]> git.wh0rd.org - tt-rss.git/blob - include/rssfuncs.php
git.wh0rd.org / tt-rss.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Merge pull request #114 from ifireball/tunable-fetches
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
5
6 function update_feedbrowser_cache($link) {
7
8 $result = db_query($link, "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($link, "BEGIN");
15
16 db_query($link, "DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($link, $line["subscribers"]);
22 $feed_url = db_escape_string($link, $line["feed_url"]);
23 $title = db_escape_string($link, $line["title"]);
24 $site_url = db_escape_string($link, $line["site_url"]);
25
26 $tmp_result = db_query($link, "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($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query($link, "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 $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 define('PREFS_NO_CACHE', true);
64
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
67 if (DB_TYPE == "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
69 } else {
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
71 }
72 } else {
73 $login_thresh_qpart = "";
74 }
75
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE == "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_user_prefs.value != '-1'
81 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
82 ) OR (
83 ttrss_feeds.update_interval > 0
84 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
85 ) OR ttrss_feeds.last_updated IS NULL
86 OR last_updated = '1970-01-01 00:00:00')";
87 } else {
88 $update_limit_qpart = "AND ((
89 ttrss_feeds.update_interval = 0
90 AND ttrss_user_prefs.value != '-1'
91 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
92 ) OR (
93 ttrss_feeds.update_interval > 0
94 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
95 ) OR ttrss_feeds.last_updated IS NULL
96 OR last_updated = '1970-01-01 00:00:00')";
97 }
98
99 // Test if feed is currently being updated by another process.
100 if (DB_TYPE == "pgsql") {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
102 } else {
103 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
104 }
105
106 // Test if there is a limit to number of updated feeds
107 $query_limit = "";
108 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
109
110 $random_qpart = sql_random_function();
111
112 // We search for feed needing update.
113 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
114 FROM
115 ttrss_feeds, ttrss_users, ttrss_user_prefs
116 WHERE
117 ttrss_feeds.owner_uid = ttrss_users.id
118 AND ttrss_users.id = ttrss_user_prefs.owner_uid
119 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
120 $login_thresh_qpart $update_limit_qpart
121 $updstart_thresh_qpart
122 ORDER BY $random_qpart $query_limit");
123
124 $user_prefs_cache = array();
125
126 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
127
128 // Here is a little cache magic in order to minimize risk of double feed updates.
129 $feeds_to_update = array();
130 while ($line = db_fetch_assoc($result)) {
131 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
132 }
133
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
137
138 if(count($feeds_to_update) > 0) {
139 $feeds_quoted = array();
140
141 foreach ($feeds_to_update as $feed) {
142 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
143 }
144
145 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
146 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
147 }
148
149 expire_cached_files($debug);
150 expire_lock_files($debug);
151
152 $nf = 0;
153
154 // For each feed, we call the feed update function.
155 foreach ($feeds_to_update as $feed) {
156 if($debug) _debug("Base feed: $feed");
157
158 //update_rss_feed($link, $line["id"], true);
159
160 // since we have the data cached, we can deal with other feeds with the same url
161
162 $tmp_result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id,last_updated
163 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
164 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
165 ttrss_users.id = ttrss_user_prefs.owner_uid AND
166 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
167 feed_url = '".db_escape_string($link, $feed)."' AND
168 (ttrss_feeds.update_interval > 0 OR
169 ttrss_user_prefs.value != '-1')
170 $login_thresh_qpart
171 ORDER BY feed_url $query_limit");
172
173 if (db_num_rows($tmp_result) > 0) {
174 while ($tline = db_fetch_assoc($tmp_result)) {
175 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
176 update_rss_feed($link, $tline["id"], true);
177 ++$nf;
178 }
179 }
180 }
181
182 require_once "digest.php";
183
184 // Send feed digests by email if needed.
