]> 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 #108 from scarabeusiv/master
[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, $no_cache ? 15 : 45,
290 $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
291
292 if ($debug_enabled) {
293 _debug("update_rss_feed: fetch done.");
294 }
295
296 }
297
298 if (!$feed_data) {
299 global $fetch_last_error;
300 global $fetch_last_error_code;
301
302 if ($debug_enabled) {
303 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
304 }
305
306 $error_escaped = '';
307
308 // If-Modified-Since
309 if ($fetch_last_error_code != 304) {
310 $error_escaped = db_escape_string($link, $fetch_last_error);
311 } else {
312 if ($debug_enabled) {
313 _debug("update_rss_feed: source claims data not modified, nothing to do.");
314 }
315 }
316
317 db_query($link,
318 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
319 last_updated = NOW() WHERE id = '$feed'");
320
321 return;
322 }
323 }
324
325 $pluginhost = new PluginHost($link);
326 $pluginhost->set_debug($debug_enabled);
327 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
328
329 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
330 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
331 $pluginhost->load_data();
332
333 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
334 $feed_data = $plugin->hook_feed_fetched($feed_data);
335 }
336
337 if (!$rss) {
338 $rss = new SimplePie();
339 $rss->set_sanitize_class("SanitizeDummy");
340 // simplepie ignores the above and creates default sanitizer anyway,
341 // so let's override it...
342 $rss->sanitize = new SanitizeDummy();
343 $rss->set_output_encoding('UTF-8');
344 $rss->set_raw_data($feed_data);
345 $rss->enable_cache(false);
346
347 @$rss->init();
348 }
349
350 // print_r($rss);
351
352 $feed = db_escape_string($link, $feed);
353
354 if (!$rss->error()) {
355
356 // cache data for later
357 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
358 $rss_data = serialize($rss);
359 $new_rss_hash = sha1($rss_data);
360
361 if ($new_rss_hash != $rss_hash) {
362 if ($debug_enabled) {
363 _debug("update_rss_feed: saving $cache_filename");
364 }
365 @file_put_contents($cache_filename, serialize($rss));
366 }
367 }
368
369 // We use local pluginhost here because we need to load different per-user feed plugins
370 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
371
372 if ($debug_enabled) {
373 _debug("update_rss_feed: processing feed data...");
374 }
375
376 // db_query($link, "BEGIN");
377
378 if (DB_TYPE == "pgsql") {
379 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
380 } else {
381 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
382 }
383
384 $result = db_query($link, "SELECT title,site_url,owner_uid,
385 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
386 favicon_needs_check
387 FROM ttrss_feeds WHERE id = '$feed'");
388
389 $registered_title = db_fetch_result($result, 0, "title");
390 $orig_site_url = db_fetch_result($result, 0, "site_url");
391 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
392 "favicon_needs_check"));
393
394 $owner_uid = db_fetch_result($result, 0, "owner_uid");
395
396 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
397
398 if ($debug_enabled) {
399 _debug("update_rss_feed: checking favicon...");
400 }
401
402 if ($favicon_needs_check) {
403 check_feed_favicon($site_url, $feed, $link);
404
405 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
406 WHERE id = '$feed'");
407 }
408
409 if (!$registered_title || $registered_title == "[Unknown]") {
410
411 $feed_title = db_escape_string($link, $rss->get_title());
412
413 if ($debug_enabled) {
414 _debug("update_rss_feed: registering title: $feed_title");
415 }
416
417 db_query($link, "UPDATE ttrss_feeds SET
418 title = '$feed_title' WHERE id = '$feed'");
419 }
420
421 if ($site_url && $orig_site_url != $site_url) {
422 db_query($link, "UPDATE ttrss_feeds SET
423 site_url = '$site_url' WHERE id = '$feed'");
424 }
425
426 if ($debug_enabled) {
427 _debug("update_rss_feed: loading filters & labels...");
428 }
429
430 $filters = load_filters($link, $feed, $owner_uid);
431 $labels = get_all_labels($link, $owner_uid);
432
433 if ($debug_enabled) {
434 //print_r($filters);
435 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
436 }
437
438 $items = $rss->get_items();
439
440 if (!is_array($items)) {
441 if ($debug_enabled) {
442 _debug("update_rss_feed: no articles found.");
443 }
444
445 db_query($link, "UPDATE ttrss_feeds
446 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
447
448 return; // no articles
449 }
450
451 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
452
453 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
454
455 $feed_hub_url = false;
456
457 $links = $rss->get_links('hub');
458
459 if ($links && is_array($links)) {
460 foreach ($links as $l) {
461 $feed_hub_url = $l;
462 break;
463 }
464 }
465
466 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
467
468 if ($feed_hub_url && function_exists('curl_init') &&
469 !ini_get("open_basedir")) {
470
471 require_once 'lib/pubsubhubbub/subscriber.php';
472
473 $callback_url = get_self_url_prefix() .
