]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
remove cache valid bailout clause
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5
6 function update_feedbrowser_cache() {
7
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");
13
14 db_query("BEGIN");
15
16 db_query("DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($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"]);
25
26 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query("COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
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
52 * by another process.
53 *
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 $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 $schema_version = get_schema_version();
64
65 if ($schema_version != SCHEMA_VERSION) {
66 die("Schema version is wrong, please upgrade the database.\n");
67 }
68
69 define('PREFS_NO_CACHE', true);
70
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'";
75 } else {
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
77 }
78 } else {
79 $login_thresh_qpart = "";
80 }
81
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)
88 ) OR (
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')";
93 } else {
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)
98 ) OR (
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')";
103 }
104
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')";
108 } else {
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
110 }
111
112 // Test if there is a limit to number of updated feeds
113 $query_limit = "";
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
115
116 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
117 FROM
118 ttrss_feeds, ttrss_users, ttrss_user_prefs
119 WHERE
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";
126
127 // We search for feed needing update.
128 $result = db_query($query);
129
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
131
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($line['feed_url']));
136 }
137
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
141
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
144
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
147 }
148
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
151 }
152
153 $nf = 0;
154
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
158
159 //update_rss_feed($line["id"], true);
160
161 // since we have the data cached, we can deal with other feeds with the same url
162
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')
171 $login_thresh_qpart
172 ORDER BY ttrss_feeds.id $query_limit");
173
174 if (db_num_rows($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);
178 ++$nf;
179 }
180 }
181 }
182
183 require_once "digest.php";
184
185 // Send feed digests by email if needed.
186 send_headlines_digests($debug);
187
188 return $nf;
189
190 } // function update_daemon_common
191
192 // ignore_daemon is not used
193 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
194 $override_url = false) {
195
196 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
197
198 _debug("start", $debug_enabled);
199
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 = '$feed'");
207
208 if (db_num_rows($result) == 0) {
209 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
210 return false;
211 }
212
213 // set last update to now so if anything *simplepie* crashes later we won't be
214 // continuously failing on the same feed
215 db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
216
217 $last_updated = db_fetch_result($result, 0, "last_updated");
218 $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
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"));
225
226 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
227 WHERE id = '$feed'");
228
229 $auth_login = db_fetch_result($result, 0, "auth_login");
230 $auth_pass = db_fetch_result($result, 0, "auth_pass");
231
232 if ($auth_pass_encrypted) {
233 require_once "crypt.php";
234 $auth_pass = decrypt_string($auth_pass);
235 }
236
237 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
238 $fetch_url = db_fetch_result($result, 0, "feed_url");
239
240 $feed = db_escape_string($feed);
241
242 if ($override_url) $fetch_url = $override_url;
243
244 $date_feed_processed = date('Y-m-d H:i');
245
246 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
247
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;
250
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);
253
254 $cached_feed_data_hash = false;
255
256 $rss = false;
257 $rss_hash = false;
258 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
259
260 $force_refetch = isset($_REQUEST["force_refetch"]);
261
262 if (file_exists($cache_filename) &&
263 is_readable($cache_filename) &&
264 !$auth_login && !$auth_pass &&
265 filemtime($cache_filename) > time() - $cache_age) {
266
267 _debug("using local cache.", $debug_enabled);
268
269 @$feed_data = file_get_contents($cache_filename);
270
271 if ($feed_data) {
272 $rss_hash = sha1($feed_data);
273 }
274
275 } else {
276 _debug("local cache will not be used for this feed", $debug_enabled);
277 }
278
279 if (!$rss) {
280
281 if (!$feed_data) {
282 _debug("fetching [$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);
284
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);
289
290 global $fetch_curl_used;
291
292 if (!$fetch_curl_used) {
293 $tmp = @gzdecode($feed_data);
294
295 if ($tmp) $feed_data = $tmp;
296 }
297
298 $feed_data = trim($feed_data);
299
300 _debug("fetch done.", $debug_enabled);
301
302 if ($feed_data) {
303 $error = verify_feed_xml($feed_data);