185 send_headlines_digests($link, $debug);
186
187 return $nf;
188
189 } // function update_daemon_common
190
191 // ignore_daemon is not used
192 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
193 $override_url = false) {
194
195 require_once "lib/simplepie/simplepie.inc";
196
197 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
198
199 if ($debug_enabled) {
200 _debug("update_rss_feed: start");
201 }
202
203 $result = db_query($link, "SELECT id,update_interval,auth_login,
204 feed_url,auth_pass,cache_images,last_updated,
205 mark_unread_on_update, owner_uid,
206 pubsub_state
207 FROM ttrss_feeds WHERE id = '$feed'");
208
209 if (db_num_rows($result) == 0) {
210 if ($debug_enabled) {
211 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
212 }
213 return false;
214 }
215
216 $last_updated = db_fetch_result($result, 0, "last_updated");
217 $owner_uid = db_fetch_result($result, 0, "owner_uid");
218 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
219 0, "mark_unread_on_update"));
220 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
221
222 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
223 WHERE id = '$feed'");
224
225 $auth_login = db_fetch_result($result, 0, "auth_login");
226 $auth_pass = db_fetch_result($result, 0, "auth_pass");
227
228 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
229 $fetch_url = db_fetch_result($result, 0, "feed_url");
230
231 $feed = db_escape_string($link, $feed);
232
233 if ($override_url) $fetch_url = $override_url;
234
235 $date_feed_processed = date('Y-m-d H:i');
236
237 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
238
239 // Ignore cache if new feed or manual update.
240 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
241 30 : get_feed_update_interval($link, $feed) * 60;
242
243 if ($debug_enabled) {
244 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
245 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
246 }
247
248 $cached_feed_data_hash = false;
249
250 $rss = false;
251 $rss_hash = false;
252 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
253 $last_updated_timestamp = strtotime($last_updated);
254
255 if (file_exists($cache_filename) &&
256 is_readable($cache_filename) &&
257 !$auth_login && !$auth_pass &&
258 filemtime($cache_filename) > time() - $cache_age) {
259
260 if ($debug_enabled) {
261 _debug("update_rss_feed: using local cache.");
262 }
263
264 if ($cache_timestamp > $last_updated_timestamp) {
265 @$rss_data = file_get_contents($cache_filename);
266
267 if ($rss_data) {
268 $rss_hash = sha1($rss_data);
269 @$rss = unserialize($rss_data);
270 }
271 } else {
272 if ($debug_enabled) {
273 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
274 }
275 return;
276 }
277 }
278
279 if (!$rss) {
280
281 if (!$feed_data) {
282 if ($debug_enabled) {
283 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
284 }
285
286 $force_refetch = isset($_REQUEST["force_refetch"]);
287
288 $feed_data = fetch_file_contents($fetch_url, false,
289 $auth_login, $auth_pass, false,
290 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
291 $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
292
293 if ($debug_enabled) {
294 _debug("update_rss_feed: fetch done.");
295 }
296
297 }
298
299 if (!$feed_data) {
300 global $fetch_last_error;
301 global $fetch_last_error_code;
302
303 if ($debug_enabled) {
304 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
305 }
306
307 $error_escaped = '';
308
309 // If-Modified-Since
310 if ($fetch_last_error_code != 304) {
311 $error_escaped = db_escape_string($link, $fetch_last_error);
312 } else {
313 if ($debug_enabled) {
314 _debug("update_rss_feed: source claims data not modified, nothing to do.");
315 }
316 }
317
318 db_query($link,
319 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
320 last_updated = NOW() WHERE id = '$feed'");
321
322 return;
323 }
324 }
325
326 $pluginhost = new PluginHost($link);
327 $pluginhost->set_debug($debug_enabled);
328 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
329
330 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
331 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
332 $pluginhost->load_data();
333
334 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
335 $feed_data = $plugin->hook_feed_fetched($feed_data);
336 }
337
338 if (!$rss) {
339 $rss = new SimplePie();
340 $rss->set_sanitize_class("SanitizeDummy");
341 // simplepie ignores the above and creates default sanitizer anyway,
342 // so let's override it...
