]> 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 branch 'master' into css-feedtree-counter
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define_default('DAEMON_FEED_LIMIT', 500);
4 define_default('DAEMON_SLEEP_INTERVAL', 120);
5
6 function update_feedbrowser_cache() {
7
8 $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
13
14 db_query("BEGIN");
15
16 db_query("DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($line["subscribers"]);
22 $feed_url = db_escape_string($line["feed_url"]);
23 $title = db_escape_string($line["title"]);
24 $site_url = db_escape_string($line["site_url"]);
25
26 $tmp_result = db_query("SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query("INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query("COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 $schema_version = get_schema_version();
64
65 if ($schema_version != SCHEMA_VERSION) {
66 die("Schema version is wrong, please upgrade the database.\n");
67 }
68
69 define('PREFS_NO_CACHE', true);
70
71 // Test if the user has loggued in recently. If not, it does not update its feeds.
72 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
73 if (DB_TYPE == "pgsql") {
74 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
75 } else {
76 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
77 }
78 } else {
79 $login_thresh_qpart = "";
80 }
81
82 // Test if the feed need a update (update interval exceded).
83 if (DB_TYPE == "pgsql") {
84 $update_limit_qpart = "AND ((
85 ttrss_feeds.update_interval = 0
86 AND ttrss_user_prefs.value != '-1'
87 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
88 ) OR (
89 ttrss_feeds.update_interval > 0
90 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
91 ) OR ttrss_feeds.last_updated IS NULL
92 OR last_updated = '1970-01-01 00:00:00')";
93 } else {
94 $update_limit_qpart = "AND ((
95 ttrss_feeds.update_interval = 0
96 AND ttrss_user_prefs.value != '-1'
97 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
98 ) OR (
99 ttrss_feeds.update_interval > 0
100 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
101 ) OR ttrss_feeds.last_updated IS NULL
102 OR last_updated = '1970-01-01 00:00:00')";
103 }
104
105 // Test if feed is currently being updated by another process.
106 if (DB_TYPE == "pgsql") {
107 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')";
108 } else {
109 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))";
110 }
111
112 // Test if there is a limit to number of updated feeds
113 $query_limit = "";
114 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
115
116 $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated
117 FROM
118 ttrss_feeds, ttrss_users, ttrss_user_prefs
119 WHERE
120 ttrss_feeds.owner_uid = ttrss_users.id
121 AND ttrss_users.id = ttrss_user_prefs.owner_uid
122 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
123 $login_thresh_qpart $update_limit_qpart
124 $updstart_thresh_qpart
125 ORDER BY last_updated $query_limit";
126
127 // We search for feed needing update.
128 $result = db_query($query);
129
130 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
131
132 // Here is a little cache magic in order to minimize risk of double feed updates.
133 $feeds_to_update = array();
134 while ($line = db_fetch_assoc($result)) {
135 array_push($feeds_to_update, db_escape_string($line['feed_url']));
136 }
137
138 // We update the feed last update started date before anything else.
139 // There is no lag due to feed contents downloads
140 // It prevent an other process to update the same feed.
141
142 if(count($feeds_to_update) > 0) {
143 $feeds_quoted = array();
144
145 foreach ($feeds_to_update as $feed) {
146 array_push($feeds_quoted, "'" . db_escape_string($feed) . "'");
147 }
148
149 db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
150 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
151 }
152
153 $nf = 0;
154
155 // For each feed, we call the feed update function.
156 foreach ($feeds_to_update as $feed) {
157 if($debug) _debug("Base feed: $feed");
158
159 //update_rss_feed($line["id"], true);
160
161 // since we have the data cached, we can deal with other feeds with the same url
162
163 $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid
164 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
165 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
166 ttrss_users.id = ttrss_user_prefs.owner_uid AND
167 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
168 feed_url = '".db_escape_string($feed)."' AND
169 (ttrss_feeds.update_interval > 0 OR
170 ttrss_user_prefs.value != '-1')
171 $login_thresh_qpart
172 ORDER BY ttrss_feeds.id $query_limit");
173
174 if (db_num_rows($tmp_result) > 0) {
175 while ($tline = db_fetch_assoc($tmp_result)) {
176 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]);
177 update_rss_feed($tline["id"], true);
178 ++$nf;
179 }
180 }
181 }
182
183 require_once "digest.php";
184
185 // Send feed digests by email if needed.
