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 $tmp .= sha1("$k:" . (is_array($v) ?
implode(",", $v) : $v));
16 return sha1(implode(",", $pluginhost->get_plugin_names()) . $tmp);
19 function update_feedbrowser_cache() {
21 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
22 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
23 WHERE tf.feed_url = ttrss_feeds.feed_url
24 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
25 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
29 db_query("DELETE FROM ttrss_feedbrowser_cache");
33 while ($line = db_fetch_assoc($result)) {
34 $subscribers = db_escape_string($line["subscribers"]);
35 $feed_url = db_escape_string($line["feed_url"]);
36 $title = db_escape_string($line["title"]);
37 $site_url = db_escape_string($line["site_url"]);
39 $tmp_result = db_query("SELECT subscribers FROM
40 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
42 if (db_num_rows($tmp_result) == 0) {
44 db_query("INSERT INTO ttrss_feedbrowser_cache
45 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
46 '$site_url', '$title', '$subscribers')");
62 * Update a feed batch.
63 * Used by daemons to update n feeds by run.
64 * Only update feed needing a update, and not being processed
67 * @param mixed $link Database link
68 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
69 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
70 * @param boolean $debug Set to false to disable debug output. Default to true.
73 function update_daemon_common($limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
74 // Process all other feeds using last_updated and interval parameters
76 $schema_version = get_schema_version();
78 if ($schema_version != SCHEMA_VERSION
) {
79 die("Schema version is wrong, please upgrade the database.\n");
82 define('PREFS_NO_CACHE', true);
84 // Test if the user has loggued in recently. If not, it does not update its feeds.
85 if (!SINGLE_USER_MODE
&& DAEMON_UPDATE_LOGIN_LIMIT
> 0) {
86 if (DB_TYPE
== "pgsql") {
87 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT
." days'";
89 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
92 $login_thresh_qpart = "";
95 // Test if the feed need a update (update interval exceeded).
96 if (DB_TYPE
== "pgsql") {
97 $update_limit_qpart = "AND ((
98 ttrss_feeds.update_interval = 0
99 AND ttrss_user_prefs.value != '-1'
100 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
102 ttrss_feeds.update_interval > 0
103 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
104 ) OR (ttrss_feeds.last_updated IS NULL
105 AND ttrss_user_prefs.value != '-1')
106 OR (last_updated = '1970-01-01 00:00:00'
107 AND ttrss_user_prefs.value != '-1'))";
109 $update_limit_qpart = "AND ((
110 ttrss_feeds.update_interval = 0
111 AND ttrss_user_prefs.value != '-1'
112 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
114 ttrss_feeds.update_interval > 0
115 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
116 ) OR (ttrss_feeds.last_updated IS NULL
117 AND ttrss_user_prefs.value != '-1')
118 OR (last_updated = '1970-01-01 00:00:00'
119 AND ttrss_user_prefs.value != '-1'))";
122 // Test if feed is currently being updated by another process.
123 if (DB_TYPE
== "pgsql") {
124 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
126 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
129 // Test if there is a limit to number of updated feeds
131 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
133 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
135 ttrss_feeds, ttrss_users, ttrss_user_prefs
137 ttrss_feeds.owner_uid = ttrss_users.id
138 AND ttrss_user_prefs.profile IS NULL
139 AND ttrss_users.id = ttrss_user_prefs.owner_uid
140 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
141 $login_thresh_qpart $update_limit_qpart
142 $updstart_thresh_qpart
143 ORDER BY last_updated $query_limit";
145 // We search for feed needing update.
146 $result = db_query($query);
148 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
150 // Here is a little cache magic in order to minimize risk of double feed updates.
151 $feeds_to_update = array();
152 while ($line = db_fetch_assoc($result)) {
153 array_push($feeds_to_update, db_escape_string($line['feed_url']));
156 // We update the feed last update started date before anything else.
157 // There is no lag due to feed contents downloads
158 // It prevent an other process to update the same feed.
160 if(count($feeds_to_update) > 0) {
161 $feeds_quoted = array();
163 foreach ($feeds_to_update as $feed) {
164 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
167 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
168 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
172 $bstarted = microtime(true);
174 // For each feed, we call the feed update function.
