]> 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
support If-Modified-Since
[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_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
81 ) OR (
82 ttrss_feeds.update_interval > 0
83 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
84 ) OR ttrss_feeds.last_updated IS NULL
85 OR last_updated = '1970-01-01 00:00:00')";
86 } else {
87 $update_limit_qpart = "AND ((
88 ttrss_feeds.update_interval = 0
89 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
90 ) OR (
91 ttrss_feeds.update_interval > 0
92 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
93 ) OR ttrss_feeds.last_updated IS NULL
94 OR last_updated = '1970-01-01 00:00:00')";
95 }
96
97 // Test if feed is currently being updated by another process.
98 if (DB_TYPE == "pgsql") {
99 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
100 } else {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
102 }
103
104 // Test if there is a limit to number of updated feeds
105 $query_limit = "";
106 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
107
108 $random_qpart = sql_random_function();
109
110 // We search for feed needing update.
111 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url
112 FROM
113 ttrss_feeds, ttrss_users, ttrss_user_prefs
114 WHERE
115 ttrss_feeds.owner_uid = ttrss_users.id
116 AND ttrss_users.id = ttrss_user_prefs.owner_uid
117 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
118 $login_thresh_qpart $update_limit_qpart
119 $updstart_thresh_qpart
120 ORDER BY feed_url $query_limit");
121
122 $user_prefs_cache = array();
123
124 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
125
126 // Here is a little cache magic in order to minimize risk of double feed updates.
127 $feeds_to_update = array();
128 while ($line = db_fetch_assoc($result)) {
129 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
130 }
131
132 // We update the feed last update started date before anything else.
133 // There is no lag due to feed contents downloads
134 // It prevent an other process to update the same feed.
135
136 if(count($feeds_to_update) > 0) {
137 $feeds_quoted = array();
138
139 foreach ($feeds_to_update as $feed) {
140 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
141 }
142
143 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
144 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
145 }
146
147 expire_cached_files($debug);
148 expire_lock_files($debug);
149
150 // For each feed, we call the feed update function.
151 foreach ($feeds_to_update as $feed) {
152 if($debug) _debug("Base feed: $feed");
153
154 //update_rss_feed($link, $line["id"], true);
155
156 // since we have the data cached, we can deal with other feeds with the same url
157
158 $tmp_result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id,last_updated
159 FROM ttrss_feeds, ttrss_users WHERE
160 ttrss_users.id = ttrss_feeds.owner_uid AND
161 feed_url = '".db_escape_string($link, $feed)."' AND
162 ttrss_feeds.update_interval != -1
163 $login_thresh_qpart
164 ORDER BY feed_url $query_limit");
165
166 if (db_num_rows($tmp_result) > 0) {
167 while ($tline = db_fetch_assoc($tmp_result)) {
168 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
169 update_rss_feed($link, $tline["id"], true);
170 }
171 }
172 }
173
174 require_once "digest.php";
175
176 // Send feed digests by email if needed.
177 send_headlines_digests($link, $debug);
178
179 } // function update_daemon_common
180
181 // ignore_daemon is not used
182 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
183 $override_url = false) {
184
185 require_once "lib/simplepie/simplepie.inc";
186
187 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
188
189 if ($debug_enabled) {
190 _debug("update_rss_feed: start");
191 }
192
193 $result = db_query($link, "SELECT id,update_interval,auth_login,
194 feed_url,auth_pass,cache_images,last_updated,
195 mark_unread_on_update, owner_uid,
196 pubsub_state
197 FROM ttrss_feeds WHERE id = '$feed'");
198
199 if (db_num_rows($result) == 0) {
200 if ($debug_enabled) {
201 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
202 }
203 return false;
204 }
205
206 $last_updated = db_fetch_result($result, 0, "last_updated");
207 $owner_uid = db_fetch_result($result, 0, "owner_uid");
208 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
209 0, "mark_unread_on_update"));
210 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
211
212 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
213 WHERE id = '$feed'");
214
215 $auth_login = db_fetch_result($result, 0, "auth_login");
216 $auth_pass = db_fetch_result($result, 0, "auth_pass");
217
218 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
219 $fetch_url = db_fetch_result($result, 0, "feed_url");
220
221 $feed = db_escape_string($link, $feed);
222
223 if ($override_url) $fetch_url = $override_url;
224
225 $date_feed_processed = date('Y-m-d H:i');
226
227 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
228
229 // Ignore cache if new feed or manual update.