186 send_headlines_digests($debug);
187
188 return $nf;
189
190 } // function update_daemon_common
191
192 // ignore_daemon is not used
193 function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false,
194 $override_url = false, $override_data = false) {
195
196 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
197
198 _debug("start", $debug_enabled);
199
200 $result = db_query("SELECT id,update_interval,auth_login,
201 feed_url,auth_pass,cache_images,last_updated,
202 mark_unread_on_update, owner_uid,
203 pubsub_state, auth_pass_encrypted,
204 (SELECT max(date_entered) FROM
205 ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp
206 FROM ttrss_feeds WHERE id = '$feed'");
207
208 if (db_num_rows($result) == 0) {
209 _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled);
210 return false;
211 }
212
213 $last_updated = db_fetch_result($result, 0, "last_updated");
214 $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp"));
215
216 if (defined('_DISABLE_HTTP_304'))
217 $last_article_timestamp = 0;
218
219 $owner_uid = db_fetch_result($result, 0, "owner_uid");
220 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
221 0, "mark_unread_on_update"));
222 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
223 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
224 0, "auth_pass_encrypted"));
225
226 db_query("UPDATE ttrss_feeds SET last_update_started = NOW()
227 WHERE id = '$feed'");
228
229 $auth_login = db_fetch_result($result, 0, "auth_login");
230 $auth_pass = db_fetch_result($result, 0, "auth_pass");
231
232 if ($auth_pass_encrypted) {
233 require_once "crypt.php";
234 $auth_pass = decrypt_string($auth_pass);
235 }
236
237 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
238 $fetch_url = db_fetch_result($result, 0, "feed_url");
239
240 $feed = db_escape_string($feed);
241
242 if ($override_url) $fetch_url = $override_url;
243
244 $date_feed_processed = date('Y-m-d H:i');
245
246 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml";
247
248 $rss = false;
249 $rss_hash = false;
250 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
251
252 $force_refetch = isset($_REQUEST["force_refetch"]);
253
254 if ($override_data) {
255 $feed_data = $override_data;
256 }
257
258 if (!$feed_data && file_exists($cache_filename) &&
259 is_readable($cache_filename) &&
260 !$auth_login && !$auth_pass &&
261 filemtime($cache_filename) > time() - 30) {
262
263 _debug("using local cache.", $debug_enabled);
264
265 @$feed_data = file_get_contents($cache_filename);
266
267 if ($feed_data) {
268 $rss_hash = sha1($feed_data);
269 }
270
271 } else {
272 _debug("local cache will not be used for this feed", $debug_enabled);
273 }
274
275 if (!$feed_data) {
276 _debug("fetching [$fetch_url]...", $debug_enabled);
277 _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled);
278
279 $feed_data = fetch_file_contents($fetch_url, false,
280 $auth_login, $auth_pass, false,
281 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
282 $force_refetch ? 0 : $last_article_timestamp);
283
284 global $fetch_curl_used;
285
286 if (!$fetch_curl_used) {
287 $tmp = @gzdecode($feed_data);
288
289 if ($tmp) $feed_data = $tmp;
290 }
291
292 $feed_data = trim($feed_data);
293
294 _debug("fetch done.", $debug_enabled);
295
296 if ($feed_data) {
297 $error = verify_feed_xml($feed_data);
298
299 if ($error) {
300 _debug("error verifying XML, code: " . $error->code, $debug_enabled);
301
302 if ($error->code == 26) {
303 _debug("got error 26, trying to decode entities...", $debug_enabled);
304
305 $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8');
306
307 $error = verify_feed_xml($feed_data);
308
309 if ($error) $feed_data = '';
310 }
311 }
312 }
313 }
314
315 if (!$feed_data) {
316 global $fetch_last_error;
317 global $fetch_last_error_code;
318
319 _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled);
320
321 $error_escaped = '';
322
323 // If-Modified-Since
324 if ($fetch_last_error_code != 304) {
325 $error_escaped = db_escape_string($fetch_last_error);
326 } else {
327 _debug("source claims data not modified, nothing to do.", $debug_enabled);
328 }
329
330 db_query(
331 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
332 last_updated = NOW() WHERE id = '$feed'");
333
334 return;
335 }
336
337 $pluginhost = new PluginHost();
338 $pluginhost->set_debug($debug_enabled);
339 $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid);
340
341 $pluginhost->load(PLUGINS, PluginHost::KIND_ALL);
342 $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid);
343 $pluginhost->load_data();
344