175 foreach ($feeds_to_update as $feed) {
176 if($debug) _debug("Base feed: $feed");
178 //update_rss_feed($line["id"], true);
180 // since we have the data cached, we can deal with other feeds with the same url
182 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
183 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
184 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
185 ttrss_users.id = ttrss_user_prefs.owner_uid AND
186 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
187 ttrss_user_prefs.profile IS NULL AND
188 feed_url = '".db_escape_string($feed)."' AND
189 (ttrss_feeds.update_interval > 0 OR
190 ttrss_user_prefs.value != '-1')
192 ORDER BY ttrss_feeds.id $query_limit");
194 if (db_num_rows($tmp_result) > 0) {
197 while ($tline = db_fetch_assoc($tmp_result)) {
198 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
200 $fstarted = microtime(true);
201 $rss = update_rss_feed($tline["id"], true, false);
202 _debug_suppress(false);
204 _debug(sprintf(" %.4f (sec)", microtime(true) - $fstarted));
212 _debug(sprintf("Processed %d feeds in %.4f (sec), %.4f (sec/feed avg)", $nf,
213 microtime(true) - $bstarted, (microtime(true) - $bstarted) / $nf));
216 require_once "digest.php";
218 // Send feed digests by email if needed.
219 send_headlines_digests($debug);
223 } // function update_daemon_common
225 // this is used when subscribing
226 function set_basic_feed_info($feed) {
228 $feed = db_escape_string($feed);
230 $result = db_query("SELECT feed_url,auth_pass,auth_pass_encrypted
231 FROM ttrss_feeds WHERE id = '$feed'");
233 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
234 0, "auth_pass_encrypted"));
236 $auth_login = db_fetch_result($result, 0, "auth_login");
237 $auth_pass = db_fetch_result($result, 0, "auth_pass");
239 if ($auth_pass_encrypted) {
240 require_once "crypt.php";
241 $auth_pass = decrypt_string($auth_pass);
244 $fetch_url = db_fetch_result($result, 0, "feed_url");
246 $feed_data = fetch_file_contents($fetch_url, false,
247 $auth_login, $auth_pass, false,
248 FEED_FETCH_TIMEOUT_TIMEOUT
,
251 global $fetch_curl_used;
253 if (!$fetch_curl_used) {
254 $tmp = @gzdecode
($feed_data);
256 if ($tmp) $feed_data = $tmp;
259 $feed_data = trim($feed_data);
261 $rss = new FeedParser($feed_data);
264 if (!$rss->error()) {
266 $result = db_query("SELECT title, site_url FROM ttrss_feeds WHERE id = '$feed'");
268 $registered_title = db_fetch_result($result, 0, "title");
269 $orig_site_url = db_fetch_result($result, 0, "site_url");
271 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
272 $feed_title = db_escape_string(mb_substr($rss->get_title(), 0, 199));
274 if ($feed_title && (!$registered_title ||
$registered_title == "[Unknown]")) {
275 db_query("UPDATE ttrss_feeds SET
276 title = '$feed_title' WHERE id = '$feed'");
279 if ($site_url && $orig_site_url != $site_url) {
280 db_query("UPDATE ttrss_feeds SET
281 site_url = '$site_url' WHERE id = '$feed'");
286 // ignore_daemon is not used
287 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) {
289 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
291 _debug_suppress(!$debug_enabled);
292 _debug("start", $debug_enabled);
294 $result = db_query("SELECT title FROM ttrss_feeds
295 WHERE id = '$feed'");
296 $title = db_fetch_result($result, 0, "title");
298 // feed was batch-subscribed or something, we need to get basic info
299 // this is not optimal currently as it fetches stuff separately TODO: optimize
300 if ($title == "[Unknown]") {
301 _debug("setting basic feed info for $feed...");
302 set_basic_feed_info($feed);
305 $result = db_query("SELECT id,update_interval,auth_login,
306 feed_url,auth_pass,cache_images,
307 mark_unread_on_update, owner_uid,
308 pubsub_state, auth_pass_encrypted,
310 (SELECT max(date_entered) FROM
311 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
312 FROM ttrss_feeds WHERE id = '$feed'");
314 if (db_num_rows($result) == 0) {
315 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
319 $last_article_timestamp = @strtotime
(db_fetch_result($result, 0, "last_article_timestamp"));
321 if (defined('_DISABLE_HTTP_304'))
322 $last_article_timestamp = 0;
324 $owner_uid = db_fetch_result($result, 0, "owner_uid");
325 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
326 0, "mark_unread_on_update"));
327 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
328 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
329 0, "auth_pass_encrypted"));
331 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
332 WHERE id = '$feed'");
334 $auth_login = db_fetch_result($result, 0, "auth_login");
335 $auth_pass = db_fetch_result($result, 0, "auth_pass");
337 if ($auth_pass_encrypted) {
338 require_once "crypt.php";
339 $auth_pass = decrypt_string($auth_pass);
342 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
343 $fetch_url = db_fetch_result($result, 0, "feed_url");
344 $feed_language = db_escape_string(mb_strtolower(db_fetch_result($result, 0, "feed_language")));
345 if (!$feed_language) $feed_language = 'english';
347 $feed = db_escape_string($feed);
349 $date_feed_processed = date('Y-m-d H:i');
351 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".xml";
353 $pluginhost = new PluginHost();
354 $pluginhost->set_debug($debug_enabled);
355 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
357 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
358 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
359 $pluginhost->load_data();