230 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
231 30 : get_feed_update_interval($link, $feed) * 60;
232
233 if ($debug_enabled) {
234 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
235 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
236 }
237
238 $cached_feed_data_hash = false;
239
240 $rss = false;
241 $rss_hash = false;
242 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
243
244 if (file_exists($cache_filename) &&
245 is_readable($cache_filename) &&
246 !$auth_login && !$auth_pass &&
247 filemtime($cache_filename) > time() - $cache_age) {
248
249 if ($debug_enabled) {
250 _debug("update_rss_feed: using local cache.");
251 }
252
253 @$rss_data = file_get_contents($cache_filename);
254
255 if ($rss_data) {
256 $rss_hash = sha1($rss_data);
257 @$rss = unserialize($rss_data);
258 }
259 }
260
261 if (!$rss) {
262
263 if (!$feed_data) {
264 if ($debug_enabled) {
265 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp)");
266 }
267
268 $feed_data = fetch_file_contents($fetch_url, false,
269 $auth_login, $auth_pass, false, $no_cache ? 15 : 45, $cache_timestamp);
270
271 if ($debug_enabled) {
272 _debug("update_rss_feed: fetch done.");
273 }
274
275 }
276
277 if (!$feed_data) {
278 global $fetch_last_error;
279 global $fetch_last_error_code;
280
281 if ($debug_enabled) {
282 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
283 }
284
285 $error_escaped = '';
286
287 // If-Modified-Since
288 if ($fetch_last_error_code != 304) {
289 $error_escaped = db_escape_string($link, $fetch_last_error);
290 } else {
291 if ($debug_enabled) {
292 _debug("update_rss_feed: source claims data not modified, nothing to do.");
293 }
294 }
295
296 db_query($link,
297 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
298 last_updated = NOW() WHERE id = '$feed'");
299
300 return;
301 }
302 }
303
304 $pluginhost = new PluginHost($link);
305 $pluginhost->set_debug($debug_enabled);
306 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
307
308 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
309 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
310 $pluginhost->load_data();
311
312 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
313 $feed_data = $plugin->hook_feed_fetched($feed_data);
314 }
315
316 if (!$rss) {
317 $rss = new SimplePie();
318 $rss->set_sanitize_class("SanitizeDummy");
319 // simplepie ignores the above and creates default sanitizer anyway,
320 // so let's override it...
321 $rss->sanitize = new SanitizeDummy();
322 $rss->set_output_encoding('UTF-8');
323 $rss->set_raw_data($feed_data);
324 $rss->enable_cache(false);
325
326 @$rss->init();
327 }
328
329 // print_r($rss);
330
331 $feed = db_escape_string($link, $feed);
332
333 if (!$rss->error()) {
334
335 // cache data for later
336 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
337 $rss_data = serialize($rss);
338 $new_rss_hash = sha1($rss_data);
339
340 if ($new_rss_hash != $rss_hash) {
341 if ($debug_enabled) {
342 _debug("update_rss_feed: saving $cache_filename");
343 }
344 @file_put_contents($cache_filename, serialize($rss));
345 }
346 }
347
348 // We use local pluginhost here because we need to load different per-user feed plugins
349 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
350
351 if ($debug_enabled) {
352 _debug("update_rss_feed: processing feed data...");
353 }
354
355 // db_query($link, "BEGIN");
356
357 if (DB_TYPE == "pgsql") {
358 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
359 } else {
360 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
361 }
362
363 $result = db_query($link, "SELECT title,site_url,owner_uid,
364 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
365 favicon_needs_check
366 FROM ttrss_feeds WHERE id = '$feed'");
367
368 $registered_title = db_fetch_result($result, 0, "title");
369 $orig_site_url = db_fetch_result($result, 0, "site_url");
370 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
371 "favicon_needs_check"));
372
373 $owner_uid = db_fetch_result($result, 0, "owner_uid");
374
375 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
376
377 if ($debug_enabled) {
378 _debug("update_rss_feed: checking favicon...");
379 }
380
381 if ($favicon_needs_check) {
382 check_feed_favicon($site_url, $feed, $link);