343 $rss->sanitize = new SanitizeDummy();
344 $rss->set_output_encoding('UTF-8');
345 $rss->set_raw_data($feed_data);
346 $rss->enable_cache(false);
347
348 @$rss->init();
349 }
350
351 // print_r($rss);
352
353 $feed = db_escape_string($link, $feed);
354
355 if (!$rss->error()) {
356
357 // cache data for later
358 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
359 $rss_data = serialize($rss);
360 $new_rss_hash = sha1($rss_data);
361
362 if ($new_rss_hash != $rss_hash) {
363 if ($debug_enabled) {
364 _debug("update_rss_feed: saving $cache_filename");
365 }
366 @file_put_contents($cache_filename, serialize($rss));
367 }
368 }
369
370 // We use local pluginhost here because we need to load different per-user feed plugins
371 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
372
373 if ($debug_enabled) {
374 _debug("update_rss_feed: processing feed data...");
375 }
376
377 // db_query($link, "BEGIN");
378
379 if (DB_TYPE == "pgsql") {
380 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
381 } else {
382 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
383 }
384
385 $result = db_query($link, "SELECT title,site_url,owner_uid,
386 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
387 favicon_needs_check
388 FROM ttrss_feeds WHERE id = '$feed'");
389
390 $registered_title = db_fetch_result($result, 0, "title");
391 $orig_site_url = db_fetch_result($result, 0, "site_url");
392 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
393 "favicon_needs_check"));
394
395 $owner_uid = db_fetch_result($result, 0, "owner_uid");
396
397 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
398
399 if ($debug_enabled) {
400 _debug("update_rss_feed: checking favicon...");
401 }
402
403 if ($favicon_needs_check) {
404 check_feed_favicon($site_url, $feed, $link);
405
406 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
407 WHERE id = '$feed'");
408 }
409
410 if (!$registered_title || $registered_title == "[Unknown]") {
411
412 $feed_title = db_escape_string($link, $rss->get_title());
413
414 if ($debug_enabled) {
415 _debug("update_rss_feed: registering title: $feed_title");
416 }
417
418 db_query($link, "UPDATE ttrss_feeds SET
419 title = '$feed_title' WHERE id = '$feed'");
420 }
421
422 if ($site_url && $orig_site_url != $site_url) {
423 db_query($link, "UPDATE ttrss_feeds SET
424 site_url = '$site_url' WHERE id = '$feed'");
425 }
426
427 if ($debug_enabled) {
428 _debug("update_rss_feed: loading filters & labels...");
429 }
430
431 $filters = load_filters($link, $feed, $owner_uid);
432 $labels = get_all_labels($link, $owner_uid);
433
434 if ($debug_enabled) {
435 //print_r($filters);
436 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
437 }
438
439 $items = $rss->get_items();
440
441 if (!is_array($items)) {
442 if ($debug_enabled) {
443 _debug("update_rss_feed: no articles found.");
444 }
445
446 db_query($link, "UPDATE ttrss_feeds
447 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
448
449 return; // no articles
450 }
451
452 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
453
454 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
455
456 $feed_hub_url = false;
457
458 $links = $rss->get_links('hub');
459
460 if ($links && is_array($links)) {
461 foreach ($links as $l) {
462 $feed_hub_url = $l;
463 break;
464 }
465 }
466
467 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
468
469 if ($feed_hub_url && function_exists('curl_init') &&
470 !ini_get("open_basedir")) {
471
472 require_once 'lib/pubsubhubbub/subscriber.php';
473
474 $callback_url = get_self_url_prefix() .
475 "/public.php?op=pubsub&id=$feed";
476
477 $s = new Subscriber($feed_hub_url, $callback_url);
478
479 $rc = $s->subscribe($fetch_url);
480
481 if ($debug_enabled)
482 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
483
484 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
485 WHERE id = '$feed'");
486 }
487 }
488
489 if ($debug_enabled) {
490 _debug("update_rss_feed: processing articles...");
491 }
492
493 foreach ($items as $item) {
494 if ($_REQUEST['xdebug'] == 3) {
495 print_r($item);
496 }
497
498 $entry_guid = $item->get_id();
499 if (!$entry_guid) $entry_guid = $item->get_link();
500 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
501
502 if ($debug_enabled) {
503 _debug("update_rss_feed: guid $entry_guid");
504 }
505
506 if (!$entry_guid) continue;
507
508 $entry_guid = "$owner_uid,$entry_guid";
509
510 $entry_timestamp = "";
511
512 $entry_timestamp = strtotime($item->get_date());
513
514 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
515 $entry_timestamp = time();
516 $no_orig_date = 'true';
517 } else {
518 $no_orig_date = 'false';
519 }
520
521 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
522
523 if ($debug_enabled) {
524 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
525 }
526
527 $entry_title = $item->get_title();
528
529 $entry_link = rewrite_relative_url($site_url, $item->get_link());
530
531 if ($debug_enabled) {
532 _debug("update_rss_feed: title $entry_title");
533 _debug("update_rss_feed: link $entry_link");
534 }
535
536 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
537