474 "/public.php?op=pubsub&id=$feed";
475
476 $s = new Subscriber($feed_hub_url, $callback_url);
477
478 $rc = $s->subscribe($fetch_url);
479
480 if ($debug_enabled)
481 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
482
483 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
484 WHERE id = '$feed'");
485 }
486 }
487
488 if ($debug_enabled) {
489 _debug("update_rss_feed: processing articles...");
490 }
491
492 foreach ($items as $item) {
493 if ($_REQUEST['xdebug'] == 3) {
494 print_r($item);
495 }
496
497 $entry_guid = $item->get_id();
498 if (!$entry_guid) $entry_guid = $item->get_link();
499 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
500
501 if ($debug_enabled) {
502 _debug("update_rss_feed: guid $entry_guid");
503 }
504
505 if (!$entry_guid) continue;
506
507 $entry_guid = "$owner_uid,$entry_guid";
508
509 $entry_timestamp = "";
510
511 $entry_timestamp = strtotime($item->get_date());
512
513 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
514 $entry_timestamp = time();
515 $no_orig_date = 'true';
516 } else {
517 $no_orig_date = 'false';
518 }
519
520 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
521
522 if ($debug_enabled) {
523 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
524 }
525
526 $entry_title = $item->get_title();
527
528 $entry_link = rewrite_relative_url($site_url, $item->get_link());
529
530 if ($debug_enabled) {
531 _debug("update_rss_feed: title $entry_title");
532 _debug("update_rss_feed: link $entry_link");
533 }
534
535 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
536
537 $entry_content = $item->get_content();
538 if (!$entry_content) $entry_content = $item->get_description();
539
540 if ($_REQUEST["xdebug"] == 2) {
541 print "update_rss_feed: content: ";
542 print $entry_content;
543 print "\n";
544 }
545
546 $entry_comments = $item->data["comments"];
547
548 if ($item->get_author()) {
549 $entry_author_item = $item->get_author();
550 $entry_author = $entry_author_item->get_name();
551 if (!$entry_author) $entry_author = $entry_author_item->get_email();
552
553 $entry_author = db_escape_string($link, $entry_author);
554 }
555
556 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
557
558 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
559 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
560
561 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
562
563 if (is_array($num_comments) && is_array($num_comments[0])) {
564 $num_comments = (int) $num_comments[0]["data"];
565 } else {
566 $num_comments = 0;
567 }
568
569 if ($debug_enabled) {
570 _debug("update_rss_feed: num_comments: $num_comments");
571 _debug("update_rss_feed: looking for tags [1]...");
572 }
573
574 // parse <category> entries into tags
575
576 $additional_tags = array();
577
578 $additional_tags_src = $item->get_categories();
579
580 if (is_array($additional_tags_src)) {
581 foreach ($additional_tags_src as $tobj) {
582 array_push($additional_tags, $tobj->get_term());
583 }
584 }
585
586 if ($debug_enabled) {
587 _debug("update_rss_feed: category tags:");
588 print_r($additional_tags);
589 }
590
591 if ($debug_enabled) {
592 _debug("update_rss_feed: looking for tags [2]...");
593 }
594
595 $entry_tags = array_unique($additional_tags);
596
597 for ($i = 0; $i < count($entry_tags); $i++)
598 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
599
600 if ($debug_enabled) {
601 //_debug("update_rss_feed: unfiltered tags found:");
602 //print_r($entry_tags);
603 }
604
605 if ($debug_enabled) {
606 _debug("update_rss_feed: done collecting data.");
607 }
608
609 // TODO: less memory-hungry implementation
610
611 if ($debug_enabled) {
612 _debug("update_rss_feed: applying plugin filters..");
613 }
614
615 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
616 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
617 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
618
619 if (db_num_rows($result) != 0) {
620 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
621 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