361 if ($rss && is_object($rss) && get_class($rss) == "FeedParser") {
362 _debug("using previously initialized parser object");
366 $force_refetch = isset($_REQUEST["force_refetch"]);
368 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FETCH_FEED
) as $plugin) {
369 $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, $last_article_timestamp, $auth_login, $auth_pass);
374 file_exists($cache_filename) &&
375 is_readable($cache_filename) &&
376 !$auth_login && !$auth_pass &&
377 filemtime($cache_filename) > time() - 30) {
379 _debug("using local cache [$cache_filename].", $debug_enabled);
381 @$feed_data = file_get_contents($cache_filename);
384 $rss_hash = sha1($feed_data);
388 _debug("local cache will not be used for this feed", $debug_enabled);
391 // fetch feed from source
393 _debug("fetching [$fetch_url]...", $debug_enabled);
394 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
396 $feed_data = fetch_file_contents($fetch_url, false,
397 $auth_login, $auth_pass, false,
398 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
399 $force_refetch ?
0 : $last_article_timestamp);
401 global $fetch_curl_used;
403 if (!$fetch_curl_used) {
404 $tmp = @gzdecode
($feed_data);
406 if ($tmp) $feed_data = $tmp;
409 $feed_data = trim($feed_data);
411 _debug("fetch done.", $debug_enabled);
413 // cache vanilla feed data for re-use
414 if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
415 $new_rss_hash = sha1($feed_data);
417 if ($new_rss_hash != $rss_hash) {
418 _debug("saving $cache_filename", $debug_enabled);
419 @file_put_contents
($cache_filename, $feed_data);
425 global $fetch_last_error;
426 global $fetch_last_error_code;
428 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
433 if ($fetch_last_error_code != 304) {
434 $error_escaped = db_escape_string($fetch_last_error);
436 _debug("source claims data not modified, nothing to do.", $debug_enabled);
440 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
441 last_updated = NOW() WHERE id = '$feed'");
447 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
448 $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed);
451 // set last update to now so if anything *simplepie* crashes later we won't be
452 // continuously failing on the same feed
453 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
456 $rss = new FeedParser($feed_data);
462 $feed = db_escape_string($feed);
464 if (!$rss->error()) {
466 // We use local pluginhost here because we need to load different per-user feed plugins
467 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
469 _debug("language: $feed_language", $debug_enabled);
470 _debug("processing feed data...", $debug_enabled);
472 // db_query("BEGIN");
474 if (DB_TYPE
== "pgsql") {
475 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
477 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
480 $result = db_query("SELECT owner_uid,favicon_avg_color,
481 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
483 FROM ttrss_feeds WHERE id = '$feed'");
485 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
486 "favicon_needs_check"));
487 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
489 $owner_uid = db_fetch_result($result, 0, "owner_uid");
491 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
493 _debug("site_url: $site_url", $debug_enabled);
494 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
496 if ($favicon_needs_check ||
$force_refetch) {
498 /* terrible hack: if we crash on floicon shit here, we won't check
499 * the icon avgcolor again (unless the icon got updated) */
501 $favicon_file = ICONS_DIR
. "/$feed.ico";
502 $favicon_modified = @filemtime
($favicon_file);
504 _debug("checking favicon...", $debug_enabled);
506 check_feed_favicon($site_url, $feed);
507 $favicon_modified_new = @filemtime
($favicon_file);
509 if ($favicon_modified_new > $favicon_modified)
510 $favicon_avg_color = '';
512 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
513 require_once "colors.php";
515 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
518 $favicon_color = db_escape_string(
519 calculate_avg_color($favicon_file));
521 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
522 } else if ($favicon_avg_color == 'fail') {
523 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
526 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
528 WHERE id = '$feed'");
531 _debug("loading filters & labels...", $debug_enabled);
533 $filters = load_filters($feed, $owner_uid);
535 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
537 $items = $rss->get_items();
539 if (!is_array($items)) {
540 _debug("no articles found.", $debug_enabled);
542 db_query("UPDATE ttrss_feeds
543 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
545 return; // no articles
548 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
550 _debug("checking for PUSH hub...", $debug_enabled);
552 $feed_hub_url = false;
554 $links = $rss->get_links('hub');
556 if ($links && is_array($links)) {
557 foreach ($links as $l) {
563 _debug("feed hub url: $feed_hub_url", $debug_enabled);
565 $feed_self_url = $fetch_url;
567 $links = $rss->get_links('self');
569 if ($links && is_array($links)) {
570 foreach ($links as $l) {
576 _debug("feed self url = $feed_self_url");
578 if ($feed_hub_url && $feed_self_url && function_exists('curl_init') &&
579 !ini_get("open_basedir")) {
581 require_once 'lib/pubsubhubbub/subscriber.php';
583 $callback_url = get_self_url_prefix() .
