friendica/include/Scrape.php

924 lines
26 KiB
PHP
Raw Normal View History

2010-07-01 19:48:07 -04:00
<?php
require_once('library/HTML5/Parser.php');
2011-08-12 06:01:11 -04:00
require_once('include/crypto.php');
require_once('include/feed.php');
2010-07-01 19:48:07 -04:00
if(! function_exists('scrape_dfrn')) {
function scrape_dfrn($url, $dont_probe = false) {
2010-07-01 19:48:07 -04:00
$a = get_app();
2010-07-01 19:48:07 -04:00
$ret = array();
logger('scrape_dfrn: url=' . $url);
// Try to fetch the data from noscrape. This is faster than parsing the HTML
$noscrape = str_replace("/hcard/", "/noscrape/", $url);
$noscrapejson = fetch_url($noscrape);
$noscrapedata = array();
if ($noscrapejson) {
$noscrapedata = json_decode($noscrapejson, true);
if (is_array($noscrapedata)) {
if ($noscrapedata["nick"] != "")
return($noscrapedata);
else
unset($noscrapedata["nick"]);
} else
$noscrapedata = array();
}
2010-07-01 19:48:07 -04:00
$s = fetch_url($url);
if (!$s)
2010-07-01 19:48:07 -04:00
return $ret;
if (!$dont_probe) {
$probe = probe_url($url);
if (isset($probe["addr"]))
$ret["addr"] = $probe["addr"];
}
$headers = $a->get_curl_headers();
logger('scrape_dfrn: headers=' . $headers, LOGGER_DEBUG);
$lines = explode("\n",$headers);
if(count($lines)) {
foreach($lines as $line) {
// don't try and run feeds through the html5 parser
if(stristr($line,'content-type:') && ((stristr($line,'application/atom+xml')) || (stristr($line,'application/rss+xml'))))
return ret;
}
}
2011-10-20 19:48:07 -04:00
try {
$dom = HTML5_Parser::parse($s);
} catch (DOMException $e) {
logger('scrape_dfrn: parse error: ' . $e);
}
2010-07-01 19:48:07 -04:00
if(! $dom)
return $ret;
$items = $dom->getElementsByTagName('link');
// get DFRN link elements
foreach($items as $item) {
$x = $item->getAttribute('rel');
2011-01-23 23:09:34 -05:00
if(($x === 'alternate') && ($item->getAttribute('type') === 'application/atom+xml'))
$ret['feed_atom'] = $item->getAttribute('href');
2011-08-18 02:01:44 -04:00
if(substr($x,0,5) == "dfrn-") {
2010-07-01 19:48:07 -04:00
$ret[$x] = $item->getAttribute('href');
2011-08-18 02:01:44 -04:00
}
if($x === 'lrdd') {
$decoded = urldecode($item->getAttribute('href'));
if(preg_match('/acct:([^@]*)@/',$decoded,$matches))
$ret['nick'] = $matches[1];
}
2010-07-01 19:48:07 -04:00
}
// Pull out hCard profile elements
2011-08-18 02:01:44 -04:00
$largest_photo = 0;
2010-07-01 19:48:07 -04:00
$items = $dom->getElementsByTagName('*');
foreach($items as $item) {
if(attribute_contains($item->getAttribute('class'), 'vcard')) {
$level2 = $item->getElementsByTagName('*');
foreach($level2 as $x) {
2011-08-18 02:01:44 -04:00
if(attribute_contains($x->getAttribute('class'),'fn')) {
2010-07-01 19:48:07 -04:00
$ret['fn'] = $x->textContent;
2011-08-18 02:01:44 -04:00
}
if((attribute_contains($x->getAttribute('class'),'photo'))
|| (attribute_contains($x->getAttribute('class'),'avatar'))) {
$size = intval($x->getAttribute('width'));
// dfrn prefers 175, so if we find this, we set largest_size so it can't be topped.
