url' || $p == 'photo') {
if(self::isURL($v))
$data[$p] = $v;
} else {
$data[$p] = $v;
}
}
}
// If no URL property was found, use the $authorURL provided
if(!$data['url'])
$data['url'] = $authorURL;
$response = [
'data' => $data
];
return $response;
}
private static function findAuthor($mf2, $item, $http) {
$author = [
'type' => 'card',
'name' => null,
'url' => null,
'photo' => null
];
// Author Discovery
// http://indiewebcamp.com/authorship
$authorPage = false;
if(array_key_exists('author', $item['properties'])) {
// Check if any of the values of the author property are an h-card
foreach($item['properties']['author'] as $a) {
if(self::isHCard($a)) {
// 5.1 "if it has an h-card, use it, exit."
return self::parseAsHCard($a, $http)['data'];
} elseif(is_string($a)) {
if(self::isURL($a)) {
// 5.2 "otherwise if author property is an http(s) URL, let the author-page have that URL"
$authorPage = $a;
} else {
// 5.3 "otherwise use the author property as the author name, exit"
// We can only set the name, no h-card or URL was found
$author['name'] = self::getPlaintext($item, 'author');
return $author;
}
} else {
// This case is only hit when the author property is an mf2 object that is not an h-card
$author['name'] = self::getPlaintext($item, 'author');
return $author;
}
}
}
// 6. "if no author page was found" ... check for rel-author link
if(!$authorPage) {
if(isset($mf2['rels']) && isset($mf2['rels']['author']))
$authorPage = $mf2['rels']['author'][0];
}
// 7. "if there is an author-page URL" ...
if($authorPage) {
// 7.1 "get the author-page from that URL and parse it for microformats2"
$authorPageContents = self::getURL($authorPage, $http);
if($authorPageContents) {
foreach($authorPageContents['items'] as $i) {
if(self::isHCard($i)) {
// 7.2 "if author-page has 1+ h-card with url == uid == author-page's URL, then use first such h-card, exit."
if(array_key_exists('url', $i['properties'])
and in_array($authorPage, $i['properties']['url'])
and array_key_exists('uid', $i['properties'])
and in_array($authorPage, $i['properties']['uid'])
) {
return self::parseAsHCard($i, $http, $authorPage)['data'];
}
// 7.3 "else if author-page has 1+ h-card with url property which matches the href of a rel-me link on the author-page"
$relMeLinks = (isset($authorPageContents['rels']) && isset($authorPageContents['rels']['me'])) ? $authorPageContents['rels']['me'] : [];
if(count($relMeLinks) > 0
and array_key_exists('url', $i['properties'])
and count(array_intersect($i['properties']['url'], $relMeLinks)) > 0
) {
return self::parseAsHCard($i, $http, $authorPage)['data'];
}
}
}
}
// 7.4 "if the h-entry's page has 1+ h-card with url == author-page URL, use first such h-card, exit."
foreach($mf2['items'] as $i) {
if(self::isHCard($i)) {
if(array_key_exists('url', $i['properties'])
and in_array($authorPage, $i['properties']['url'])
) {
return self::parseAsHCard($i, $http)['data'];
}
}
}
}
if(!$author['name'] && !$author['photo'] && !$author['url'])
return null;
return $author;
}
private static function sanitizeHTML($html) {
$config = HTMLPurifier_Config::createDefault();
$config->set('Cache.DefinitionImpl', null);
$config->set('HTML.AllowedElements', [
'a',
'abbr',
'b',
'code',
'del',
'em',
'i',
'img',
'q',
'strike',
'strong',
'time',
'blockquote',
'pre',
'p',
'h1',
'h2',
'h3',
'h4',
'h5',
'h6',
'ul',
'li',
'ol'
]);
$def = $config->getHTMLDefinition(true);
$def->addElement(
'time',
'Inline',
'Inline',
'Common',
[
'datetime' => 'Text'
]
);
// Override the allowed classes to only support Microformats2 classes
$def->manager->attrTypes->set('Class', new HTMLPurifier_AttrDef_HTML_Microformats2());
$purifier = new HTMLPurifier($config);
$sanitized = $purifier->purify($html);
$sanitized = str_replace("
","\r",$sanitized);
return $sanitized;
}
private static function hasNumericKeys(array $arr) {
foreach($arr as $key=>$val)
if (is_numeric($key))
return true;
return false;
}
private static function isMicroformat($mf) {
return is_array($mf)
and !self::hasNumericKeys($mf)
and !empty($mf['type'])
and isset($mf['properties']);
}
private static function isHCard($mf) {
return is_array($mf)
and !empty($mf['type'])
and is_array($mf['type'])
and in_array('h-card', $mf['type']);
}
private static function isURL($string) {
return preg_match('/^https?:\/\/.+\..+$/', $string);
}
// Given an array of microformats properties and a key name, return the plaintext value
// at that property
// e.g.
// {"properties":{"published":["foo"]}} results in "foo"
private static function getPlaintext($mf2, $k, $fallback=null) {
if(!empty($mf2['properties'][$k]) and is_array($mf2['properties'][$k])) {
// $mf2['properties'][$v] will always be an array since the input was from the mf2 parser
$value = $mf2['properties'][$k][0];
if(is_string($value)) {
return $value;
} elseif(self::isMicroformat($value) && array_key_exists('value', $value)) {
return $value['value'];
}
}
return $fallback;
}
private static function getURL($url, $http) {
if(!$url) return null;
// TODO: consider adding caching here
$result = $http->get($url);
if($result['error'] || !$result['body']) {
return null;
}
return \mf2\Parse($result['body'], $url);
}
private static function normalize_url($url) {
$parts = parse_url($url);
if(empty($parts['path']))
$parts['path'] = '/';
$parts['host'] = strtolower($parts['host']);
return self::build_url($parts);
}
private static function build_url($parsed_url) {
$scheme = isset($parsed_url['scheme']) ? $parsed_url['scheme'] . '://' : '';
$host = isset($parsed_url['host']) ? $parsed_url['host'] : '';
$port = isset($parsed_url['port']) ? ':' . $parsed_url['port'] : '';
$user = isset($parsed_url['user']) ? $parsed_url['user'] : '';
$pass = isset($parsed_url['pass']) ? ':' . $parsed_url['pass'] : '';
$pass = ($user || $pass) ? "$pass@" : '';
$path = isset($parsed_url['path']) ? $parsed_url['path'] : '';
$query = isset($parsed_url['query']) ? '?' . $parsed_url['query'] : '';
$fragment = isset($parsed_url['fragment']) ? '#' . $parsed_url['fragment'] : '';
return "$scheme$user$pass$host$port$path$query$fragment";
}
}