304
305 if ($error) {
306 _debug("error verifying XML, code: " . $error->code, $debug_enabled);
307
308 if ($error->code == 26) {
309 _debug("got error 26, trying to decode entities...", $debug_enabled);
310
311 $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
312
313 $error = verify_feed_xml($feed_data);
314
315 if ($error) $feed_data = '';
316 }
317 }
318 }
319 }
320
321 if (!$feed_data) {
322 global $fetch_last_error;
323 global $fetch_last_error_code;
324
325 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
326
327 $error_escaped = '';
328
329 // If-Modified-Since
330 if ($fetch_last_error_code != 304) {
331 $error_escaped = db_escape_string($fetch_last_error);
332 } else {
333 _debug("source claims data not modified, nothing to do.", $debug_enabled);
334 }
335
336 db_query(
337 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
338 last_updated = NOW() WHERE id = '$feed'");
339
340 return;
341 }
342 }
343
344 $pluginhost = new PluginHost();
345 $pluginhost->set_debug($debug_enabled, $debug_enabled);
346 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
347
348 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
349 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
350 $pluginhost->load_data();
351
352 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
353 $feed_data = $plugin->hook_feed_fetched($feed_data);
354 }
355
356 if (!$rss) {
357 $rss = new FeedParser($feed_data);
358 $rss->init();
359 }
360
361 // print_r($rss);
362
363 $feed = db_escape_string($feed);
364
365 if (!$rss->error()) {
366
367 // cache data for later
368 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
369 $new_rss_hash = sha1($rss_data);
370
371 if ($new_rss_hash != $rss_hash && count($rss->get_items()) > 0 ) {
372 _debug("saving $cache_filename", $debug_enabled);
373 @file_put_contents($cache_filename, $feed_data);
374 }
375 }
376
377 // We use local pluginhost here because we need to load different per-user feed plugins
378 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
379
380 _debug("processing feed data...", $debug_enabled);
381
382 // db_query("BEGIN");
383
384 if (DB_TYPE == "pgsql") {
385 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
386 } else {
387 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
388 }
389
390 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
391 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
392 favicon_needs_check
393 FROM ttrss_feeds WHERE id = '$feed'");
394
395 $registered_title = db_fetch_result($result, 0, "title");
396 $orig_site_url = db_fetch_result($result, 0, "site_url");
397 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
398 "favicon_needs_check"));
399 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
400
401 $owner_uid = db_fetch_result($result, 0, "owner_uid");
402
403 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
404
405 _debug("site_url: $site_url", $debug_enabled);
406 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
407
408 if ($favicon_needs_check || $force_refetch) {
409
410 /* terrible hack: if we crash on floicon shit here, we won't check
411 * the icon avgcolor again (unless the icon got updated) */
412
413 $favicon_file = ICONS_DIR . "/$feed.ico";
414 $favicon_modified = @filemtime($favicon_file);
415
416 _debug("checking favicon...", $debug_enabled);
417
418 check_feed_favicon($site_url, $feed, $link);
419 $favicon_modified_new = @filemtime($favicon_file);
420
421 if ($favicon_modified_new > $favicon_modified)
422 $favicon_avg_color = '';
423
424 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
425 require_once "colors.php";
426
427 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
428 id = '$feed'");
429
430 $favicon_color = db_escape_string(
431 calculate_avg_color($favicon_file));
432
433 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
434 } else if ($favicon_avg_color == 'fail') {
435 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
436 }
437
438 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
439 $favicon_colorstring
440 WHERE id = '$feed'");
441 }
442
443 if (!$registered_title || $registered_title == "[Unknown]") {
444
445 $feed_title = db_escape_string($rss->get_title());
446
447 if ($feed_title) {
448 _debug("registering title: $feed_title", $debug_enabled);
449
450 db_query("UPDATE ttrss_feeds SET
451 title = '$feed_title' WHERE id = '$feed'");
452 }
453 }
454
455 if ($site_url && $orig_site_url != $site_url) {
456 db_query("UPDATE ttrss_feeds SET
457 site_url = '$site_url' WHERE id = '$feed'");
458 }
459
460 _debug("loading filters & labels...", $debug_enabled);
461
462 $filters = load_filters($feed, $owner_uid);
463 $labels = get_all_labels($owner_uid);
464
465 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
466
467 $items = $rss->get_items();
468
469 if (!is_array($items)) {
470 _debug("no articles found.", $debug_enabled);
471
472 db_query("UPDATE ttrss_feeds
473 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
474
475 return; // no articles
476 }
477
478 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
479
480 _debug("checking for PUSH hub...", $debug_enabled);
481
482 $feed_hub_url = false;
483
484 $links = $rss->get_links('hub');
485
486 if ($links && is_array($links)) {
487 foreach ($links as $l) {
488 $feed_hub_url = $l;
489 break;
490 }
491 }
492
493 _debug("feed hub url: $feed_hub_url", $debug_enabled);
494
495 if ($feed_hub_url && function_exists('curl_init') &&
496 !ini_get("open_basedir")) {
497
498 require_once 'lib/pubsubhubbub/subscriber.php';
499
500 $callback_url = get_self_url_prefix() .