if(($size > $largest_photo) || ($size == 175) || (! $largest_photo)) {
$ret['photo'] = $x->getAttribute('src');
$largest_photo = (($size == 175) ? 9999 : $size);
}
}
if(attribute_contains($x->getAttribute('class'),'key')) {
2010-07-01 19:48:07 -04:00
$ret['key'] = $x->textContent;
2011-08-18 02:01:44 -04:00
}
}
2010-07-01 19:48:07 -04:00
}
}
return array_merge($ret, $noscrapedata);
2010-07-01 19:48:07 -04:00
}}
if(! function_exists('validate_dfrn')) {
function validate_dfrn($a) {
$errors = 0;
if(! x($a,'key'))
$errors ++;
if(! x($a,'dfrn-request'))
$errors ++;
if(! x($a,'dfrn-confirm'))
$errors ++;
if(! x($a,'dfrn-notify'))
$errors ++;
if(! x($a,'dfrn-poll'))
$errors ++;
return $errors;
}}
if(! function_exists('scrape_meta')) {
function scrape_meta($url) {
2010-07-01 19:48:07 -04:00
$a = get_app();
$ret = array();
logger('scrape_meta: url=' . $url);
$s = fetch_url($url);
if(! $s)
return $ret;
$headers = $a->get_curl_headers();
logger('scrape_meta: headers=' . $headers, LOGGER_DEBUG);
$lines = explode("\n",$headers);
if(count($lines)) {
foreach($lines as $line) {
// don't try and run feeds through the html5 parser
if(stristr($line,'content-type:') && ((stristr($line,'application/atom+xml')) || (stristr($line,'application/rss+xml'))))
return ret;
}
}
2011-10-20 19:48:07 -04:00
try {
$dom = HTML5_Parser::parse($s);
} catch (DOMException $e) {
logger('scrape_meta: parse error: ' . $e);
}
if(! $dom)
return $ret;
2010-07-01 19:48:07 -04:00
$items = $dom->getElementsByTagName('meta');
// get DFRN link elements
foreach($items as $item) {
$x = $item->getAttribute('name');
if(substr($x,0,5) == "dfrn-")
$ret[$x] = $item->getAttribute('content');
}
return $ret;
}}
if(! function_exists('scrape_vcard')) {
function scrape_vcard($url) {
$a = get_app();
$ret = array();
logger('scrape_vcard: url=' . $url);
$s = fetch_url($url);
if(! $s)
return $ret;
$headers = $a->get_curl_headers();
$lines = explode("\n",$headers);
if(count($lines)) {
foreach($lines as $line) {
// don't try and run feeds through the html5 parser
if(stristr($line,'content-type:') && ((stristr($line,'application/atom+xml')) || (stristr($line,'application/rss+xml'))))
return ret;
}
}
2011-10-20 19:48:07 -04:00
try {
$dom = HTML5_Parser::parse($s);
} catch (DOMException $e) {
logger('scrape_vcard: parse error: ' . $e);
}
if(! $dom)
return $ret;
// Pull out hCard profile elements
2011-08-12 06:01:11 -04:00
$largest_photo = 0;
$items = $dom->getElementsByTagName('*');
foreach($items as $item) {
if(attribute_contains($item->getAttribute('class'), 'vcard')) {
$level2 = $item->getElementsByTagName('*');
foreach($level2 as $x) {
if(attribute_contains($x->getAttribute('class'),'fn'))
$ret['fn'] = $x->textContent;
if((attribute_contains($x->getAttribute('class'),'photo'))
2011-08-12 06:01:11 -04:00
|| (attribute_contains($x->getAttribute('class'),'avatar'))) {
$size = intval($x->getAttribute('width'));
if(($size > $largest_photo) || (! $largest_photo)) {
$ret['photo'] = $x->getAttribute('src');
$largest_photo = $size;
}
}
if((attribute_contains($x->getAttribute('class'),'nickname'))
2011-08-18 02:01:44 -04:00
|| (attribute_contains($x->getAttribute('class'),'uid'))) {