622 "content" => db_fetch_result($result, 0, "content"),
623 "link" => db_fetch_result($result, 0, "link"),
624 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
625 "author" => db_fetch_result($result, 0, "author"));
626 } else {
627 $entry_plugin_data = "";
628 $stored_article = array();
629 }
630
631 $article = array("owner_uid" => $owner_uid, // read only
632 "guid" => $entry_guid, // read only
633 "title" => $entry_title,
634 "content" => $entry_content,
635 "link" => $entry_link,
636 "tags" => $entry_tags,
637 "plugin_data" => $entry_plugin_data,
638 "author" => $entry_author,
639 "stored" => $stored_article);
640
641 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
642 $article = $plugin->hook_article_filter($article);
643 }
644
645 $entry_tags = $article["tags"];
646 $entry_guid = db_escape_string($link, $entry_guid);
647 $entry_title = db_escape_string($link, $article["title"]);
648 $entry_author = db_escape_string($link, $article["author"]);
649 $entry_link = db_escape_string($link, $article["link"]);
650 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
651 $entry_content = $article["content"]; // escaped below
652
653
654 if ($debug_enabled) {
655 _debug("update_rss_feed: plugin data: $entry_plugin_data");
656 }
657
658 if ($cache_images && is_writable(CACHE_DIR . '/images'))
659 cache_images($entry_content, $site_url, $debug_enabled);
660
661 $entry_content = db_escape_string($link, $entry_content, false);
662
663 $content_hash = "SHA1:" . sha1($entry_content);
664
665 db_query($link, "BEGIN");
666
667 $result = db_query($link, "SELECT id FROM ttrss_entries
668 WHERE guid = '$entry_guid'");
669
670 if (db_num_rows($result) == 0) {
671
672 if ($debug_enabled) {
673 _debug("update_rss_feed: base guid [$entry_guid] not found");
674 }
675
676 // base post entry does not exist, create it
677
678 $result = db_query($link,
679 "INSERT INTO ttrss_entries
680 (title,
681 guid,
682 link,
683 updated,
684 content,
685 content_hash,
686 cached_content,
687 no_orig_date,
688 date_updated,
689 date_entered,
690 comments,
691 num_comments,
692 plugin_data,
693 author)
694 VALUES
695 ('$entry_title',
696 '$entry_guid',
697 '$entry_link',
698 '$entry_timestamp_fmt',
699 '$entry_content',
700 '$content_hash',
701 '',
702 $no_orig_date,
703 NOW(),
704 '$date_feed_processed',
705 '$entry_comments',
706 '$num_comments',
707 '$entry_plugin_data',
708 '$entry_author')");
709
710 $article_labels = array();
711
712 } else {
713 // we keep encountering the entry in feeds, so we need to
714 // update date_updated column so that we don't get horrible
715 // dupes when the entry gets purged and reinserted again e.g.
716 // in the case of SLOW SLOW OMG SLOW updating feeds
717
718 $base_entry_id = db_fetch_result($result, 0, "id");
719
720 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
721 WHERE id = '$base_entry_id'");
722
723 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
724 }
725
726 // now it should exist, if not - bad luck then
727
728 $result = db_query($link, "SELECT
729 id,content_hash,no_orig_date,title,plugin_data,
730 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
731 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
732 num_comments
733 FROM
734 ttrss_entries
735 WHERE guid = '$entry_guid'");
736
737 $entry_ref_id = 0;
738 $entry_int_id = 0;
739
740 if (db_num_rows($result) == 1) {
741
742 if ($debug_enabled) {
743 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
744 }
745
746 // this will be used below in update handler
747 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
748 $orig_title = db_fetch_result($result, 0, "title");
749 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
750 $orig_date_updated = strtotime(db_fetch_result($result,
751 0, "date_updated"));
752 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
753
754 $ref_id = db_fetch_result($result, 0, "id");
755 $entry_ref_id = $ref_id;
756
757 // check for user post link to main table
758
759 // do we allow duplicate posts with same GUID in different feeds?
