X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=include%2FScrape.php;h=0169f55fd5d93db839d0dfa77250a8fc81fac0be;hb=be70dde61f14514d31da090e44546ce021b0a60a;hp=99784af336120cf08e0a76ac306947712309de3e;hpb=1940905454b311fc5999a0d8020832bb92ee52f5;p=friendica.git diff --git a/include/Scrape.php b/include/Scrape.php index 99784af336..0169f55fd5 100644 --- a/include/Scrape.php +++ b/include/Scrape.php @@ -249,7 +249,7 @@ function scrape_feed($url) { $ret['feed_atom'] = $url; return $ret; } - if(stristr($line,'application/rss+xml') || stristr($s,'loadHTML($s); + $xpath = new DomXPath($doc); + $base = $xpath->query("//base"); + foreach ($base as $node) { + $attr = array(); - $head = $dom->getElementsByTagName('base'); - if($head) { - foreach($head as $head0) { - $basename = $head0->getAttribute('href'); - break; - } + if ($node->attributes->length) + foreach ($node->attributes as $attribute) + $attr[$attribute->name] = $attribute->value; + + if ($attr["href"] != "") + $basename = $attr["href"] ; } - if(! $basename) - $basename = implode('/', array_slice(explode('/',$url),0,3)) . '/'; - $items = $dom->getElementsByTagName('link'); + $list = $xpath->query("//link"); + foreach ($list as $node) { + $attr = array(); - // get Atom/RSS link elements, take the first one of either. + if ($node->attributes->length) + foreach ($node->attributes as $attribute) + $attr[$attribute->name] = $attribute->value; - if($items) { - foreach($items as $item) { - $x = $item->getAttribute('rel'); - if(($x === 'alternate') && ($item->getAttribute('type') === 'application/atom+xml')) { - if(! x($ret,'feed_atom')) - $ret['feed_atom'] = $item->getAttribute('href'); - } - if(($x === 'alternate') && ($item->getAttribute('type') === 'application/rss+xml')) { - if(! x($ret,'feed_rss')) - $ret['feed_rss'] = $item->getAttribute('href'); - } - } + if (($attr["rel"] == "alternate") AND ($attr["type"] == "application/atom+xml")) + $ret["feed_atom"] = $attr["href"]; + + if (($attr["rel"] == "alternate") AND ($attr["type"] == "application/rss+xml")) + $ret["feed_rss"] = $attr["href"]; } - // Drupal and perhaps others only provide relative URL's. Turn them into absolute. + // Drupal and perhaps others only provide relative URLs. Turn them into absolute. if(x($ret,'feed_atom') && (! strstr($ret['feed_atom'],'://'))) $ret['feed_atom'] = $basename . $ret['feed_atom']; @@ -335,7 +328,7 @@ function scrape_feed($url) { define ( 'PROBE_NORMAL', 0); define ( 'PROBE_DIASPORA', 1); -function probe_url($url, $mode = PROBE_NORMAL) { +function probe_url($url, $mode = PROBE_NORMAL, $level = 1) { require_once('include/email.php'); $result = array(); @@ -343,6 +336,12 @@ function probe_url($url, $mode = PROBE_NORMAL) { if(! $url) return $result; + $result = Cache::get("probe_url:".$mode.":".$url); + if (!is_null($result)) { + $result = unserialize($result); + return $result; + } + $network = null; $diaspora = false; $diaspora_base = ''; @@ -350,6 +349,13 @@ function probe_url($url, $mode = PROBE_NORMAL) { $diaspora_key = ''; $has_lrdd = false; $email_conversant = false; + $connectornetworks = false; + $appnet = false; + + if (strpos($url,'twitter.com')) { + $connectornetworks = true; + $network = NETWORK_TWITTER; + } // Twitter is deactivated since twitter closed its old API //$twitter = ((strpos($url,'twitter.com') !== false) ? true : false); @@ -357,7 +363,7 @@ function probe_url($url, $mode = PROBE_NORMAL) { $at_addr = ((strpos($url,'@') !