$ret['nick'] = $x->textContent;
2011-08-18 02:01:44 -04:00
}
}
}
}
return $ret;
}}
2011-01-23 23:09:34 -05:00
if(! function_exists('scrape_feed')) {
function scrape_feed($url) {
$a = get_app();
2011-01-23 23:09:34 -05:00
$ret = array();
$cookiejar = tempnam(get_temppath(), 'cookiejar-scrape-feed-');
$s = fetch_url($url, false, $redirects, 0, Null, $cookiejar);
unlink($cookiejar);
2011-01-23 23:09:34 -05:00
2012-02-26 17:34:01 -05:00
$headers = $a->get_curl_headers();
$code = $a->get_curl_code();
logger('scrape_feed: returns: ' . $code . ' headers=' . $headers, LOGGER_DEBUG);
if(! $s) {
logger('scrape_feed: no data returned for ' . $url);
2011-01-23 23:09:34 -05:00
return $ret;
2012-02-26 17:34:01 -05:00
}
2011-01-23 23:09:34 -05:00
$lines = explode("\n",$headers);
if(count($lines)) {
foreach($lines as $line) {
if(stristr($line,'content-type:')) {
if(stristr($line,'application/atom+xml') || stristr($s,'<feed')) {
$ret['feed_atom'] = $url;
return $ret;
}
2015-07-22 05:51:37 -04:00
if(stristr($line,'application/rss+xml') || stristr($s,'<rss')) {
$ret['feed_rss'] = $url;
2011-04-04 22:36:18 -04:00
return $ret;
}
}
}
// perhaps an RSS version 1 feed with a generic or incorrect content-type?
if(stristr($s,'</item>')) {
$ret['feed_rss'] = $url;
return $ret;
}
}
2015-07-22 05:51:37 -04:00
$basename = implode('/', array_slice(explode('/',$url),0,3)) . '/';
2011-01-23 23:09:34 -05:00
2015-07-22 05:51:37 -04:00
$doc = new DOMDocument();
@$doc->loadHTML($s);
$xpath = new DomXPath($doc);
2011-01-23 23:09:34 -05:00
2015-07-22 05:51:37 -04:00
$base = $xpath->query("//base");
foreach ($base as $node) {
$attr = array();
2011-04-04 23:40:31 -04:00
2015-07-22 05:51:37 -04:00
if ($node->attributes->length)
foreach ($node->attributes as $attribute)
$attr[$attribute->name] = $attribute->value;
if ($attr["href"] != "")
$basename = $attr["href"] ;
}
2015-07-22 05:51:37 -04:00
$list = $xpath->query("//link");
foreach ($list as $node) {
$attr = array();
2011-01-23 23:09:34 -05:00
2015-07-22 05:51:37 -04:00
if ($node->attributes->length)
foreach ($node->attributes as $attribute)
$attr[$attribute->name] = $attribute->value;
2011-01-23 23:09:34 -05:00
2015-07-22 05:51:37 -04:00
if (($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"] = $attr["href"];
2011-01-23 23:09:34 -05:00
}
2015-07-22 05:51:37 -04:00
// Drupal and perhaps others only provide relative URLs. Turn them into absolute.
if(x($ret,'feed_atom') && (! strstr($ret['feed_atom'],'://')))
$ret['feed_atom'] = $basename . $ret['feed_atom'];
if(x($ret,'feed_rss') && (! strstr($ret['feed_rss'],'://')))
$ret['feed_rss'] = $basename . $ret['feed_rss'];
2011-01-23 23:09:34 -05:00
return $ret;
}}
2011-08-18 02:10:55 -04:00
/**
*
* Probe a network address to discover what kind of protocols we need to communicate with it.
*
* Warning: this function is a bit touchy and there are some subtle dependencies within the logic flow.
* Edit with care.
*
*/
/**
*
* PROBE_DIASPORA has a bias towards returning Diaspora information
* while PROBE_NORMAL has a bias towards dfrn/zot - in the case where
2012-02-19 13:27:54 -05:00
* an address (such as a Friendica address) supports more than one type
* of network.