760 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
761 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
762 } else {
763 $dupcheck_qpart = "";
764 }
765
766 /* Collect article tags here so we could filter by them: */
767
768 $article_filters = get_article_filters($filters, $entry_title,
769 $entry_content, $entry_link, $entry_timestamp, $entry_author,
770 $entry_tags);
771
772 if ($debug_enabled) {
773 _debug("update_rss_feed: article filters: ");
774 if (count($article_filters) != 0) {
775 print_r($article_filters);
776 }
777 }
778
779 if (find_article_filter($article_filters, "filter")) {
780 db_query($link, "COMMIT"); // close transaction in progress
781 continue;
782 }
783
784 $score = calculate_article_score($article_filters);
785
786 if ($debug_enabled) {
787 _debug("update_rss_feed: initial score: $score");
788 }
789
790 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
791 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
792 $dupcheck_qpart";
793
794 // if ($_REQUEST["xdebug"]) print "$query\n";
795
796 $result = db_query($link, $query);
797
798 // okay it doesn't exist - create user entry
799 if (db_num_rows($result) == 0) {
800
801 if ($debug_enabled) {
802 _debug("update_rss_feed: user record not found, creating...");
803 }
804
805 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
806 $unread = 'true';
807 $last_read_qpart = 'NULL';
808 } else {
809 $unread = 'false';
810 $last_read_qpart = 'NOW()';
811 }
812
813 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
814 $marked = 'true';
815 } else {
816 $marked = 'false';
817 }
818
819 if (find_article_filter($article_filters, 'publish')) {
820 $published = 'true';
821 } else {
822 $published = 'false';
823 }
824
825 // N-grams
826
827 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
828
829 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
830 ttrss_entries,ttrss_user_entries
831 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
832 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
833 AND owner_uid = $owner_uid");
834
835 $ngram_similar = db_fetch_result($result, 0, "similar");
836
837 if ($debug_enabled) {
838 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
839 }
840
841 if ($ngram_similar > 0) {
842 $unread = 'false';
843 }
844 }
845
846 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
847 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
848
849 $result = db_query($link,
850 "INSERT INTO ttrss_user_entries
851 (ref_id, owner_uid, feed_id, unread, last_read, marked,
852 published, score, tag_cache, label_cache, uuid,
853 last_marked, last_published)
854 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
855 $last_read_qpart, $marked, $published, '$score', '', '',
856 '', $last_marked, $last_published)");
857
858 if (PUBSUBHUBBUB_HUB && $published == 'true') {
859 $rss_link = get_self_url_prefix() .
860 "/public.php?op=rss&id=-2&key=" .
861 get_feed_access_key($link, -2, false, $owner_uid);
862
863 $p = new Publisher(PUBSUBHUBBUB_HUB);
864
865 $pubsub_result = $p->publish_update($rss_link);
866 }
867
868 $result = db_query($link,
869 "SELECT int_id FROM ttrss_user_entries WHERE
870 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
871 feed_id = '$feed' LIMIT 1");
872
873 if (db_num_rows($result) == 1) {
874 $entry_int_id = db_fetch_result($result, 0, "int_id");
875 }
876 } else {
877 if ($debug_enabled) {
878 _debug("update_rss_feed: user record FOUND");
879 }
880
881 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
882 $entry_int_id = db_fetch_result($result, 0, "int_id");
883 }
884
885 if ($debug_enabled) {
886 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
887 }
888
889 $post_needs_update = false;
890 $update_insignificant = false;
891
892 if ($orig_num_comments != $num_comments) {
893 $post_needs_update = true;
894 $update_insignificant = true;
895 }
896
897 if ($entry_plugin_data != $orig_plugin_data) {