383
384 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
385 WHERE id = '$feed'");
386 }
387
388 if (!$registered_title || $registered_title == "[Unknown]") {
389
390 $feed_title = db_escape_string($link, $rss->get_title());
391
392 if ($debug_enabled) {
393 _debug("update_rss_feed: registering title: $feed_title");
394 }
395
396 db_query($link, "UPDATE ttrss_feeds SET
397 title = '$feed_title' WHERE id = '$feed'");
398 }
399
400 if ($site_url && $orig_site_url != $site_url) {
401 db_query($link, "UPDATE ttrss_feeds SET
402 site_url = '$site_url' WHERE id = '$feed'");
403 }
404
405 if ($debug_enabled) {
406 _debug("update_rss_feed: loading filters & labels...");
407 }
408
409 $filters = load_filters($link, $feed, $owner_uid);
410 $labels = get_all_labels($link, $owner_uid);
411
412 if ($debug_enabled) {
413 //print_r($filters);
414 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
415 }
416
417 $items = $rss->get_items();
418
419 if (!is_array($items)) {
420 if ($debug_enabled) {
421 _debug("update_rss_feed: no articles found.");
422 }
423
424 db_query($link, "UPDATE ttrss_feeds
425 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
426
427 return; // no articles
428 }
429
430 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
431
432 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
433
434 $feed_hub_url = false;
435
436 $links = $rss->get_links('hub');
437
438 if ($links && is_array($links)) {
439 foreach ($links as $l) {
440 $feed_hub_url = $l;
441 break;
442 }
443 }
444
445 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
446
447 if ($feed_hub_url && function_exists('curl_init') &&
448 !ini_get("open_basedir")) {
449
450 require_once 'lib/pubsubhubbub/subscriber.php';
451
452 $callback_url = get_self_url_prefix() .
453 "/public.php?op=pubsub&id=$feed";
454
455 $s = new Subscriber($feed_hub_url, $callback_url);
456
457 $rc = $s->subscribe($fetch_url);
458
459 if ($debug_enabled)
460 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
461
462 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
463 WHERE id = '$feed'");
464 }
465 }
466
467 if ($debug_enabled) {
468 _debug("update_rss_feed: processing articles...");
469 }
470
471 foreach ($items as $item) {
472 if ($_REQUEST['xdebug'] == 3) {
473 print_r($item);
474 }
475
476 $entry_guid = $item->get_id();
477 if (!$entry_guid) $entry_guid = $item->get_link();
478 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
479
480 if ($debug_enabled) {
481 _debug("update_rss_feed: guid $entry_guid");
482 }
483
484 if (!$entry_guid) continue;
485
486 $entry_guid = "$owner_uid,$entry_guid";
487
488 $entry_timestamp = "";
489
490 $entry_timestamp = strtotime($item->get_date());
491
492 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
493 $entry_timestamp = time();
494 $no_orig_date = 'true';
495 } else {
496 $no_orig_date = 'false';
497 }
498
499 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
500
501 if ($debug_enabled) {
502 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
503 }
504
505 $entry_title = $item->get_title();
506
507 $entry_link = rewrite_relative_url($site_url, $item->get_link());
508
509 if ($debug_enabled) {
510 _debug("update_rss_feed: title $entry_title");
511 _debug("update_rss_feed: link $entry_link");
512 }
513
514 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
515
516 $entry_content = $item->get_content();
517 if (!$entry_content) $entry_content = $item->get_description();
518
519 if ($_REQUEST["xdebug"] == 2) {
520 print "update_rss_feed: content: ";
521 print $entry_content;
522 print "\n";
523 }
524
525 $entry_comments = $item->data["comments"];
526
527 if ($item->get_author()) {
528 $entry_author_item = $item->get_author();
529 $entry_author = $entry_author_item->get_name();
530 if (!$entry_author) $entry_author = $entry_author_item->get_email();
531
532 $entry_author = db_escape_string($link, $entry_author);
533 }
534
535 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
536
537 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
538 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
539
540 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
541
542 if (is_array($num_comments) && is_array($num_comments[0])) {