538 $entry_content = $item->get_content();
539 if (!$entry_content) $entry_content = $item->get_description();
540
541 if ($_REQUEST["xdebug"] == 2) {
542 print "update_rss_feed: content: ";
543 print $entry_content;
544 print "\n";
545 }
546
547 $entry_comments = $item->data["comments"];
548
549 if ($item->get_author()) {
550 $entry_author_item = $item->get_author();
551 $entry_author = $entry_author_item->get_name();
552 if (!$entry_author) $entry_author = $entry_author_item->get_email();
553
554 $entry_author = db_escape_string($link, $entry_author);
555 }
556
557 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
558
559 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
560 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
561
562 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
563
564 if (is_array($num_comments) && is_array($num_comments[0])) {
565 $num_comments = (int) $num_comments[0]["data"];
566 } else {
567 $num_comments = 0;
568 }
569
570 if ($debug_enabled) {
571 _debug("update_rss_feed: num_comments: $num_comments");
572 _debug("update_rss_feed: looking for tags [1]...");
573 }
574
575 // parse <category> entries into tags
576
577 $additional_tags = array();
578
579 $additional_tags_src = $item->get_categories();
580
581 if (is_array($additional_tags_src)) {
582 foreach ($additional_tags_src as $tobj) {
583 array_push($additional_tags, $tobj->get_term());
584 }
585 }
586
587 if ($debug_enabled) {
588 _debug("update_rss_feed: category tags:");
589 print_r($additional_tags);
590 }
591
592 if ($debug_enabled) {
593 _debug("update_rss_feed: looking for tags [2]...");
594 }
595
596 $entry_tags = array_unique($additional_tags);
597
598 for ($i = 0; $i < count($entry_tags); $i++)
599 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
600
601 if ($debug_enabled) {
602 //_debug("update_rss_feed: unfiltered tags found:");
603 //print_r($entry_tags);
604 }
605
606 if ($debug_enabled) {
607 _debug("update_rss_feed: done collecting data.");
608 }
609
610 // TODO: less memory-hungry implementation
611
612 if ($debug_enabled) {
613 _debug("update_rss_feed: applying plugin filters..");
614 }
615
616 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
617 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
618 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
619
620 if (db_num_rows($result) != 0) {
621 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
622 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
623 "content" => db_fetch_result($result, 0, "content"),
624 "link" => db_fetch_result($result, 0, "link"),
625 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
626 "author" => db_fetch_result($result, 0, "author"));
627 } else {
628 $entry_plugin_data = "";
629 $stored_article = array();
630 }
631
632 $article = array("owner_uid" => $owner_uid, // read only
633 "guid" => $entry_guid, // read only
634 "title" => $entry_title,
635 "content" => $entry_content,
636 "link" => $entry_link,
637 "tags" => $entry_tags,
638 "plugin_data" => $entry_plugin_data,
639 "author" => $entry_author,
640 "stored" => $stored_article);
641
642 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
643 $article = $plugin->hook_article_filter($article);
644 }
645
646 $entry_tags = $article["tags"];
647 $entry_guid = db_escape_string($link, $entry_guid);
648 $entry_title = db_escape_string($link, $article["title"]);
649 $entry_author = db_escape_string($link, $article["author"]);
650 $entry_link = db_escape_string($link, $article["link"]);
651 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
652 $entry_content = $article["content"]; // escaped below
653
654
655 if ($debug_enabled) {
656 _debug("update_rss_feed: plugin data: $entry_plugin_data");
657 }
658
659 if ($cache_images && is_writable(CACHE_DIR . '/images'))
660 cache_images($entry_content, $site_url, $debug_enabled);
661
662 $entry_content = db_escape_string($link, $entry_content, false);
663
664 $content_hash = "SHA1:" . sha1($entry_content);
665
666 db_query($link, "BEGIN");
667
668 $result = db_query($link, "SELECT id FROM ttrss_entries
669 WHERE guid = '$entry_guid'");
670
671 if (db_num_rows($result) == 0) {
672
673 if ($debug_enabled) {
674 _debug("update_rss_feed: base guid [$entry_guid] not found");
675 }
676
677 // base post entry does not exist, create it
678
679 $result = db_query($link,
680 "INSERT INTO ttrss_entries
681 (title,
682 guid,
683 link,
684 updated,
685 content,
686 content_hash,
687 cached_content,
688 no_orig_date,
689 date_updated,
690 date_entered,
691 comments,
692 num_comments,
693 plugin_data,
694 author)
695 VALUES
696 ('$entry_title',
697 '$entry_guid',
698 '$entry_link',
699 '$entry_timestamp_fmt',
700 '$entry_content',
701 '$content_hash',
702 '',
703 $no_orig_date,
704 NOW(),
705 '$date_feed_processed',
706 '$entry_comments',
707 '$num_comments',
708 '$entry_plugin_data',
709 '$entry_author')");
710
711 $article_labels = array();
712
713 } else {
714 // we keep encountering the entry in feeds, so we need to
715 // update date_updated column so that we don't get horrible
716 // dupes when the entry gets purged and reinserted again e.g.