898 $post_needs_update = true;
899 $update_insignificant = true;
900 }
901
902 if ($content_hash != $orig_content_hash) {
903 $post_needs_update = true;
904 $update_insignificant = false;
905 }
906
907 if (db_escape_string($link, $orig_title) != $entry_title) {
908 $post_needs_update = true;
909 $update_insignificant = false;
910 }
911
912 // if post needs update, update it and mark all user entries
913 // linking to this post as updated
914 if ($post_needs_update) {
915
916 if (defined('DAEMON_EXTENDED_DEBUG')) {
917 _debug("update_rss_feed: post $entry_guid needs update...");
918 }
919
920 // print "<!-- post $orig_title needs update : $post_needs_update -->";
921
922 db_query($link, "UPDATE ttrss_entries
923 SET title = '$entry_title', content = '$entry_content',
924 content_hash = '$content_hash',
925 updated = '$entry_timestamp_fmt',
926 num_comments = '$num_comments',
927 plugin_data = '$entry_plugin_data'
928 WHERE id = '$ref_id'");
929
930 if (!$update_insignificant) {
931 if ($mark_unread_on_update) {
932 db_query($link, "UPDATE ttrss_user_entries
933 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
934 }
935 }
936 }
937 }
938
939 db_query($link, "COMMIT");
940
941 if ($debug_enabled) {
942 _debug("update_rss_feed: assigning labels...");
943 }
944
945 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
946 $owner_uid, $article_labels);
947
948 if ($debug_enabled) {
949 _debug("update_rss_feed: looking for enclosures...");
950 }
951
952 // enclosures
953
954 $enclosures = array();
955
956 $encs = $item->get_enclosures();
957
958 if (is_array($encs)) {
959 foreach ($encs as $e) {
960 $e_item = array(
961 $e->link, $e->type, $e->length);
962 array_push($enclosures, $e_item);
963 }
964 }
965
966 if ($debug_enabled) {
967 _debug("update_rss_feed: article enclosures:");
968 print_r($enclosures);
969 }
970
971 db_query($link, "BEGIN");
972
973 foreach ($enclosures as $enc) {
974 $enc_url = db_escape_string($link, $enc[0]);
975 $enc_type = db_escape_string($link, $enc[1]);
976 $enc_dur = db_escape_string($link, $enc[2]);
977
978 $result = db_query($link, "SELECT id FROM ttrss_enclosures
979 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
980
981 if (db_num_rows($result) == 0) {
982 db_query($link, "INSERT INTO ttrss_enclosures
983 (content_url, content_type, title, duration, post_id) VALUES
984 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
985 }
986 }
987
988 db_query($link, "COMMIT");
989
990 // check for manual tags (we have to do it here since they're loaded from filters)
991
992 foreach ($article_filters as $f) {
993 if ($f["type"] == "tag") {
994
995 $manual_tags = trim_array(explode(",", $f["param"]));
996
997 foreach ($manual_tags as $tag) {
998 if (tag_is_valid($tag)) {
999 array_push($entry_tags, $tag);
1000 }
1001 }
1002 }
1003 }
1004
1005 // Skip boring tags
1006
1007 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1008 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1009
1010 $filtered_tags = array();
1011 $tags_to_cache = array();
1012
1013 if ($entry_tags && is_array($entry_tags)) {
1014 foreach ($entry_tags as $tag) {
1015 if (array_search($tag, $boring_tags) === false) {
1016 array_push($filtered_tags, $tag);
1017 }
1018 }
1019 }
1020
1021 $filtered_tags = array_unique($filtered_tags);
1022
1023 if ($debug_enabled) {
1024 _debug("update_rss_feed: filtered article tags:");
1025 print_r($filtered_tags);
1026 }
1027
1028 // Save article tags in the database
1029
1030 if (count($filtered_tags) > 0) {
1031
1032 db_query($link, "BEGIN");
1033
1034 foreach ($filtered_tags as $tag) {
1035
1036 $tag = sanitize_tag($tag);
1037 $tag = db_escape_string($link, $tag);
1038
1039 if (!tag_is_valid($tag)) continue;
1040
1041 $result = db_query($link, "SELECT id FROM ttrss_tags
1042 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1043 owner_uid = '$owner_uid' LIMIT 1");
1044
1045 if ($result && db_num_rows($result) == 0) {