584 "/public.php?op=pubsub&id=$feed";
586 $s = new Subscriber($feed_hub_url, $callback_url);
588 $rc = $s->subscribe($feed_self_url);
590 _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled);
592 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
593 WHERE id = '$feed'");
597 _debug("processing articles...", $debug_enabled);
601 foreach ($items as $item) {
602 if ($_REQUEST['xdebug'] == 3) {
606 if (ini_get("max_execution_time") > 0 && time() - $tstart >= ini_get("max_execution_time") * 0.7) {
607 _debug("looks like there's too many articles to process at once, breaking out", $debug_enabled);
611 $entry_guid = $item->get_id();
612 if (!$entry_guid) $entry_guid = $item->get_link();
613 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
614 if (!$entry_guid) continue;
616 $entry_guid = "$owner_uid,$entry_guid";
618 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
620 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
622 $entry_timestamp = "";
624 $entry_timestamp = $item->get_date();
626 _debug("orig date: " . $item->get_date(), $debug_enabled);
628 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
629 $entry_timestamp = time();
632 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
634 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
636 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
637 // $entry_title = decode_numeric_entities($entry_title);
638 $entry_title = $item->get_title();
640 $entry_link = rewrite_relative_url($site_url, $item->get_link());
642 _debug("title $entry_title", $debug_enabled);
643 _debug("link $entry_link", $debug_enabled);
645 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
647 $entry_content = $item->get_content();
648 if (!$entry_content) $entry_content = $item->get_description();
650 if ($_REQUEST["xdebug"] == 2) {
652 print $entry_content;
656 $entry_comments = $item->get_comments_url();
657 $entry_author = $item->get_author();
659 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
661 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
662 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
664 $num_comments = (int) $item->get_comments_count();
666 _debug("author $entry_author", $debug_enabled);
667 _debug("num_comments: $num_comments", $debug_enabled);
668 _debug("looking for tags...", $debug_enabled);
670 // parse <category> entries into tags
672 $additional_tags = array();
674 $additional_tags_src = $item->get_categories();
676 if (is_array($additional_tags_src)) {
677 foreach ($additional_tags_src as $tobj) {
678 array_push($additional_tags, $tobj);
682 $entry_tags = array_unique($additional_tags);
684 for ($i = 0; $i < count($entry_tags); $i++
)
685 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
687 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
689 _debug("done collecting data.", $debug_enabled);
691 $result = db_query("SELECT id, content_hash, lang FROM ttrss_entries
692 WHERE guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed'");
694 if (db_num_rows($result) != 0) {
695 $base_entry_id = db_fetch_result($result, 0, "id");
696 $entry_stored_hash = db_fetch_result($result, 0, "content_hash");
697 $article_labels = get_article_labels($base_entry_id, $owner_uid);
698 $entry_language = db_fetch_result($result, 0, "lang");
700 $base_entry_id = false;
701 $entry_stored_hash = "";
702 $article_labels = array();
703 $entry_language = "";
706 $article = array("owner_uid" => $owner_uid, // read only
707 "guid" => $entry_guid, // read only
708 "guid_hashed" => $entry_guid_hashed, // read only
709 "title" => $entry_title,
710 "content" => $entry_content,
711 "link" => $entry_link,
712 "labels" => $article_labels, // current limitation: can add labels to article, can't remove them
713 "tags" => $entry_tags,
714 "author" => $entry_author,
715 "force_catchup" => false, // ugly hack for the time being
716 "score_modifier" => 0, // no previous value, plugin should recalculate score modifier based on content if needed
717 "language" => $entry_language,
718 "feed" => array("id" => $feed,
719 "fetch_url" => $fetch_url,
720 "site_url" => $site_url)
723 $entry_plugin_data = "";
724 $entry_current_hash = calculate_article_hash($article, $pluginhost);
726 _debug("article hash: $entry_current_hash [stored=$entry_stored_hash]", $debug_enabled);
728 if ($entry_current_hash == $entry_stored_hash && !isset($_REQUEST["force_rehash"])) {
729 _debug("stored article seems up to date [IID: $base_entry_id], updating timestamp only", $debug_enabled);
731 // we keep encountering the entry in feeds, so we need to
732 // update date_updated column so that we don't get horrible
733 // dupes when the entry gets purged and reinserted again e.g.
