2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5 define_default('_MIN_CACHE_IMAGE_SIZE', 1024);
7 function calculate_article_hash($article, $pluginhost) {
10 foreach ($article as $k => $v) {
11 if ($k != "feed" && isset($v)) {
12 $x = strip_tags(is_array($v) ?
implode(",", $v) : $v);
14 //_debug("$k:" . sha1($x) . ":" . htmlspecialchars($x), true);
16 $tmp .= sha1("$k:" . sha1($x));
20 return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
23 function update_feedbrowser_cache() {
25 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
26 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
27 WHERE tf.feed_url = ttrss_feeds.feed_url
28 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
29 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
33 db_query("DELETE FROM ttrss_feedbrowser_cache");
37 while ($line = db_fetch_assoc($result)) {
38 $subscribers = db_escape_string($line["subscribers"]);
39 $feed_url = db_escape_string($line["feed_url"]);
40 $title = db_escape_string($line["title"]);
41 $site_url = db_escape_string($line["site_url"]);
43 $tmp_result = db_query("SELECT subscribers FROM
44 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
46 if (db_num_rows($tmp_result) == 0) {
48 db_query("INSERT INTO ttrss_feedbrowser_cache
49 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
50 '$site_url', '$title', '$subscribers')");
66 * Update a feed batch.
67 * Used by daemons to update n feeds by run.
68 * Only update feed needing a update, and not being processed
71 * @param mixed $link Database link
72 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
73 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
74 * @param boolean $debug Set to false to disable debug output. Default to true.
77 function update_daemon_common($limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
78 // Process all other feeds using last_updated and interval parameters
80 $schema_version = get_schema_version();
82 if ($schema_version != SCHEMA_VERSION
) {
83 die("Schema version is wrong, please upgrade the database.\n");
86 define('PREFS_NO_CACHE', true);
88 // Test if the user has loggued in recently. If not, it does not update its feeds.
89 if (!SINGLE_USER_MODE
&& DAEMON_UPDATE_LOGIN_LIMIT
> 0) {
90 if (DB_TYPE
== "pgsql") {
91 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT
." days'";
93 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
96 $login_thresh_qpart = "";
99 // Test if the feed need a update (update interval exceeded).
100 if (DB_TYPE
== "pgsql") {
101 $update_limit_qpart = "AND ((
102 ttrss_feeds.update_interval = 0
103 AND ttrss_user_prefs.value != '-1'
104 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
106 ttrss_feeds.update_interval > 0
107 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
108 ) OR (ttrss_feeds.last_updated IS NULL
109 AND ttrss_user_prefs.value != '-1')
110 OR (last_updated = '1970-01-01 00:00:00'
111 AND ttrss_user_prefs.value != '-1'))";
113 $update_limit_qpart = "AND ((
114 ttrss_feeds.update_interval = 0
115 AND ttrss_user_prefs.value != '-1'
116 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
118 ttrss_feeds.update_interval > 0
119 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
120 ) OR (ttrss_feeds.last_updated IS NULL
121 AND ttrss_user_prefs.value != '-1')
122 OR (last_updated = '1970-01-01 00:00:00'
123 AND ttrss_user_prefs.value != '-1'))";
126 // Test if feed is currently being updated by another process.
127 if (DB_TYPE
== "pgsql") {
128 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
130 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
133 // Test if there is a limit to number of updated feeds
135 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
137 // Update the least recently updated feeds first
138 $query_order = "ORDER BY last_updated";
139 if (DB_TYPE
== "pgsql") $query_order .= " NULLS FIRST";
141 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
143 ttrss_feeds, ttrss_users, ttrss_user_prefs
145 ttrss_feeds.owner_uid = ttrss_users.id
146 AND ttrss_user_prefs.profile IS NULL
147 AND ttrss_users.id = ttrss_user_prefs.owner_uid
148 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
149 $login_thresh_qpart $update_limit_qpart
150 $updstart_thresh_qpart
151 $query_order $query_limit";
153 // We search for feed needing update.
154 $result = db_query($query);
156 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
158 // Here is a little cache magic in order to minimize risk of double feed updates.
159 $feeds_to_update = array();
160 while ($line = db_fetch_assoc($result)) {
161 array_push($feeds_to_update, db_escape_string($line['feed_url']));
164 // We update the feed last update started date before anything else.
165 // There is no lag due to feed contents downloads
166 // It prevent an other process to update the same feed.