345 foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) {
346 $feed_data = $plugin->hook_feed_fetched($feed_data);
347 }
348
349 // set last update to now so if anything *simplepie* crashes later we won't be
350 // continuously failing on the same feed
351 //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'");
352
353 if (!$rss) {
354 $rss = new FeedParser($feed_data);
355 $rss->init();
356 }
357
358 // print_r($rss);
359
360 $feed = db_escape_string($feed);
361
362 if (!$rss->error()) {
363
364 // cache data for later
365 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
366 $new_rss_hash = sha1($rss_data);
367
368 if ($new_rss_hash != $rss_hash && count($rss->get_items()) > 0 ) {
369 _debug("saving $cache_filename", $debug_enabled);
370 @file_put_contents($cache_filename, $feed_data);
371 }
372 }
373
374 // We use local pluginhost here because we need to load different per-user feed plugins
375 $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
376
377 _debug("processing feed data...", $debug_enabled);
378
379 // db_query("BEGIN");
380
381 if (DB_TYPE == "pgsql") {
382 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
383 } else {
384 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
385 }
386
387 $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color,
388 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
389 favicon_needs_check
390 FROM ttrss_feeds WHERE id = '$feed'");
391
392 $registered_title = db_fetch_result($result, 0, "title");
393 $orig_site_url = db_fetch_result($result, 0, "site_url");
394 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
395 "favicon_needs_check"));
396 $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color");
397
398 $owner_uid = db_fetch_result($result, 0, "owner_uid");
399
400 $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
401
402 _debug("site_url: $site_url", $debug_enabled);
403 _debug("feed_title: " . $rss->get_title(), $debug_enabled);
404
405 if ($favicon_needs_check || $force_refetch) {
406
407 /* terrible hack: if we crash on floicon shit here, we won't check
408 * the icon avgcolor again (unless the icon got updated) */
409
410 $favicon_file = ICONS_DIR . "/$feed.ico";
411 $favicon_modified = @filemtime($favicon_file);
412
413 _debug("checking favicon...", $debug_enabled);
414
415 check_feed_favicon($site_url, $feed);
416 $favicon_modified_new = @filemtime($favicon_file);
417
418 if ($favicon_modified_new > $favicon_modified)
419 $favicon_avg_color = '';
420
421 if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') {
422 require_once "colors.php";
423
424 db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE
425 id = '$feed'");
426
427 $favicon_color = db_escape_string(
428 calculate_avg_color($favicon_file));
429
430 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
431 } else if ($favicon_avg_color == 'fail') {
432 _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled);
433 }
434
435 db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW()
436 $favicon_colorstring
437 WHERE id = '$feed'");
438 }
439
440 if (!$registered_title || $registered_title == "[Unknown]") {
441
442 $feed_title = db_escape_string($rss->get_title());
443
444 if ($feed_title) {
445 _debug("registering title: $feed_title", $debug_enabled);
446
447 db_query("UPDATE ttrss_feeds SET
448 title = '$feed_title' WHERE id = '$feed'");
449 }
450 }
451
452 if ($site_url && $orig_site_url != $site_url) {
453 db_query("UPDATE ttrss_feeds SET
454 site_url = '$site_url' WHERE id = '$feed'");
455 }
456
457 _debug("loading filters & labels...", $debug_enabled);
458
459 $filters = load_filters($feed, $owner_uid);
460 $labels = get_all_labels($owner_uid);
461
462 _debug("" . count($filters) . " filters loaded.", $debug_enabled);
463
464 $items = $rss->get_items();
465
466 if (!is_array($items)) {
467 _debug("no articles found.", $debug_enabled);
468
469 db_query("UPDATE ttrss_feeds
470 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
471
472 return; // no articles
473 }
474
475 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
476
477 _debug("checking for PUSH hub...", $debug_enabled);
478
479 $feed_hub_url = false;
480
481 $links = $rss->get_links('hub');
482
483 if ($links && is_array($links)) {
484 foreach ($links as $l) {
485 $feed_hub_url = $l;
486 break;
487 }
488 }
489
490 _debug("feed hub url: $feed_hub_url", $debug_enabled);
491
492 if ($feed_hub_url && function_exists('curl_init') &&
493 !ini_get("open_basedir")) {
494
495 require_once 'lib/pubsubhubbub/subscriber.php';
496
497 $callback_url = get_self_url_prefix() .