501 "/public.php?op=pubsub&id=$feed";
502
503 $s = new Subscriber($feed_hub_url, $callback_url);
504
505 $rc = $s->subscribe($fetch_url);
506
507 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
508
509 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
510 WHERE id = '$feed'");
511 }
512 }
513
514 _debug("processing articles...", $debug_enabled);
515
516 foreach ($items as $item) {
517 if ($_REQUEST['xdebug'] == 3) {
518 print_r($item);
519 }
520
521 $entry_guid = $item->get_id();
522 if (!$entry_guid) $entry_guid = $item->get_link();
523 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
524
525 _debug("f_guid $entry_guid", $debug_enabled);
526
527 if (!$entry_guid) continue;
528
529 $entry_guid = "$owner_uid,$entry_guid";
530
531 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
532
533 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
534
535 $entry_timestamp = "";
536
537 $entry_timestamp = $item->get_date();
538
539 _debug("orig date: " . $item->get_date(), $debug_enabled);
540
541 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
542 $entry_timestamp = time();
543 $no_orig_date = 'true';
544 } else {
545 $no_orig_date = 'false';
546 }
547
548 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
549
550 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
551
552 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
553 // $entry_title = decode_numeric_entities($entry_title);
554 $entry_title = $item->get_title();
555
556 $entry_link = rewrite_relative_url($site_url, $item->get_link());
557
558 _debug("title $entry_title", $debug_enabled);
559 _debug("link $entry_link", $debug_enabled);
560
561 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
562
563 $entry_content = $item->get_content();
564 if (!$entry_content) $entry_content = $item->get_description();
565
566 if ($_REQUEST["xdebug"] == 2) {
567 print "content: ";
568 print $entry_content;
569 print "\n";
570 }
571
572 $entry_comments = $item->get_comments_url();
573 $entry_author = $item->get_author();
574
575 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
576
577 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
578 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
579
580 $num_comments = (int) $item->get_comments_count();
581
582 _debug("author $entry_author", $debug_enabled);
583 _debug("num_comments: $num_comments", $debug_enabled);
584 _debug("looking for tags...", $debug_enabled);
585
586 // parse <category> entries into tags
587
588 $additional_tags = array();
589
590 $additional_tags_src = $item->get_categories();
591
592 if (is_array($additional_tags_src)) {
593 foreach ($additional_tags_src as $tobj) {
594 array_push($additional_tags, $tobj);
595 }
596 }
597
598 $entry_tags = array_unique($additional_tags);
599
600 for ($i = 0; $i < count($entry_tags); $i++)
601 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
602
603 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
604
605 _debug("done collecting data.", $debug_enabled);
606
607 // TODO: less memory-hungry implementation
608
609 _debug("applying plugin filters..", $debug_enabled);
610
611 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
612 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
613 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
614
615 if (db_num_rows($result) != 0) {
616 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
617 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
618 "content" => db_fetch_result($result, 0, "content"),
619 "link" => db_fetch_result($result, 0, "link"),
620 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
621 "author" => db_fetch_result($result, 0, "author"));
622 } else {
623 $entry_plugin_data = "";
624 $stored_article = array();
625 }
626
627 $article = array("owner_uid" => $owner_uid, // read only
628 "guid" => $entry_guid, // read only