168 if(count($feeds_to_update) > 0) {
169 $feeds_quoted = array();
171 foreach ($feeds_to_update as $feed) {
172 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
175 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
176 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
180 $bstarted = microtime(true);
182 $batch_owners = array();
184 // For each feed, we call the feed update function.
185 foreach ($feeds_to_update as $feed) {
186 if($debug) _debug("Base feed: $feed");
188 //update_rss_feed($line["id"], true);
190 // since we have the data cached, we can deal with other feeds with the same url
192 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
193 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
194 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
195 ttrss_users.id = ttrss_user_prefs.owner_uid AND
196 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
197 ttrss_user_prefs.profile IS NULL AND
198 feed_url = '".db_escape_string($feed)."'
201 ORDER BY ttrss_feeds.id $query_limit");
203 if (db_num_rows($tmp_result) > 0) {
206 while ($tline = db_fetch_assoc($tmp_result)) {
207 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
209 if (array_search($tline["owner_uid"], $batch_owners) === FALSE)
210 array_push($batch_owners, $tline["owner_uid"]);
212 $fstarted = microtime(true);
213 $rss = update_rss_feed($tline["id"], true, false);
214 _debug_suppress(false);
216 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
224 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
225 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
228 foreach ($batch_owners as $owner_uid) {
229 _debug("Running housekeeping tasks for user $owner_uid...");
231 housekeeping_user($owner_uid);
234 require_once "digest.php";
236 // Send feed digests by email if needed.
237 send_headlines_digests($debug);
241 } // function update_daemon_common
243 // this is used when subscribing
244 function set_basic_feed_info($feed) {
246 $feed = db_escape_string($feed);
248 $result = db_query("SELECT feed_url,auth_pass,auth_login,auth_pass_encrypted
249 FROM ttrss_feeds WHERE id = '$feed'");
251 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
252 0, "auth_pass_encrypted"));
254 $auth_login = db_fetch_result($result, 0, "auth_login");
255 $auth_pass = db_fetch_result($result, 0, "auth_pass");
257 if ($auth_pass_encrypted) {
258 require_once "crypt.php";
259 $auth_pass = decrypt_string($auth_pass);
262 $fetch_url = db_fetch_result($result, 0, "feed_url");
264 $feed_data = fetch_file_contents($fetch_url, false,
265 $auth_login, $auth_pass, false,
269 global $fetch_curl_used;
271 if (!$fetch_curl_used) {
272 $tmp = @gzdecode
($feed_data);
274 if ($tmp) $feed_data = $tmp;
277 $feed_data = trim($feed_data);
279 $rss = new FeedParser($feed_data);
282 if (!$rss->error()) {
284 $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
286 $registered_title = db_fetch_result($result, 0, "title");
287 $orig_site_url = db_fetch_result($result, 0, "site_url");
289 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
290 $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
292 if ($feed_title && (!$registered_title ||
$registered_title == "[Unknown]")) {
293 db_query("UPDATE ttrss_feeds SET
294 title = '$feed_title' WHERE id = '$feed'");
297 if ($site_url && $orig_site_url != $site_url) {
298 db_query("UPDATE ttrss_feeds SET
299 site_url = '$site_url' WHERE id = '$feed'");
304 // ignore_daemon is not used
305 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) {
307 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
309 _debug_suppress(!$debug_enabled);
310 _debug("start", $debug_enabled);
312 $result = db_query("SELECT title FROM ttrss_feeds
313 WHERE id = '$feed'");
315 if (db_num_rows($result) == 0) {
316 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
317 user_error("Attempt to update unknown/invalid feed $feed", E_USER_WARNING
);
321 $title = db_fetch_result($result, 0, "title");
323 // feed was batch-subscribed or something, we need to get basic info
324 // this is not optimal currently as it fetches stuff separately TODO: optimize
325 if ($title == "[Unknown]") {
326 _debug("setting basic feed info for $feed...");
327 set_basic_feed_info($feed);
330 $result = db_query("SELECT id,update_interval,auth_login,
331 feed_url,auth_pass,cache_images,
332 mark_unread_on_update, owner_uid,
333 pubsub_state, auth_pass_encrypted,
335 FROM ttrss_feeds WHERE id = '$feed'");
337 $owner_uid = db_fetch_result($result, 0, "owner_uid");
338 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
339 0, "mark_unread_on_update"));
340 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
341 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
342 0, "auth_pass_encrypted"));
344 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
345 WHERE id = '$feed'");
347 $auth_login = db_fetch_result($result, 0, "auth_login");
348 $auth_pass = db_fetch_result($result, 0, "auth_pass");
350 if ($auth_pass_encrypted) {
351 require_once "crypt.php";
352 $auth_pass = decrypt_string($auth_pass);
355 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
356 $fetch_url = db_fetch_result($result, 0, "feed_url");
357 $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
358 if (!$feed_language) $feed_language = 'english';
360 $feed = db_escape_string($feed);
362 $date_feed_processed = date('Y-m-d H:i');
364 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".xml";
366 $pluginhost = new PluginHost();
367 $pluginhost->set_debug($debug_enabled);
368 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
