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 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated
164 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
165 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
166 ttrss_users.id = ttrss_user_prefs.owner_uid AND
167 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
168 feed_url = '".db_escape_string($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"] . ", " . $tline["id"]);
177 update_rss_feed($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 require_once "lib/simplepie/simplepie.inc";
198 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') ||
$_REQUEST['xdebug'];
200 _debug("start", $debug_enabled);
202 $result = db_query("SELECT id,update_interval,auth_login,
203 feed_url,auth_pass,cache_images,last_updated,
204 mark_unread_on_update, owner_uid,
205 pubsub_state, auth_pass_encrypted
206 FROM ttrss_feeds WHERE id = '$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 $owner_uid = db_fetch_result($result, 0, "owner_uid");
215 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
216 0, "mark_unread_on_update"));
217 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
218 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
219 0, "auth_pass_encrypted"));
221 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
222 WHERE id = '$feed'");
224 $auth_login = db_fetch_result($result, 0, "auth_login");
225 $auth_pass = db_fetch_result($result, 0, "auth_pass");
227 if ($auth_pass_encrypted) {
228 require_once "crypt.php";
229 $auth_pass = decrypt_string($auth_pass);
232 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
233 $fetch_url = db_fetch_result($result, 0, "feed_url");
235 $feed = db_escape_string($feed);
237 if ($override_url) $fetch_url = $override_url;
239 $date_feed_processed = date('Y-m-d H:i');
241 $cache_filename = CACHE_DIR
. "/simplepie/" . sha1($fetch_url) . ".feed";
243 // Ignore cache if new feed or manual update.
244 $cache_age = ($no_cache ||
is_null($last_updated) ||
strpos($last_updated, '1970-01-01') === 0) ?
30 : get_feed_update_interval($feed) * 60;
246 _debug("cache filename: $cache_filename exists: " . file_exists($cache_filename), $debug_enabled);
247 _debug("cache age: $cache_age; no cache: $no_cache", $debug_enabled);
249 $cached_feed_data_hash = false;
253 $cache_timestamp = file_exists($cache_filename) ?
filemtime($cache_filename) : 0;
254 $last_updated_timestamp = strtotime($last_updated);
256 $force_refetch = isset($_REQUEST["force_refetch"]);
258 if (file_exists($cache_filename) &&
259 is_readable($cache_filename) &&
260 !$auth_login && !$auth_pass &&
261 filemtime($cache_filename) > time() - $cache_age) {
263 _debug("using local cache.", $debug_enabled);
265 if ($cache_timestamp > $last_updated_timestamp) {
266 @$rss_data = file_get_contents($cache_filename);
269 $rss_hash = sha1($rss_data);
270 @$rss = unserialize($rss_data);
272 } else if (!$force_refetch) {
273 _debug("local cache valid and older than last_updated, nothing to do.", $debug_enabled);
277 _debug("local cache will not be used for this feed", $debug_enabled);
278 $cache_timestamp = 0;
284 _debug("fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)", $debug_enabled);
286 $feed_data = fetch_file_contents($fetch_url, false,
287 $auth_login, $auth_pass, false,
288 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT
: FEED_FETCH_TIMEOUT
,
289 $force_refetch ?
0 : max($last_updated_timestamp, $cache_timestamp));
291 global $fetch_curl_used;
293 if (!$fetch_curl_used) {
294 $tmp = @gzdecode
($feed_data);
296 if ($tmp) $feed_data = $tmp;
299 $feed_data = trim($feed_data);
301 _debug("fetch done.", $debug_enabled);
304 $error = verify_feed_xml($feed_data);
307 _debug("error verifying XML, code: " . $error->code
, $debug_enabled);
309 if ($error->code
== 26) {
310 _debug("got error 26, trying to decode entities...", $debug_enabled);
312 $feed_data = html_entity_decode($feed_data, ENT_COMPAT
, 'UTF-8');
314 $error = verify_feed_xml($feed_data);
316 if ($error) $feed_data = '';
323 global $fetch_last_error;
324 global $fetch_last_error_code;
326 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
331 if ($fetch_last_error_code != 304) {
332 $error_escaped = db_escape_string($fetch_last_error);
334 _debug("source claims data not modified, nothing to do.", $debug_enabled);
338 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
339 last_updated = NOW() WHERE id = '$feed'");
345 $pluginhost = new PluginHost();
346 $pluginhost->set_debug($debug_enabled, $debug_enabled);
347 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
349 $pluginhost->load(PLUGINS
, PluginHost
::KIND_ALL
);
350 $pluginhost->load($user_plugins, PluginHost
::KIND_USER
, $owner_uid);
351 $pluginhost->load_data();
353 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_FEED_FETCHED
) as $plugin) {
354 $feed_data = $plugin->hook_feed_fetched($feed_data);
358 $rss = new SimplePie();
359 $rss->set_sanitize_class("SanitizeDummy");
360 // simplepie ignores the above and creates default sanitizer anyway,
361 // so let's override it...