543 $num_comments = (int) $num_comments[0]["data"];
544 } else {
545 $num_comments = 0;
546 }
547
548 if ($debug_enabled) {
549 _debug("update_rss_feed: num_comments: $num_comments");
550 _debug("update_rss_feed: looking for tags [1]...");
551 }
552
553 // parse <category> entries into tags
554
555 $additional_tags = array();
556
557 $additional_tags_src = $item->get_categories();
558
559 if (is_array($additional_tags_src)) {
560 foreach ($additional_tags_src as $tobj) {
561 array_push($additional_tags, $tobj->get_term());
562 }
563 }
564
565 if ($debug_enabled) {
566 _debug("update_rss_feed: category tags:");
567 print_r($additional_tags);
568 }
569
570 if ($debug_enabled) {
571 _debug("update_rss_feed: looking for tags [2]...");
572 }
573
574 $entry_tags = array_unique($additional_tags);
575
576 for ($i = 0; $i < count($entry_tags); $i++)
577 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
578
579 if ($debug_enabled) {
580 //_debug("update_rss_feed: unfiltered tags found:");
581 //print_r($entry_tags);
582 }
583
584 if ($debug_enabled) {
585 _debug("update_rss_feed: done collecting data.");
586 }
587
588 // TODO: less memory-hungry implementation
589
590 if ($debug_enabled) {
591 _debug("update_rss_feed: applying plugin filters..");
592 }
593
594 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
595 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
596 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
597
598 if (db_num_rows($result) != 0) {
599 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
600 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
601 "content" => db_fetch_result($result, 0, "content"),
602 "link" => db_fetch_result($result, 0, "link"),
603 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
604 "author" => db_fetch_result($result, 0, "author"));
605 } else {
606 $entry_plugin_data = "";
607 $stored_article = array();
608 }
609
610 $article = array("owner_uid" => $owner_uid, // read only
611 "guid" => $entry_guid, // read only
612 "title" => $entry_title,
613 "content" => $entry_content,
614 "link" => $entry_link,
615 "tags" => $entry_tags,
616 "plugin_data" => $entry_plugin_data,
617 "author" => $entry_author,
618 "stored" => $stored_article);
619
620 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
621 $article = $plugin->hook_article_filter($article);
622 }
623
624 $entry_tags = $article["tags"];
625 $entry_guid = db_escape_string($link, $entry_guid);
626 $entry_title = db_escape_string($link, $article["title"]);
627 $entry_author = db_escape_string($link, $article["author"]);
628 $entry_link = db_escape_string($link, $article["link"]);
629 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
630 $entry_content = $article["content"]; // escaped below
631
632
633 if ($debug_enabled) {
634 _debug("update_rss_feed: plugin data: $entry_plugin_data");
635 }
636
637 if ($cache_images && is_writable(CACHE_DIR . '/images'))
638 cache_images($entry_content, $site_url, $debug_enabled);
639
640 $entry_content = db_escape_string($link, $entry_content, false);
641
642 $content_hash = "SHA1:" . sha1($entry_content);
643
644 db_query($link, "BEGIN");
645
646 $result = db_query($link, "SELECT id FROM ttrss_entries
647 WHERE guid = '$entry_guid'");
648
649 if (db_num_rows($result) == 0) {
650
651 if ($debug_enabled) {
652 _debug("update_rss_feed: base guid [$entry_guid] not found");
653 }
654
655 // base post entry does not exist, create it
656
657 $result = db_query($link,
658 "INSERT INTO ttrss_entries
659 (title,
660 guid,
661 link,
662 updated,
663 content,
664 content_hash,
665 cached_content,
666 no_orig_date,
667 date_updated,
668 date_entered,
669 comments,
670 num_comments,
671 plugin_data,
672 author)
673 VALUES
674 ('$entry_title',
675 '$entry_guid',
676 '$entry_link',
677 '$entry_timestamp_fmt',
678 '$entry_content',
679 '$content_hash',
680 '',
681 $no_orig_date,
682 NOW(),
683 '$date_feed_processed',
684 '$entry_comments',
685 '$num_comments',
686 '$entry_plugin_data',
687 '$entry_author')");
688
689 $article_labels = array();
690
691 } else {
692 // we keep encountering the entry in feeds, so we need to
693 // update date_updated column so that we don't get horrible
694 // dupes when the entry gets purged and reinserted again e.g.