2011-08-18 02:10:55 -04:00
*
*/
2011-08-18 02:01:44 -04:00
define ( 'PROBE_NORMAL', 0);
define ( 'PROBE_DIASPORA', 1);
function probe_url($url, $mode = PROBE_NORMAL, $level = 1) {
2011-04-16 11:45:08 -04:00
require_once('include/email.php');
$result = array();
if (!$url)
return $result;
$result = Cache::get("probe_url:".$mode.":".$url);
if (!is_null($result)) {
$result = unserialize($result);
return $result;
}
$original_url = $url;
$network = null;
2011-08-12 06:01:11 -04:00
$diaspora = false;
$diaspora_base = '';
$diaspora_guid = '';
2011-08-12 06:01:11 -04:00
$diaspora_key = '';
$has_lrdd = false;
$email_conversant = false;
$connectornetworks = false;
$appnet = false;
if (strpos($url,'twitter.com')) {
$connectornetworks = true;
$network = NETWORK_TWITTER;
}
2012-06-21 01:59:01 -04:00
$lastfm = ((strpos($url,'last.fm/user') !== false) ? true : false);
$at_addr = ((strpos($url,'@') !== false) ? true : false);
if((!$appnet) && (!$lastfm) && !$connectornetworks) {
if(strpos($url,'mailto:') !== false && $at_addr) {
$url = str_replace('mailto:','',$url);
$links = array();
}
else
$links = lrdd($url);
if ((count($links) == 0) AND strstr($url, "/index.php")) {
$url = str_replace("/index.php", "", $url);
$links = lrdd($url);
}
if (count($links)) {
$has_lrdd = true;
logger('probe_url: found lrdd links: ' . print_r($links,true), LOGGER_DATA);
foreach($links as $link) {
2011-07-11 21:28:13 -04:00
if($link['@attributes']['rel'] === NAMESPACE_ZOT)
$zot = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === NAMESPACE_DFRN)
$dfrn = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === 'salmon')
$notify = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === NAMESPACE_FEED)
$poll = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === 'http://microformats.org/profile/hcard')
$hcard = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === 'http://webfinger.net/rel/profile-page')
$profile = unamp($link['@attributes']['href']);
if($link['@attributes']['rel'] === 'http://portablecontacts.net/spec/1.0')
$poco = unamp($link['@attributes']['href']);
2011-08-12 06:01:11 -04:00
if($link['@attributes']['rel'] === 'http://joindiaspora.com/seed_location') {
$diaspora_base = unamp($link['@attributes']['href']);
$diaspora = true;
}
if($link['@attributes']['rel'] === 'http://joindiaspora.com/guid') {
$diaspora_guid = unamp($link['@attributes']['href']);
$diaspora = true;
2011-08-12 06:01:11 -04:00
}
if($link['@attributes']['rel'] === 'diaspora-public-key') {
$diaspora_key = base64_decode(unamp($link['@attributes']['href']));
if(strstr($diaspora_key,'RSA '))
$pubkey = rsatopem($diaspora_key);
else
$pubkey = $diaspora_key;
2011-08-12 06:01:11 -04:00
$diaspora = true;
}
if(($link['@attributes']['rel'] === 'http://ostatus.org/schema/1.0/subscribe') AND ($mode == PROBE_NORMAL)) {
$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
// otherwise we're screwed.