734 // in the case of SLOW SLOW OMG SLOW updating feeds
736 $base_entry_id = db_fetch_result($result, 0, "id");
738 db_query("UPDATE ttrss_entries SET date_updated = NOW()
739 WHERE id = '$base_entry_id'");
741 // if we allow duplicate posts, we have to continue to
742 // create the user entries for this feed
743 if (!get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
748 _debug("hash differs, applying plugin filters:", $debug_enabled);
750 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
751 _debug("... " . get_class($plugin), $debug_enabled);
753 $start = microtime(true);
754 $article = $plugin->hook_article_filter($article);
756 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
758 $entry_plugin_data .= mb_strtolower(get_class($plugin)) . ",";
761 $entry_plugin_data = db_escape_string($entry_plugin_data);
763 _debug("plugin data: $entry_plugin_data", $debug_enabled);
765 // Workaround: 4-byte unicode requires utf8mb4 in MySQL. See https://tt-rss.org/forum/viewtopic.php?f=1&t=3377&p=20077#p20077
766 if (DB_TYPE
== "mysql") {
767 foreach ($article as $k => $v) {
769 // i guess we'll have to take the risk of 4byte unicode labels & tags here
770 if (!is_array($article[$k])) {
771 $article[$k] = preg_replace('/[\x{10000}-\x{10FFFF}]/u', "\xEF\xBF\xBD", $v);
776 /* Collect article tags here so we could filter by them: */
778 $article_filters = get_article_filters($filters, $article["title"],
779 $article["content"], $article["link"], 0, $article["author"],
782 if ($debug_enabled) {
783 _debug("article filters: ", $debug_enabled);
784 if (count($article_filters) != 0) {
785 print_r($article_filters);
789 $plugin_filter_names = find_article_filters($article_filters, "plugin");
790 $plugin_filter_actions = $pluginhost->get_filter_actions();
792 if (count($plugin_filter_names) > 0) {
793 _debug("applying plugin filter actions...", $debug_enabled);
795 foreach ($plugin_filter_names as $pfn) {
796 list($pfclass,$pfaction) = explode(":", $pfn["param"]);
798 if (isset($plugin_filter_actions[$pfclass])) {
799 $plugin = $pluginhost->get_plugin($pfclass);
801 _debug("... $pfclass: $pfaction", $debug_enabled);
804 $start = microtime(true);
805 $article = $plugin->hook_article_filter_action($article, $pfaction);
807 _debug("=== " . sprintf("%.4f (sec)", microtime(true) - $start), $debug_enabled);
809 _debug("??? $pfclass: plugin object not found.");
812 _debug("??? $pfclass: filter plugin not registered.");
817 $entry_tags = $article["tags"];
818 $entry_guid = db_escape_string($entry_guid);
819 $entry_title = db_escape_string($article["title"]);
820 $entry_author = db_escape_string($article["author"]);
821 $entry_link = db_escape_string($article["link"]);
822 $entry_content = $article["content"]; // escaped below
823 $entry_force_catchup = $article["force_catchup"];
824 $article_labels = $article["labels"];
825 $entry_score_modifier = (int) $article["score_modifier"];
826 $entry_language = db_escape_string($article["language"]);
828 if ($debug_enabled) {
829 _debug("article labels:", $debug_enabled);
830 print_r($article_labels);
833 _debug("force catchup: $entry_force_catchup");
835 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
836 cache_images($entry_content, $site_url, $debug_enabled);
838 $entry_content = db_escape_string($entry_content, false);
842 $result = db_query("SELECT id FROM ttrss_entries
843 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
845 if (db_num_rows($result) == 0) {
847 _debug("base guid [$entry_guid] not found", $debug_enabled);
849 // base post entry does not exist, create it
852 "INSERT INTO ttrss_entries
869 '$entry_guid_hashed',
871 '$entry_timestamp_fmt',
873 '$entry_current_hash',
876 '$date_feed_processed',
879 '$entry_plugin_data',
884 $base_entry_id = db_fetch_result($result, 0, "id");
887 // now it should exist, if not - bad luck then
889 $result = db_query("SELECT id FROM ttrss_entries
890 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
895 if (db_num_rows($result) == 1) {
897 _debug("base guid found, checking for user record", $debug_enabled);
899 $ref_id = db_fetch_result($result, 0, "id");
900 $entry_ref_id = $ref_id;
902 /* $stored_guid = db_fetch_result($result, 0, "guid");
903 if ($stored_guid != $entry_guid_hashed) {
904 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
906 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
910 // check for user post link to main table
912 // do we allow duplicate posts with same GUID in different feeds?