629 "title" => $entry_title,
630 "content" => $entry_content,
631 "link" => $entry_link,
632 "tags" => $entry_tags,
633 "plugin_data" => $entry_plugin_data,
634 "author" => $entry_author,
635 "stored" => $stored_article);
636
637 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
638 $article = $plugin->hook_article_filter($article);
639 }
640
641 $entry_tags = $article["tags"];
642 $entry_guid = db_escape_string($entry_guid);
643 $entry_title = db_escape_string($article["title"]);
644 $entry_author = db_escape_string($article["author"]);
645 $entry_link = db_escape_string($article["link"]);
646 $entry_plugin_data = db_escape_string($article["plugin_data"]);
647 $entry_content = $article["content"]; // escaped below
648
649
650 _debug("plugin data: $entry_plugin_data", $debug_enabled);
651
652 if ($cache_images && is_writable(CACHE_DIR . '/images'))
653 cache_images($entry_content, $site_url, $debug_enabled);
654
655 $entry_content = db_escape_string($entry_content, false);
656
657 $content_hash = "SHA1:" . sha1($entry_content);
658
659 db_query("BEGIN");
660
661 $result = db_query("SELECT id FROM ttrss_entries
662 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
663
664 if (db_num_rows($result) == 0) {
665
666 _debug("base guid [$entry_guid] not found", $debug_enabled);
667
668 // base post entry does not exist, create it
669
670 $result = db_query(
671 "INSERT INTO ttrss_entries
672 (title,
673 guid,
674 link,
675 updated,
676 content,
677 content_hash,
678 cached_content,
679 no_orig_date,
680 date_updated,
681 date_entered,
682 comments,
683 num_comments,
684 plugin_data,
685 author)
686 VALUES
687 ('$entry_title',
688 '$entry_guid_hashed',
689 '$entry_link',
690 '$entry_timestamp_fmt',
691 '$entry_content',
692 '$content_hash',
693 '',
694 $no_orig_date,
695 NOW(),
696 '$date_feed_processed',
697 '$entry_comments',
698 '$num_comments',
699 '$entry_plugin_data',
700 '$entry_author')");
701
702 $article_labels = array();
703
704 } else {
705 // we keep encountering the entry in feeds, so we need to
706 // update date_updated column so that we don't get horrible
707 // dupes when the entry gets purged and reinserted again e.g.
708 // in the case of SLOW SLOW OMG SLOW updating feeds
709
710 $base_entry_id = db_fetch_result($result, 0, "id");
711
712 db_query("UPDATE ttrss_entries SET date_updated = NOW()
713 WHERE id = '$base_entry_id'");
714
715 $article_labels = get_article_labels($base_entry_id, $owner_uid);
716 }
717
718 // now it should exist, if not - bad luck then
719
720 $result = db_query("SELECT
721 id,content_hash,no_orig_date,title,plugin_data,guid,
722 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
723 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
724 num_comments
725 FROM
726 ttrss_entries
727 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
728
729 $entry_ref_id = 0;
730 $entry_int_id = 0;
731
732 if (db_num_rows($result) == 1) {
733
734 _debug("base guid found, checking for user record", $debug_enabled);
735
736 // this will be used below in update handler
737 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
738 $orig_title = db_fetch_result($result, 0, "title");
739 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
740 $orig_date_updated = strtotime(db_fetch_result($result,
741 0, "date_updated"));
742 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
743
744 $ref_id = db_fetch_result($result, 0, "id");
745 $entry_ref_id = $ref_id;
746
747 /* $stored_guid = db_fetch_result($result, 0, "guid");
748 if ($stored_guid != $entry_guid_hashed) {
749 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
750
751 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
752 id = '$ref_id'");
753 } */
754
755 // check for user post link to main table
756
757 // do we allow duplicate posts with same GUID in different feeds?