695 // in the case of SLOW SLOW OMG SLOW updating feeds
696
697 $base_entry_id = db_fetch_result($result, 0, "id");
698
699 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
700 WHERE id = '$base_entry_id'");
701
702 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
703 }
704
705 // now it should exist, if not - bad luck then
706
707 $result = db_query($link, "SELECT
708 id,content_hash,no_orig_date,title,plugin_data,
709 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
710 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
711 num_comments
712 FROM
713 ttrss_entries
714 WHERE guid = '$entry_guid'");
715
716 $entry_ref_id = 0;
717 $entry_int_id = 0;
718
719 if (db_num_rows($result) == 1) {
720
721 if ($debug_enabled) {
722 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
723 }
724
725 // this will be used below in update handler
726 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
727 $orig_title = db_fetch_result($result, 0, "title");
728 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
729 $orig_date_updated = strtotime(db_fetch_result($result,
730 0, "date_updated"));
731 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
732
733 $ref_id = db_fetch_result($result, 0, "id");
734 $entry_ref_id = $ref_id;
735
736 // check for user post link to main table
737
738 // do we allow duplicate posts with same GUID in different feeds?
739 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
740 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
741 } else {
742 $dupcheck_qpart = "";
743 }
744
745 /* Collect article tags here so we could filter by them: */
746
747 $article_filters = get_article_filters($filters, $entry_title,
748 $entry_content, $entry_link, $entry_timestamp, $entry_author,
749 $entry_tags);
750
751 if ($debug_enabled) {
752 _debug("update_rss_feed: article filters: ");
753 if (count($article_filters) != 0) {
754 print_r($article_filters);
755 }
756 }
757
758 if (find_article_filter($article_filters, "filter")) {
759 db_query($link, "COMMIT"); // close transaction in progress
760 continue;
761 }
762
763 $score = calculate_article_score($article_filters);
764
765 if ($debug_enabled) {
766 _debug("update_rss_feed: initial score: $score");
767 }
768
769 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
770 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
771 $dupcheck_qpart";
772
773 // if ($_REQUEST["xdebug"]) print "$query\n";
774
775 $result = db_query($link, $query);
776
777 // okay it doesn't exist - create user entry
778 if (db_num_rows($result) == 0) {
779
780 if ($debug_enabled) {
781 _debug("update_rss_feed: user record not found, creating...");
782 }
783
784 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
785 $unread = 'true';
786 $last_read_qpart = 'NULL';
787 } else {
788 $unread = 'false';
789 $last_read_qpart = 'NOW()';
790 }
791
792 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
793 $marked = 'true';
794 } else {
795 $marked = 'false';
796 }
797
798 if (find_article_filter($article_filters, 'publish')) {
799 $published = 'true';
800 } else {
801 $published = 'false';
802 }
803
804 // N-grams
805
806 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
807
808 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
809 ttrss_entries,ttrss_user_entries
810 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
811 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
812 AND owner_uid = $owner_uid");
813
814 $ngram_similar = db_fetch_result($result, 0, "similar");
815
816 if ($debug_enabled) {
817 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
818 }
819
820 if ($ngram_similar > 0) {
821 $unread = 'false';
822 }
823 }
824
825 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
826 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
827
828 $result = db_query($link,
829 "INSERT INTO ttrss_user_entries
830 (ref_id, owner_uid, feed_id, unread, last_read, marked,
831 published, score, tag_cache, label_cache, uuid,
832 last_marked, last_published)
833 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
834 $last_read_qpart, $marked, $published, '$score', '', '',
835 '', $last_marked, $last_published)");
836
837 if (PUBSUBHUBBUB_HUB && $published == 'true') {
838 $rss_link = get_self_url_prefix() .
839 "/public.php?op=rss&id=-2&key=" .