913 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
914 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
916 $dupcheck_qpart = "";
919 if (find_article_filter($article_filters, "filter")) {
920 //db_query("COMMIT"); // close transaction in progress
924 $score = calculate_article_score($article_filters) +
$entry_score_modifier;
926 _debug("initial score: $score [including plugin modifier: $entry_score_modifier]", $debug_enabled);
928 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
929 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
932 // if ($_REQUEST["xdebug"]) print "$query\n";
934 $result = db_query($query);
936 // okay it doesn't exist - create user entry
937 if (db_num_rows($result) == 0) {
939 _debug("user record not found, creating...", $debug_enabled);
941 if ($score >= -500 && !find_article_filter($article_filters, 'catchup') && !$entry_force_catchup) {
943 $last_read_qpart = 'NULL';
946 $last_read_qpart = 'NOW()';
949 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
955 if (find_article_filter($article_filters, 'publish')) {
958 $published = 'false';
963 /* if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
965 $result = db_query("SELECT COUNT(*) AS similar FROM
966 ttrss_entries,ttrss_user_entries
967 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
968 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
969 AND owner_uid = $owner_uid");
971 $ngram_similar = db_fetch_result($result, 0, "similar");
973 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
975 if ($ngram_similar > 0) {
980 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
981 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
984 "INSERT INTO ttrss_user_entries
985 (ref_id, owner_uid, feed_id, unread, last_read, marked,
986 published, score, tag_cache, label_cache, uuid,
987 last_marked, last_published)
988 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
989 $last_read_qpart, $marked, $published, '$score', '', '',
990 '', $last_marked, $last_published)");
992 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
993 $rss_link = get_self_url_prefix() .
994 "/public.php?op=rss&id=-2&key=" .
995 get_feed_access_key(-2, false, $owner_uid);
997 $p = new Publisher(PUBSUBHUBBUB_HUB
);
999 /* $pubsub_result = */ $p->publish_update($rss_link);
1003 "SELECT int_id FROM ttrss_user_entries WHERE
1004 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
1005 feed_id = '$feed' LIMIT 1");
1007 if (db_num_rows($result) == 1) {
1008 $entry_int_id = db_fetch_result($result, 0, "int_id");
1011 _debug("user record FOUND", $debug_enabled);
1013 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
1014 $entry_int_id = db_fetch_result($result, 0, "int_id");
1017 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
1019 if (DB_TYPE
== "pgsql") {
1020 $tsvector_combined = db_escape_string(mb_substr($entry_title . ' ' . strip_tags(str_replace('<', ' <', $entry_content)),
1023 $tsvector_qpart = "tsvector_combined = to_tsvector('$feed_language', '$tsvector_combined'),";
1026 $tsvector_qpart = "";
1029 db_query("UPDATE ttrss_entries
1030 SET title = '$entry_title',
1031 content = '$entry_content',
1032 content_hash = '$entry_current_hash',
1033 updated = '$entry_timestamp_fmt',
1035 num_comments = '$num_comments',
1036 plugin_data = '$entry_plugin_data',
1037 author = '$entry_author',
1038 lang = '$entry_language'
1039 WHERE id = '$ref_id'");
1042 db_query("UPDATE ttrss_user_entries
1043 SET score = '$score' WHERE ref_id = '$ref_id'");
1045 if ($mark_unread_on_update) {
1046 db_query("UPDATE ttrss_user_entries
1047 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1051 //db_query("COMMIT");
1053 _debug("assigning labels [other]...", $debug_enabled);
1055 foreach ($article_labels as $label) {
1056 label_add_article($entry_ref_id, $label[1], $owner_uid);
1059 _debug("assigning labels [filters]...", $debug_enabled);
1061 assign_article_to_label_filters($entry_ref_id, $article_filters,
1062 $owner_uid, $article_labels);
1064 _debug("looking for enclosures...", $debug_enabled);
1068 $enclosures = array();
1070 $encs = $item->get_enclosures();
1072 if (is_array($encs)) {
1073 foreach ($encs as $e) {
1075 $e->link
, $e->type
, $e->length
, $e->title
, $e->width
, $e->height
);
1076 array_push($enclosures, $e_item);
1080 if ($debug_enabled) {
1081 _debug("article enclosures:", $debug_enabled);
1082 print_r($enclosures);
1085 //db_query("BEGIN");
1088 // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'");