370 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
371 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
372 $pluginhost->load_data();
374 if ($rss && is_object($rss) && get_class($rss) == "FeedParser") {
375 _debug("using previously initialized parser object");
379 $force_refetch = isset($_REQUEST["force_refetch"]);
382 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FETCH_FEED
) as $plugin) {
383 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, 0, $auth_login, $auth_pass);
388 file_exists($cache_filename) &&
389 is_readable($cache_filename) &&
390 !$auth_login && !$auth_pass &&
391 filemtime($cache_filename) > time() - 30) {
393 _debug("using local cache [$cache_filename].", $debug_enabled);
395 @$feed_data = file_get_contents($cache_filename);
398 $rss_hash = sha1($feed_data);
402 _debug("local cache will not be used for this feed", $debug_enabled);
405 // fetch feed from source
407 _debug("fetching [$fetch_url]...", $debug_enabled);
409 if (ini_get("open_basedir") && function_exists("curl_init")) {
410 _debug("not using CURL due to open_basedir restrictions");
413 $feed_data = fetch_file_contents($fetch_url, false,
414 $auth_login, $auth_pass, false,
415 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
418 global $fetch_curl_used;
420 if (!$fetch_curl_used) {
421 $tmp = @gzdecode
($feed_data);
423 if ($tmp) $feed_data = $tmp;
426 $feed_data = trim($feed_data);
428 _debug("fetch done.", $debug_enabled);
430 // cache vanilla feed data for re-use
431 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
432 $new_rss_hash = sha1($feed_data);
434 if ($new_rss_hash != $rss_hash) {
435 _debug("saving $cache_filename", $debug_enabled);
436 @file_put_contents
($cache_filename, $feed_data);
442 global $fetch_last_error;
443 global $fetch_last_error_code;
445 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
450 if ($fetch_last_error_code != 304) {
451 $error_escaped = db_escape_string($fetch_last_error);
453 _debug("source claims data not modified, nothing to do.", $debug_enabled);
457 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
458 last_updated = NOW() WHERE id = '$feed'");
464 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
465 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
468 // set last update to now so if anything *simplepie* crashes later we won't be
469 // continuously failing on the same feed
470 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
473 $rss = new FeedParser($feed_data);
479 $feed = db_escape_string($feed);
481 if (!$rss->error()) {
483 // We use local pluginhost here because we need to load different per-user feed plugins
484 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
486 _debug("language: $feed_language", $debug_enabled);
487 _debug("processing feed data...", $debug_enabled);
489 // db_query("BEGIN");
491 if (DB_TYPE
== "pgsql") {
492 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
494 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
497 $result = db_query("SELECT owner_uid,favicon_avg_color,
498 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
500 FROM ttrss_feeds WHERE id = '$feed'");
502 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
503 "favicon_needs_check"));
504 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
506 $owner_uid = db_fetch_result($result, 0, "owner_uid");
508 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
510 _debug("site_url: $site_url", $debug_enabled);
511 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
513 if ($favicon_needs_check ||
$force_refetch) {
515 /* terrible hack: if we crash on floicon shit here, we won't check
516 * the icon avgcolor again (unless the icon got updated) */
518 $favicon_file = ICONS_DIR
. "/$feed.ico";
519 $favicon_modified = @filemtime
($favicon_file);
521 _debug("checking favicon...", $debug_enabled);
523 check_feed_favicon($site_url, $feed);
524 $favicon_modified_new = @filemtime
($favicon_file);
526 if ($favicon_modified_new > $favicon_modified)
527 $favicon_avg_color = '';
529 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
530 require_once "colors.php";
532 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
535 $favicon_color = db_escape_string(
536 calculate_avg_color($favicon_file));
538 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
539 } else if ($favicon_avg_color == 'fail') {
540 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
543 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
545 WHERE id = '$feed'");
548 _debug("loading filters & labels...", $debug_enabled);
550 $filters = load_filters($feed, $owner_uid);
552 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
554 $items = $rss->get_items();
556 if (!is_array($items)) {
557 _debug("no articles found.", $debug_enabled);
559 db_query("UPDATE ttrss_feeds
560 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
562 return; // no articles
565 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
567 _debug("checking for PUSH hub...", $debug_enabled);
569 $feed_hub_url = false;
571 $links = $rss->get_links('hub');
573 if ($links && is_array($links)) {
574 foreach ($links as $l) {
580 _debug("feed hub url: $feed_hub_url", $debug_enabled);
582 $feed_self_url = $fetch_url;
584 $links = $rss->get_links('self');
586 if ($links && is_array($links)) {
587 foreach ($links as $l) {
593 _debug("feed self url = $feed_self_url");
595 if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
596 !ini_get("open_basedir")) {
598 require_once 'lib/pubsubhubbub/subscriber.php';
600 $callback_url = get_self_url_prefix() .