758 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
759 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
760 } else {
761 $dupcheck_qpart = "";
762 }
763
764 /* Collect article tags here so we could filter by them: */
765
766 $article_filters = get_article_filters($filters, $entry_title,
767 $entry_content, $entry_link, $entry_timestamp, $entry_author,
768 $entry_tags);
769
770 if ($debug_enabled) {
771 _debug("article filters: ", $debug_enabled);
772 if (count($article_filters) != 0) {
773 print_r($article_filters);
774 }
775 }
776
777 if (find_article_filter($article_filters, "filter")) {
778 db_query("COMMIT"); // close transaction in progress
779 continue;
780 }
781
782 $score = calculate_article_score($article_filters);
783
784 _debug("initial score: $score", $debug_enabled);
785
786 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
787 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
788 $dupcheck_qpart";
789
790 // if ($_REQUEST["xdebug"]) print "$query\n";
791
792 $result = db_query($query);
793
794 // okay it doesn't exist - create user entry
795 if (db_num_rows($result) == 0) {
796
797 _debug("user record not found, creating...", $debug_enabled);
798
799 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
800 $unread = 'true';
801 $last_read_qpart = 'NULL';
802 } else {
803 $unread = 'false';
804 $last_read_qpart = 'NOW()';
805 }
806
807 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
808 $marked = 'true';
809 } else {
810 $marked = 'false';
811 }
812
813 if (find_article_filter($article_filters, 'publish')) {
814 $published = 'true';
815 } else {
816 $published = 'false';
817 }
818
819 // N-grams
820
821 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
822
823 $result = db_query("SELECT COUNT(*) AS similar FROM
824 ttrss_entries,ttrss_user_entries
825 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
826 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
827 AND owner_uid = $owner_uid");
828
829 $ngram_similar = db_fetch_result($result, 0, "similar");
830
831 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
832
833 if ($ngram_similar > 0) {
834 $unread = 'false';
835 }
836 }
837
838 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
839 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
840
841 $result = db_query(
842 "INSERT INTO ttrss_user_entries
843 (ref_id, owner_uid, feed_id, unread, last_read, marked,
844 published, score, tag_cache, label_cache, uuid,
845 last_marked, last_published)
846 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
847 $last_read_qpart, $marked, $published, '$score', '', '',
848 '', $last_marked, $last_published)");
849
850 if (PUBSUBHUBBUB_HUB && $published == 'true') {
851 $rss_link = get_self_url_prefix() .
852 "/public.php?op=rss&id=-2&key=" .
853 get_feed_access_key(-2, false, $owner_uid);
854
855 $p = new Publisher(PUBSUBHUBBUB_HUB);
856
857 $pubsub_result = $p->publish_update($rss_link);
858 }
859
860 $result = db_query(
861 "SELECT int_id FROM ttrss_user_entries WHERE
862 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
863 feed_id = '$feed' LIMIT 1");
864
865 if (db_num_rows($result) == 1) {
866 $entry_int_id = db_fetch_result($result, 0, "int_id");
867 }
868 } else {
869 _debug("user record FOUND", $debug_enabled);
870
871 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
872 $entry_int_id = db_fetch_result($result, 0, "int_id");
873 }
874
875 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
876
877 $post_needs_update = false;
878 $update_insignificant = false;
879
880 if ($orig_num_comments != $num_comments) {
881 $post_needs_update = true;
882 $update_insignificant = true;
883 }
884
885 if ($entry_plugin_data != $orig_plugin_data) {
886 $post_needs_update = true;
887 $update_insignificant = true;
888 }
889
890 if ($content_hash != $orig_content_hash) {
891 $post_needs_update = true;
892 $update_insignificant = false;
893 }
894
895 if (db_escape_string($orig_title) != $entry_title) {
896 $post_needs_update = true;