1046
1047 db_query($link, "INSERT INTO ttrss_tags
1048 (owner_uid,tag_name,post_int_id)
1049 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1050 }
1051
1052 array_push($tags_to_cache, $tag);
1053 }
1054
1055 /* update the cache */
1056
1057 $tags_to_cache = array_unique($tags_to_cache);
1058
1059 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1060
1061 db_query($link, "UPDATE ttrss_user_entries
1062 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1063 AND owner_uid = $owner_uid");
1064
1065 db_query($link, "COMMIT");
1066 }
1067
1068 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1069 if ($debug_enabled) {
1070 _debug("update_rss_feed: auto-assigning labels...");
1071 }
1072
1073 foreach ($labels as $label) {
1074 $caption = preg_quote($label["caption"]);
1075
1076 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1077 if (!labels_contains_caption($article_labels, $caption)) {
1078 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1079 }
1080 }
1081 }
1082 }
1083
1084 if ($debug_enabled) {
1085 _debug("update_rss_feed: article processed");
1086 }
1087 }
1088
1089 if (!$last_updated) {
1090 if ($debug_enabled) {
1091 _debug("update_rss_feed: new feed, catching it up...");
1092 }
1093 catchup_feed($link, $feed, false, $owner_uid);
1094 }
1095
1096 if ($debug_enabled) {
1097 _debug("purging feed...");
1098 }
1099
1100 purge_feed($link, $feed, 0, $debug_enabled);
1101
1102 db_query($link, "UPDATE ttrss_feeds
1103 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1104
1105 // db_query($link, "COMMIT");
1106
1107 } else {
1108
1109 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1110
1111 if ($debug_enabled) {
1112 _debug("update_rss_feed: error fetching feed: $error_msg");
1113 }
1114
1115 db_query($link,
1116 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1117 last_updated = NOW() WHERE id = '$feed'");
1118 }
1119
1120 unset($rss);
1121
1122 if ($debug_enabled) {
1123 _debug("update_rss_feed: done");
1124 }
1125
1126 }
1127
1128 function cache_images($html, $site_url, $debug) {
1129 $cache_dir = CACHE_DIR . "/images";
1130
1131 libxml_use_internal_errors(true);
1132
1133 $charset_hack = '<head>
1134 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1135 </head>';
1136
1137 $doc = new DOMDocument();
1138 $doc->loadHTML($charset_hack . $html);
1139 $xpath = new DOMXPath($doc);
1140
1141 $entries = $xpath->query('(//img[@src])');
1142
1143 foreach ($entries as $entry) {
1144 if ($entry->hasAttribute('src')) {
1145 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1146
1147 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1148
1149 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1150
1151 if (!file_exists($local_filename)) {
1152 $file_content = fetch_file_contents($src);
1153
1154 if ($file_content && strlen($file_content) > 1024) {
1155 file_put_contents($local_filename, $file_content);
1156 }
1157 }
1158
1159 if (file_exists($local_filename)) {
1160 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1161 base64_encode($src));
1162 }
1163 }
1164 }
1165
1166 $node = $doc->getElementsByTagName('body')->item(0);
1167
1168 return $doc->saveXML($node);
1169 }
1170
1171 function expire_lock_files($debug) {
1172 if ($debug) _debug("Removing old lock files...");
1173
1174 $num_deleted = 0;
1175
1176 if (is_writable(LOCK_DIRECTORY)) {
1177 $files = glob(LOCK_DIRECTORY . "/*.lock");
1178
1179 if ($files) {
1180 foreach ($files as $file) {
1181 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1182 unlink($file);
1183 ++$num_deleted;
1184 }
1185 }
1186 }
1187 }
1188
1189 if ($debug) _debug("Removed $num_deleted files.");
1190 }
1191
1192 function expire_cached_files($debug) {
1193 foreach (array("simplepie", "images", "export") as $dir) {
1194 $cache_dir = CACHE_DIR . "/$dir";
1195
1196 if ($debug) _debug("Expiring $cache_dir");
1197
1198 $num_deleted = 0;
1199
1200 if (is_writable($cache_dir)) {