840 get_feed_access_key($link, -2, false, $owner_uid);
841
842 $p = new Publisher(PUBSUBHUBBUB_HUB);
843
844 $pubsub_result = $p->publish_update($rss_link);
845 }
846
847 $result = db_query($link,
848 "SELECT int_id FROM ttrss_user_entries WHERE
849 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
850 feed_id = '$feed' LIMIT 1");
851
852 if (db_num_rows($result) == 1) {
853 $entry_int_id = db_fetch_result($result, 0, "int_id");
854 }
855 } else {
856 if ($debug_enabled) {
857 _debug("update_rss_feed: user record FOUND");
858 }
859
860 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
861 $entry_int_id = db_fetch_result($result, 0, "int_id");
862 }
863
864 if ($debug_enabled) {
865 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
866 }
867
868 $post_needs_update = false;
869 $update_insignificant = false;
870
871 if ($orig_num_comments != $num_comments) {
872 $post_needs_update = true;
873 $update_insignificant = true;
874 }
875
876 if ($entry_plugin_data != $orig_plugin_data) {
877 $post_needs_update = true;
878 $update_insignificant = true;
879 }
880
881 if ($content_hash != $orig_content_hash) {
882 $post_needs_update = true;
883 $update_insignificant = false;
884 }
885
886 if (db_escape_string($link, $orig_title) != $entry_title) {
887 $post_needs_update = true;
888 $update_insignificant = false;
889 }
890
891 // if post needs update, update it and mark all user entries
892 // linking to this post as updated
893 if ($post_needs_update) {
894
895 if (defined('DAEMON_EXTENDED_DEBUG')) {
896 _debug("update_rss_feed: post $entry_guid needs update...");
897 }
898
899 // print "<!-- post $orig_title needs update : $post_needs_update -->";
900
901 db_query($link, "UPDATE ttrss_entries
902 SET title = '$entry_title', content = '$entry_content',
903 content_hash = '$content_hash',
904 updated = '$entry_timestamp_fmt',
905 num_comments = '$num_comments',
906 plugin_data = '$entry_plugin_data'
907 WHERE id = '$ref_id'");
908
909 if (!$update_insignificant) {
910 if ($mark_unread_on_update) {
911 db_query($link, "UPDATE ttrss_user_entries
912 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
913 }
914 }
915 }
916 }
917
918 db_query($link, "COMMIT");
919
920 if ($debug_enabled) {
921 _debug("update_rss_feed: assigning labels...");
922 }
923
924 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
925 $owner_uid, $article_labels);
926
927 if ($debug_enabled) {
928 _debug("update_rss_feed: looking for enclosures...");
929 }
930
931 // enclosures
932
933 $enclosures = array();
934
935 $encs = $item->get_enclosures();
936
937 if (is_array($encs)) {
938 foreach ($encs as $e) {
939 $e_item = array(
940 $e->link, $e->type, $e->length);
941 array_push($enclosures, $e_item);
942 }
943 }
944
945 if ($debug_enabled) {
946 _debug("update_rss_feed: article enclosures:");
947 print_r($enclosures);
948 }
949
950 db_query($link, "BEGIN");
951
952 foreach ($enclosures as $enc) {
953 $enc_url = db_escape_string($link, $enc[0]);
954 $enc_type = db_escape_string($link, $enc[1]);
955 $enc_dur = db_escape_string($link, $enc[2]);
956
957 $result = db_query($link, "SELECT id FROM ttrss_enclosures
958 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
959
960 if (db_num_rows($result) == 0) {
961 db_query($link, "INSERT INTO ttrss_enclosures
962 (content_url, content_type, title, duration, post_id) VALUES
963 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
964 }
965 }
966
967 db_query($link, "COMMIT");
968
969 // check for manual tags (we have to do it here since they're loaded from filters)
970
971 foreach ($article_filters as $f) {
972 if ($f["type"] == "tag") {
973
974 $manual_tags = trim_array(explode(",", $f["param"]));
975
976 foreach ($manual_tags as $tag) {
977 if (tag_is_valid($tag)) {
978 array_push($entry_tags, $tag);
979 }
980 }
981 }
982 }
983
984 // Skip boring tags
985
986 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
987 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
988
989 $filtered_tags = array();
990 $tags_to_cache = array();
991
992 if ($entry_tags && is_array($entry_tags)) {
993 foreach ($entry_tags as $tag) {
994 if (array_search($tag, $boring_tags) === false) {
995 array_push($filtered_tags, $tag);
996 }