$backup_alias = "";
foreach($links as $link) {
if($link['@attributes']['rel'] === 'alias') {
if(strpos($link['@attributes']['href'],'@') === false) {
if(isset($profile)) {
$alias_url = $link['@attributes']['href'];
if(($alias_url !== $profile) AND ($backup_alias == "") AND
($alias_url !== str_replace("/index.php", "", $profile)))
$backup_alias = $alias_url;
if(($alias_url !== $profile) AND !strstr($alias_url, "index.php") AND
($alias_url !== str_replace("/index.php", "", $profile)))
$alias = $alias_url;
}
else
$profile = unamp($link['@attributes']['href']);
}
}
}
if ($alias == "")
$alias = $backup_alias;
// If the profile is different from the url then the url is abviously an alias
if (($alias == "") AND ($profile != "") AND !$at_addr AND (normalise_link($profile) != normalise_link($url)))
$alias = $url;
}
2011-08-18 02:01:44 -04:00
elseif($mode == PROBE_NORMAL) {
2011-04-16 11:45:08 -04:00
// Check email
$orig_url = $url;
if((strpos($orig_url,'@')) && validate_email($orig_url)) {
2011-04-16 11:45:08 -04:00
$x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1",
intval(local_user())
);
2011-04-18 02:27:11 -04:00
$r = q("SELECT * FROM `mailacct` WHERE `uid` = %d AND `server` != '' LIMIT 1",
2011-04-16 11:45:08 -04:00
intval(local_user())
);
if(count($x) && count($r)) {
$mailbox = construct_mailbox_name($r[0]);
2011-04-16 11:45:08 -04:00
$password = '';
openssl_private_decrypt(hex2bin($r[0]['pass']),$password,$x[0]['prvkey']);
$mbox = email_connect($mailbox,$r[0]['user'],$password);
if(! $mbox)
logger('probe_url: email_connect failed.');
2011-04-16 11:45:08 -04:00
unset($password);
}
if($mbox) {
$msgs = email_poll($mbox,$orig_url);
logger('probe_url: searching ' . $orig_url . ', ' . count($msgs) . ' messages found.', LOGGER_DEBUG);
2011-04-16 11:45:08 -04:00
if(count($msgs)) {
$addr = $orig_url;
$network = NETWORK_MAIL;
$name = substr($url,0,strpos($url,'@'));
2011-09-13 01:24:36 -04:00
$phost = substr($url,strpos($url,'@')+1);
$profile = 'http://' . $phost;
2011-04-17 18:24:24 -04:00
// fix nick character range
$vcard = array('fn' => $name, 'nick' => $name, 'photo' => avatar_img($url));
2011-04-18 02:27:11 -04:00
$notify = 'smtp ' . random_string();
$poll = 'email ' . random_string();
2011-04-16 11:45:08 -04:00
$priority = 0;
2011-04-17 18:24:24 -04:00
$x = email_msg_meta($mbox,$msgs[0]);
if(stristr($x[0]->from,$orig_url))
$adr = imap_rfc822_parse_adrlist($x[0]->from,'');
elseif(stristr($x[0]->to,$orig_url))
$adr = imap_rfc822_parse_adrlist($x[0]->to,'');
2011-09-13 01:24:36 -04:00
if(isset($adr)) {
foreach($adr as $feadr) {
if((strcasecmp($feadr->mailbox,$name) == 0)
&&(strcasecmp($feadr->host,$phost) == 0)
2011-09-13 01:24:36 -04:00
&& (strlen($feadr->personal))) {
$personal = imap_mime_header_decode($feadr->personal);
$vcard['fn'] = "";
foreach($personal as $perspart)
if ($perspart->charset != "default")
$vcard['fn'] .= iconv($perspart->charset, 'UTF-8//IGNORE', $perspart->text);
else
$vcard['fn'] .= $perspart->text;
$vcard['fn'] = notags($vcard['fn']);
2011-09-13 01:24:36 -04:00
}
}
}
2011-04-16 11:45:08 -04:00
}
imap_close($mbox);
}
}
}
}
2011-08-18 02:01:44 -04:00
if($mode == PROBE_NORMAL) {
2015-07-22 05:51:37 -04:00
2011-08-18 02:01:44 -04:00
if(strlen($zot)) {
$s = fetch_url($zot);
if($s) {
$j = json_decode($s);
if($j) {
$network = NETWORK_ZOT;
$vcard = array(
'fn' => $j->fullname,
'nick' => $j->nickname,
2011-08-18 02:01:44 -04:00
'photo' => $j->photo
);
$profile = $j->url;
$notify = $j->post;
$pubkey = $j->pubkey;
$poll = 'N/A';
}
2011-07-11 21:28:13 -04:00
}
}
2015-12-28 11:09:22 -05:00
2011-08-18 02:01:44 -04:00
if(strlen($dfrn)) {
$ret = scrape_dfrn(($hcard) ? $hcard : $dfrn, true);
2011-08-18 02:01:44 -04:00
if(is_array($ret) && x($ret,'dfrn-request')) {
$network = NETWORK_DFRN;
$request = $ret['dfrn-request'];
$confirm = $ret['dfrn-confirm'];
$notify = $ret['dfrn-notify'];
$poll = $ret['dfrn-poll'];
$vcard = array();
$vcard['fn'] = $ret['fn'];
$vcard['nick'] = $ret['nick'];
$vcard['photo'] = $ret['photo'];
}
}
}
// Scrape the public key from the hcard.