897 $update_insignificant = false;
898 }
899
900 // if post needs update, update it and mark all user entries
901 // linking to this post as updated
902 if ($post_needs_update) {
903
904 if (defined('DAEMON_EXTENDED_DEBUG')) {
905 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
906 }
907
908 // print "<!-- post $orig_title needs update : $post_needs_update -->";
909
910 db_query("UPDATE ttrss_entries
911 SET title = '$entry_title', content = '$entry_content',
912 content_hash = '$content_hash',
913 updated = '$entry_timestamp_fmt',
914 num_comments = '$num_comments',
915 plugin_data = '$entry_plugin_data'
916 WHERE id = '$ref_id'");
917
918 if (!$update_insignificant) {
919 if ($mark_unread_on_update) {
920 db_query("UPDATE ttrss_user_entries
921 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
922 }
923 }
924 }
925 }
926
927 db_query("COMMIT");
928
929 _debug("assigning labels...", $debug_enabled);
930
931 assign_article_to_label_filters($entry_ref_id, $article_filters,
932 $owner_uid, $article_labels);
933
934 _debug("looking for enclosures...", $debug_enabled);
935
936 // enclosures
937
938 $enclosures = array();
939
940 $encs = $item->get_enclosures();
941
942 if (is_array($encs)) {
943 foreach ($encs as $e) {
944 $e_item = array(
945 $e->link, $e->type, $e->length);
946 array_push($enclosures, $e_item);
947 }
948 }
949
950 if ($debug_enabled) {
951 _debug("article enclosures:", $debug_enabled);
952 print_r($enclosures);
953 }
954
955 db_query("BEGIN");
956
957 foreach ($enclosures as $enc) {
958 $enc_url = db_escape_string($enc[0]);
959 $enc_type = db_escape_string($enc[1]);
960 $enc_dur = db_escape_string($enc[2]);
961
962 $result = db_query("SELECT id FROM ttrss_enclosures
963 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
964
965 if (db_num_rows($result) == 0) {
966 db_query("INSERT INTO ttrss_enclosures
967 (content_url, content_type, title, duration, post_id) VALUES
968 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
969 }
970 }
971
972 db_query("COMMIT");
973
974 // check for manual tags (we have to do it here since they're loaded from filters)
975
976 foreach ($article_filters as $f) {
977 if ($f["type"] == "tag") {
978
979 $manual_tags = trim_array(explode(",", $f["param"]));
980
981 foreach ($manual_tags as $tag) {
982 if (tag_is_valid($tag)) {
983 array_push($entry_tags, $tag);
984 }
985 }
986 }
987 }
988
989 // Skip boring tags
990
991 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
992 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
993
994 $filtered_tags = array();
995 $tags_to_cache = array();
996
997 if ($entry_tags && is_array($entry_tags)) {
998 foreach ($entry_tags as $tag) {
999 if (array_search($tag, $boring_tags) === false) {
1000 array_push($filtered_tags, $tag);
1001 }
1002 }
1003 }
1004
1005 $filtered_tags = array_unique($filtered_tags);
1006
1007 if ($debug_enabled) {
1008 _debug("filtered article tags:", $debug_enabled);
1009 print_r($filtered_tags);
1010 }
1011
1012 // Save article tags in the database
1013
1014 if (count($filtered_tags) > 0) {
1015
1016 db_query("BEGIN");
1017
1018 foreach ($filtered_tags as $tag) {
1019
1020 $tag = sanitize_tag($tag);
1021 $tag = db_escape_string($tag);
1022
1023 if (!tag_is_valid($tag)) continue;
1024
1025 $result = db_query("SELECT id FROM ttrss_tags
1026 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1027 owner_uid = '$owner_uid' LIMIT 1");
1028
1029 if ($result && db_num_rows($result) == 0) {
1030
1031 db_query("INSERT INTO ttrss_tags
1032 (owner_uid,tag_name,post_int_id)
1033 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1034 }
1035
1036 array_push($tags_to_cache, $tag);
1037 }
1038
1039 /* update the cache */
1040
1041 $tags_to_cache = array_unique($tags_to_cache);
1042
1043 $tags_str = db_escape_string(join(",", $tags_to_cache));
1044
1045 db_query("UPDATE ttrss_user_entries
1046 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1047 AND owner_uid = $owner_uid");
1048
1049 db_query("COMMIT");
1050 }
1051
