2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
6 function update_feedbrowser_cache() {
8 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
16 db_query("DELETE FROM ttrss_feedbrowser_cache");
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($line["subscribers"]);
22 $feed_url = db_escape_string($line["feed_url"]);
23 $title = db_escape_string($line["title"]);
24 $site_url = db_escape_string($line["site_url"]);
26 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
29 if (db_num_rows($tmp_result) == 0) {
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT
, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
63 $schema_version = get_schema_version();
65 if ($schema_version != SCHEMA_VERSION
) {
66 die("Schema version is wrong, please upgrade the database.\n");
69 define('PREFS_NO_CACHE', true);
71 // Test if the user has loggued in recently. If not, it does not update its feeds.
72 if (!SINGLE_USER_MODE
&& DAEMON_UPDATE_LOGIN_LIMIT
> 0) {
73 if (DB_TYPE
== "pgsql") {
74 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT
." days'";
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT
." DAY)";
79 $login_thresh_qpart = "";
82 // Test if the feed need a update (update interval exceded).
83 if (DB_TYPE
== "pgsql") {
84 $update_limit_qpart = "AND ((
85 ttrss_feeds.update_interval = 0
86 AND ttrss_user_prefs.value != '-1'
87 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
89 ttrss_feeds.update_interval > 0
90 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
91 ) OR ttrss_feeds.last_updated IS NULL
92 OR last_updated = '1970-01-01 00:00:00')";
94 $update_limit_qpart = "AND ((
95 ttrss_feeds.update_interval = 0
96 AND ttrss_user_prefs.value != '-1'
97 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
99 ttrss_feeds.update_interval > 0
100 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
101 ) OR ttrss_feeds.last_updated IS NULL
102 OR last_updated = '1970-01-01 00:00:00')";
105 // Test if feed is currently being updated by another process.
106 if (DB_TYPE
== "pgsql") {
107 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
112 // Test if there is a limit to number of updated feeds
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
116 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
118 ttrss_feeds, ttrss_users, ttrss_user_prefs
120 ttrss_feeds.owner_uid = ttrss_users.id
121 AND ttrss_users.id = ttrss_user_prefs.owner_uid
122 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
123 $login_thresh_qpart $update_limit_qpart
124 $updstart_thresh_qpart
125 ORDER BY last_updated $query_limit";
127 // We search for feed needing update.
128 $result = db_query($query);
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
132 // Here is a little cache magic in order to minimize risk of double feed updates.
133 $feeds_to_update = array();
134 while ($line = db_fetch_assoc($result)) {
135 array_push($feeds_to_update, db_escape_string($line['feed_url']));
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
159 //update_rss_feed($line["id"], true);
161 // since we have the data cached, we can deal with other feeds with the same url
163 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
164 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
165 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
166 ttrss_users.id = ttrss_user_prefs.owner_uid AND
167 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
168 feed_url = '".db_escape_string($feed)."' AND
169 (ttrss_feeds.update_interval > 0 OR
170 ttrss_user_prefs.value != '-1')
172 ORDER BY ttrss_feeds.id $query_limit");
174 if (db_num_rows($tmp_result) > 0) {
175 while ($tline = db_fetch_assoc($tmp_result)) {
176 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
177 update_rss_feed($tline["id"], true);
183 require_once "digest.php";
185 // Send feed digests by email if needed.
186 send_headlines_digests($debug);
190 } // function update_daemon_common
192 // ignore_daemon is not used
193 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
194 $override_url = false) {
196 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
198 _debug("start", $debug_enabled);
200 $result = db_query("SELECT id,update_interval,auth_login,
201 feed_url,auth_pass,cache_images,last_updated,
202 mark_unread_on_update, owner_uid,
203 pubsub_state, auth_pass_encrypted,
204 (SELECT max(date_entered) FROM
205 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
206 FROM ttrss_feeds WHERE id = '$feed'");
208 if (db_num_rows($result) == 0) {
209 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
213 $last_updated = db_fetch_result($result, 0, "last_updated");
214 $last_article_timestamp = @strtotime
(db_fetch_result($result, 0, "last_article_timestamp"));
216 if (defined('_DISABLE_HTTP_304'))
217 $last_article_timestamp = 0;
219 $owner_uid = db_fetch_result($result, 0, "owner_uid");
220 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
221 0, "mark_unread_on_update"));
222 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
223 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
224 0, "auth_pass_encrypted"));
226 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
227 WHERE id = '$feed'");
229 $auth_login = db_fetch_result($result, 0, "auth_login");
230 $auth_pass = db_fetch_result($result, 0, "auth_pass");
232 if ($auth_pass_encrypted) {
233 require_once "crypt.php";
234 $auth_pass = decrypt_string($auth_pass);
237 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
238 $fetch_url = db_fetch_result($result, 0, "feed_url");
240 $feed = db_escape_string($feed);
242 if ($override_url) $fetch_url = $override_url;
244 $date_feed_processed = date('Y-m-d H:i');
246 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".xml";
248 // Ignore cache if new feed or manual update.