== false) ? true : false); - if((! $twitter) && (! $lastfm)) { + if((!$appnet) && (!$lastfm) && !$connectornetworks) { if(strpos($url,'mailto:') !== false && $at_addr) { $url = str_replace('mailto:','',$url); @@ -401,13 +407,16 @@ function probe_url($url, $mode = PROBE_NORMAL) { $pubkey = $diaspora_key; $diaspora = true; } + if($link['@attributes']['rel'] === 'http://ostatus.org/schema/1.0/subscribe') { + $diaspora = false; + } } // Status.Net can have more than one profile URL. We need to match the profile URL // to a contact on incoming messages to prevent spam, and we won't know which one // to match. So in case of two, one of them is stored as an alias. Only store URL's // and not webfinger user@host aliases. If they've got more than two non-email style - // aliases, let's hope we're lucky and get one that matches the feed author-uri because + // aliases, let's hope we're lucky and get one that matches the feed author-uri because // otherwise we're screwed. foreach($links as $link) { @@ -422,6 +431,10 @@ function probe_url($url, $mode = PROBE_NORMAL) { } } } + + // If the profile is different from the url then the url is abviously an alias + if (($alias == "") AND ($profile != "") AND !$at_addr AND (normalise_link($profile) != normalise_link($url))) + $alias = $url; } elseif($mode == PROBE_NORMAL) { @@ -489,6 +502,7 @@ function probe_url($url, $mode = PROBE_NORMAL) { } if($mode == PROBE_NORMAL) { + if(strlen($zot)) { $s = fetch_url($zot); if($s) { @@ -496,8 +510,8 @@ function probe_url($url, $mode = PROBE_NORMAL) { if($j) { $network = NETWORK_ZOT; $vcard = array( - 'fn' => $j->fullname, - 'nick' => $j->nickname, + 'fn' => $j->fullname, + 'nick' => $j->nickname, 'photo' => $j->photo ); $profile = $j->url; @@ -508,6 +522,7 @@ function probe_url($url, $mode = PROBE_NORMAL) { } } + if(strlen($dfrn)) { $ret = scrape_dfrn(($hcard) ? $hcard : $dfrn); if(is_array($ret) && x($ret,'dfrn-request')) { @@ -537,8 +552,12 @@ function probe_url($url, $mode = PROBE_NORMAL) { if($network !== NETWORK_ZOT && $network !== NETWORK_DFRN && $network !== NETWORK_MAIL) { if($diaspora) $network = NETWORK_DIASPORA; - elseif($has_lrdd) + elseif($has_lrdd AND ($notify)) $network = NETWORK_OSTATUS; + + if(strpos($url,'@')) + $addr = str_replace('acct:', '', $url); + $priority = 0; if($hcard && ! $vcard) { @@ -597,16 +616,20 @@ function probe_url($url, $mode = PROBE_NORMAL) { // Will leave it to others to figure out how to grab the avatar, which is on the $url page in the open graph meta links } - if($twitter || ! $poll) + if($appnet || ! $poll) $check_feed = true; if((! isset($vcard)) || (! x($vcard,'fn')) || (! $profile)) $check_feed = true; if(($at_addr) && (! count($links))) $check_feed = false; + if ($connectornetworks) + $check_feed = false; + if($check_feed) { $feedret = scrape_feed(($poll) ? $poll : $url); + logger('probe_url: scrape_feed ' . (($poll)? $poll : $url) . ' returns: ' . print_r($feedret,true), LOGGER_DATA); if(count($feedret) && ($feedret['feed_atom'] || $feedret['feed_rss'])) { $poll = ((x($feedret,'feed_atom')) ? unamp($feedret['feed_atom']) : unamp($feedret['feed_rss'])); @@ -626,12 +649,13 @@ function probe_url($url, $mode = PROBE_NORMAL) { logger('probe_url: scrape_feed: headers: ' . $a->get_curl_headers(), LOGGER_DATA); // Don't try and parse an empty string - $feed->set_raw_data(($xml) ? $xml : ''); + $feed->set_raw_data(($xml) ? $xml : ''); $feed->init(); - if($feed->error()) + if($feed->error()) { logger('probe_url: scrape_feed: Error parsing XML: ' . $feed->error()); - + $network = NETWORK_PHANTOM; + } if(! x($vcard,'photo')) $vcard['photo'] = $feed->get_image_url(); @@ -643,6 +667,7 @@ function probe_url($url, $mode = PROBE_NORMAL) { $vcard['fn'] = trim(unxmlify($author->get_email())); if(strpos($vcard['fn'],'@') !