362 $rss->sanitize
= new SanitizeDummy();
363 $rss->set_output_encoding('UTF-8');
364 $rss->set_raw_data($feed_data);
365 $rss->enable_cache(false);
372 $feed = db_escape_string($feed);
374 if (!$rss->error()) {
376 // cache data for later
377 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR
. "/simplepie")) {
378 $rss_data = serialize($rss);
379 $new_rss_hash = sha1($rss_data);
381 if ($new_rss_hash != $rss_hash) {
382 _debug("saving $cache_filename", $debug_enabled);
383 @file_put_contents
($cache_filename, serialize($rss));
387 // We use local pluginhost here because we need to load different per-user feed plugins
388 $pluginhost->run_hooks(PluginHost
::HOOK_FEED_PARSED
, "hook_feed_parsed", $rss);
390 _debug("processing feed data...", $debug_enabled);
392 // db_query("BEGIN");
394 if (DB_TYPE
== "pgsql") {
395 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
397 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
400 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
401 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
403 FROM ttrss_feeds WHERE id = '$feed'");
405 $registered_title = db_fetch_result($result, 0, "title");
406 $orig_site_url = db_fetch_result($result, 0, "site_url");
407 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
408 "favicon_needs_check"));
409 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
411 $owner_uid = db_fetch_result($result, 0, "owner_uid");
413 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
415 if ($favicon_needs_check ||
$force_refetch) {
417 /* terrible hack: if we crash on floicon shit here, we won't check
418 * the icon avgcolor again (unless the icon got updated) */
420 $favicon_file = ICONS_DIR
. "/$feed.ico";
421 $favicon_modified = @filemtime
($favicon_file);
423 _debug("checking favicon...", $debug_enabled);
425 check_feed_favicon($site_url, $feed, $link);
426 $favicon_modified_new = @filemtime
($favicon_file);
428 if ($favicon_modified_new > $favicon_modified)
429 $favicon_avg_color = '';
431 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
432 require_once "colors.php";
434 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
437 $favicon_color = db_escape_string(
438 calculate_avg_color($favicon_file));
440 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
441 } else if ($favicon_avg_color == 'fail') {
442 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
445 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
447 WHERE id = '$feed'");
450 if (!$registered_title ||
$registered_title == "[Unknown]") {
452 $feed_title = db_escape_string($rss->get_title());
454 _debug("registering title: $feed_title", $debug_enabled);
456 db_query("UPDATE ttrss_feeds SET
457 title = '$feed_title' WHERE id = '$feed'");
460 if ($site_url && $orig_site_url != $site_url) {
461 db_query("UPDATE ttrss_feeds SET
462 site_url = '$site_url' WHERE id = '$feed'");
465 _debug("loading filters & labels...", $debug_enabled);
467 $filters = load_filters($feed, $owner_uid);
468 $labels = get_all_labels($owner_uid);
470 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
472 $items = $rss->get_items();
474 if (!is_array($items)) {
475 _debug("no articles found.", $debug_enabled);
477 db_query("UPDATE ttrss_feeds
478 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
480 return; // no articles
483 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED
) {
485 _debug("checking for PUSH hub...", $debug_enabled);
487 $feed_hub_url = false;
489 $links = $rss->get_links('hub');
491 if ($links && is_array($links)) {
492 foreach ($links as $l) {
498 _debug("feed hub url: $feed_hub_url", $debug_enabled);
500 if ($feed_hub_url && function_exists('curl_init') &&
501 !ini_get("open_basedir")) {
503 require_once 'lib/pubsubhubbub/subscriber.php';
505 $callback_url = get_self_url_prefix() .