249 $cache_age = ($no_cache ||
is_null($last_updated) ||
strpos($last_updated, '1970-01-01') === 0) ?
30 : get_feed_update_interval($feed) * 60;
251 _debug("cache filename: $cache_filename exists: " . file_exists($cache_filename), $debug_enabled);
252 _debug("cache age: $cache_age; no cache: $no_cache", $debug_enabled);
254 $cached_feed_data_hash = false;
258 $cache_timestamp = file_exists($cache_filename) ?
filemtime($cache_filename) : 0;
260 $force_refetch = isset($_REQUEST["force_refetch"]);
262 if (file_exists($cache_filename) &&
263 is_readable($cache_filename) &&
264 !$auth_login && !$auth_pass &&
265 filemtime($cache_filename) > time() - $cache_age) {
267 _debug("using local cache.", $debug_enabled);
269 @$feed_data = file_get_contents($cache_filename);
272 $rss_hash = sha1($feed_data);
276 _debug("local cache will not be used for this feed", $debug_enabled);
282 _debug("fetching [$fetch_url]...", $debug_enabled);
283 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
285 $feed_data = fetch_file_contents($fetch_url, false,
286 $auth_login, $auth_pass, false,
287 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
288 $force_refetch ?
0 : $last_article_timestamp);
290 global $fetch_curl_used;
292 if (!$fetch_curl_used) {
293 $tmp = @gzdecode
($feed_data);
295 if ($tmp) $feed_data = $tmp;
298 $feed_data = trim($feed_data);
300 _debug("fetch done.", $debug_enabled);
303 $error = verify_feed_xml($feed_data);
306 _debug("error verifying XML, code: " . $error->code
, $debug_enabled);
308 if ($error->code
== 26) {
309 _debug("got error 26, trying to decode entities...", $debug_enabled);
311 $feed_data = html_entity_decode($feed_data, ENT_COMPAT
, 'UTF-8');
313 $error = verify_feed_xml($feed_data);
315 if ($error) $feed_data = '';
322 global $fetch_last_error;
323 global $fetch_last_error_code;
325 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
330 if ($fetch_last_error_code != 304) {
331 $error_escaped = db_escape_string($fetch_last_error);
333 _debug("source claims data not modified, nothing to do.", $debug_enabled);
337 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
338 last_updated = NOW() WHERE id = '$feed'");
344 $pluginhost = new PluginHost();
345 $pluginhost->set_debug($debug_enabled, $debug_enabled);
346 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
348 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
349 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
350 $pluginhost->load_data();
352 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
353 $feed_data = $plugin->hook_feed_fetched($feed_data);
356 // set last update to now so if anything *simplepie* crashes later we won't be
357 // continuously failing on the same feed
358 db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
361 $rss = new FeedParser($feed_data);
367 $feed = db_escape_string($feed);
369 if (!$rss->error()) {
371 // cache data for later
372 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
373 $new_rss_hash = sha1($rss_data);
375 if ($new_rss_hash != $rss_hash && count($rss->get_items()) > 0 ) {
376 _debug("saving $cache_filename", $debug_enabled);
377 @file_put_contents
($cache_filename, $feed_data);
381 // We use local pluginhost here because we need to load different per-user feed plugins
382 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
384 _debug("processing feed data...", $debug_enabled);
386 // db_query("BEGIN");
388 if (DB_TYPE
== "pgsql") {
389 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
391 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
394 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
395 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
397 FROM ttrss_feeds WHERE id = '$feed'");
399 $registered_title = db_fetch_result($result, 0, "title");
400 $orig_site_url = db_fetch_result($result, 0, "site_url");
401 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
402 "favicon_needs_check"));
403 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
405 $owner_uid = db_fetch_result($result, 0, "owner_uid");
407 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
409 _debug("site_url: $site_url", $debug_enabled);
410 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
412 if ($favicon_needs_check ||
$force_refetch) {
414 /* terrible hack: if we crash on floicon shit here, we won't check
415 * the icon avgcolor again (unless the icon got updated) */