498 "/public.php?op=pubsub&id=$feed";
499
500 $s = new Subscriber($feed_hub_url, $callback_url);
501
502 $rc = $s->subscribe($fetch_url);
503
504 _debug("feed hub url found, subscribe request sent.", $debug_enabled);
505
506 db_query("UPDATE ttrss_feeds SET pubsub_state = 1
507 WHERE id = '$feed'");
508 }
509 }
510
511 _debug("processing articles...", $debug_enabled);
512
513 foreach ($items as $item) {
514 if ($_REQUEST['xdebug'] == 3) {
515 print_r($item);
516 }
517
518 $entry_guid = $item->get_id();
519 if (!$entry_guid) $entry_guid = $item->get_link();
520 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
521
522 _debug("f_guid $entry_guid", $debug_enabled);
523
524 if (!$entry_guid) continue;
525
526 $entry_guid = "$owner_uid,$entry_guid";
527
528 $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid));
529
530 _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled);
531
532 $entry_timestamp = "";
533
534 $entry_timestamp = $item->get_date();
535
536 _debug("orig date: " . $item->get_date(), $debug_enabled);
537
538 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
539 $entry_timestamp = time();
540 $no_orig_date = 'true';
541 } else {
542 $no_orig_date = 'false';
543 }
544
545 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
546
547 _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled);
548
549 // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8');
550 // $entry_title = decode_numeric_entities($entry_title);
551 $entry_title = $item->get_title();
552
553 $entry_link = rewrite_relative_url($site_url, $item->get_link());
554
555 _debug("title $entry_title", $debug_enabled);
556 _debug("link $entry_link", $debug_enabled);
557
558 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
559
560 $entry_content = $item->get_content();
561 if (!$entry_content) $entry_content = $item->get_description();
562
563 if ($_REQUEST["xdebug"] == 2) {
564 print "content: ";
565 print $entry_content;
566 print "\n";
567 }
568
569 $entry_comments = $item->get_comments_url();
570 $entry_author = $item->get_author();
571
572 $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245));
573
574 $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245));
575 $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245));
576
577 $num_comments = (int) $item->get_comments_count();
578
579 _debug("author $entry_author", $debug_enabled);
580 _debug("num_comments: $num_comments", $debug_enabled);
581 _debug("looking for tags...", $debug_enabled);
582
583 // parse <category> entries into tags
584
585 $additional_tags = array();
586
587 $additional_tags_src = $item->get_categories();
588
589 if (is_array($additional_tags_src)) {
590 foreach ($additional_tags_src as $tobj) {
591 array_push($additional_tags, $tobj);
592 }
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 _debug("tags found: " . join(",", $entry_tags), $debug_enabled);
601
602 _debug("done collecting data.", $debug_enabled);
603
604 // TODO: less memory-hungry implementation
605
606 _debug("applying plugin filters..", $debug_enabled);
607
608 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
609 $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
610 WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
611
612 if (db_num_rows($result) != 0) {
613 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
614 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
615 "content" => db_fetch_result($result, 0, "content"),
616 "link" => db_fetch_result($result, 0, "link"),
617 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
618 "author" => db_fetch_result($result, 0, "author"));
619 } else {
620 $entry_plugin_data = "";
621 $stored_article = array();
622 }
623
624 $article = array("owner_uid" => $owner_uid, // read only
625 "guid" => $entry_guid, // read only
626 "title" => $entry_title,
627 "content" => $entry_content,
628 "link" => $entry_link,
629 "tags" => $entry_tags,
630 "plugin_data" => $entry_plugin_data,
631 "author" => $entry_author,
632 "stored" => $stored_article);
633
634 foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) {
635 $article = $plugin->hook_article_filter($article);
636 }
637
638 $entry_tags = $article["tags"];
639 $entry_guid = db_escape_string($entry_guid);
640 $entry_title = db_escape_string($article["title"]);
641 $entry_author = db_escape_string($article["author"]);
642 $entry_link = db_escape_string($article["link"]);
643 $entry_plugin_data = db_escape_string($article["plugin_data"]);
644 $entry_content = $article["content"]; // escaped below
645
646
647 _debug("plugin data: $entry_plugin_data", $debug_enabled);
648
649 if ($cache_images && is_writable(CACHE_DIR . '/images'))
650 cache_images($entry_content, $site_url, $debug_enabled);
651
652 $entry_content = db_escape_string($entry_content, false);
653
654 $content_hash = "SHA1:" . sha1($entry_content);
655
656 db_query("BEGIN");
657
658 $result = db_query("SELECT id FROM ttrss_entries
659 WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
660
661 if (db_num_rows($result) == 0) {
662
663 _debug("base guid [$entry_guid] not found", $debug_enabled);
664
665 // base post entry does not exist, create it
666
667 $result = db_query(
668 "INSERT INTO ttrss_entries
669 (title,
670 guid,
671 link,
672 updated,
673 content,
674 content_hash,
675 cached_content,
676 no_orig_date,
677 date_updated,
678 date_entered,
679 comments,
680 num_comments,
681 plugin_data,
682 author)
683 VALUES
684 ('$entry_title',
685 '$entry_guid_hashed',
686 '$entry_link',
687 '$entry_timestamp_fmt',
688 '$entry_content',
689 '$content_hash',
690 '',
691 $no_orig_date,
692 NOW(),
693 '$date_feed_processed',
694 '$entry_comments',
695 '$num_comments',
696 '$entry_plugin_data',
697 '$entry_author')");
698
699 $article_labels = array();
700
701 } else {
702 // we keep encountering the entry in feeds, so we need to
703 // update date_updated column so that we don't get horrible
704 // dupes when the entry gets purged and reinserted again e.g.