// Diaspora will remove it from the webfinger somewhere in the future.
if (($hcard != "") AND ($pubkey == "")) {
$ret = scrape_dfrn(($hcard) ? $hcard : $dfrn, true);
if (isset($ret["key"])) {
$hcard_key = $ret["key"];
if(strstr($hcard_key,'RSA '))
$pubkey = rsatopem($hcard_key);
else
$pubkey = $hcard_key;
}
}
2011-08-12 06:01:11 -04:00
if($diaspora && $diaspora_base && $diaspora_guid) {
$diaspora_notify = $diaspora_base.'receive/users/'.$diaspora_guid;
if($mode == PROBE_DIASPORA || ! $notify || ($notify == $diaspora_notify)) {
$notify = $diaspora_notify;
$batch = $diaspora_base . 'receive/public' ;
}
2011-08-12 06:01:11 -04:00
if(strpos($url,'@'))
$addr = str_replace('acct:', '', $url);
}
2011-08-12 06:01:11 -04:00
2011-07-11 21:28:13 -04:00
if($network !== NETWORK_ZOT && $network !== NETWORK_DFRN && $network !== NETWORK_MAIL) {
2011-08-12 06:01:11 -04:00
if($diaspora)
$network = NETWORK_DIASPORA;
elseif($has_lrdd AND ($notify))
2011-08-12 06:01:11 -04:00
$network = NETWORK_OSTATUS;
if(strpos($url,'@'))
$addr = str_replace('acct:', '', $url);
$priority = 0;
2011-08-18 02:01:44 -04:00
if($hcard && ! $vcard) {
$vcard = scrape_vcard($hcard);
// Google doesn't use absolute url in profile photos
if((x($vcard,'photo')) && substr($vcard['photo'],0,1) == '/') {
$h = @parse_url($hcard);
if($h)
$vcard['photo'] = $h['scheme'] . '://' . $h['host'] . $vcard['photo'];
}
logger('probe_url: scrape_vcard: ' . print_r($vcard,true), LOGGER_DATA);
}
if($diaspora && $addr) {
// Diaspora returns the name as the nick. As the nick will never be updated,
// let's use the Diaspora nickname (the first part of the handle) as the nick instead
$addr_parts = explode('@', $addr);
$vcard['nick'] = $addr_parts[0];
}
2012-06-21 01:59:01 -04:00
if($lastfm) {
$profile = $url;
$poll = str_replace(array('www.','last.fm/'),array('','ws.audioscrobbler.com/1.0/'),$url) . '/recenttracks.rss';
$vcard['nick'] = basename($url);
2012-06-21 02:04:55 -04:00
$vcard['fn'] = $vcard['nick'] . t(' on Last.fm');
2012-06-21 01:59:01 -04:00
$network = NETWORK_FEED;
}
if(! x($vcard,'fn'))
if(x($vcard,'nick'))
$vcard['fn'] = $vcard['nick'];
2011-08-14 21:13:52 -04:00
$check_feed = false;
if(stristr($url,'tumblr.com') && (! stristr($url,'/rss'))) {
$poll = $url . '/rss';
$check_feed = true;
// Will leave it to others to figure out how to grab the avatar, which is on the $url page in the open graph meta links
}
if($appnet || ! $poll)
2011-08-14 21:13:52 -04:00
$check_feed = true;
if((! isset($vcard)) || (! x($vcard,'fn')) || (! $profile))
2011-08-14 21:13:52 -04:00
$check_feed = true;
if(($at_addr) && (! count($links)))
$check_feed = false;
if ($connectornetworks)
$check_feed = false;
2015-12-06 12:57:23 -05:00
2011-08-14 21:13:52 -04:00
if($check_feed) {
2015-12-06 12:57:23 -05:00
2011-08-15 08:27:24 -04:00
$feedret = scrape_feed(($poll) ? $poll : $url);
2015-07-22 05:51:37 -04:00
2012-02-26 17:34:01 -05:00
logger('probe_url: scrape_feed ' . (($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']));
if(! x($vcard))
2011-08-18 00:14:21 -04:00
$vcard = array();
}
2011-08-18 00:14:21 -04:00