417 $favicon_file = ICONS_DIR
. "/$feed.ico";
418 $favicon_modified = @filemtime
($favicon_file);
420 _debug("checking favicon...", $debug_enabled);
422 check_feed_favicon($site_url, $feed, $link);
423 $favicon_modified_new = @filemtime
($favicon_file);
425 if ($favicon_modified_new > $favicon_modified)
426 $favicon_avg_color = '';
428 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
429 require_once "colors.php";
431 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
434 $favicon_color = db_escape_string(
435 calculate_avg_color($favicon_file));
437 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
438 } else if ($favicon_avg_color == 'fail') {
439 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
442 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
444 WHERE id = '$feed'");
447 if (!$registered_title ||
$registered_title == "[Unknown]") {
449 $feed_title = db_escape_string($rss->get_title());
452 _debug("registering title: $feed_title", $debug_enabled);
454 db_query("UPDATE ttrss_feeds SET
455 title = '$feed_title' WHERE id = '$feed'");
459 if ($site_url && $orig_site_url != $site_url) {
460 db_query("UPDATE ttrss_feeds SET
461 site_url = '$site_url' WHERE id = '$feed'");
464 _debug("loading filters & labels...", $debug_enabled);
466 $filters = load_filters($feed, $owner_uid);
467 $labels = get_all_labels($owner_uid);
469 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
471 $items = $rss->get_items();
473 if (!is_array($items)) {
474 _debug("no articles found.", $debug_enabled);
476 db_query("UPDATE ttrss_feeds
477 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
479 return; // no articles
482 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
484 _debug("checking for PUSH hub...", $debug_enabled);
486 $feed_hub_url = false;
488 $links = $rss->get_links('hub');
490 if ($links && is_array($links)) {
491 foreach ($links as $l) {
497 _debug("feed hub url: $feed_hub_url", $debug_enabled);
499 if ($feed_hub_url && function_exists('curl_init') &&
500 !ini_get("open_basedir")) {
502 require_once 'lib/pubsubhubbub/subscriber.php';
504 $callback_url = get_self_url_prefix() .
505 "/public.php?op=pubsub&id=$feed";
507 $s = new Subscriber($feed_hub_url, $callback_url);
509 $rc = $s->subscribe($fetch_url);
511 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
513 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
514 WHERE id = '$feed'");
518 _debug("processing articles...", $debug_enabled);
520 foreach ($items as $item) {
521 if ($_REQUEST['xdebug'] == 3) {
525 $entry_guid = $item->get_id();
526 if (!$entry_guid) $entry_guid = $item->get_link();
527 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
529 _debug("f_guid $entry_guid", $debug_enabled);
531 if (!$entry_guid) continue;
533 $entry_guid = "$owner_uid,$entry_guid";
535 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
537 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
539 $entry_timestamp = "";
541 $entry_timestamp = $item->get_date();
543 _debug("orig date: " . $item->get_date(), $debug_enabled);
545 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
546 $entry_timestamp = time();
547 $no_orig_date = 'true';
549 $no_orig_date = 'false';
552 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
554 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
556 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
557 // $entry_title = decode_numeric_entities($entry_title);
558 $entry_title = $item->get_title();
560 $entry_link = rewrite_relative_url($site_url, $item->get_link());
562 _debug("title $entry_title", $debug_enabled);
563 _debug("link $entry_link", $debug_enabled);
565 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
567 $entry_content = $item->get_content();
568 if (!$entry_content) $entry_content = $item->get_description();
570 if ($_REQUEST["xdebug"] == 2) {
572 print $entry_content;
576 $entry_comments = $item->get_comments_url();
577 $entry_author = $item->get_author();
579 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
581 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
582 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
584 $num_comments = (int) $item->get_comments_count();
586 _debug("author $entry_author", $debug_enabled);
587 _debug("num_comments: $num_comments", $debug_enabled);
588 _debug("looking for tags...", $debug_enabled);
590 // parse <category> entries into tags