717 // in the case of SLOW SLOW OMG SLOW updating feeds
718
719 $base_entry_id = db_fetch_result($result, 0, "id");
720
721 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
722 WHERE id = '$base_entry_id'");
723
724 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
725 }
726
727 // now it should exist, if not - bad luck then
728
729 $result = db_query($link, "SELECT
730 id,content_hash,no_orig_date,title,plugin_data,
731 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
732 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
733 num_comments
734 FROM
735 ttrss_entries
736 WHERE guid = '$entry_guid'");
737
738 $entry_ref_id = 0;
739 $entry_int_id = 0;
740
741 if (db_num_rows($result) == 1) {
742
743 if ($debug_enabled) {
744 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
745 }
746
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,
752 0, "date_updated"));
753 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
754
755 $ref_id = db_fetch_result($result, 0, "id");
756 $entry_ref_id = $ref_id;
757
758 // check for user post link to main table
759
760 // do we allow duplicate posts with same GUID in different feeds?
761 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
762 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
763 } else {
764 $dupcheck_qpart = "";
765 }
766
767 /* Collect article tags here so we could filter by them: */
768
769 $article_filters = get_article_filters($filters, $entry_title,
770 $entry_content, $entry_link, $entry_timestamp, $entry_author,
771 $entry_tags);
772
773 if ($debug_enabled) {
774 _debug("update_rss_feed: article filters: ");
775 if (count($article_filters) != 0) {
776 print_r($article_filters);
777 }
778 }
779
780 if (find_article_filter($article_filters, "filter")) {
781 db_query($link, "COMMIT"); // close transaction in progress
782 continue;
783 }
784
785 $score = calculate_article_score($article_filters);
786
787 if ($debug_enabled) {
788 _debug("update_rss_feed: initial score: $score");
789 }
790
791 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
792 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
793 $dupcheck_qpart";
794
795 // if ($_REQUEST["xdebug"]) print "$query\n";
796
797 $result = db_query($link, $query);
798
799 // okay it doesn't exist - create user entry
800 if (db_num_rows($result) == 0) {
801
802 if ($debug_enabled) {
803 _debug("update_rss_feed: user record not found, creating...");
804 }
805
806 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
807 $unread = 'true';
808 $last_read_qpart = 'NULL';
809 } else {
810 $unread = 'false';
811 $last_read_qpart = 'NOW()';
812 }
813
814 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
815 $marked = 'true';
816 } else {
817 $marked = 'false';
818 }
819
820 if (find_article_filter($article_filters, 'publish')) {
821 $published = 'true';
822 } else {
823 $published = 'false';
824 }
825
826 // N-grams
827
828 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
829
830 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
831 ttrss_entries,ttrss_user_entries
832 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
833 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
834 AND owner_uid = $owner_uid");
835
836 $ngram_similar = db_fetch_result($result, 0, "similar");
837
838 if ($debug_enabled) {
839 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
840 }
841
842 if ($ngram_similar > 0) {
843 $unread = 'false';
844 }
845 }
846
847 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
848 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
849
850 $result = db_query($link,
851 "INSERT INTO ttrss_user_entries
852 (ref_id, owner_uid, feed_id, unread, last_read, marked,
853 published, score, tag_cache, label_cache, uuid,
854 last_marked, last_published)
855 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
856 $last_read_qpart, $marked, $published, '$score', '', '',
857 '', $last_marked, $last_published)");
858
859 if (PUBSUBHUBBUB_HUB && $published == 'true') {
860 $rss_link = get_self_url_prefix() .
861 "/public.php?op=rss&id=-2&key=" .