705 // in the case of SLOW SLOW OMG SLOW updating feeds
706
707 $base_entry_id = db_fetch_result($result, 0, "id");
708
709 db_query("UPDATE ttrss_entries SET date_updated = NOW()
710 WHERE id = '$base_entry_id'");
711
712 $article_labels = get_article_labels($base_entry_id, $owner_uid);
713 }
714
715 // now it should exist, if not - bad luck then
716
717 $result = db_query("SELECT
718 id,content_hash,no_orig_date,title,plugin_data,guid,
719 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
720 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
721 num_comments
722 FROM
723 ttrss_entries
724 WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
725
726 $entry_ref_id = 0;
727 $entry_int_id = 0;
728
729 if (db_num_rows($result) == 1) {
730
731 _debug("base guid found, checking for user record", $debug_enabled);
732
733 // this will be used below in update handler
734 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
735 $orig_title = db_fetch_result($result, 0, "title");
736 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
737 $orig_date_updated = strtotime(db_fetch_result($result,
738 0, "date_updated"));
739 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
740
741 $ref_id = db_fetch_result($result, 0, "id");
742 $entry_ref_id = $ref_id;
743
744 /* $stored_guid = db_fetch_result($result, 0, "guid");
745 if ($stored_guid != $entry_guid_hashed) {
746 if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled);
747
748 db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
749 id = '$ref_id'");
750 } */
751
752 // check for user post link to main table
753
754 // do we allow duplicate posts with same GUID in different feeds?
755 if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
756 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
757 } else {
758 $dupcheck_qpart = "";
759 }
760
761 /* Collect article tags here so we could filter by them: */
762
763 $article_filters = get_article_filters($filters, $entry_title,
764 $entry_content, $entry_link, $entry_timestamp, $entry_author,
765 $entry_tags);
766
767 if ($debug_enabled) {
768 _debug("article filters: ", $debug_enabled);
769 if (count($article_filters) != 0) {
770 print_r($article_filters);
771 }
772 }
773
774 if (find_article_filter($article_filters, "filter")) {
775 db_query("COMMIT"); // close transaction in progress
776 continue;
777 }
778
779 $score = calculate_article_score($article_filters);
780
781 _debug("initial score: $score", $debug_enabled);
782
783 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
784 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
785 $dupcheck_qpart";
786
787 // if ($_REQUEST["xdebug"]) print "$query\n";
788
789 $result = db_query($query);
790
791 // okay it doesn't exist - create user entry
792 if (db_num_rows($result) == 0) {
793
794 _debug("user record not found, creating...", $debug_enabled);
795
796 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
797 $unread = 'true';
798 $last_read_qpart = 'NULL';
799 } else {
800 $unread = 'false';
801 $last_read_qpart = 'NOW()';
802 }
803
804 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
805 $marked = 'true';
806 } else {
807 $marked = 'false';
808 }
809
810 if (find_article_filter($article_filters, 'publish')) {
811 $published = 'true';
812 } else {
813 $published = 'false';
814 }
815
816 // N-grams
817
818 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
819
820 $result = db_query("SELECT COUNT(*) AS similar FROM
821 ttrss_entries,ttrss_user_entries
822 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
823 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
824 AND owner_uid = $owner_uid");
825
826 $ngram_similar = db_fetch_result($result, 0, "similar");
827
828 _debug("N-gram similar results: $ngram_similar", $debug_enabled);
829
830 if ($ngram_similar > 0) {
831 $unread = 'false';
832 }
833 }
834
835 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
836 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
837
838 $result = db_query(
839 "INSERT INTO ttrss_user_entries
840 (ref_id, owner_uid, feed_id, unread, last_read, marked,
841 published, score, tag_cache, label_cache, uuid,
842 last_marked, last_published)
843 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
844 $last_read_qpart, $marked, $published, '$score', '', '',
845 '', $last_marked, $last_published)");
846
847 if (PUBSUBHUBBUB_HUB && $published == 'true') {
848 $rss_link = get_self_url_prefix() .
849 "/public.php?op=rss&id=-2&key=" .