601 "/public.php?op=pubsub&id=$feed";
603 $s = new Subscriber($feed_hub_url, $callback_url);
605 $rc = $s->subscribe($feed_self_url);
607 _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
609 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
610 WHERE id = '$feed'");
614 _debug("processing articles...", $debug_enabled);
618 foreach ($items as $item) {
619 if ($_REQUEST['xdebug'] == 3) {
623 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
624 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
628 $entry_guid = $item->get_id();
629 if (!$entry_guid) $entry_guid = $item->get_link();
630 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
631 if (!$entry_guid) continue;
633 $entry_guid = "$owner_uid,$entry_guid";
635 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
637 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
639 $entry_timestamp = "";
641 $entry_timestamp = $item->get_date();
643 _debug("orig date: " . $item->get_date(), $debug_enabled);
645 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
646 $entry_timestamp = time();
649 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
651 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
653 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
654 // $entry_title = decode_numeric_entities($entry_title);
655 $entry_title = $item->get_title();
657 $entry_link = rewrite_relative_url($site_url, $item->get_link());
659 _debug("title $entry_title", $debug_enabled);
660 _debug("link $entry_link", $debug_enabled);
662 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
664 $entry_content = $item->get_content();
665 if (!$entry_content) $entry_content = $item->get_description();
667 if ($_REQUEST["xdebug"] == 2) {
669 print htmlspecialchars($entry_content);
673 $entry_comments = db_escape_string(mb_substr($item->get_comments_url(), 0, 245));
674 $num_comments = (int) $item->get_comments_count();
676 $entry_author = $item->get_author(); // escaped later
677 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
679 _debug("author $entry_author", $debug_enabled);
680 _debug("num_comments: $num_comments", $debug_enabled);
681 _debug("looking for tags...", $debug_enabled);
683 // parse <category> entries into tags
685 $additional_tags = array();
687 $additional_tags_src = $item->get_categories();
689 if (is_array($additional_tags_src)) {
690 foreach ($additional_tags_src as $tobj) {
691 array_push($additional_tags, $tobj);
695 $entry_tags = array_unique($additional_tags);
697 for ($i = 0; $i < count($entry_tags); $i++
)
698 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
700 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
702 _debug("done collecting data.", $debug_enabled);
704 $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
705 WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
707 if (db_num_rows($result) != 0) {
708 $base_entry_id = db_fetch_result($result, 0, "id");
709 $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
710 $article_labels = get_article_labels($base_entry_id, $owner_uid);
711 $entry_language = db_fetch_result($result, 0, "lang");
713 $existing_tags = get_article_tags($base_entry_id, $owner_uid);
714 $entry_tags = array_unique(array_merge($entry_tags, $existing_tags));
717 $base_entry_id = false;
718 $entry_stored_hash = "";
719 $article_labels = array();
720 $entry_language = "";
723 $article = array("owner_uid" => $owner_uid, // read only
724 "guid" => $entry_guid, // read only
725 "guid_hashed" => $entry_guid_hashed, // read only
726 "title" => $entry_title,
727 "content" => $entry_content,
728 "link" => $entry_link,
729 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
730 "tags" => $entry_tags,
731 "author" => $entry_author,
732 "force_catchup" => false, // ugly hack for the time being
733 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
734 "language" => $entry_language,
735 "feed" => array("id" => $feed,
736 "fetch_url" => $fetch_url,
737 "site_url" => $site_url,
738 "cache_images" => $cache_images)
741 $entry_plugin_data = "";
742 $entry_current_hash = calculate_article_hash($article, $pluginhost);
744 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
746 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
747 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
749 // we keep encountering the entry in feeds, so we need to
750 // update date_updated column so that we don't get horrible
751 // dupes when the entry gets purged and reinserted again e.g.