862 get_feed_access_key($link, -2, false, $owner_uid);
863
864 $p = new Publisher(PUBSUBHUBBUB_HUB);
865
866 $pubsub_result = $p->publish_update($rss_link);
867 }
868
869 $result = db_query($link,
870 "SELECT int_id FROM ttrss_user_entries WHERE
871 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
872 feed_id = '$feed' LIMIT 1");
873
874 if (db_num_rows($result) == 1) {
875 $entry_int_id = db_fetch_result($result, 0, "int_id");
876 }
877 } else {
878 if ($debug_enabled) {
879 _debug("update_rss_feed: user record FOUND");
880 }
881
882 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
883 $entry_int_id = db_fetch_result($result, 0, "int_id");
884 }
885
886 if ($debug_enabled) {
887 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
888 }
889
890 $post_needs_update = false;
891 $update_insignificant = false;
892
893 if ($orig_num_comments != $num_comments) {
894 $post_needs_update = true;
895 $update_insignificant = true;
896 }
897
898 if ($entry_plugin_data != $orig_plugin_data) {
899 $post_needs_update = true;
900 $update_insignificant = true;
901 }
902
903 if ($content_hash != $orig_content_hash) {
904 $post_needs_update = true;
905 $update_insignificant = false;
906 }
907
908 if (db_escape_string($link, $orig_title) != $entry_title) {
909 $post_needs_update = true;
910 $update_insignificant = false;
911 }
912
913 // if post needs update, update it and mark all user entries
914 // linking to this post as updated
915 if ($post_needs_update) {
916
917 if (defined('DAEMON_EXTENDED_DEBUG')) {
918 _debug("update_rss_feed: post $entry_guid needs update...");
919 }
920
921 // print "<!-- post $orig_title needs update : $post_needs_update -->";
922
923 db_query($link, "UPDATE ttrss_entries
924 SET title = '$entry_title', content = '$entry_content',
925 content_hash = '$content_hash',
926 updated = '$entry_timestamp_fmt',
927 num_comments = '$num_comments',
928 plugin_data = '$entry_plugin_data'
929 WHERE id = '$ref_id'");
930
931 if (!$update_insignificant) {
932 if ($mark_unread_on_update) {
933 db_query($link, "UPDATE ttrss_user_entries
934 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
935 }
936 }
937 }
938 }
939
940 db_query($link, "COMMIT");
941
942 if ($debug_enabled) {
943 _debug("update_rss_feed: assigning labels...");
944 }
945
946 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
947 $owner_uid, $article_labels);
948
949 if ($debug_enabled) {
950 _debug("update_rss_feed: looking for enclosures...");
951 }
952
953 // enclosures
954
955 $enclosures = array();
956
957 $encs = $item->get_enclosures();
958
959 if (is_array($encs)) {
960 foreach ($encs as $e) {
961 $e_item = array(
962 $e->link, $e->type, $e->length);
963 array_push($enclosures, $e_item);
964 }
965 }
966
967 if ($debug_enabled) {
968 _debug("update_rss_feed: article enclosures:");
969 print_r($enclosures);
970 }
971
972 db_query($link, "BEGIN");
973
974 foreach ($enclosures as $enc) {
975 $enc_url = db_escape_string($link, $enc[0]);
976 $enc_type = db_escape_string($link, $enc[1]);
977 $enc_dur = db_escape_string($link, $enc[2]);
978
979 $result = db_query($link, "SELECT id FROM ttrss_enclosures
980 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
981
982 if (db_num_rows($result) == 0) {
983 db_query($link, "INSERT INTO ttrss_enclosures
984 (content_url, content_type, title, duration, post_id) VALUES
985 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
986 }
987 }
988
989 db_query($link, "COMMIT");
990
991 // check for manual tags (we have to do it here since they're loaded from filters)
992
993 foreach ($article_filters as $f) {
994 if ($f["type"] == "tag") {
995
996 $manual_tags = trim_array(explode(",", $f["param"]));
997
998 foreach ($manual_tags as $tag) {
999 if (tag_is_valid($tag)) {
1000 array_push($entry_tags, $tag);
1001 }
1002 }
1003 }
1004 }
1005
1006 // Skip boring tags
1007
1008 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1009 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1010
1011 $filtered_tags = array();
1012 $tags_to_cache = array();
1013
1014 if ($entry_tags && is_array($entry_tags)) {
1015 foreach ($entry_tags as $tag) {
1016 if (array_search($tag, $boring_tags) === false) {
1017 array_push($filtered_tags, $tag);
1018 }
1019 }
1020 }
1021