1090 foreach ($enclosures as $enc) {
1091 $enc_url = db_escape_string($enc[0]);
1092 $enc_type = db_escape_string($enc[1]);
1093 $enc_dur = db_escape_string($enc[2]);
1094 $enc_title = db_escape_string($enc[3]);
1095 $enc_width = intval($enc[4]);
1096 $enc_height = intval($enc[5]);
1098 $result = db_query("SELECT id FROM ttrss_enclosures
1099 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1101 if (db_num_rows($result) == 0) {
1102 db_query("INSERT INTO ttrss_enclosures
1103 (content_url, content_type, title, duration, post_id, width, height) VALUES
1104 ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)");
1108 //db_query("COMMIT");
1110 // check for manual tags (we have to do it here since they're loaded from filters)
1112 foreach ($article_filters as $f) {
1113 if ($f["type"] == "tag") {
1115 $manual_tags = trim_array(explode(",", $f["param"]));
1117 foreach ($manual_tags as $tag) {
1118 if (tag_is_valid($tag)) {
1119 array_push($entry_tags, $tag);
1127 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1128 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1130 $filtered_tags = array();
1131 $tags_to_cache = array();
1133 if ($entry_tags && is_array($entry_tags)) {
1134 foreach ($entry_tags as $tag) {
1135 if (array_search($tag, $boring_tags) === false) {
1136 array_push($filtered_tags, $tag);
1141 $filtered_tags = array_unique($filtered_tags);
1143 if ($debug_enabled) {
1144 _debug("filtered article tags:", $debug_enabled);
1145 print_r($filtered_tags);
1148 // Save article tags in the database
1150 if (count($filtered_tags) > 0) {
1152 //db_query("BEGIN");
1154 foreach ($filtered_tags as $tag) {
1156 $tag = sanitize_tag($tag);
1157 $tag = db_escape_string($tag);
1159 if (!tag_is_valid($tag)) continue;
1161 $result = db_query("SELECT id FROM ttrss_tags
1162 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1163 owner_uid = '$owner_uid' LIMIT 1");
1165 if ($result && db_num_rows($result) == 0) {
1167 db_query("INSERT INTO ttrss_tags
1168 (owner_uid,tag_name,post_int_id)
1169 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1172 array_push($tags_to_cache, $tag);
1175 /* update the cache */
1177 $tags_to_cache = array_unique($tags_to_cache);
1179 $tags_str = db_escape_string(join(",", $tags_to_cache));
1181 db_query("UPDATE ttrss_user_entries
1182 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1183 AND owner_uid = $owner_uid");
1185 //db_query("COMMIT");
1188 _debug("article processed", $debug_enabled);
1191 _debug("purging feed...", $debug_enabled);
1193 purge_feed($feed, 0, $debug_enabled);
1195 db_query("UPDATE ttrss_feeds
1196 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1198 // db_query("COMMIT");
1202 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1204 _debug("fetch error: $error_msg", $debug_enabled);
1206 if (count($rss->errors()) > 1) {
1207 foreach ($rss->errors() as $error) {
1213 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1214 last_updated = NOW() WHERE id = '$feed'");
1219 _debug("done", $debug_enabled);
1224 function cache_images($html, $site_url, $debug) {
1225 libxml_use_internal_errors(true);
1227 $charset_hack = '<head>
1228 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1231 $doc = new DOMDocument();
1232 $doc->loadHTML($charset_hack . $html);
1233 $xpath = new DOMXPath($doc);
1235 $entries = $xpath->query('(//img[@src])');
1237 foreach ($entries as $entry) {
1238 if ($entry->hasAttribute('src')) {
1239 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1241 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1243 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1245 if (!file_exists($local_filename)) {
1246 $file_content = fetch_file_contents($src);
1248 if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE
) {
1249 file_put_contents($local_filename, $file_content);
1252 touch($local_filename);
1258 function expire_error_log($debug) {
1259 if ($debug) _debug("Removing old error log entries...");
1261 if (DB_TYPE
== "pgsql") {
1262 db_query("DELETE FROM ttrss_error_log
1263 WHERE created_at < NOW() - INTERVAL '7 days'");
1265 db_query("DELETE FROM ttrss_error_log
1266 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1271 function expire_lock_files($debug) {
1272 //if ($debug) _debug("Removing old lock files...");
1276 if (is_writable(LOCK_DIRECTORY
)) {
1277 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1280 foreach ($files as $file) {
1281 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1289 if ($debug) _debug("Removed $num_deleted old lock files.");