592 $additional_tags = array();
594 $additional_tags_src = $item->get_categories();
596 if (is_array($additional_tags_src)) {
597 foreach ($additional_tags_src as $tobj) {
598 array_push($additional_tags, $tobj);
602 $entry_tags = array_unique($additional_tags);
604 for ($i = 0; $i < count($entry_tags); $i++
)
605 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
607 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
609 _debug("done collecting data.", $debug_enabled);
611 // TODO: less memory-hungry implementation
613 _debug("applying plugin filters..", $debug_enabled);
615 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
616 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
617 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
619 if (db_num_rows($result) != 0) {
620 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
621 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
622 "content" => db_fetch_result($result, 0, "content"),
623 "link" => db_fetch_result($result, 0, "link"),
624 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
625 "author" => db_fetch_result($result, 0, "author"));
627 $entry_plugin_data = "";
628 $stored_article = array();
631 $article = array("owner_uid" => $owner_uid, // read only
632 "guid" => $entry_guid, // read only
633 "title" => $entry_title,
634 "content" => $entry_content,
635 "link" => $entry_link,
636 "tags" => $entry_tags,
637 "plugin_data" => $entry_plugin_data,
638 "author" => $entry_author,
639 "stored" => $stored_article);
641 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
642 $article = $plugin->hook_article_filter($article);
645 $entry_tags = $article["tags"];
646 $entry_guid = db_escape_string($entry_guid);
647 $entry_title = db_escape_string($article["title"]);
648 $entry_author = db_escape_string($article["author"]);
649 $entry_link = db_escape_string($article["link"]);
650 $entry_plugin_data = db_escape_string($article["plugin_data"]);
651 $entry_content = $article["content"]; // escaped below
654 _debug("plugin data: $entry_plugin_data", $debug_enabled);
656 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
657 cache_images($entry_content, $site_url, $debug_enabled);
659 $entry_content = db_escape_string($entry_content, false);
661 $content_hash = "SHA1:" . sha1($entry_content);
665 $result = db_query("SELECT id FROM ttrss_entries
666 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
668 if (db_num_rows($result) == 0) {
670 _debug("base guid [$entry_guid] not found", $debug_enabled);
672 // base post entry does not exist, create it
675 "INSERT INTO ttrss_entries
692 '$entry_guid_hashed',
694 '$entry_timestamp_fmt',
700 '$date_feed_processed',
703 '$entry_plugin_data',
706 $article_labels = array();
709 // we keep encountering the entry in feeds, so we need to
710 // update date_updated column so that we don't get horrible
711 // dupes when the entry gets purged and reinserted again e.g.
712 // in the case of SLOW SLOW OMG SLOW updating feeds
714 $base_entry_id = db_fetch_result($result, 0, "id");
716 db_query("UPDATE ttrss_entries SET date_updated = NOW()
717 WHERE id = '$base_entry_id'");
719 $article_labels = get_article_labels($base_entry_id, $owner_uid);
722 // now it should exist, if not - bad luck then
724 $result = db_query("SELECT
725 id,content_hash,no_orig_date,title,plugin_data,guid,
726 ".SUBSTRING_FOR_DATE
."(date_updated,1,19) as date_updated,
727 ".SUBSTRING_FOR_DATE
."(updated,1,19) as updated,
731 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
736 if (db_num_rows($result) == 1) {
738 _debug("base guid found, checking for user record", $debug_enabled);
740 // this will be used below in update handler
741 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
742 $orig_title = db_fetch_result($result, 0, "title");
743 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
744 $orig_date_updated = strtotime(db_fetch_result($result,
746 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
748 $ref_id = db_fetch_result($result, 0, "id");
749 $entry_ref_id = $ref_id;
751 /* $stored_guid = db_fetch_result($result, 0, "guid");
752 if ($stored_guid != $entry_guid_hashed) {
753 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
755 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
759 // check for user post link to main table
761 // do we allow duplicate posts with same GUID in different feeds?