1022 $filtered_tags = array_unique($filtered_tags);
1023
1024 if ($debug_enabled) {
1025 _debug("update_rss_feed: filtered article tags:");
1026 print_r($filtered_tags);
1027 }
1028
1029 // Save article tags in the database
1030
1031 if (count($filtered_tags) > 0) {
1032
1033 db_query($link, "BEGIN");
1034
1035 foreach ($filtered_tags as $tag) {
1036
1037 $tag = sanitize_tag($tag);
1038 $tag = db_escape_string($link, $tag);
1039
1040 if (!tag_is_valid($tag)) continue;
1041
1042 $result = db_query($link, "SELECT id FROM ttrss_tags
1043 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1044 owner_uid = '$owner_uid' LIMIT 1");
1045
1046 if ($result && db_num_rows($result) == 0) {
1047
1048 db_query($link, "INSERT INTO ttrss_tags
1049 (owner_uid,tag_name,post_int_id)
1050 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1051 }
1052
1053 array_push($tags_to_cache, $tag);
1054 }
1055
1056 /* update the cache */
1057
1058 $tags_to_cache = array_unique($tags_to_cache);
1059
1060 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1061
1062 db_query($link, "UPDATE ttrss_user_entries
1063 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1064 AND owner_uid = $owner_uid");
1065
1066 db_query($link, "COMMIT");
1067 }
1068
1069 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1070 if ($debug_enabled) {
1071 _debug("update_rss_feed: auto-assigning labels...");
1072 }
1073
1074 foreach ($labels as $label) {
1075 $caption = preg_quote($label["caption"]);
1076
1077 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1078 if (!labels_contains_caption($article_labels, $caption)) {
1079 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1080 }
1081 }
1082 }
1083 }
1084
1085 if ($debug_enabled) {
1086 _debug("update_rss_feed: article processed");
1087 }
1088 }
1089
1090 if (!$last_updated) {
1091 if ($debug_enabled) {
1092 _debug("update_rss_feed: new feed, catching it up...");
1093 }
1094 catchup_feed($link, $feed, false, $owner_uid);
1095 }
1096
1097 if ($debug_enabled) {
1098 _debug("purging feed...");
1099 }
1100
1101 purge_feed($link, $feed, 0, $debug_enabled);
1102
1103 db_query($link, "UPDATE ttrss_feeds
1104 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1105
1106 // db_query($link, "COMMIT");
1107
1108 } else {
1109
1110 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1111
1112 if ($debug_enabled) {
1113 _debug("update_rss_feed: error fetching feed: $error_msg");
1114 }
1115
1116 db_query($link,
1117 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1118 last_updated = NOW() WHERE id = '$feed'");
1119 }
1120
1121 unset($rss);
1122
1123 if ($debug_enabled) {
1124 _debug("update_rss_feed: done");
1125 }
1126
1127 }
1128
1129 function cache_images($html, $site_url, $debug) {
1130 $cache_dir = CACHE_DIR . "/images";
1131
1132 libxml_use_internal_errors(true);
1133
1134 $charset_hack = '<head>
1135 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1136 </head>';
1137
1138 $doc = new DOMDocument();
1139 $doc->loadHTML($charset_hack . $html);
1140 $xpath = new DOMXPath($doc);
1141
1142 $entries = $xpath->query('(//img[@src])');
1143
1144 foreach ($entries as $entry) {
1145 if ($entry->hasAttribute('src')) {
1146 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1147
1148 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1149
1150 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1151
1152 if (!file_exists($local_filename)) {
1153 $file_content = fetch_file_contents($src);
1154
1155 if ($file_content && strlen($file_content) > 1024) {
1156 file_put_contents($local_filename, $file_content);
1157 }
1158 }
1159
1160 if (file_exists($local_filename)) {
1161 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1162 base64_encode($src));
1163 }
1164 }
1165 }
1166
1167 $node = $doc->getElementsByTagName('body')->item(0);
1168
1169 return $doc->saveXML($node);
1170 }
1171
1172 function expire_lock_files($debug) {
1173 if ($debug) _debug("Removing old lock files...");
1174
1175 $num_deleted = 0;
1176
1177 if (is_writable(LOCK_DIRECTORY)) {
1178 $files = glob(LOCK_DIRECTORY . "/*.lock");
1179
1180 if ($files) {
1181 foreach ($files as $file) {
1182 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1183 unlink($file);
1184 ++$num_deleted;
1185 }
1186 }
1187 }
1188 }
1189
