X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=include%2Fpoller.php;h=89a3408ec90efcf04deca6aa7f5e1dd00afecee2;hb=4d5580ab7781eea81dc4cb3d67b7376f49357bef;hp=4567a5cfc3d6d04a65f54677957a6d32c57b8a69;hpb=3eefe8b50003c858d4930c03cc06d2679a14347c;p=friendica.git diff --git a/include/poller.php b/include/poller.php index 4567a5cfc3..89a3408ec9 100644 --- a/include/poller.php +++ b/include/poller.php @@ -1,279 +1,528 @@ set_baseurl(get_config('system','url')); + load_hooks(); + logger('poller: start'); // run queue delivery process in the background - $php_path = ((x($a->config,'php_path') && strlen($a->config['php_path'])) ? $a->config['php_path'] : 'php'); - //proc_close(proc_open("\"$php_path\" \"include/queue.php\" &", array(), $foo)); - proc_run($php_path,"include/queue.php"); + proc_run('php',"include/queue.php"); + + // expire any expired accounts + + q("UPDATE user SET `account_expired` = 1 where `account_expired` = 0 + AND `account_expires_on` != '0000-00-00 00:00:00' + AND `account_expires_on` < UTC_TIMESTAMP() "); + + $abandon_days = intval(get_config('system','account_abandon_days')); + if($abandon_days < 1) + $abandon_days = 0; + + + // once daily run expire in background + $d1 = get_config('system','last_expire_day'); + $d2 = intval(datetime_convert('UTC','UTC','now','d')); + + if($d2 != intval($d1)) { + set_config('system','last_expire_day',$d2); + proc_run('php','include/expire.php'); + } + + // clear old cache + q("DELETE FROM `cache` WHERE `updated` < '%s'", + dbesc(datetime_convert('UTC','UTC',"now - 30 days"))); + + $manual_id = 0; + $generation = 0; $hub_update = false; - $force = false; + $force = false; + $restart = false; if(($argc > 1) && ($argv[1] == 'force')) $force = true; + if(($argc > 1) && ($argv[1] == 'restart')) { + $restart = true; + $generation = intval($argv[2]); + if(! $generation) + killme(); + } + if(($argc > 1) && intval($argv[1])) { $manual_id = intval($argv[1]); - $force = true; + $force = true; } $sql_extra = (($manual_id) ? " AND `id` = $manual_id " : ""); - // 'stat' clause is a temporary measure until we have federation subscriptions working both directions - $contacts = q("SELECT * FROM `contact` - WHERE ( ( `network` = 'dfrn' AND ( `dfrn-id` != '' OR (`issued-id` != '' AND `duplex` = 1))) - OR ( `network` IN ( 'stat', 'feed' ) AND `poll` != '' )) - $sql_extra - AND `self` = 0 AND `blocked` = 0 AND `readonly` = 0 ORDER BY RAND()"); + reload_plugins(); + + $d = datetime_convert(); + + if(! $restart) + proc_run('php','include/cronhooks.php'); + + // Only poll from those with suitable relationships, + // and which have a polling address and ignore Diaspora since + // we are unable to match those posts with a Diaspora GUID and prevent duplicates. - if(! count($contacts)){ + $abandon_sql = (($abandon_days) + ? sprintf(" AND `user`.`login_date` > UTC_TIMESTAMP() - INTERVAL %d DAY ", intval($abandon_days)) + : '' + ); + + $contacts = q("SELECT `contact`.`id` FROM `contact` LEFT JOIN `user` ON `user`.`uid` = `contact`.`uid` + WHERE ( `rel` = %d OR `rel` = %d ) AND `poll` != '' + AND `network` != '%s' + $sql_extra + AND `self` = 0 AND `contact`.`blocked` = 0 AND `contact`.`readonly` = 0 + AND `user`.`account_expired` = 0 $abandon_sql ORDER BY RAND()", + intval(CONTACT_IS_SHARING), + intval(CONTACT_IS_FRIEND), + dbesc(NETWORK_DIASPORA) + ); + + if(! count($contacts)) { return; } - foreach($contacts as $contact) { + foreach($contacts as $c) { + + $res = q("SELECT * FROM `contact` WHERE `id` = %d LIMIT 1", + intval($c['id']) + ); + + if((! $res) || (! count($res))) + continue; + + foreach($res as $contact) { + + $xml = false; if($manual_id) $contact['last-update'] = '0000-00-00 00:00:00'; - if($contact['priority'] || $contact['subhub']) { - - $hub_update = true; - $update = false; + if($contact['priority'] || $contact['subhub']) { - $t = $contact['last-update']; + $hub_update = true; + $update = false; - // We should be getting everything via a hub. But just to be sure, let's check once a day. - // (You can make this more or less frequent if desired by setting 'pushpoll_frequency' appropriately) - // This also lets us update our subscription to the hub, and add or replace hubs in case it - // changed. We will only update hubs once a day, regardless of 'pushpoll_frequency'. + $t = $contact['last-update']; + // We should be getting everything via a hub. But just to be sure, let's check once a day. + // (You can make this more or less frequent if desired by setting 'pushpoll_frequency' appropriately) + // This also lets us update our subscription to the hub, and add or replace hubs in case it + // changed. We will only update hubs once a day, regardless of 'pushpoll_frequency'. - if($contact['subhub']) { - $interval = get_config('system','pushpoll_frequency'); - $contact['priority'] = (($interval !== false) ? intval($interval) : 3); - $hub_update = false; - if((datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 day")) || $force) - $hub_update = true; + if($contact['subhub']) { + $interval = get_config('system','pushpoll_frequency'); + $contact['priority'] = (($interval !== false) ? intval($interval) : 3); + $hub_update = false; + + if((datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 day")) || $force) + $hub_update = true; + } + else + $hub_update = false; + + /** + * Based on $contact['priority'], should we poll this site now? Or later? + */ + + switch ($contact['priority']) { + case 5: + if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 month")) + $update = true; + break; + case 4: + if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 week")) + $update = true; + break; + case 3: + if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 day")) + $update = true; + break; + case 2: + if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 12 hour")) + $update = true; + break; + case 1: + default: + if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 hour")) + $update = true; + break; + } + if((! $update) && (! $force)) + continue; } - - /** - * Based on $contact['priority'], should we poll this site now? Or later? - */ - - switch ($contact['priority']) { - case 5: - if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 month")) - $update = true; - break; - case 4: - if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 week")) - $update = true; - break; - case 3: - if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 day")) - $update = true; - break; - case 2: - if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 12 hour")) - $update = true; - break; - case 1: - default: - if(datetime_convert('UTC','UTC', 'now') > datetime_convert('UTC','UTC', $t . " + 1 hour")) - $update = true; - break; + // Check to see if we are running out of memory - if so spawn a new process and kill this one + + $avail_memory = return_bytes(ini_get('memory_limit')); + $memused = memory_get_peak_usage(true); + if(intval($avail_memory)) { + if(($memused / $avail_memory) > 0.95) { + if($generation + 1 > 10) { + logger('poller: maximum number of spawns exceeded. Terminating.'); + killme(); + } + logger('poller: memory exceeded. ' . $memused . ' bytes used. Spawning new poll.'); + proc_run('php', 'include/poller.php', 'restart', (string) $generation + 1); + killme(); + } } - if((! $update) && (! $force)) - continue; - } - $importer_uid = $contact['uid']; + $importer_uid = $contact['uid']; + + $r = q("SELECT * FROM `contact` WHERE `uid` = %d AND `self` = 1 LIMIT 1", + intval($importer_uid) + ); + if(! count($r)) + continue; - $r = q("SELECT * FROM `contact` WHERE `uid` = %d AND `self` = 1 LIMIT 1", - intval($importer_uid) - ); - if(! count($r)) - continue; + $importer = $r[0]; - $importer = $r[0]; + logger("poller: poll: IMPORTER: {$importer['name']}, CONTACT: {$contact['name']}"); - logger("poller: poll: IMPORTER: {$importer['name']}, CONTACT: {$contact['name']}"); + $last_update = (($contact['last-update'] === '0000-00-00 00:00:00') + ? datetime_convert('UTC','UTC','now - 30 days', ATOM_TIME) + : datetime_convert('UTC','UTC',$contact['last-update'], ATOM_TIME) + ); - $last_update = (($contact['last-update'] === '0000-00-00 00:00:00') - ? datetime_convert('UTC','UTC','now - 30 days', ATOM_TIME) - : datetime_convert('UTC','UTC',$contact['last-update'], ATOM_TIME) - ); + if($contact['network'] === NETWORK_DFRN) { - if($contact['network'] === 'dfrn') { + $idtosend = $orig_id = (($contact['dfrn-id']) ? $contact['dfrn-id'] : $contact['issued-id']); - $idtosend = $orig_id = (($contact['dfrn-id']) ? $contact['dfrn-id'] : $contact['issued-id']); + if(intval($contact['duplex']) && $contact['dfrn-id']) + $idtosend = '0:' . $orig_id; + if(intval($contact['duplex']) && $contact['issued-id']) + $idtosend = '1:' . $orig_id; - if(intval($contact['duplex']) && $contact['dfrn-id']) - $idtosend = '0:' . $orig_id; - if(intval($contact['duplex']) && $contact['issued-id']) - $idtosend = '1:' . $orig_id; + // they have permission to write to us. We already filtered this in the contact query. + $perm = 'rw'; - $url = $contact['poll'] . '?dfrn_id=' . $idtosend - . '&dfrn_version=' . DFRN_PROTOCOL_VERSION - . '&type=data&last_update=' . $last_update ; + $url = $contact['poll'] . '?dfrn_id=' . $idtosend + . '&dfrn_version=' . DFRN_PROTOCOL_VERSION + . '&type=data&last_update=' . $last_update + . '&perm=' . $perm ; - $xml = fetch_url($url); + $handshake_xml = fetch_url($url); - logger('poller: handshake with url ' . $url . ' returns xml: ' . $xml, LOGGER_DATA); + logger('poller: handshake with url ' . $url . ' returns xml: ' . $handshake_xml, LOGGER_DATA); - if(! $xml) { - logger("poller: $url appears to be dead - marking for death "); - // dead connection - might be a transient event, or this might - // mean the software was uninstalled or the domain expired. - // Will keep trying for one month. - mark_for_death($contact); + if(! $handshake_xml) { + logger("poller: $url appears to be dead - marking for death "); + // dead connection - might be a transient event, or this might + // mean the software was uninstalled or the domain expired. + // Will keep trying for one month. + mark_for_death($contact); - // set the last-update so we don't keep polling + // set the last-update so we don't keep polling - $r = q("UPDATE `contact` SET `last-update` = '%s' WHERE `id` = %d LIMIT 1", - dbesc(datetime_convert()), - intval($contact['id']) - ); + $r = q("UPDATE `contact` SET `last-update` = '%s' WHERE `id` = %d LIMIT 1", + dbesc(datetime_convert()), + intval($contact['id']) + ); - continue; - } - - if(! strstr($xml,'status) == 1) { - logger("poller: $url replied status 1 - marking for death "); + $res = parse_xml_string($handshake_xml); + + if(intval($res->status) == 1) { + logger("poller: $url replied status 1 - marking for death "); - // we may not be friends anymore. Will keep trying for one month. - // set the last-update so we don't keep polling + // we may not be friends anymore. Will keep trying for one month. + // set the last-update so we don't keep polling - $r = q("UPDATE `contact` SET `last-update` = '%s' WHERE `id` = %d LIMIT 1", - dbesc(datetime_convert()), - intval($contact['id']) - ); + $r = q("UPDATE `contact` SET `last-update` = '%s' WHERE `id` = %d LIMIT 1", + dbesc(datetime_convert()), + intval($contact['id']) + ); - mark_for_death($contact); - } - else { - if($contact['term-date'] != '0000-00-00 00:00:00') { - logger("poller: $url back from the dead - removing mark for death"); - unmark_for_death($contact); + mark_for_death($contact); + } + else { + if($contact['term-date'] != '0000-00-00 00:00:00') { + logger("poller: $url back from the dead - removing mark for death"); + unmark_for_death($contact); + } } - } - if((intval($res->status) != 0) || (! strlen($res->challenge)) || (! strlen($res->dfrn_id))) - continue; + if((intval($res->status) != 0) || (! strlen($res->challenge)) || (! strlen($res->dfrn_id))) + continue; - $postvars = array(); + $postvars = array(); - $sent_dfrn_id = hex2bin((string) $res->dfrn_id); - $challenge = hex2bin((string) $res->challenge); + $sent_dfrn_id = hex2bin((string) $res->dfrn_id); + $challenge = hex2bin((string) $res->challenge); - $final_dfrn_id = ''; + $final_dfrn_id = ''; - if(($contact['duplex']) && strlen($contact['prvkey'])) { - openssl_private_decrypt($sent_dfrn_id,$final_dfrn_id,$contact['prvkey']); - openssl_private_decrypt($challenge,$postvars['challenge'],$contact['prvkey']); - } - else { - openssl_public_decrypt($sent_dfrn_id,$final_dfrn_id,$contact['pubkey']); - openssl_public_decrypt($challenge,$postvars['challenge'],$contact['pubkey']); - } + if(($contact['duplex']) && strlen($contact['prvkey'])) { + openssl_private_decrypt($sent_dfrn_id,$final_dfrn_id,$contact['prvkey']); + openssl_private_decrypt($challenge,$postvars['challenge'],$contact['prvkey']); + } + else { + openssl_public_decrypt($sent_dfrn_id,$final_dfrn_id,$contact['pubkey']); + openssl_public_decrypt($challenge,$postvars['challenge'],$contact['pubkey']); + } - $final_dfrn_id = substr($final_dfrn_id, 0, strpos($final_dfrn_id, '.')); + $final_dfrn_id = substr($final_dfrn_id, 0, strpos($final_dfrn_id, '.')); - if(strpos($final_dfrn_id,':') == 1) - $final_dfrn_id = substr($final_dfrn_id,2); + if(strpos($final_dfrn_id,':') == 1) + $final_dfrn_id = substr($final_dfrn_id,2); - if($final_dfrn_id != $orig_id) { + if($final_dfrn_id != $orig_id) { + logger('poller: ID did not decode: ' . $contact['id'] . ' orig: ' . $orig_id . ' final: ' . $final_dfrn_id); + // did not decode properly - cannot trust this site + continue; + } - // did not decode properly - cannot trust this site - continue; + $postvars['dfrn_id'] = $idtosend; + $postvars['dfrn_version'] = DFRN_PROTOCOL_VERSION; + $postvars['perm'] = 'rw'; + + $xml = post_url($contact['poll'],$postvars); } + elseif(($contact['network'] === NETWORK_OSTATUS) + || ($contact['network'] === NETWORK_DIASPORA) + || ($contact['network'] === NETWORK_FEED) ) { - $postvars['dfrn_id'] = $idtosend; - $postvars['dfrn_version'] = DFRN_PROTOCOL_VERSION; + // Upgrading DB fields from an older Friendika version + // Will only do this once per notify-enabled OStatus contact + // or if relationship changes - $xml = post_url($contact['poll'],$postvars); - } - else { - // $contact['network'] !== 'dfrn' + $stat_writeable = ((($contact['notify']) && ($contact['rel'] == CONTACT_IS_FOLLOWER || $contact['rel'] == CONTACT_IS_FRIEND)) ? 1 : 0); - $xml = fetch_url($contact['poll']); - } + if($stat_writeable != $contact['writable']) { + q("UPDATE `contact` SET `writable` = %d WHERE `id` = %d LIMIT 1", + intval($stat_writeable), + intval($contact['id']) + ); + } - logger('poller: received xml : ' . $xml, LOGGER_DATA); + // Are we allowed to import from this person? - if(! strlen($xml)) - continue; + if($contact['rel'] == CONTACT_IS_FOLLOWER || $contact['blocked'] || $contact['readonly']) + continue; + + $xml = fetch_url($contact['poll']); + } + elseif($contact['network'] === NETWORK_MAIL) { + + $mail_disabled = ((function_exists('imap_open') && (! get_config('system','imap_disabled'))) ? 