850 get_feed_access_key(-2, false, $owner_uid);
851
852 $p = new Publisher(PUBSUBHUBBUB_HUB);
853
854 $pubsub_result = $p->publish_update($rss_link);
855 }
856
857 $result = db_query(
858 "SELECT int_id FROM ttrss_user_entries WHERE
859 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
860 feed_id = '$feed' LIMIT 1");
861
862 if (db_num_rows($result) == 1) {
863 $entry_int_id = db_fetch_result($result, 0, "int_id");
864 }
865 } else {
866 _debug("user record FOUND", $debug_enabled);
867
868 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
869 $entry_int_id = db_fetch_result($result, 0, "int_id");
870 }
871
872 _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled);
873
874 $post_needs_update = false;
875 $update_insignificant = false;
876
877 if ($orig_num_comments != $num_comments) {
878 $post_needs_update = true;
879 $update_insignificant = true;
880 }
881
882 if ($entry_plugin_data != $orig_plugin_data) {
883 $post_needs_update = true;
884 $update_insignificant = true;
885 }
886
887 if ($content_hash != $orig_content_hash) {
888 $post_needs_update = true;
889 $update_insignificant = false;
890 }
891
892 if (db_escape_string($orig_title) != $entry_title) {
893 $post_needs_update = true;
894 $update_insignificant = false;
895 }
896
897 // if post needs update, update it and mark all user entries
898 // linking to this post as updated
899 if ($post_needs_update) {
900
901 if (defined('DAEMON_EXTENDED_DEBUG')) {
902 _debug("post $entry_guid_hashed needs update...", $debug_enabled);
903 }
904
905 // print "<!-- post $orig_title needs update : $post_needs_update -->";
906
907 db_query("UPDATE ttrss_entries
908 SET title = '$entry_title', content = '$entry_content',
909 content_hash = '$content_hash',
910 updated = '$entry_timestamp_fmt',
911 num_comments = '$num_comments',
912 plugin_data = '$entry_plugin_data'
913 WHERE id = '$ref_id'");
914
915 if (!$update_insignificant) {
916 if ($mark_unread_on_update) {
917 db_query("UPDATE ttrss_user_entries
918 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
919 }
920 }
921 }
922 }
923
924 db_query("COMMIT");
925
926 _debug("assigning labels...", $debug_enabled);
927
928 assign_article_to_label_filters($entry_ref_id, $article_filters,
929 $owner_uid, $article_labels);
930
931 _debug("looking for enclosures...", $debug_enabled);
932
933 // enclosures
934
935 $enclosures = array();
936
937 $encs = $item->get_enclosures();
938
939 if (is_array($encs)) {
940 foreach ($encs as $e) {
941 $e_item = array(
942 $e->link, $e->type, $e->length);
943 array_push($enclosures, $e_item);
944 }
945 }
946
947 if ($debug_enabled) {
948 _debug("article enclosures:", $debug_enabled);
949 print_r($enclosures);
950 }
951
952 db_query("BEGIN");
953
954 foreach ($enclosures as $enc) {
955 $enc_url = db_escape_string($enc[0]);
956 $enc_type = db_escape_string($enc[1]);
957 $enc_dur = db_escape_string($enc[2]);
958
959 $result = db_query("SELECT id FROM ttrss_enclosures
960 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
961
962 if (db_num_rows($result) == 0) {
963 db_query("INSERT INTO ttrss_enclosures
964 (content_url, content_type, title, duration, post_id) VALUES
965 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
966 }
967 }
968
969 db_query("COMMIT");
970
971 // check for manual tags (we have to do it here since they're loaded from filters)
972
973 foreach ($article_filters as $f) {
974 if ($f["type"] == "tag") {
975
976 $manual_tags = trim_array(explode(",", $f["param"]));
977
978 foreach ($manual_tags as $tag) {
979 if (tag_is_valid($tag)) {
980 array_push($entry_tags, $tag);
981 }
982 }
983 }
984 }
985
986 // Skip boring tags
987
988 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref(
989 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
990
991 $filtered_tags = array();
992 $tags_to_cache = array();
993
994 if ($entry_tags && is_array($entry_tags)) {
995 foreach ($entry_tags as $tag) {
996 if (array_search($tag, $boring_tags) === false) {
997 array_push($filtered_tags, $tag);
998 }
999 }
1000 }
1001
1002 $filtered_tags = array_unique($filtered_tags);
1003
1004 if ($debug_enabled) {
1005 _debug("filtered article tags:", $debug_enabled);
1006 print_r($filtered_tags);
1007 }
1008
1009 // Save article tags in the database
1010
1011 if (count($filtered_tags) > 0) {
1012
1013 db_query("BEGIN");
1014
1015 foreach ($filtered_tags as $tag) {
1016
1017 $tag = sanitize_tag($tag);