if(x($feedret,'photo') && (! x($vcard,'photo')))
$vcard['photo'] = $feedret['photo'];
$cookiejar = tempnam(get_temppath(), 'cookiejar-scrape-feed-');
$xml = fetch_url($poll, false, $redirects, 0, Null, $cookiejar);
unlink($cookiejar);
logger('probe_url: fetch feed: ' . $poll . ' returns: ' . $xml, LOGGER_DATA);
if ($xml == "") {
logger("scrape_feed: XML is empty for feed ".$poll);
$network = NETWORK_PHANTOM;
} else {
$data = feed_import($xml,$dummy1,$dummy2, $dummy3, true);
if (!is_array($data)) {
logger("scrape_feed: This doesn't seem to be a feed: ".$poll);
$network = NETWORK_PHANTOM;
} else {
if (($vcard["photo"] == "") AND ($data["header"]["author-avatar"] != ""))
$vcard["photo"] = $data["header"]["author-avatar"];
if (($vcard["fn"] == "") AND ($data["header"]["author-name"] != ""))
$vcard["fn"] = $data["header"]["author-name"];
2011-08-18 00:14:21 -04:00
if (($vcard["nick"] == "") AND ($data["header"]["author-nick"] != ""))
$vcard["nick"] = $data["header"]["author-nick"];
if (($network == NETWORK_OSTATUS) AND ($data["header"]["author-id"] != ""))
$alias = $data["header"]["author-id"];
if(!$profile AND ($data["header"]["author-link"] != "") AND !in_array($network, array("", NETWORK_FEED)))
$profile = $data["header"]["author-link"];
}
}
2011-08-18 00:14:21 -04:00
// Workaround for misconfigured Friendica servers
if (($network == "") AND (strstr($url, "/profile/"))) {
$noscrape = str_replace("/profile/", "/noscrape/", $url);
$noscrapejson = fetch_url($noscrape);
if ($noscrapejson) {
$network = NETWORK_DFRN;
$poco = str_replace("/profile/", "/poco/", $url);
$noscrapedata = json_decode($noscrapejson, true);
if (isset($noscrapedata["addr"]))
$addr = $noscrapedata["addr"];
if (isset($noscrapedata["fn"]))
$vcard["fn"] = $noscrapedata["fn"];
if (isset($noscrapedata["key"]))
$pubkey = $noscrapedata["key"];
if (isset($noscrapedata["photo"]))
$vcard["photo"] = $noscrapedata["photo"];
if (isset($noscrapedata["dfrn-request"]))
$request = $noscrapedata["dfrn-request"];
if (isset($noscrapedata["dfrn-confirm"]))
$confirm = $noscrapedata["dfrn-confirm"];
if (isset($noscrapedata["dfrn-notify"]))
$notify = $noscrapedata["dfrn-notify"];
if (isset($noscrapedata["dfrn-poll"]))
$poll = $noscrapedata["dfrn-poll"];
}
}
if(! $network)
$network = NETWORK_FEED;
if(! x($vcard,'nick')) {
$vcard['nick'] = strtolower(notags(unxmlify($vcard['fn'])));
if(strpos($vcard['nick'],' '))
$vcard['nick'] = trim(substr($vcard['nick'],0,strpos($vcard['nick'],' ')));
}
2011-08-14 21:13:52 -04:00
if(! $priority)
$priority = 2;
}
}
if(! x($vcard,'photo')) {
$a = get_app();
$vcard['photo'] = App::get_baseurl() . '/images/person-175.jpg' ;
}
2011-08-14 21:13:52 -04:00
if(! $profile)
$profile = $url;
// No human could be associated with this link, use the URL as the contact name
if(($network === NETWORK_FEED) && ($poll) && (! x($vcard,'fn')))
$vcard['fn'] = $url;
2011-04-16 11:45:08 -04:00
if (($notify != "") AND ($poll != "")) {
$baseurl = matching_url(normalise_link($notify), normalise_link($poll));
$baseurl2 = matching_url($baseurl, normalise_link($profile));
if ($baseurl2 != "")