1052 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1053 _debug("auto-assigning labels...", $debug_enabled);
1054
1055 foreach ($labels as $label) {
1056 $caption = preg_quote($label["caption"]);
1057
1058 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1059 if (!labels_contains_caption($article_labels, $caption)) {
1060 label_add_article($entry_ref_id, $caption, $owner_uid);
1061 }
1062 }
1063 }
1064 }
1065
1066 _debug("article processed", $debug_enabled);
1067 }
1068
1069 if (!$last_updated) {
1070 _debug("new feed, catching it up...", $debug_enabled);
1071 catchup_feed($feed, false, $owner_uid);
1072 }
1073
1074 _debug("purging feed...", $debug_enabled);
1075
1076 purge_feed($feed, 0, $debug_enabled);
1077
1078 db_query("UPDATE ttrss_feeds
1079 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1080
1081 // db_query("COMMIT");
1082
1083 } else {
1084
1085 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1086
1087 _debug("error fetching feed: $error_msg", $debug_enabled);
1088
1089 db_query(
1090 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1091 last_updated = NOW() WHERE id = '$feed'");
1092 }
1093
1094 unset($rss);
1095
1096 _debug("done", $debug_enabled);
1097 }
1098
1099 function cache_images($html, $site_url, $debug) {
1100 $cache_dir = CACHE_DIR . "/images";
1101
1102 libxml_use_internal_errors(true);
1103
1104 $charset_hack = '<head>
1105 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1106 </head>';
1107
1108 $doc = new DOMDocument();
1109 $doc->loadHTML($charset_hack . $html);
1110 $xpath = new DOMXPath($doc);
1111
1112 $entries = $xpath->query('(//img[@src])');
1113
1114 foreach ($entries as $entry) {
1115 if ($entry->hasAttribute('src')) {
1116 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1117
1118 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1119
1120 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1121
1122 if (!file_exists($local_filename)) {
1123 $file_content = fetch_file_contents($src);
1124
1125 if ($file_content && strlen($file_content) > 1024) {
1126 file_put_contents($local_filename, $file_content);
1127 }
1128 }
1129
1130 if (file_exists($local_filename)) {
1131 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1132 base64_encode($src));
1133 }
1134 }
1135 }
1136
1137 $node = $doc->getElementsByTagName('body')->item(0);
1138
1139 return $doc->saveXML($node);
1140 }
1141
1142 function expire_error_log($debug) {
1143 if ($debug) _debug("Removing old error log entries...");
1144
1145 if (DB_TYPE == "pgsql") {
1146 db_query("DELETE FROM ttrss_error_log
1147 WHERE created_at < NOW() - INTERVAL '7 days'");
1148 } else {
1149 db_query("DELETE FROM ttrss_error_log
1150 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1151 }
1152
1153 }
1154
1155 function expire_lock_files($debug) {
1156 //if ($debug) _debug("Removing old lock files...");
1157
1158 $num_deleted = 0;
1159
1160 if (is_writable(LOCK_DIRECTORY)) {
1161 $files = glob(LOCK_DIRECTORY . "/*.lock");
1162
1163 if ($files) {
1164 foreach ($files as $file) {
1165 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1166 unlink($file);
1167 ++$num_deleted;
1168 }
1169 }
1170 }
1171 }
1172
1173 if ($debug) _debug("Removed $num_deleted old lock files.");
1174 }
1175
1176 function expire_cached_files($debug) {
1177 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1178 $cache_dir = CACHE_DIR . "/$dir";
1179
1180 // if ($debug) _debug("Expiring $cache_dir");
1181
1182 $num_deleted = 0;
1183
1184 if (is_writable($cache_dir)) {
1185 $files = glob("$cache_dir/*");
1186
1187 if ($files) {
1188 foreach ($files as $file) {
1189 if (time() - filemtime($file) > 86400*7) {
1190 unlink($file);
1191
1192 ++$num_deleted;
1193 }
1194 }
1195 }
1196 }
1197
1198 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1199 }
1200 }
1201
1202 /**
1203 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1204 * Returns the url query as associative array
1205 *
1206 * @param string query
1207 * @return array params
1208 */
1209 function convertUrlQuery($query) {