== false) $vcard['fn'] = substr($vcard['fn'],0,strpos($vcard['fn'],'@')); + $email = unxmlify($author->get_email()); if(! $profile && $author->get_link()) $profile = trim(unxmlify($author->get_link())); @@ -654,6 +679,15 @@ function probe_url($url, $mode = PROBE_NORMAL) { $vcard['photo'] = $elems['link'][0]['attribs']['']['href']; } } + // Fetch fullname via poco:displayName + $pocotags = $feed->get_feed_tags(SIMPLEPIE_NAMESPACE_ATOM_10, 'author'); + if ($pocotags) { + $elems = $pocotags[0]['child']['http://portablecontacts.net/spec/1.0']; + if (isset($elems["displayName"])) + $vcard['fn'] = $elems["displayName"][0]["data"]; + if (isset($elems["preferredUsername"])) + $vcard['nick'] = $elems["preferredUsername"][0]["data"]; + } } else { $item = $feed->get_item(0); @@ -729,6 +763,22 @@ function probe_url($url, $mode = PROBE_NORMAL) { if(($network === NETWORK_FEED) && ($poll) && (! x($vcard,'fn'))) $vcard['fn'] = $url; + if (($notify != "") AND ($poll != "")) { + $baseurl = matching(normalise_link($notify), normalise_link($poll)); + + $baseurl2 = matching($baseurl, normalise_link($profile)); + if ($baseurl2 != "") + $baseurl = $baseurl2; + } + + if (($baseurl == "") AND ($notify != "")) + $baseurl = matching(normalise_link($profile), normalise_link($notify)); + + if (($baseurl == "") AND ($poll != "")) + $baseurl = matching(normalise_link($profile), normalise_link($poll)); + + $baseurl = rtrim($baseurl, "/"); + $vcard['fn'] = notags($vcard['fn']); $vcard['nick'] = str_replace(' ','',notags($vcard['nick'])); @@ -747,8 +797,49 @@ function probe_url($url, $mode = PROBE_NORMAL) { $result['network'] = $network; $result['alias'] = $alias; $result['pubkey'] = $pubkey; + $result['baseurl'] = $baseurl; logger('probe_url: ' . print_r($result,true), LOGGER_DEBUG); + if ($level == 1) { + // Trying if it maybe a diaspora account + if (($result['network'] == NETWORK_FEED) OR ($result['addr'] == "")) { + require_once('include/bbcode.php'); + $address = GetProfileUsername($url, "", true); + $result2 = probe_url($address, $mode, ++$level); + if ($result2['network'] != "") + $result = $result2; + } + + // Maybe it's some non standard GNU Social installation (Single user, subfolder or no uri rewrite) + if (($result['network'] == NETWORK_FEED) AND ($result['baseurl'] != "") AND ($result['nick'] != "")) { + $addr = $result['nick'].'@'.str_replace("http://", "", $result['baseurl']); + $result2 = probe_url($addr, $mode, ++$level); + if (($result2['network'] != "") AND ($result2['network'] != NETWORK_FEED)) + $result = $result2; + } + } + + // Only store into the cache if the value seems to be valid + if ($result['network'] != NETWORK_FEED) + Cache::set("probe_url:".$mode.":".$url,serialize($result), CACHE_DAY); + return $result; } + +function matching($part1, $part2) { + $len = min(strlen($part1), strlen($part2)); + + $match = ""; + $matching = true; + $i = 0; + while (($i <= $len) AND $matching) { + if (substr($part1, $i, 1) == substr($part2, $i, 1)) + $match .= substr($part1, $i, 1); + else + $matching = false; + + $i++; + } + return($match); +}