752 // in the case of SLOW SLOW OMG SLOW updating feeds
754 $base_entry_id = db_fetch_result($result, 0, "id");
756 db_query("UPDATE ttrss_entries SET date_updated = NOW()
757 WHERE id = '$base_entry_id'");
762 _debug("hash differs, applying plugin filters:", $debug_enabled);
764 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
765 _debug("... " . get_class($plugin), $debug_enabled);
767 $start = microtime(true);
768 $article = $plugin->hook_article_filter($article);
770 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
772 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
775 if ($_REQUEST["xdebug"] == 2) {
776 print "processed content: ";
777 print htmlspecialchars($article["content"]);
781 $entry_plugin_data = db_escape_string($entry_plugin_data);
783 _debug("plugin data: $entry_plugin_data", $debug_enabled);
785 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
786 if (DB_TYPE
== "mysql") {
787 foreach ($article as $k => $v) {
789 // i guess we'll have to take the risk of 4byte unicode labels & tags here
790 if (is_string($article[$k])) {
791 $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
796 /* Collect article tags here so we could filter by them: */
798 $matched_rules = array();
800 $article_filters = get_article_filters($filters, $article["title"],
801 $article["content"], $article["link"], 0, $article["author"],
802 $article["tags"], $matched_rules);
804 if ($debug_enabled) {
805 _debug("matched filter rules: ", $debug_enabled);
807 if (count($matched_rules) != 0) {
808 print_r($matched_rules);
811 _debug("filter actions: ", $debug_enabled);
813 if (count($article_filters) != 0) {
814 print_r($article_filters);
818 $plugin_filter_names = find_article_filters($article_filters, "plugin");
819 $plugin_filter_actions = $pluginhost->get_filter_actions();
821 if (count($plugin_filter_names) > 0) {
822 _debug("applying plugin filter actions...", $debug_enabled);
824 foreach ($plugin_filter_names as $pfn) {
825 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
827 if (isset($plugin_filter_actions[$pfclass])) {
828 $plugin = $pluginhost->get_plugin($pfclass);
830 _debug("... $pfclass: $pfaction", $debug_enabled);
833 $start = microtime(true);
834 $article = $plugin->hook_article_filter_action($article, $pfaction);
836 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
838 _debug("??? $pfclass: plugin object not found.");
841 _debug("??? $pfclass: filter plugin not registered.");
846 $entry_tags = $article["tags"];
847 $entry_guid = db_escape_string($entry_guid);
848 $entry_title = db_escape_string($article["title"]);
849 $entry_author = db_escape_string(mb_substr($article["author"], 0, 245));
850 $entry_link = db_escape_string($article["link"]);
851 $entry_content = $article["content"]; // escaped below
852 $entry_force_catchup = $article["force_catchup"];
853 $article_labels = $article["labels"];
854 $entry_score_modifier = (int) $article["score_modifier"];
855 $entry_language = db_escape_string($article["language"]);
857 if ($debug_enabled) {
858 _debug("article labels:", $debug_enabled);
860 if (count($article_labels) != 0) {
861 print_r($article_labels);
865 _debug("force catchup: $entry_force_catchup");
867 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
868 cache_images($entry_content, $site_url, $debug_enabled);
870 $entry_content = db_escape_string($entry_content, false);
874 $result = db_query("SELECT id FROM ttrss_entries
875 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
877 if (db_num_rows($result) == 0) {
879 _debug("base guid [$entry_guid] not found", $debug_enabled);
881 // base post entry does not exist, create it
884 "INSERT INTO ttrss_entries
901 '$entry_guid_hashed',
903 '$entry_timestamp_fmt',
905 '$entry_current_hash',
908 '$date_feed_processed',
911 '$entry_plugin_data',
916 $base_entry_id = db_fetch_result($result, 0, "id");
919 // now it should exist, if not - bad luck then
921 $result = db_query("SELECT id FROM ttrss_entries
922 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
927 if (db_num_rows($result) == 1) {
929 _debug("base guid found, checking for user record", $debug_enabled);
931 $ref_id = db_fetch_result($result, 0, "id");
932 $entry_ref_id = $ref_id;
934 /* $stored_guid = db_fetch_result($result, 0, "guid");
935 if ($stored_guid != $entry_guid_hashed) {
936 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