997 }
998 }
999
1000 $filtered_tags = array_unique($filtered_tags);
1001
1002 if ($debug_enabled) {
1003 _debug("update_rss_feed: filtered article tags:");
1004 print_r($filtered_tags);
1005 }
1006
1007 // Save article tags in the database
1008
1009 if (count($filtered_tags) > 0) {
1010
1011 db_query($link, "BEGIN");
1012
1013 foreach ($filtered_tags as $tag) {
1014
1015 $tag = sanitize_tag($tag);
1016 $tag = db_escape_string($link, $tag);
1017
1018 if (!tag_is_valid($tag)) continue;
1019
1020 $result = db_query($link, "SELECT id FROM ttrss_tags
1021 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1022 owner_uid = '$owner_uid' LIMIT 1");
1023
1024 if ($result && db_num_rows($result) == 0) {
1025
1026 db_query($link, "INSERT INTO ttrss_tags
1027 (owner_uid,tag_name,post_int_id)
1028 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1029 }
1030
1031 array_push($tags_to_cache, $tag);
1032 }
1033
1034 /* update the cache */
1035
1036 $tags_to_cache = array_unique($tags_to_cache);
1037
1038 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1039
1040 db_query($link, "UPDATE ttrss_user_entries
1041 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1042 AND owner_uid = $owner_uid");
1043
1044 db_query($link, "COMMIT");
1045 }
1046
1047 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1048 if ($debug_enabled) {
1049 _debug("update_rss_feed: auto-assigning labels...");
1050 }
1051
1052 foreach ($labels as $label) {
1053 $caption = preg_quote($label["caption"]);
1054
1055 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1056 if (!labels_contains_caption($article_labels, $caption)) {
1057 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1058 }
1059 }
1060 }
1061 }
1062
1063 if ($debug_enabled) {
1064 _debug("update_rss_feed: article processed");
1065 }
1066 }
1067
1068 if (!$last_updated) {
1069 if ($debug_enabled) {
1070 _debug("update_rss_feed: new feed, catching it up...");
1071 }
1072 catchup_feed($link, $feed, false, $owner_uid);
1073 }
1074
1075 if ($debug_enabled) {
1076 _debug("purging feed...");
1077 }
1078
1079 purge_feed($link, $feed, 0, $debug_enabled);
1080
1081 db_query($link, "UPDATE ttrss_feeds
1082 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1083
1084 // db_query($link, "COMMIT");
1085
1086 } else {
1087
1088 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1089
1090 if ($debug_enabled) {
1091 _debug("update_rss_feed: error fetching feed: $error_msg");
1092 }
1093
1094 db_query($link,
1095 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1096 last_updated = NOW() WHERE id = '$feed'");
1097 }
1098
1099 unset($rss);
1100
1101 if ($debug_enabled) {
1102 _debug("update_rss_feed: done");
1103 }
1104
1105 }
1106
1107 function cache_images($html, $site_url, $debug) {
1108 $cache_dir = CACHE_DIR . "/images";
1109
1110 libxml_use_internal_errors(true);
1111
1112 $charset_hack = '<head>
1113 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1114 </head>';
1115
1116 $doc = new DOMDocument();
1117 $doc->loadHTML($charset_hack . $html);
1118 $xpath = new DOMXPath($doc);
1119
1120 $entries = $xpath->query('(//img[@src])');
1121
1122 foreach ($entries as $entry) {
1123 if ($entry->hasAttribute('src')) {
1124 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1125
1126 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1127
1128 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1129
1130 if (!file_exists($local_filename)) {
1131 $file_content = fetch_file_contents($src);
1132
1133 if ($file_content && strlen($file_content) > 1024) {
1134 file_put_contents($local_filename, $file_content);
1135 }
1136 }
1137
1138 if (file_exists($local_filename)) {
1139 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1140 base64_encode($src));
1141 }
1142 }
1143 }
1144
1145 $node = $doc->getElementsByTagName('body')->item(0);
1146
1147 return $doc->saveXML($node);
1148 }
1149
1150 function expire_lock_files($debug) {
1151 if ($debug) _debug("Removing old lock files...");
1152
1153 $num_deleted = 0;
1154
1155 if (is_writable(LOCK_DIRECTORY)) {
1156 $files = glob(LOCK_DIRECTORY . "/*.lock");
1157
1158 if ($files) {
1159 foreach ($files as $file) {
1160 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1161 unlink($file);