1201 $files = glob("$cache_dir/*");
1202
1203 if ($files) {
1204 foreach ($files as $file) {
1205 if (time() - filemtime($file) > 86400*7) {
1206 unlink($file);
1207
1208 ++$num_deleted;
1209 }
1210 }
1211 }
1212 }
1213
1214 if ($debug) _debug("Removed $num_deleted files.");
1215 }
1216 }
1217
1218 /**
1219 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1220 * Returns the url query as associative array
1221 *
1222 * @param string query
1223 * @return array params
1224 */
1225 function convertUrlQuery($query) {
1226 $queryParts = explode('&', $query);
1227
1228 $params = array();
1229
1230 foreach ($queryParts as $param) {
1231 $item = explode('=', $param);
1232 $params[$item[0]] = $item[1];
1233 }
1234
1235 return $params;
1236 }
1237
1238 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1239 $matches = array();
1240
1241 foreach ($filters as $filter) {
1242 $match_any_rule = $filter["match_any_rule"];
1243 $inverse = $filter["inverse"];
1244 $filter_match = false;
1245
1246 foreach ($filter["rules"] as $rule) {
1247 $match = false;
1248 $reg_exp = $rule["reg_exp"];
1249 $rule_inverse = $rule["inverse"];
1250
1251 if (!$reg_exp)
1252 continue;
1253
1254 switch ($rule["type"]) {
1255 case "title":
1256 $match = @preg_match("/$reg_exp/i", $title);
1257 break;
1258 case "content":
1259 // we don't need to deal with multiline regexps
1260 $content = preg_replace("/[\r\n\t]/", "", $content);
1261
1262 $match = @preg_match("/$reg_exp/i", $content);
1263 break;
1264 case "both":
1265 // we don't need to deal with multiline regexps
1266 $content = preg_replace("/[\r\n\t]/", "", $content);
1267
1268 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1269 break;
1270 case "link":
1271 $match = @preg_match("/$reg_exp/i", $link);
1272 break;
1273 case "author":
1274 $match = @preg_match("/$reg_exp/i", $author);
1275 break;
1276 case "tag":
1277 $tag_string = join(",", $tags);
1278 $match = @preg_match("/$reg_exp/i", $tag_string);
1279 break;
1280 }
1281
1282 if ($rule_inverse) $match = !$match;
1283
1284 if ($match_any_rule) {
1285 if ($match) {
1286 $filter_match = true;
1287 break;
1288 }
1289 } else {
1290 $filter_match = $match;
1291 if (!$match) {
1292 break;
1293 }
1294 }
1295 }
1296
1297 if ($inverse) $filter_match = !$filter_match;
1298
1299 if ($filter_match) {
1300 foreach ($filter["actions"] AS $action) {
1301 array_push($matches, $action);
1302
1303 // if Stop action encountered, perform no further processing
1304 if ($action["type"] == "stop") return $matches;
1305 }
1306 }
1307 }
1308
1309 return $matches;
1310 }
1311
1312 function find_article_filter($filters, $filter_name) {
1313 foreach ($filters as $f) {
1314 if ($f["type"] == $filter_name) {
1315 return $f;
1316 };
1317 }
1318 return false;
1319 }
1320
1321 function find_article_filters($filters, $filter_name) {
1322 $results = array();
1323
1324 foreach ($filters as $f) {
1325 if ($f["type"] == $filter_name) {
1326 array_push($results, $f);
1327 };
1328 }
1329 return $results;
1330 }
1331
1332 function calculate_article_score($filters) {
1333 $score = 0;
1334
1335 foreach ($filters as $f) {
1336 if ($f["type"] == "score") {
1337 $score += $f["param"];
1338 };
1339 }
1340 return $score;
1341 }
1342
1343 function labels_contains_caption($labels, $caption) {
1344 foreach ($labels as $label) {
1345 if ($label[1] == $caption) {
1346 return true;
1347 }
1348 }
1349
1350 return false;
1351 }
1352
1353 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1354 foreach ($filters as $f) {
1355 if ($f["type"] == "label") {
1356 if (!labels_contains_caption($article_labels, $f["param"])) {
1357 label_add_article($link, $id, $f["param"], $owner_uid);
1358 }
1359 }
1360 }
1361 }
1362
1363 function make_guid_from_title($title) {
1364 return preg_replace("/[ \"\',.:;]/", "-",
1365 mb_strtolower(strip_tags($title), 'utf-8'));
1366 }
1367
1368
1369 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/