1018 $tag = db_escape_string($tag);
1019
1020 if (!tag_is_valid($tag)) continue;
1021
1022 $result = db_query("SELECT id FROM ttrss_tags
1023 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1024 owner_uid = '$owner_uid' LIMIT 1");
1025
1026 if ($result && db_num_rows($result) == 0) {
1027
1028 db_query("INSERT INTO ttrss_tags
1029 (owner_uid,tag_name,post_int_id)
1030 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1031 }
1032
1033 array_push($tags_to_cache, $tag);
1034 }
1035
1036 /* update the cache */
1037
1038 $tags_to_cache = array_unique($tags_to_cache);
1039
1040 $tags_str = db_escape_string(join(",", $tags_to_cache));
1041
1042 db_query("UPDATE ttrss_user_entries
1043 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1044 AND owner_uid = $owner_uid");
1045
1046 db_query("COMMIT");
1047 }
1048
1049 if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1050 _debug("auto-assigning labels...", $debug_enabled);
1051
1052 foreach ($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, $caption)) {
1057 label_add_article($entry_ref_id, $caption, $owner_uid);
1058 }
1059 }
1060 }
1061 }
1062
1063 _debug("article processed", $debug_enabled);
1064 }
1065
1066 if (!$last_updated) {
1067 _debug("new feed, catching it up...", $debug_enabled);
1068 catchup_feed($feed, false, $owner_uid);
1069 }
1070
1071 _debug("purging feed...", $debug_enabled);
1072
1073 purge_feed($feed, 0, $debug_enabled);
1074
1075 db_query("UPDATE ttrss_feeds
1076 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1077
1078 // db_query("COMMIT");
1079
1080 } else {
1081
1082 $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245));
1083
1084 _debug("error fetching feed: $error_msg", $debug_enabled);
1085
1086 db_query(
1087 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1088 last_updated = NOW() WHERE id = '$feed'");
1089 }
1090
1091 unset($rss);
1092
1093 _debug("done", $debug_enabled);
1094 }
1095
1096 function cache_images($html, $site_url, $debug) {
1097 $cache_dir = CACHE_DIR . "/images";
1098
1099 libxml_use_internal_errors(true);
1100
1101 $charset_hack = '<head>
1102 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1103 </head>';
1104
1105 $doc = new DOMDocument();
1106 $doc->loadHTML($charset_hack . $html);
1107 $xpath = new DOMXPath($doc);
1108
1109 $entries = $xpath->query('(//img[@src])');
1110
1111 foreach ($entries as $entry) {
1112 if ($entry->hasAttribute('src')) {
1113 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1114
1115 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1116
1117 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1118
1119 if (!file_exists($local_filename)) {
1120 $file_content = fetch_file_contents($src);
1121
1122 if ($file_content && strlen($file_content) > 1024) {
1123 file_put_contents($local_filename, $file_content);
1124 }
1125 }
1126
1127 if (file_exists($local_filename)) {
1128 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1129 base64_encode($src));
1130 }
1131 }
1132 }
1133
1134 $node = $doc->getElementsByTagName('body')->item(0);
1135
1136 return $doc->saveXML($node);
1137 }
1138
1139 function expire_error_log($debug) {
1140 if ($debug) _debug("Removing old error log entries...");
1141
1142 if (DB_TYPE == "pgsql") {
1143 db_query("DELETE FROM ttrss_error_log
1144 WHERE created_at < NOW() - INTERVAL '7 days'");
1145 } else {
1146 db_query("DELETE FROM ttrss_error_log
1147 WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
1148 }
1149
1150 }
1151
1152 function expire_lock_files($debug) {
1153 //if ($debug) _debug("Removing old lock files...");
1154
1155 $num_deleted = 0;
1156
1157 if (is_writable(LOCK_DIRECTORY)) {
1158 $files = glob(LOCK_DIRECTORY . "/*.lock");
1159
1160 if ($files) {
1161 foreach ($files as $file) {
1162 if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) {
1163 unlink($file);
1164 ++$num_deleted;
1165 }
1166 }
1167 }
1168 }
1169
1170 if ($debug) _debug("Removed $num_deleted old lock files.");
1171 }
1172
1173 function expire_cached_files($debug) {
1174 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1175 $cache_dir = CACHE_DIR . "/$dir";
1176
1177 // if ($debug) _debug("Expiring $cache_dir");
1178
1179 $num_deleted = 0;
1180
1181 if (is_writable($cache_dir)) {
1182 $files = glob("$cache_dir/*");
1183
1184 if ($files) {
1185 foreach ($files as $file) {
1186 if (time() - filemtime($file) > 86400*7) {
1187 unlink($file);
1188
1189 ++$num_deleted;