1190 if ($debug) _debug("Removed $num_deleted files.");
1191 }
1192
1193 function expire_cached_files($debug) {
1194 foreach (array("simplepie", "images", "export") as $dir) {
1195 $cache_dir = CACHE_DIR . "/$dir";
1196
1197 if ($debug) _debug("Expiring $cache_dir");
1198
1199 $num_deleted = 0;
1200
1201 if (is_writable($cache_dir)) {
1202 $files = glob("$cache_dir/*");
1203
1204 if ($files) {
1205 foreach ($files as $file) {
1206 if (time() - filemtime($file) > 86400*7) {
1207 unlink($file);
1208
1209 ++$num_deleted;
1210 }
1211 }
1212 }
1213 }
1214
1215 if ($debug) _debug("Removed $num_deleted files.");
1216 }
1217 }
1218
1219 /**
1220 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1221 * Returns the url query as associative array
1222 *
1223 * @param string query
1224 * @return array params
1225 */
1226 function convertUrlQuery($query) {
1227 $queryParts = explode('&', $query);
1228
1229 $params = array();
1230
1231 foreach ($queryParts as $param) {
1232 $item = explode('=', $param);
1233 $params[$item[0]] = $item[1];
1234 }
1235
1236 return $params;
1237 }
1238
1239 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1240 $matches = array();
1241
1242 foreach ($filters as $filter) {
1243 $match_any_rule = $filter["match_any_rule"];
1244 $inverse = $filter["inverse"];
1245 $filter_match = false;
1246
1247 foreach ($filter["rules"] as $rule) {
1248 $match = false;
1249 $reg_exp = $rule["reg_exp"];
1250 $rule_inverse = $rule["inverse"];
1251
1252 if (!$reg_exp)
1253 continue;
1254
1255 switch ($rule["type"]) {
1256 case "title":
1257 $match = @preg_match("/$reg_exp/i", $title);
1258 break;
1259 case "content":
1260 // we don't need to deal with multiline regexps
1261 $content = preg_replace("/[\r\n\t]/", "", $content);
1262
1263 $match = @preg_match("/$reg_exp/i", $content);
1264 break;
1265 case "both":
1266 // we don't need to deal with multiline regexps
1267 $content = preg_replace("/[\r\n\t]/", "", $content);
1268
1269 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1270 break;
1271 case "link":
1272 $match = @preg_match("/$reg_exp/i", $link);
1273 break;
1274 case "author":
1275 $match = @preg_match("/$reg_exp/i", $author);
1276 break;
1277 case "tag":
1278 $tag_string = join(",", $tags);
1279 $match = @preg_match("/$reg_exp/i", $tag_string);
1280 break;
1281 }
1282
1283 if ($rule_inverse) $match = !$match;
1284
1285 if ($match_any_rule) {
1286 if ($match) {
1287 $filter_match = true;
1288 break;
1289 }
1290 } else {
1291 $filter_match = $match;
1292 if (!$match) {
1293 break;
1294 }
1295 }
1296 }
1297
1298 if ($inverse) $filter_match = !$filter_match;
1299
1300 if ($filter_match) {
1301 foreach ($filter["actions"] AS $action) {
1302 array_push($matches, $action);
1303
1304 // if Stop action encountered, perform no further processing
1305 if ($action["type"] == "stop") return $matches;
1306 }
1307 }
1308 }
1309
1310 return $matches;
1311 }
1312
1313 function find_article_filter($filters, $filter_name) {
1314 foreach ($filters as $f) {
1315 if ($f["type"] == $filter_name) {
1316 return $f;
1317 };
1318 }
1319 return false;
1320 }
1321
1322 function find_article_filters($filters, $filter_name) {
1323 $results = array();
1324
1325 foreach ($filters as $f) {
1326 if ($f["type"] == $filter_name) {
1327 array_push($results, $f);
1328 };
1329 }
1330 return $results;
1331 }
1332
1333 function calculate_article_score($filters) {
1334 $score = 0;
1335
1336 foreach ($filters as $f) {
1337 if ($f["type"] == "score") {
1338 $score += $f["param"];
1339 };
1340 }
1341 return $score;
1342 }
1343
1344 function labels_contains_caption($labels, $caption) {
1345 foreach ($labels as $label) {
1346 if ($label[1] == $caption) {
1347 return true;
1348 }
1349 }
1350
1351 return false;
1352 }
1353
1354 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1355 foreach ($filters as $f) {
1356 if ($f["type"] == "label") {
1357 if (!labels_contains_caption($article_labels, $f["param"])) {
1358 label_add_article($link, $id, $f["param"], $owner_uid);
1359 }
1360 }
1361 }
1362 }
1363
1364 function make_guid_from_title($title) {
1365 return preg_replace("/[ \"\',.:;]/", "-",
1366 mb_strtolower(strip_tags($title), 'utf-8'));
1367 }
1368
1369
1370 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/