1210 $queryParts = explode('&', $query);
1211
1212 $params = array();
1213
1214 foreach ($queryParts as $param) {
1215 $item = explode('=', $param);
1216 $params[$item[0]] = $item[1];
1217 }
1218
1219 return $params;
1220 }
1221
1222 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1223 $matches = array();
1224
1225 foreach ($filters as $filter) {
1226 $match_any_rule = $filter["match_any_rule"];
1227 $inverse = $filter["inverse"];
1228 $filter_match = false;
1229
1230 foreach ($filter["rules"] as $rule) {
1231 $match = false;
1232 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1233 $rule_inverse = $rule["inverse"];
1234
1235 if (!$reg_exp)
1236 continue;
1237
1238 switch ($rule["type"]) {
1239 case "title":
1240 $match = @preg_match("/$reg_exp/i", $title);
1241 break;
1242 case "content":
1243 // we don't need to deal with multiline regexps
1244 $content = preg_replace("/[\r\n\t]/", "", $content);
1245
1246 $match = @preg_match("/$reg_exp/i", $content);
1247 break;
1248 case "both":
1249 // we don't need to deal with multiline regexps
1250 $content = preg_replace("/[\r\n\t]/", "", $content);
1251
1252 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1253 break;
1254 case "link":
1255 $match = @preg_match("/$reg_exp/i", $link);
1256 break;
1257 case "author":
1258 $match = @preg_match("/$reg_exp/i", $author);
1259 break;
1260 case "tag":
1261 foreach ($tags as $tag) {
1262 if (@preg_match("/$reg_exp/i", $tag)) {
1263 $match = true;
1264 break;
1265 }
1266 }
1267 break;
1268 }
1269
1270 if ($rule_inverse) $match = !$match;
1271
1272 if ($match_any_rule) {
1273 if ($match) {
1274 $filter_match = true;
1275 break;
1276 }
1277 } else {
1278 $filter_match = $match;
1279 if (!$match) {
1280 break;
1281 }
1282 }
1283 }
1284
1285 if ($inverse) $filter_match = !$filter_match;
1286
1287 if ($filter_match) {
1288 foreach ($filter["actions"] AS $action) {
1289 array_push($matches, $action);
1290
1291 // if Stop action encountered, perform no further processing
1292 if ($action["type"] == "stop") return $matches;
1293 }
1294 }
1295 }
1296
1297 return $matches;
1298 }
1299
1300 function find_article_filter($filters, $filter_name) {
1301 foreach ($filters as $f) {
1302 if ($f["type"] == $filter_name) {
1303 return $f;
1304 };
1305 }
1306 return false;
1307 }
1308
1309 function find_article_filters($filters, $filter_name) {
1310 $results = array();
1311
1312 foreach ($filters as $f) {
1313 if ($f["type"] == $filter_name) {
1314 array_push($results, $f);
1315 };
1316 }
1317 return $results;
1318 }
1319
1320 function calculate_article_score($filters) {
1321 $score = 0;
1322
1323 foreach ($filters as $f) {
1324 if ($f["type"] == "score") {
1325 $score += $f["param"];
1326 };
1327 }
1328 return $score;
1329 }
1330
1331 function labels_contains_caption($labels, $caption) {
1332 foreach ($labels as $label) {
1333 if ($label[1] == $caption) {
1334 return true;
1335 }
1336 }
1337
1338 return false;
1339 }
1340
1341 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1342 foreach ($filters as $f) {
1343 if ($f["type"] == "label") {
1344 if (!labels_contains_caption($article_labels, $f["param"])) {
1345 label_add_article($id, $f["param"], $owner_uid);
1346 }
1347 }
1348 }
1349 }
1350
1351 function make_guid_from_title($title) {
1352 return preg_replace("/[ \"\',.:;]/", "-",
1353 mb_strtolower(strip_tags($title), 'utf-8'));
1354 }
1355
1356 function verify_feed_xml($feed_data) {
1357 libxml_use_internal_errors(true);
1358 $doc = new DOMDocument();
1359 $doc->loadXML($feed_data);
1360 $error = libxml_get_last_error();
1361 libxml_clear_errors();
1362 return $error;
1363 }
1364
1365 function housekeeping_common($debug) {
1366 expire_cached_files($debug);
1367 expire_lock_files($debug);
1368 expire_error_log($debug);
1369
1370 $count = update_feedbrowser_cache();
1371 _debug("Feedbrowser updated, $count feeds processed.");
1372
1373 purge_orphans( true);
1374 $rc = cleanup_tags( 14, 50000);
1375
1376 _debug("Cleaned $rc cached tags.");
1377 }
1378 ?>