506 "/public.php?op=pubsub&id=$feed";
508 $s = new Subscriber($feed_hub_url, $callback_url);
510 $rc = $s->subscribe($fetch_url);
512 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
514 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
515 WHERE id = '$feed'");
519 _debug("processing articles...", $debug_enabled);
521 foreach ($items as $item) {
522 if ($_REQUEST['xdebug'] == 3) {
526 $entry_guid = $item->get_id();
527 if (!$entry_guid) $entry_guid = $item->get_link();
528 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
530 if (!$entry_guid) continue;
532 $entry_guid = "$owner_uid,$entry_guid";
534 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
536 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
538 $entry_timestamp = "";
540 $entry_timestamp = strtotime($item->get_date());
542 if ($entry_timestamp == -1 ||
!$entry_timestamp ||
$entry_timestamp > time()) {
543 $entry_timestamp = time();
544 $no_orig_date = 'true';
546 $no_orig_date = 'false';
549 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
551 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
553 $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT
, 'UTF-8');
555 $entry_link = rewrite_relative_url($site_url, $item->get_link());
557 _debug("title $entry_title", $debug_enabled);
558 _debug("link $entry_link", $debug_enabled);
560 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
562 $entry_content = $item->get_content();
563 if (!$entry_content) $entry_content = $item->get_description();
565 if ($_REQUEST["xdebug"] == 2) {
567 print $entry_content;
571 $entry_comments = $item->data
["comments"];
573 if ($item->get_author()) {
574 $entry_author_item = $item->get_author();
575 $entry_author = $entry_author_item->get_name();
576 if (!$entry_author) $entry_author = $entry_author_item->get_email();
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 = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
586 if (is_array($num_comments) && is_array($num_comments[0])) {
587 $num_comments = (int) $num_comments[0]["data"];
592 _debug("author $entry_author", $debug_enabled);
593 _debug("num_comments: $num_comments", $debug_enabled);
594 _debug("looking for tags [1]...", $debug_enabled);
596 // parse <category> entries into tags
598 $additional_tags = array();
600 $additional_tags_src = $item->get_categories();
602 if (is_array($additional_tags_src)) {
603 foreach ($additional_tags_src as $tobj) {
604 array_push($additional_tags, $tobj->get_term());
608 _debug("category tags:", $debug_enabled);
609 _debug("looking for tags [2]...", $debug_enabled);
611 $entry_tags = array_unique($additional_tags);
613 for ($i = 0; $i < count($entry_tags); $i++
)
614 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
616 _debug("done collecting data.", $debug_enabled);
618 // TODO: less memory-hungry implementation
620 _debug("applying plugin filters..", $debug_enabled);
622 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
623 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
624 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
626 if (db_num_rows($result) != 0) {
627 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
628 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
629 "content" => db_fetch_result($result, 0, "content"),
630 "link" => db_fetch_result($result, 0, "link"),
631 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
632 "author" => db_fetch_result($result, 0, "author"));
634 $entry_plugin_data = "";
635 $stored_article = array();
638 $article = array("owner_uid" => $owner_uid, // read only
639 "guid" => $entry_guid, // read only
640 "title" => $entry_title,
641 "content" => $entry_content,
642 "link" => $entry_link,
643 "tags" => $entry_tags,
644 "plugin_data" => $entry_plugin_data,
645 "author" => $entry_author,
646 "stored" => $stored_article);
648 foreach ($pluginhost->get_hooks(PluginHost
::HOOK_ARTICLE_FILTER
) as $plugin) {
649 $article = $plugin->hook_article_filter($article);
652 $entry_tags = $article["tags"];
653 $entry_guid = db_escape_string($entry_guid);
654 $entry_title = db_escape_string($article["title"]);
655 $entry_author = db_escape_string($article["author"]);
656 $entry_link = db_escape_string($article["link"]);
657 $entry_plugin_data = db_escape_string($article["plugin_data"]);