762 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
763 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
765 $dupcheck_qpart = "";
768 /* Collect article tags here so we could filter by them: */
770 $article_filters = get_article_filters($filters, $entry_title,
771 $entry_content, $entry_link, $entry_timestamp, $entry_author,
774 if ($debug_enabled) {
775 _debug("article filters: ", $debug_enabled);
776 if (count($article_filters) != 0) {
777 print_r($article_filters);
781 if (find_article_filter($article_filters, "filter")) {
782 db_query("COMMIT"); // close transaction in progress
786 $score = calculate_article_score($article_filters);
788 _debug("initial score: $score", $debug_enabled);
790 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
791 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
794 // if ($_REQUEST["xdebug"]) print "$query\n";
796 $result = db_query($query);
798 // okay it doesn't exist - create user entry
799 if (db_num_rows($result) == 0) {
801 _debug("user record not found, creating...", $debug_enabled);
803 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
805 $last_read_qpart = 'NULL';
808 $last_read_qpart = 'NOW()';
811 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
817 if (find_article_filter($article_filters, 'publish')) {
820 $published = 'false';
825 if (DB_TYPE
== "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
827 $result = db_query("SELECT COUNT(*) AS similar FROM
828 ttrss_entries,ttrss_user_entries
829 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
830 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD
."
831 AND owner_uid = $owner_uid");
833 $ngram_similar = db_fetch_result($result, 0, "similar");
835 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
837 if ($ngram_similar > 0) {
842 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
843 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
846 "INSERT INTO ttrss_user_entries
847 (ref_id, owner_uid, feed_id, unread, last_read, marked,
848 published, score, tag_cache, label_cache, uuid,
849 last_marked, last_published)
850 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
851 $last_read_qpart, $marked, $published, '$score', '', '',
852 '', $last_marked, $last_published)");
854 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
855 $rss_link = get_self_url_prefix() .
856 "/public.php?op=rss&id=-2&key=" .
857 get_feed_access_key(-2, false, $owner_uid);
859 $p = new Publisher(PUBSUBHUBBUB_HUB
);
861 $pubsub_result = $p->publish_update($rss_link);
865 "SELECT int_id FROM ttrss_user_entries WHERE
866 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
867 feed_id = '$feed' LIMIT 1");
869 if (db_num_rows($result) == 1) {
870 $entry_int_id = db_fetch_result($result, 0, "int_id");
873 _debug("user record FOUND", $debug_enabled);
875 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
876 $entry_int_id = db_fetch_result($result, 0, "int_id");
879 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
881 $post_needs_update = false;
882 $update_insignificant = false;
884 if ($orig_num_comments != $num_comments) {
885 $post_needs_update = true;
886 $update_insignificant = true;
889 if ($entry_plugin_data != $orig_plugin_data) {
890 $post_needs_update = true;
891 $update_insignificant = true;
894 if ($content_hash != $orig_content_hash) {
895 $post_needs_update = true;
896 $update_insignificant = false;
899 if (db_escape_string($orig_title) != $entry_title) {
900 $post_needs_update = true;
901 $update_insignificant = false;
904 // if post needs update, update it and mark all user entries
905 // linking to this post as updated
906 if ($post_needs_update) {
908 if (defined('DAEMON_EXTENDED_DEBUG')) {
909 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
912 // print "<!-- post $orig_title needs update : $post_needs_update -->";
914 db_query("UPDATE ttrss_entries
915 SET title = '$entry_title', content = '$entry_content',
916 content_hash = '$content_hash',
917 updated = '$entry_timestamp_fmt',
918 num_comments = '$num_comments',
919 plugin_data = '$entry_plugin_data'
920 WHERE id = '$ref_id'");
922 if (!$update_insignificant) {
923 if ($mark_unread_on_update) {
924 db_query("UPDATE ttrss_user_entries
925 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
933 _debug("assigning labels...", $debug_enabled);
935 assign_article_to_label_filters($entry_ref_id, $article_filters,
936 $owner_uid, $article_labels);
938 _debug("looking for enclosures...", $debug_enabled);