938 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
942 if (find_article_filter($article_filters, "filter")) {
943 //db_query("COMMIT"); // close transaction in progress
947 $score = calculate_article_score($article_filters) +
$entry_score_modifier;
949 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
951 // check for user post link to main table
953 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
954 ref_id = '$ref_id' AND owner_uid = '$owner_uid'";
956 // if ($_REQUEST["xdebug"]) print "$query\n";
958 $result = db_query($query);
960 // okay it doesn't exist - create user entry
961 if (db_num_rows($result) == 0) {
963 _debug("user record not found, creating...", $debug_enabled);
965 if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
967 $last_read_qpart = 'NULL';
970 $last_read_qpart = 'NOW()';
973 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
979 if (find_article_filter($article_filters, 'publish')) {
982 $published = 'false';
985 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
986 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
989 "INSERT INTO ttrss_user_entries
990 (ref_id, owner_uid, feed_id, unread, last_read, marked,
991 published, score, tag_cache, label_cache, uuid,
992 last_marked, last_published)
993 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
994 $last_read_qpart, $marked, $published, '$score', '', '',
995 '', $last_marked, $last_published)");
997 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
998 $rss_link = get_self_url_prefix() .
999 "/public.php?op=rss&id=-2&key=" .
1000 get_feed_access_key(-2, false, $owner_uid);
1002 $p = new Publisher(PUBSUBHUBBUB_HUB
);
1004 /* $pubsub_result = */ $p->publish_update($rss_link);
1008 "SELECT int_id FROM ttrss_user_entries WHERE
1009 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
1010 feed_id = '$feed' LIMIT 1");
1012 if (db_num_rows($result) == 1) {
1013 $entry_int_id = db_fetch_result($result, 0, "int_id");
1016 _debug("user record FOUND", $debug_enabled);
1018 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
1019 $entry_int_id = db_fetch_result($result, 0, "int_id");
1022 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
1024 if (DB_TYPE
== "pgsql") {
1025 $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
1028 $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
1031 $tsvector_qpart = "";
1034 db_query("UPDATE ttrss_entries
1035 SET title = '$entry_title',
1036 content = '$entry_content',
1037 content_hash = '$entry_current_hash',
1038 updated = '$entry_timestamp_fmt',
1040 num_comments = '$num_comments',
1041 plugin_data = '$entry_plugin_data',
1042 author = '$entry_author',
1043 lang = '$entry_language'
1044 WHERE id = '$ref_id'");
1047 db_query("UPDATE ttrss_user_entries
1048 SET score = '$score' WHERE ref_id = '$ref_id'");
1050 if ($mark_unread_on_update) {
1051 _debug("article updated, marking unread as requested.", $debug_enabled);
1053 db_query("UPDATE ttrss_user_entries
1054 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1058 //db_query("COMMIT");
1060 _debug("assigning labels [other]...", $debug_enabled);
1062 foreach ($article_labels as $label) {
1063 label_add_article($entry_ref_id, $label[1], $owner_uid);
1066 _debug("assigning labels [filters]...", $debug_enabled);
1068 assign_article_to_label_filters($entry_ref_id, $article_filters,
1069 $owner_uid, $article_labels);
1071 _debug("looking for enclosures...", $debug_enabled);
1075 $enclosures = array();
1077 $encs = $item->get_enclosures();
1079 if (is_array($encs)) {
1080 foreach ($encs as $e) {
1082 $e->link
, $e->type
, $e->length
, $e->title
, $e->width
, $e->height
);
1083 array_push($enclosures, $e_item);
1087 if ($debug_enabled) {
1088 _debug("article enclosures:", $debug_enabled);
1089 print_r($enclosures);
1092 //db_query("BEGIN");
1095 // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
1097 foreach ($enclosures as $enc) {
1098 $enc_url = db_escape_string($enc[0]);
1099 $enc_type = db_escape_string($enc[1]);
1100 $enc_dur = db_escape_string($enc[2]);
1101 $enc_title = db_escape_string($enc[3]);
1102 $enc_width = intval($enc[4]);
1103 $enc_height = intval($enc[5]);
1105 $result = db_query("SELECT id FROM ttrss_enclosures
1106 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1108 if (db_num_rows($result) == 0) {
1109 db_query("INSERT INTO ttrss_enclosures