658 $entry_content = $article["content"]; // escaped below
661 _debug("plugin data: $entry_plugin_data", $debug_enabled);
663 if ($cache_images && is_writable(CACHE_DIR
. '/images'))
664 cache_images($entry_content, $site_url, $debug_enabled);
666 $entry_content = db_escape_string($entry_content, false);
668 $content_hash = "SHA1:" . sha1($entry_content);
672 $result = db_query("SELECT id FROM ttrss_entries
673 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
675 if (db_num_rows($result) == 0) {
677 _debug("base guid [$entry_guid] not found", $debug_enabled);
679 // base post entry does not exist, create it
682 "INSERT INTO ttrss_entries
699 '$entry_guid_hashed',
701 '$entry_timestamp_fmt',
707 '$date_feed_processed',
710 '$entry_plugin_data',
713 $article_labels = array();
716 // we keep encountering the entry in feeds, so we need to
717 // update date_updated column so that we don't get horrible
718 // dupes when the entry gets purged and reinserted again e.g.
719 // in the case of SLOW SLOW OMG SLOW updating feeds
721 $base_entry_id = db_fetch_result($result, 0, "id");
723 db_query("UPDATE ttrss_entries SET date_updated = NOW()
724 WHERE id = '$base_entry_id'");
726 $article_labels = get_article_labels($base_entry_id, $owner_uid);
729 // now it should exist, if not - bad luck then
731 $result = db_query("SELECT
732 id,content_hash,no_orig_date,title,plugin_data,guid,
733 ".SUBSTRING_FOR_DATE
."(date_updated,1,19) as date_updated,
734 ".SUBSTRING_FOR_DATE
."(updated,1,19) as updated,
738 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
743 if (db_num_rows($result) == 1) {
745 _debug("base guid found, checking for user record", $debug_enabled);
747 // this will be used below in update handler
748 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
749 $orig_title = db_fetch_result($result, 0, "title");
750 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
751 $orig_date_updated = strtotime(db_fetch_result($result,
753 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
755 $ref_id = db_fetch_result($result, 0, "id");
756 $entry_ref_id = $ref_id;
758 /* $stored_guid = db_fetch_result($result, 0, "guid");
759 if ($stored_guid != $entry_guid_hashed) {
760 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
762 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
766 // check for user post link to main table
768 // do we allow duplicate posts with same GUID in different feeds?
769 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
770 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
772 $dupcheck_qpart = "";
775 /* Collect article tags here so we could filter by them: */
777 $article_filters = get_article_filters($filters, $entry_title,
778 $entry_content, $entry_link, $entry_timestamp, $entry_author,
781 if ($debug_enabled) {
782 _debug("article filters: ", $debug_enabled);
783 if (count($article_filters) != 0) {
784 print_r($article_filters);
788 if (find_article_filter($article_filters, "filter")) {
789 db_query("COMMIT"); // close transaction in progress
793 $score = calculate_article_score($article_filters);
795 _debug("initial score: $score", $debug_enabled);
797 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
798 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
801 // if ($_REQUEST["xdebug"]) print "$query\n";
803 $result = db_query($query);
805 // okay it doesn't exist - create user entry
806 if (db_num_rows($result) == 0) {
808 _debug("user record not found, creating...", $debug_enabled);
810 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
812 $last_read_qpart = 'NULL';
815 $last_read_qpart = 'NOW()';
818 if (find_article_filter($article_filters, 'mark') ||
$score > 1000) {
824 if (find_article_filter($article_filters, 'publish')) {
827 $published = 'false';
832 if (DB_TYPE
== "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
834 $result = db_query("SELECT COUNT(*) AS similar FROM
835 ttrss_entries,ttrss_user_entries
836 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
837 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD
."
838 AND owner_uid = $owner_uid");
840 $ngram_similar = db_fetch_result($result, 0, "similar");
842 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
844 if ($ngram_similar > 0) {
849 $last_marked = ($marked == 'true') ?