942 $enclosures = array();
944 $encs = $item->get_enclosures();
946 if (is_array($encs)) {
947 foreach ($encs as $e) {
949 $e->link
, $e->type
, $e->length
);
950 array_push($enclosures, $e_item);
954 if ($debug_enabled) {
955 _debug("article enclosures:", $debug_enabled);
956 print_r($enclosures);
961 foreach ($enclosures as $enc) {
962 $enc_url = db_escape_string($enc[0]);
963 $enc_type = db_escape_string($enc[1]);
964 $enc_dur = db_escape_string($enc[2]);
966 $result = db_query("SELECT id FROM ttrss_enclosures
967 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
969 if (db_num_rows($result) == 0) {
970 db_query("INSERT INTO ttrss_enclosures
971 (content_url, content_type, title, duration, post_id) VALUES
972 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
978 // check for manual tags (we have to do it here since they're loaded from filters)
980 foreach ($article_filters as $f) {
981 if ($f["type"] == "tag") {
983 $manual_tags = trim_array(explode(",", $f["param"]));
985 foreach ($manual_tags as $tag) {
986 if (tag_is_valid($tag)) {
987 array_push($entry_tags, $tag);
995 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
996 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
998 $filtered_tags = array();
999 $tags_to_cache = array();
1001 if ($entry_tags && is_array($entry_tags)) {
1002 foreach ($entry_tags as $tag) {
1003 if (array_search($tag, $boring_tags) === false) {
1004 array_push($filtered_tags, $tag);
1009 $filtered_tags = array_unique($filtered_tags);
1011 if ($debug_enabled) {
1012 _debug("filtered article tags:", $debug_enabled);
1013 print_r($filtered_tags);
1016 // Save article tags in the database
1018 if (count($filtered_tags) > 0) {
1022 foreach ($filtered_tags as $tag) {
1024 $tag = sanitize_tag($tag);
1025 $tag = db_escape_string($tag);
1027 if (!tag_is_valid($tag)) continue;
1029 $result = db_query("SELECT id FROM ttrss_tags
1030 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1031 owner_uid = '$owner_uid' LIMIT 1");
1033 if ($result && db_num_rows($result) == 0) {
1035 db_query("INSERT INTO ttrss_tags
1036 (owner_uid,tag_name,post_int_id)
1037 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1040 array_push($tags_to_cache, $tag);
1043 /* update the cache */
1045 $tags_to_cache = array_unique($tags_to_cache);
1047 $tags_str = db_escape_string(join(",", $tags_to_cache));
1049 db_query("UPDATE ttrss_user_entries
1050 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1051 AND owner_uid = $owner_uid");
1056 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1057 _debug("auto-assigning labels...", $debug_enabled);
1059 foreach ($labels as $label) {
1060 $caption = preg_quote($label["caption"]);
1062 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1063 if (!labels_contains_caption($article_labels, $caption)) {
1064 label_add_article($entry_ref_id, $caption, $owner_uid);
1070 _debug("article processed", $debug_enabled);
1073 if (!$last_updated) {
1074 _debug("new feed, catching it up...", $debug_enabled);
1075 catchup_feed($feed, false, $owner_uid);
1078 _debug("purging feed...", $debug_enabled);
1080 purge_feed($feed, 0, $debug_enabled);
1082 db_query("UPDATE ttrss_feeds
1083 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1085 // db_query("COMMIT");
1089 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1091 _debug("error fetching feed: $error_msg", $debug_enabled);
1094 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1095 last_updated = NOW() WHERE id = '$feed'");
1100 _debug("done", $debug_enabled);
1103 function cache_images($html, $site_url, $debug) {
1104 $cache_dir = CACHE_DIR
. "/images";
1106 libxml_use_internal_errors(true);
1108 $charset_hack = '<head>
1109 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1112 $doc = new DOMDocument();
1113 $doc->loadHTML($charset_hack . $html);
1114 $xpath = new DOMXPath($doc);
1116 $entries = $xpath->query('(//img[@src])');
1118 foreach ($entries as $entry) {
1119 if ($entry->hasAttribute('src')) {
1120 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1122 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1124 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1126 if (!file_exists($local_filename)) {
1127 $file_content = fetch_file_contents($src);
1129 if ($file_content && strlen($file_content) > 1024) {
1130 file_put_contents($local_filename, $file_content);
1134 if (file_exists($local_filename)) {
1135 $entry->setAttribute('src', SELF_URL_PATH
. '/image.php?url=' .