1162 ++$num_deleted;
1163 }
1164 }
1165 }
1166 }
1167
1168 if ($debug) _debug("Removed $num_deleted files.");
1169 }
1170
1171 function expire_cached_files($debug) {
1172 foreach (array("simplepie", "images", "export") as $dir) {
1173 $cache_dir = CACHE_DIR . "/$dir";
1174
1175 if ($debug) _debug("Expiring $cache_dir");
1176
1177 $num_deleted = 0;
1178
1179 if (is_writable($cache_dir)) {
1180 $files = glob("$cache_dir/*");
1181
1182 if ($files) {
1183 foreach ($files as $file) {
1184 if (time() - filemtime($file) > 86400*7) {
1185 unlink($file);
1186
1187 ++$num_deleted;
1188 }
1189 }
1190 }
1191 }
1192
1193 if ($debug) _debug("Removed $num_deleted files.");
1194 }
1195 }
1196
1197 /**
1198 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1199 * Returns the url query as associative array
1200 *
1201 * @param string query
1202 * @return array params
1203 */
1204 function convertUrlQuery($query) {
1205 $queryParts = explode('&', $query);
1206
1207 $params = array();
1208
1209 foreach ($queryParts as $param) {
1210 $item = explode('=', $param);
1211 $params[$item[0]] = $item[1];
1212 }
1213
1214 return $params;
1215 }
1216
1217 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1218 $matches = array();
1219
1220 foreach ($filters as $filter) {
1221 $match_any_rule = $filter["match_any_rule"];
1222 $inverse = $filter["inverse"];
1223 $filter_match = false;
1224
1225 foreach ($filter["rules"] as $rule) {
1226 $match = false;
1227 $reg_exp = $rule["reg_exp"];
1228 $rule_inverse = $rule["inverse"];
1229
1230 if (!$reg_exp)
1231 continue;
1232
1233 switch ($rule["type"]) {
1234 case "title":
1235 $match = @preg_match("/$reg_exp/i", $title);
1236 break;
1237 case "content":
1238 // we don't need to deal with multiline regexps
1239 $content = preg_replace("/[\r\n\t]/", "", $content);
1240
1241 $match = @preg_match("/$reg_exp/i", $content);
1242 break;
1243 case "both":
1244 // we don't need to deal with multiline regexps
1245 $content = preg_replace("/[\r\n\t]/", "", $content);
1246
1247 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1248 break;
1249 case "link":
1250 $match = @preg_match("/$reg_exp/i", $link);
1251 break;
1252 case "author":
1253 $match = @preg_match("/$reg_exp/i", $author);
1254 break;
1255 case "tag":
1256 $tag_string = join(",", $tags);
1257 $match = @preg_match("/$reg_exp/i", $tag_string);
1258 break;
1259 }
1260
1261 if ($rule_inverse) $match = !$match;
1262
1263 if ($match_any_rule) {
1264 if ($match) {
1265 $filter_match = true;
1266 break;
1267 }
1268 } else {
1269 $filter_match = $match;
1270 if (!$match) {
1271 break;
1272 }
1273 }
1274 }
1275
1276 if ($inverse) $filter_match = !$filter_match;
1277
1278 if ($filter_match) {
1279 foreach ($filter["actions"] AS $action) {
1280 array_push($matches, $action);
1281 }
1282 }
1283 }
1284
1285 return $matches;
1286 }
1287
1288 function find_article_filter($filters, $filter_name) {
1289 foreach ($filters as $f) {
1290 if ($f["type"] == $filter_name) {
1291 return $f;
1292 };
1293 }
1294 return false;
1295 }
1296
1297 function find_article_filters($filters, $filter_name) {
1298 $results = array();
1299
1300 foreach ($filters as $f) {
1301 if ($f["type"] == $filter_name) {
1302 array_push($results, $f);
1303 };
1304 }
1305 return $results;
1306 }
1307
1308 function calculate_article_score($filters) {
1309 $score = 0;
1310
1311 foreach ($filters as $f) {
1312 if ($f["type"] == "score") {
1313 $score += $f["param"];
1314 };
1315 }
1316 return $score;
1317 }
1318
1319 function labels_contains_caption($labels, $caption) {
1320 foreach ($labels as $label) {
1321 if ($label[1] == $caption) {
1322 return true;
1323 }
1324 }
1325
1326 return false;
1327 }
1328
1329 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1330 foreach ($filters as $f) {
1331 if ($f["type"] == "label") {
1332 if (!labels_contains_caption($article_labels, $f["param"])) {
1333 label_add_article($link, $id, $f["param"], $owner_uid);
1334 }
1335 }
1336 }
1337 }
1338
1339 function make_guid_from_title($title) {
1340 return preg_replace("/[ \"\',.:;]/", "-",
1341 mb_strtolower(strip_tags($title), 'utf-8'));
1342 }
1343
1344
1345 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/