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