1136 base64_encode($src));
1141 $node = $doc->getElementsByTagName('body')->item(0);
1143 return $doc->saveXML($node);
1146 function expire_error_log($debug) {
1147 if ($debug) _debug("Removing old error log entries...");
1149 if (DB_TYPE
== "pgsql") {
1150 db_query("DELETE FROM ttrss_error_log
1151 WHERE created_at < NOW() - INTERVAL '7 days'");
1153 db_query("DELETE FROM ttrss_error_log
1154 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1159 function expire_lock_files($debug) {
1160 //if ($debug) _debug("Removing old lock files...");
1164 if (is_writable(LOCK_DIRECTORY
)) {
1165 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1168 foreach ($files as $file) {
1169 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1177 if ($debug) _debug("Removed $num_deleted old lock files.");
1180 function expire_cached_files($debug) {
1181 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1182 $cache_dir = CACHE_DIR
. "/$dir";
1184 // if ($debug) _debug("Expiring $cache_dir");
1188 if (is_writable($cache_dir)) {
1189 $files = glob("$cache_dir/*");
1192 foreach ($files as $file) {
1193 if (time() - filemtime($file) > 86400*7) {
1202 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1207 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1208 * Returns the url query as associative array
1210 * @param string query
1211 * @return array params
1213 function convertUrlQuery($query) {
1214 $queryParts = explode('&', $query);
1218 foreach ($queryParts as $param) {
1219 $item = explode('=', $param);
1220 $params[$item[0]] = $item[1];
1226 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1229 foreach ($filters as $filter) {
1230 $match_any_rule = $filter["match_any_rule"];
1231 $inverse = $filter["inverse"];
1232 $filter_match = false;
1234 foreach ($filter["rules"] as $rule) {
1236 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1237 $rule_inverse = $rule["inverse"];
1242 switch ($rule["type"]) {
1244 $match = @preg_match
("/$reg_exp/i", $title);
1247 // we don't need to deal with multiline regexps
1248 $content = preg_replace("/[\r\n\t]/", "", $content);
1250 $match = @preg_match
("/$reg_exp/i", $content);
1253 // we don't need to deal with multiline regexps
1254 $content = preg_replace("/[\r\n\t]/", "", $content);
1256 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1259 $match = @preg_match
("/$reg_exp/i", $link);
1262 $match = @preg_match
("/$reg_exp/i", $author);
1265 foreach ($tags as $tag) {
1266 if (@preg_match
("/$reg_exp/i", $tag)) {
1274 if ($rule_inverse) $match = !$match;
1276 if ($match_any_rule) {
1278 $filter_match = true;
1282 $filter_match = $match;
1289 if ($inverse) $filter_match = !$filter_match;
1291 if ($filter_match) {
1292 foreach ($filter["actions"] AS $action) {
1293 array_push($matches, $action);
1295 // if Stop action encountered, perform no further processing
1296 if ($action["type"] == "stop") return $matches;
1304 function find_article_filter($filters, $filter_name) {
1305 foreach ($filters as $f) {
1306 if ($f["type"] == $filter_name) {
1313 function find_article_filters($filters, $filter_name) {
1316 foreach ($filters as $f) {
1317 if ($f["type"] == $filter_name) {
1318 array_push($results, $f);
1324 function calculate_article_score($filters) {
1327 foreach ($filters as $f) {
1328 if ($f["type"] == "score") {
1329 $score +
= $f["param"];
1335 function labels_contains_caption($labels, $caption) {
1336 foreach ($labels as $label) {
1337 if ($label[1] == $caption) {
1345 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1346 foreach ($filters as $f) {
1347 if ($f["type"] == "label") {
1348 if (!labels_contains_caption($article_labels, $f["param"])) {
1349 label_add_article($id, $f["param"], $owner_uid);
1355 function make_guid_from_title($title) {
1356 return preg_replace("/[ \"\',.:;]/", "-",
1357 mb_strtolower(strip_tags($title), 'utf-8'));
1360 function verify_feed_xml($feed_data) {
1361 libxml_use_internal_errors(true);
1362 $doc = new DOMDocument();
1363 $doc->loadXML($feed_data);
1364 $error = libxml_get_last_error();
1365 libxml_clear_errors();
1369 function housekeeping_common($debug) {
1370 expire_cached_files($debug);
1371 expire_lock_files($debug);
1372 expire_error_log($debug);
1374 $count = update_feedbrowser_cache();
1375 _debug("Feedbrowser updated, $count feeds processed.");
1377 purge_orphans( true);
1378 $rc = cleanup_tags( 14, 50000);
1380 _debug("Cleaned $rc cached tags.");