0 : 1); + if($mail_disabled) + continue; + + $mbox = null; + $x = q("SELECT `prvkey` FROM `user` WHERE `uid` = %d LIMIT 1", + intval($importer_uid) + ); + $mailconf = q("SELECT * FROM `mailacct` WHERE `server` != '' AND `uid` = %d LIMIT 1", + intval($importer_uid) + ); + if(count($x) && count($mailconf)) { + $mailbox = construct_mailbox_name($mailconf[0]); + $password = ''; + openssl_private_decrypt(hex2bin($mailconf[0]['pass']),$password,$x[0]['prvkey']); + $mbox = email_connect($mailbox,$mailconf[0]['user'],$password); + unset($password); + if($mbox) { + q("UPDATE `mailacct` SET `last_check` = '%s' WHERE `id` = %d AND `uid` = %d LIMIT 1", + dbesc(datetime_convert()), + intval($mailconf[0]['id']), + intval($importer_uid) + ); + } + } + if($mbox) { + + $msgs = email_poll($mbox,$contact['addr']); + + if(count($msgs)) { + foreach($msgs as $msg_uid) { + $datarray = array(); + $meta = email_msg_meta($mbox,$msg_uid); + $headers = email_msg_headers($mbox,$msg_uid); + + // look for a 'references' header and try and match with a parent item we have locally. + + $raw_refs = ((x($headers,'references')) ? str_replace("\t",'',$headers['references']) : ''); + $datarray['uri'] = trim($meta->message_id,'<>'); + + if($raw_refs) { + $refs_arr = explode(' ', $raw_refs); + if(count($refs_arr)) { + for($x = 0; $x < count($refs_arr); $x ++) + $refs_arr[$x] = "'" . str_replace(array('<','>',' '),array('','',''),dbesc($refs_arr[$x])) . "'"; + } + $qstr = implode(',',$refs_arr); + $r = q("SELECT `uri` , `parent-uri` FROM `item` WHERE `uri` IN ( $qstr ) AND `uid` = %d LIMIT 1", + intval($importer_uid) + ); + if(count($r)) + $datarray['parent-uri'] = $r[0]['uri']; + } + + + if(! x($datarray,'parent-uri')) + $datarray['parent-uri'] = $datarray['uri']; + + // Have we seen it before? + $r = q("SELECT * FROM `item` WHERE `uid` = %d AND `uri` = '%s' LIMIT 1", + intval($importer_uid), + dbesc($datarray['uri']) + ); + + if(count($r)) { + if($meta->deleted && ! $r[0]['deleted']) { + q("UPDATE `item` SET `deleted` = 1, `changed` = '%s' WHERE `id` = %d LIMIT 1", + dbesc(datetime_convert()), + intval($r[0]['id']) + ); + } + continue; + } + $datarray['title'] = notags(trim($meta->subject)); + $datarray['created'] = datetime_convert('UTC','UTC',$meta->date); + + $r = email_get_msg($mbox,$msg_uid); + if(! $r) + continue; + $datarray['body'] = escape_tags($r['body']); + + // some mailing lists have the original author as 'from' - add this sender info to msg body. + // todo: adding a gravatar for the original author would be cool + + if(! stristr($meta->from,$contact['addr'])) + $datarray['body'] = t('From: ') . escape_tags($meta->from) . "\n\n" . $datarray['body']; + + $datarray['uid'] = $importer_uid; + $datarray['contact-id'] = $contact['id']; + if($datarray['parent-uri'] === $datarray['uri']) + $datarray['private'] = 1; + if(! get_pconfig($importer_uid,'system','allow_public_email_replies')) { + $datarray['private'] = 1; + $datarray['allow_cid'] = '<' . $contact['id'] . '>'; + } + $datarray['author-name'] = $contact['name']; + $datarray['author-link'] = 'mailbox'; + $datarray['author-avatar'] = $contact['photo']; + + $stored_item = item_store($datarray); + q("UPDATE `item` SET `last-child` = 0 WHERE `parent-uri` = '%s' AND `uid` = %d", + dbesc($datarray['parent-uri']), + intval($importer_uid) + ); + q("UPDATE `item` SET `last-child` = 1 WHERE `id` = %d LIMIT 1", + intval($stored_item) + ); + } + } + + imap_close($mbox); + } + } + elseif($contact['network'] === NETWORK_FACEBOOK) { + // This is picked up by the Facebook plugin on a cron hook. + // Ignored here. + } - consume_feed($xml,$importer,$contact,$hub,1); + if($xml) { + logger('poller: received xml : ' . $xml, LOGGER_DATA); - // do it twice. Ensures that children of parents which may be later in the stream aren't tossed + if(! strstr($xml,'