'NOW()' : 'NULL';
850 $last_published = ($published == 'true') ?
'NOW()' : 'NULL';
853 "INSERT INTO ttrss_user_entries
854 (ref_id, owner_uid, feed_id, unread, last_read, marked,
855 published, score, tag_cache, label_cache, uuid,
856 last_marked, last_published)
857 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
858 $last_read_qpart, $marked, $published, '$score', '', '',
859 '', $last_marked, $last_published)");
861 if (PUBSUBHUBBUB_HUB
&& $published == 'true') {
862 $rss_link = get_self_url_prefix() .
863 "/public.php?op=rss&id=-2&key=" .
864 get_feed_access_key(-2, false, $owner_uid);
866 $p = new Publisher(PUBSUBHUBBUB_HUB
);
868 $pubsub_result = $p->publish_update($rss_link);
872 "SELECT int_id FROM ttrss_user_entries WHERE
873 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
874 feed_id = '$feed' LIMIT 1");
876 if (db_num_rows($result) == 1) {
877 $entry_int_id = db_fetch_result($result, 0, "int_id");
880 _debug("user record FOUND", $debug_enabled);
882 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
883 $entry_int_id = db_fetch_result($result, 0, "int_id");
886 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
888 $post_needs_update = false;
889 $update_insignificant = false;
891 if ($orig_num_comments != $num_comments) {
892 $post_needs_update = true;
893 $update_insignificant = true;
896 if ($entry_plugin_data != $orig_plugin_data) {
897 $post_needs_update = true;
898 $update_insignificant = true;
901 if ($content_hash != $orig_content_hash) {
902 $post_needs_update = true;
903 $update_insignificant = false;
906 if (db_escape_string($orig_title) != $entry_title) {
907 $post_needs_update = true;
908 $update_insignificant = false;
911 // if post needs update, update it and mark all user entries
912 // linking to this post as updated
913 if ($post_needs_update) {
915 if (defined('DAEMON_EXTENDED_DEBUG')) {
916 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
919 // print "<!-- post $orig_title needs update : $post_needs_update -->";
921 db_query("UPDATE ttrss_entries
922 SET title = '$entry_title', content = '$entry_content',
923 content_hash = '$content_hash',
924 updated = '$entry_timestamp_fmt',
925 num_comments = '$num_comments',
926 plugin_data = '$entry_plugin_data'
927 WHERE id = '$ref_id'");
929 if (!$update_insignificant) {
930 if ($mark_unread_on_update) {
931 db_query("UPDATE ttrss_user_entries
932 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
940 _debug("assigning labels...", $debug_enabled);
942 assign_article_to_label_filters($entry_ref_id, $article_filters,
943 $owner_uid, $article_labels);
945 _debug("looking for enclosures...", $debug_enabled);
949 $enclosures = array();
951 $encs = $item->get_enclosures();
953 if (is_array($encs)) {
954 foreach ($encs as $e) {
956 $e->link
, $e->type
, $e->length
);
957 array_push($enclosures, $e_item);
961 if ($debug_enabled) {
962 _debug("article enclosures:", $debug_enabled);
963 print_r($enclosures);
968 foreach ($enclosures as $enc) {
969 $enc_url = db_escape_string($enc[0]);
970 $enc_type = db_escape_string($enc[1]);
971 $enc_dur = db_escape_string($enc[2]);
973 $result = db_query("SELECT id FROM ttrss_enclosures
974 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
976 if (db_num_rows($result) == 0) {
977 db_query("INSERT INTO ttrss_enclosures
978 (content_url, content_type, title, duration, post_id) VALUES
979 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
985 // check for manual tags (we have to do it here since they're loaded from filters)
987 foreach ($article_filters as $f) {