$baseurl = $baseurl2;
}
if (($baseurl == "") AND ($notify != ""))
$baseurl = matching_url(normalise_link($profile), normalise_link($notify));
if (($baseurl == "") AND ($poll != ""))
$baseurl = matching_url(normalise_link($profile), normalise_link($poll));
if (substr($baseurl, -10) == "/index.php")
$baseurl = str_replace("/index.php", "", $baseurl);
$baseurl = rtrim($baseurl, "/");
if(strpos($url,'@') AND ($addr == "") AND ($network == NETWORK_DFRN))
$addr = str_replace('acct:', '', $url);
$vcard['fn'] = notags($vcard['fn']);
$vcard['nick'] = str_replace(' ','',notags($vcard['nick']));
$result['name'] = $vcard['fn'];
$result['nick'] = $vcard['nick'];
$result['url'] = $profile;
2011-04-16 11:45:08 -04:00
$result['addr'] = $addr;
$result['batch'] = $batch;
$result['notify'] = $notify;
$result['poll'] = $poll;
$result['request'] = $request;
$result['confirm'] = $confirm;
$result['poco'] = $poco;
$result['photo'] = $vcard['photo'];
$result['priority'] = $priority;
$result['network'] = $network;
$result['alias'] = $alias;
2011-07-15 06:08:43 -04:00
$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_PHANTOM) {
Cache::set("probe_url:".$mode.":".$original_url,serialize($result), CACHE_DAY);
/// @todo temporary fix - we need a real contact update function that updates only changing fields
/// The biggest problem is the avatar picture that could have a reduced image size.
/// It should only be updated if the existing picture isn't existing anymore.
if (($result['network'] != NETWORK_FEED) AND ($mode == PROBE_NORMAL) AND
$result["name"] AND $result["nick"] AND $result["url"] AND $result["addr"] AND $result["poll"])
q("UPDATE `contact` SET `name` = '%s', `nick` = '%s', `url` = '%s', `addr` = '%s',
`notify` = '%s', `poll` = '%s', `alias` = '%s', `success_update` = '%s'
WHERE `nurl` = '%s' AND NOT `self` AND `uid` = 0",
dbesc($result["name"]),
dbesc($result["nick"]),
dbesc($result["url"]),
dbesc($result["addr"]),
dbesc($result["notify"]),
dbesc($result["poll"]),
dbesc($result["alias"]),
dbesc(datetime_convert()),
dbesc(normalise_link($result['url']))
);
}
return $result;
}
/**
* @brief Find the matching part between two url
*
* @param string $url1
* @param string $url2
* @return string The matching part
*/
function matching_url($url1, $url2) {
if (($url1 == "") OR ($url2 == ""))
return "";
$url1 = normalise_link($url1);
$url2 = normalise_link($url2);
$parts1 = parse_url($url1);
$parts2 = parse_url($url2);
if (!isset($parts1["host"]) OR !isset($parts2["host"]))
return "";
if ($parts1["scheme"] != $parts2["scheme"])
return "";
if ($parts1["host"] != $parts2["host"])
return "";
if ($parts1["port"] != $parts2["port"])
return "";
$match = $parts1["scheme"]."://".$parts1["host"];
if ($parts1["port"])
$match .= ":".$parts1["port"];
$pathparts1 = explode("/", $parts1["path"]);
$pathparts2 = explode("/", $parts2["path"]);
$i = 0;
$path = "";
do {
$path1 = $pathparts1[$i];
$path2 = $pathparts2[$i];
if ($path1 == $path2)
$path .= $path1."/";
} while (($path1 == $path2) AND ($i++ <= count($pathparts1)));
$match .= $path;
return normalise_link($match);
}