1292 function expire_cached_files($debug) {
1293 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1294 $cache_dir = CACHE_DIR
. "/$dir";
1296 // if ($debug) _debug("Expiring $cache_dir");
1300 if (is_writable($cache_dir)) {
1301 $files = glob("$cache_dir/*");
1304 foreach ($files as $file) {
1305 if (time() - filemtime($file) > 86400*7) {
1314 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1319 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1320 * Returns the url query as associative array
1322 * @param string query
1323 * @return array params
1325 function convertUrlQuery($query) {
1326 $queryParts = explode('&', $query);
1330 foreach ($queryParts as $param) {
1331 $item = explode('=', $param);
1332 $params[$item[0]] = $item[1];
1338 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1341 foreach ($filters as $filter) {
1342 $match_any_rule = $filter["match_any_rule"];
1343 $inverse = $filter["inverse"];
1344 $filter_match = false;
1346 foreach ($filter["rules"] as $rule) {
1348 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1349 $rule_inverse = $rule["inverse"];
1354 switch ($rule["type"]) {
1356 $match = @preg_match
("/$reg_exp/i", $title);
1359 // we don't need to deal with multiline regexps
1360 $content = preg_replace("/[\r\n\t]/", "", $content);
1362 $match = @preg_match
("/$reg_exp/i", $content);
1365 // we don't need to deal with multiline regexps
1366 $content = preg_replace("/[\r\n\t]/", "", $content);
1368 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1371 $match = @preg_match
("/$reg_exp/i", $link);
1374 $match = @preg_match
("/$reg_exp/i", $author);
1377 foreach ($tags as $tag) {
1378 if (@preg_match
("/$reg_exp/i", $tag)) {
1386 if ($rule_inverse) $match = !$match;
1388 if ($match_any_rule) {
1390 $filter_match = true;
1394 $filter_match = $match;
1401 if ($inverse) $filter_match = !$filter_match;
1403 if ($filter_match) {
1404 foreach ($filter["actions"] AS $action) {
1405 array_push($matches, $action);
1407 // if Stop action encountered, perform no further processing
1408 if (isset($action["type"]) && $action["type"] == "stop") return $matches;
1416 function find_article_filter($filters, $filter_name) {
1417 foreach ($filters as $f) {
1418 if ($f["type"] == $filter_name) {
1425 function find_article_filters($filters, $filter_name) {
1428 foreach ($filters as $f) {
1429 if ($f["type"] == $filter_name) {
1430 array_push($results, $f);
1436 function calculate_article_score($filters) {
1439 foreach ($filters as $f) {
1440 if ($f["type"] == "score") {
1441 $score +
= $f["param"];
1447 function labels_contains_caption($labels, $caption) {
1448 foreach ($labels as $label) {
1449 if ($label[1] == $caption) {
1457 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1458 foreach ($filters as $f) {
1459 if ($f["type"] == "label") {
1460 if (!labels_contains_caption($article_labels, $f["param"])) {
1461 label_add_article($id, $f["param"], $owner_uid);
1467 function make_guid_from_title($title) {
1468 return preg_replace("/[ \"\',.:;]/", "-",
1469 mb_strtolower(strip_tags($title), 'utf-8'));
1472 /* function verify_feed_xml($feed_data) {
1473 libxml_use_internal_errors(true);
1474 $doc = new DOMDocument();
1475 $doc->loadXML($feed_data);
1476 $error = libxml_get_last_error();
1477 libxml_clear_errors();
1481 function cleanup_counters_cache($debug) {
1482 $result = db_query("DELETE FROM ttrss_counters_cache
1483 WHERE feed_id > 0 AND
1484 (SELECT COUNT(id) FROM ttrss_feeds WHERE
1486 ttrss_counters_cache.owner_uid = ttrss_feeds.owner_uid) = 0");
1487 $frows = db_affected_rows($result);
1489 $result = db_query("DELETE FROM ttrss_cat_counters_cache
1490 WHERE feed_id > 0 AND
1491 (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
1493 ttrss_cat_counters_cache.owner_uid = ttrss_feed_categories.owner_uid) = 0");
1494 $crows = db_affected_rows($result);
1496 _debug("Removed $frows (feeds) $crows (cats) orphaned counter cache entries.");
1499 function housekeeping_common($debug) {
1500 expire_cached_files($debug);
1501 expire_lock_files($debug);
1502 expire_error_log($debug);
1504 $count = update_feedbrowser_cache();
1505 _debug("Feedbrowser updated, $count feeds processed.");
1507 purge_orphans( true);
1508 cleanup_counters_cache($debug);
1509 $rc = cleanup_tags( 14, 50000);
1511 _debug("Cleaned $rc cached tags.");
1513 PluginHost
::getInstance()->run_hooks(PluginHost
::HOOK_HOUSE_KEEPING
, "hook_house_keeping", "");