988 if ($f["type"] == "tag") {
990 $manual_tags = trim_array(explode(",", $f["param"]));
992 foreach ($manual_tags as $tag) {
993 if (tag_is_valid($tag)) {
994 array_push($entry_tags, $tag);
1002 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
1003 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1005 $filtered_tags = array();
1006 $tags_to_cache = array();
1008 if ($entry_tags && is_array($entry_tags)) {
1009 foreach ($entry_tags as $tag) {
1010 if (array_search($tag, $boring_tags) === false) {
1011 array_push($filtered_tags, $tag);
1016 $filtered_tags = array_unique($filtered_tags);
1018 if ($debug_enabled) {
1019 _debug("filtered article tags:", $debug_enabled);
1020 print_r($filtered_tags);
1023 // Save article tags in the database
1025 if (count($filtered_tags) > 0) {
1029 foreach ($filtered_tags as $tag) {
1031 $tag = sanitize_tag($tag);
1032 $tag = db_escape_string($tag);
1034 if (!tag_is_valid($tag)) continue;
1036 $result = db_query("SELECT id FROM ttrss_tags
1037 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1038 owner_uid = '$owner_uid' LIMIT 1");
1040 if ($result && db_num_rows($result) == 0) {
1042 db_query("INSERT INTO ttrss_tags
1043 (owner_uid,tag_name,post_int_id)
1044 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1047 array_push($tags_to_cache, $tag);
1050 /* update the cache */
1052 $tags_to_cache = array_unique($tags_to_cache);
1054 $tags_str = db_escape_string(join(",", $tags_to_cache));
1056 db_query("UPDATE ttrss_user_entries
1057 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1058 AND owner_uid = $owner_uid");
1063 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1064 _debug("auto-assigning labels...", $debug_enabled);
1066 foreach ($labels as $label) {
1067 $caption = preg_quote($label["caption"]);
1069 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1070 if (!labels_contains_caption($article_labels, $caption)) {
1071 label_add_article($entry_ref_id, $caption, $owner_uid);
1077 _debug("article processed", $debug_enabled);
1080 if (!$last_updated) {
1081 _debug("new feed, catching it up...", $debug_enabled);
1082 catchup_feed($feed, false, $owner_uid);
1085 _debug("purging feed...", $debug_enabled);
1087 purge_feed($feed, 0, $debug_enabled);
1089 db_query("UPDATE ttrss_feeds
1090 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1092 // db_query("COMMIT");
1096 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1098 _debug("error fetching feed: $error_msg", $debug_enabled);
1101 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1102 last_updated = NOW() WHERE id = '$feed'");
1107 _debug("done", $debug_enabled);
1110 function cache_images($html, $site_url, $debug) {
1111 $cache_dir = CACHE_DIR
. "/images";
1113 libxml_use_internal_errors(true);
1115 $charset_hack = '<head>
1116 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1119 $doc = new DOMDocument();
1120 $doc->loadHTML($charset_hack . $html);
1121 $xpath = new DOMXPath($doc);
1123 $entries = $xpath->query('(//img[@src])');
1125 foreach ($entries as $entry) {
1126 if ($entry->hasAttribute('src')) {
1127 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1129 $local_filename = CACHE_DIR
. "/images/" . sha1($src) . ".png";
1131 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1133 if (!file_exists($local_filename)) {
1134 $file_content = fetch_file_contents($src);
1136 if ($file_content && strlen($file_content) > 1024) {
1137 file_put_contents($local_filename, $file_content);
1141 if (file_exists($local_filename)) {
1142 $entry->setAttribute('src', SELF_URL_PATH
. '/image.php?url=' .