1110 (content_url, content_type, title, duration, post_id, width, height) VALUES
1111 ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
1115 //db_query("COMMIT");
1117 // check for manual tags (we have to do it here since they're loaded from filters)
1119 foreach ($article_filters as $f) {
1120 if ($f["type"] == "tag") {
1122 $manual_tags = trim_array(explode(",", $f["param"]));
1124 foreach ($manual_tags as $tag) {
1125 if (tag_is_valid($tag)) {
1126 array_push($entry_tags, $tag);
1134 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1135 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1137 $filtered_tags = array();
1138 $tags_to_cache = array();
1140 if ($entry_tags && is_array($entry_tags)) {
1141 foreach ($entry_tags as $tag) {
1142 if (array_search($tag, $boring_tags) === false) {
1143 array_push($filtered_tags, $tag);
1148 $filtered_tags = array_unique($filtered_tags);
1150 if ($debug_enabled) {
1151 _debug("filtered article tags:", $debug_enabled);
1152 print_r($filtered_tags);
1155 // Save article tags in the database
1157 if (count($filtered_tags) > 0) {
1159 //db_query("BEGIN");
1161 foreach ($filtered_tags as $tag) {
1163 $tag = sanitize_tag($tag);
1164 $tag = db_escape_string($tag);
1166 if (!tag_is_valid($tag)) continue;
1168 $result = db_query("SELECT id FROM ttrss_tags
1169 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1170 owner_uid = '$owner_uid' LIMIT 1");
1172 if ($result && db_num_rows($result) == 0) {
1174 db_query("INSERT INTO ttrss_tags
1175 (owner_uid,tag_name,post_int_id)
1176 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1179 array_push($tags_to_cache, $tag);
1182 /* update the cache */
1184 $tags_to_cache = array_unique($tags_to_cache);
1186 $tags_str = db_escape_string(join(",", $tags_to_cache));
1188 db_query("UPDATE ttrss_user_entries
1189 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1190 AND owner_uid = $owner_uid");
1192 //db_query("COMMIT");
1195 _debug("article processed", $debug_enabled);
1198 _debug("purging feed...", $debug_enabled);
1200 purge_feed($feed, 0, $debug_enabled);
1202 db_query("UPDATE ttrss_feeds
1203 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1205 // db_query("COMMIT");
1209 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1211 _debug("fetch error: $error_msg", $debug_enabled);
1213 if (count($rss->errors()) > 1) {
1214 foreach ($rss->errors() as $error) {
1220 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1221 last_updated = NOW() WHERE id = '$feed'");
1226 _debug("done", $debug_enabled);
1231 function cache_images($html, $site_url, $debug) {
1232 libxml_use_internal_errors(true);
1234 $charset_hack = '<head>
1235 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1238 $doc = new DOMDocument();
1239 $doc->loadHTML($charset_hack . $html);
1240 $xpath = new DOMXPath($doc);
1242 $entries = $xpath->query('(//img[@src])');
1244 foreach ($entries as $entry) {
1245 if ($entry->hasAttribute('src')) {
1246 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1248 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1250 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1252 if (!file_exists($local_filename)) {
1253 $file_content = fetch_file_contents($src);
1255 if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE
) {
1256 file_put_contents($local_filename, $file_content);
1259 touch($local_filename);
1265 function expire_error_log($debug) {
1266 if ($debug) _debug("Removing old error log entries...");
1268 if (DB_TYPE
== "pgsql") {
1269 db_query("DELETE FROM ttrss_error_log
1270 WHERE created_at < NOW() - INTERVAL '7 days'");
1272 db_query("DELETE FROM ttrss_error_log
1273 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1278 function expire_lock_files($debug) {
1279 //if ($debug) _debug("Removing old lock files...");
1283 if (is_writable(LOCK_DIRECTORY
)) {
1284 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1287 foreach ($files as $file) {
1288 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1296 if ($debug) _debug("Removed $num_deleted old lock files.");
1299 function expire_cached_files($debug) {
1300 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1301 $cache_dir = CACHE_DIR
. "/$dir";
1303 // if ($debug) _debug("Expiring $cache_dir");
1307 if (is_writable($cache_dir)) {
1308 $files = glob("$cache_dir/*");
1311 foreach ($files as $file) {
1312 if (time() - filemtime($file) > 86400*7) {