1190 }
1191 }
1192 }
1193 }
1194
1195 if ($debug) _debug("$cache_dir: removed $num_deleted files.");
1196 }
1197 }
1198
1199 /**
1200 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1201 * Returns the url query as associative array
1202 *
1203 * @param string query
1204 * @return array params
1205 */
1206 function convertUrlQuery($query) {
1207 $queryParts = explode('&', $query);
1208
1209 $params = array();
1210
1211 foreach ($queryParts as $param) {
1212 $item = explode('=', $param);
1213 $params[$item[0]] = $item[1];
1214 }
1215
1216 return $params;
1217 }
1218
1219 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1220 $matches = array();
1221
1222 foreach ($filters as $filter) {
1223 $match_any_rule = $filter["match_any_rule"];
1224 $inverse = $filter["inverse"];
1225 $filter_match = false;
1226
1227 foreach ($filter["rules"] as $rule) {
1228 $match = false;
1229 $reg_exp = str_replace('/', '\/', $rule["reg_exp"]);
1230 $rule_inverse = $rule["inverse"];
1231
1232 if (!$reg_exp)
1233 continue;
1234
1235 switch ($rule["type"]) {
1236 case "title":
1237 $match = @preg_match("/$reg_exp/i", $title);
1238 break;
1239 case "content":
1240 // we don't need to deal with multiline regexps
1241 $content = preg_replace("/[\r\n\t]/", "", $content);
1242
1243 $match = @preg_match("/$reg_exp/i", $content);
1244 break;
1245 case "both":
1246 // we don't need to deal with multiline regexps
1247 $content = preg_replace("/[\r\n\t]/", "", $content);
1248
1249 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1250 break;
1251 case "link":
1252 $match = @preg_match("/$reg_exp/i", $link);
1253 break;
1254 case "author":
1255 $match = @preg_match("/$reg_exp/i", $author);
1256 break;
1257 case "tag":
1258 foreach ($tags as $tag) {
1259 if (@preg_match("/$reg_exp/i", $tag)) {
1260 $match = true;
1261 break;
1262 }
1263 }
1264 break;
1265 }
1266
1267 if ($rule_inverse) $match = !$match;
1268
1269 if ($match_any_rule) {
1270 if ($match) {
1271 $filter_match = true;
1272 break;
1273 }
1274 } else {
1275 $filter_match = $match;
1276 if (!$match) {
1277 break;
1278 }
1279 }
1280 }
1281
1282 if ($inverse) $filter_match = !$filter_match;
1283
1284 if ($filter_match) {
1285 foreach ($filter["actions"] AS $action) {
1286 array_push($matches, $action);
1287
1288 // if Stop action encountered, perform no further processing
1289 if ($action["type"] == "stop") return $matches;
1290 }
1291 }
1292 }
1293
1294 return $matches;
1295 }
1296
1297 function find_article_filter($filters, $filter_name) {
1298 foreach ($filters as $f) {
1299 if ($f["type"] == $filter_name) {
1300 return $f;
1301 };
1302 }
1303 return false;
1304 }
1305
1306 function find_article_filters($filters, $filter_name) {
1307 $results = array();
1308
1309 foreach ($filters as $f) {
1310 if ($f["type"] == $filter_name) {
1311 array_push($results, $f);
1312 };
1313 }
1314 return $results;
1315 }
1316
1317 function calculate_article_score($filters) {
1318 $score = 0;
1319
1320 foreach ($filters as $f) {
1321 if ($f["type"] == "score") {
1322 $score += $f["param"];
1323 };
1324 }
1325 return $score;
1326 }
1327
1328 function labels_contains_caption($labels, $caption) {
1329 foreach ($labels as $label) {
1330 if ($label[1] == $caption) {
1331 return true;
1332 }
1333 }
1334
1335 return false;
1336 }
1337
1338 function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) {
1339 foreach ($filters as $f) {
1340 if ($f["type"] == "label") {
1341 if (!labels_contains_caption($article_labels, $f["param"])) {
1342 label_add_article($id, $f["param"], $owner_uid);
1343 }
1344 }
1345 }
1346 }
1347
1348 function make_guid_from_title($title) {
1349 return preg_replace("/[ \"\',.:;]/", "-",
1350 mb_strtolower(strip_tags($title), 'utf-8'));
1351 }
1352
1353 function verify_feed_xml($feed_data) {
1354 libxml_use_internal_errors(true);
1355 $doc = new DOMDocument();
1356 $doc->loadXML($feed_data);
1357 $error = libxml_get_last_error();
1358 libxml_clear_errors();
1359 return $error;
1360 }
1361
1362 function housekeeping_common($debug) {
1363 expire_cached_files($debug);
1364 expire_lock_files($debug);
1365 expire_error_log($debug);
1366
1367 $count = update_feedbrowser_cache();
1368 _debug("Feedbrowser updated, $count feeds processed.");
1369
1370 purge_orphans( true);
1371 $rc = cleanup_tags( 14, 50000);
1372
1373 _debug("Cleaned $rc cached tags.");
1374 }
1375 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/