1143 base64_encode($src));
1148 $node = $doc->getElementsByTagName('body')->item(0);
1150 return $doc->saveXML($node);
1153 function expire_error_log($debug) {
1154 if ($debug) _debug("Removing old error log entries...");
1156 if (DB_TYPE
== "pgsql") {
1157 db_query("DELETE FROM ttrss_error_log
1158 WHERE created_at < NOW() - INTERVAL '7 days'");
1160 db_query("DELETE FROM ttrss_error_log
1161 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1166 function expire_lock_files($debug) {
1167 //if ($debug) _debug("Removing old lock files...");
1171 if (is_writable(LOCK_DIRECTORY
)) {
1172 $files = glob(LOCK_DIRECTORY
. "/*.lock");
1175 foreach ($files as $file) {
1176 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1184 if ($debug) _debug("Removed $num_deleted old lock files.");
1187 function expire_cached_files($debug) {
1188 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1189 $cache_dir = CACHE_DIR
. "/$dir";
1191 // if ($debug) _debug("Expiring $cache_dir");
1195 if (is_writable($cache_dir)) {
1196 $files = glob("$cache_dir/*");
1199 foreach ($files as $file) {
1200 if (time() - filemtime($file) > 86400*7) {
1209 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1214 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1215 * Returns the url query as associative array
1217 * @param string query
1218 * @return array params
1220 function convertUrlQuery($query) {
1221 $queryParts = explode('&', $query);
1225 foreach ($queryParts as $param) {
1226 $item = explode('=', $param);
1227 $params[$item[0]] = $item[1];
1233 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1236 foreach ($filters as $filter) {
1237 $match_any_rule = $filter["match_any_rule"];
1238 $inverse = $filter["inverse"];
1239 $filter_match = false;
1241 foreach ($filter["rules"] as $rule) {
1243 $reg_exp = $rule["reg_exp"];
1244 $rule_inverse = $rule["inverse"];
1249 switch ($rule["type"]) {
1251 $match = @preg_match
("/$reg_exp/i", $title);
1254 // we don't need to deal with multiline regexps
1255 $content = preg_replace("/[\r\n\t]/", "", $content);
1257 $match = @preg_match
("/$reg_exp/i", $content);
1260 // we don't need to deal with multiline regexps
1261 $content = preg_replace("/[\r\n\t]/", "", $content);
1263 $match = (@preg_match
("/$reg_exp/i", $title) || @preg_match
("/$reg_exp/i", $content));
1266 $match = @preg_match
("/$reg_exp/i", $link);
1269 $match = @preg_match
("/$reg_exp/i", $author);
1272 $tag_string = join(",", $tags);
1273 $match = @preg_match
("/$reg_exp/i", $tag_string);
1277 if ($rule_inverse) $match = !$match;
1279 if ($match_any_rule) {
1281 $filter_match = true;
1285 $filter_match = $match;
1292 if ($inverse) $filter_match = !$filter_match;
1294 if ($filter_match) {
1295 foreach ($filter["actions"] AS $action) {
1296 array_push($matches, $action);
1298 // if Stop action encountered, perform no further processing
1299 if ($action["type"] == "stop") return $matches;
1307 function find_article_filter($filters, $filter_name) {
1308 foreach ($filters as $f) {
1309 if ($f["type"] == $filter_name) {
1316 function find_article_filters($filters, $filter_name) {
1319 foreach ($filters as $f) {
1320 if ($f["type"] == $filter_name) {
1321 array_push($results, $f);
1327 function calculate_article_score($filters) {
1330 foreach ($filters as $f) {
1331 if ($f["type"] == "score") {
1332 $score +
= $f["param"];
1338 function labels_contains_caption($labels, $caption) {
1339 foreach ($labels as $label) {
1340 if ($label[1] == $caption) {
1348 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1349 foreach ($filters as $f) {
1350 if ($f["type"] == "label") {
1351 if (!labels_contains_caption($article_labels, $f["param"])) {
1352 label_add_article($id, $f["param"], $owner_uid);
1358 function make_guid_from_title($title) {
1359 return preg_replace("/[ \"\',.:;]/", "-",
1360 mb_strtolower(strip_tags($title), 'utf-8'));
1363 function verify_feed_xml($feed_data) {
1364 libxml_use_internal_errors(true);
1365 $doc = new DOMDocument();
1366 $doc->loadXML($feed_data);
1367 $error = libxml_get_last_error();
1368 libxml_clear_errors();
1372 function housekeeping_common($debug) {
1373 expire_cached_files($debug);
1374 expire_lock_files($debug);
1375 expire_error_log($debug);
1377 $count = update_feedbrowser_cache();
1378 _debug("Feedbrowser updated, $count feeds processed.");
1380 purge_orphans( true);
1381 $rc = cleanup_tags( 14, 50000);
1383 _debug("Cleaned $rc cached tags.");