You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 

128 lines
3.8 KiB

<?php
namespace p3k\XRay;
use p3k\XRay\Formats;
class Feeds {
private $http;
public function __construct($http) {
$this->http = $http;
}
public function find($url, $opts=[]) {
if(isset($opts['timeout']))
$this->http->set_timeout($opts['timeout']);
if(isset($opts['max_redirects']))
$this->http->set_max_redirects($opts['max_redirects']);
$scheme = parse_url($url, PHP_URL_SCHEME);
if(!in_array($scheme, ['http','https'])) {
return [
'error' => 'invalid_url',
'error_description' => 'Only http and https URLs are supported'
];
}
$host = parse_url($url, PHP_URL_HOST);
if(!$host) {
return [
'error' => 'invalid_url',
'error_description' => 'The URL provided was not valid'
];
}
$url = normalize_url($url);
$result = $this->http->get($url);
if(isset($result['error']) && $result['error']) {
return [
'error' => $result['error'],
'error_description' => $result['error_description']
];
}
$body = $result['body'];
$feeds = [];
// First check the content type of the response
$contentType = isset($result['headers']['Content-Type']) ? $result['headers']['Content-Type'] : '';
if(is_array($contentType))
$contentType = $contentType[count($contentType)-1];
if(strpos($contentType, 'application/atom+xml') !== false || strpos(substr($body, 0, 50), '<feed ') !== false) {
$feeds[] = [
'url' => $result['url'],
'type' => 'atom'
];
} elseif(strpos($contentType, 'application/rss+xml') !== false || strpos($contentType, 'text/xml') !== false
|| strpos(substr($body, 0, 50), '<rss ') !== false) {
$feeds[] = [
'url' => $result['url'],
'type' => 'rss'
];
} elseif(strpos($contentType, 'application/json') !== false && substr($body, 0, 1) == '{') {
$feeddata = json_decode($body, true);
if($feeddata && isset($feeddata['version']) && $feeddata['version'] == 'https://jsonfeed.org/version/1') {
$feeds[] = [
'url' => $result['url'],
'type' => 'jsonfeed'
];
}
} else {
// Some other document was returned, parse the HTML and look for rel alternates and Microformats
$mf2 = \mf2\Parse($body, $result['url']);
if(isset($mf2['rel-urls'])) {
foreach($mf2['rel-urls'] as $rel=>$info) {
if(isset($info['rels']) && in_array('alternate', $info['rels'])) {
if(isset($info['type'])) {
if(strpos($info['type'], 'application/json') !== false) {
$feeds[] = [
'url' => $rel,
'type' => 'jsonfeed'
];
}
if(strpos($info['type'], 'application/atom+xml') !== false) {
$feeds[] = [
'url' => $rel,
'type' => 'atom'
];
}
if(strpos($info['type'], 'application/rss+xml') !== false) {
$feeds[] = [
'url' => $rel,
'type' => 'rss'
];
}
}
}
}
}
$parsed = Formats\HTML::parse($this->http, $body, $result['url'], array_merge($opts, ['expect'=>'feed']));
if($parsed && isset($parsed['data']['type']) && $parsed['data']['type'] == 'feed') {
$feeds[] = [
'url' => $result['url'],
'type' => 'microformats'
];
}
}
// Sort feeds by priority
$rank = ['microformats'=>0,'jsonfeed'=>1,'atom'=>2,'rss'=>3];
usort($feeds, function($a, $b) use($rank) {
return $rank[$a['type']] > $rank[$b['type']];
});
return [
'url' => $result['url'],
'code' => $result['code'],
'feeds' => $feeds,
];
}
}