1321 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1326 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1327 * Returns the url query as associative array
1329 * @param string query
1330 * @return array params
1332 function convertUrlQuery($query) {
1333 $queryParts = explode('&', $query);
1337 foreach ($queryParts as $param) {
1338 $item = explode('=', $param);
1339 $params[$item[0]] = $item[1];
1345 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags, &$matched_rules = false) {
1348 foreach ($filters as $filter) {
1349 $match_any_rule = $filter["match_any_rule"];
1350 $inverse = $filter["inverse"];
1351 $filter_match = false;
1353 foreach ($filter["rules"] as $rule) {
1355 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1356 $rule_inverse = $rule["inverse"];
1361 switch ($rule["type"]) {
1363 $match = @preg_match
("/$reg_exp/iu", $title);
1366 // we don't need to deal with multiline regexps
1367 $content = preg_replace("/[\r\n\t]/", "", $content);
1369 $match = @preg_match
("/$reg_exp/iu", $content);
1372 // we don't need to deal with multiline regexps
1373 $content = preg_replace("/[\r\n\t]/", "", $content);
1375 $match = (@preg_match
("/$reg_exp/iu", $title) || @preg_match
("/$reg_exp/iu", $content));
1378 $match = @preg_match
("/$reg_exp/iu", $link);
1381 $match = @preg_match
("/$reg_exp/iu", $author);
1384 foreach ($tags as $tag) {
1385 if (@preg_match
("/$reg_exp/iu", $tag)) {
1393 if ($rule_inverse) $match = !$match;
1395 if ($match_any_rule) {
1397 $filter_match = true;
1401 $filter_match = $match;
1408 if ($inverse) $filter_match = !$filter_match;
1410 if ($filter_match) {
1411 if (is_array($matched_rules)) array_push($matched_rules, $rule);
1413 foreach ($filter["actions"] AS $action) {
1414 array_push($matches, $action);
1416 // if Stop action encountered, perform no further processing
1417 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1425 function find_article_filter($filters, $filter_name) {
1426 foreach ($filters as $f) {
1427 if ($f["type"] == $filter_name) {
1434 function find_article_filters($filters, $filter_name) {
1437 foreach ($filters as $f) {
1438 if ($f["type"] == $filter_name) {
1439 array_push($results, $f);
1445 function calculate_article_score($filters) {
1448 foreach ($filters as $f) {
1449 if ($f["type"] == "score") {
1450 $score +
= $f["param"];
1456 function labels_contains_caption($labels, $caption) {
1457 foreach ($labels as $label) {
1458 if ($label[1] == $caption) {
1466 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1467 foreach ($filters as $f) {
1468 if ($f["type"] == "label") {
1469 if (!labels_contains_caption($article_labels, $f["param"])) {
1470 label_add_article($id, $f["param"], $owner_uid);
1476 function make_guid_from_title($title) {
1477 return preg_replace("/[ \"\',.:;]/", "-",
1478 mb_strtolower(strip_tags($title), 'utf-8'));
1481 /* function verify_feed_xml($feed_data) {
1482 libxml_use_internal_errors(true);
1483 $doc = new DOMDocument();
1484 $doc->loadXML($feed_data);
1485 $error = libxml_get_last_error();
1486 libxml_clear_errors();
1490 function cleanup_counters_cache($debug) {
1491 $result = db_query("DELETE FROM ttrss_counters_cache
1492 WHERE feed_id > 0 AND
1493 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1495 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1496 $frows = db_affected_rows($result);
1498 $result = db_query("DELETE FROM ttrss_cat_counters_cache
1499 WHERE feed_id > 0 AND
1500 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1502 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1503 $crows = db_affected_rows($result);
1505 _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1508 function housekeeping_user($owner_uid) {
1509 $tmph = new PluginHost();
1511 load_user_plugins($owner_uid, $tmph);
1513 $tmph->run_hooks(PluginHost
::HOOK_HOUSE_KEEPING
, "hook_house_keeping", "");
1516 function housekeeping_common($debug) {
1517 expire_cached_files($debug);
1518 expire_lock_files($debug);
1519 expire_error_log($debug);
1521 $count = update_feedbrowser_cache();
1522 _debug("Feedbrowser updated, $count feeds processed.");
1524 purge_orphans( true);
1525 cleanup_counters_cache($debug);
1527 //$rc = cleanup_tags( 14, 50000);
1528 //_debug("Cleaned $rc cached tags.");
1530 PluginHost
::getInstance()->run_hooks(PluginHost
::HOOK_HOUSE_KEEPING
, "hook_house_keeping", "");