]> 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
remove hsl stuff from classes/feeds, we take care of this when choosing the color
[tt-rss.git] / include / rssfuncs.php
1 <?php
2 define('DAEMON_UPDATE_LOGIN_LIMIT', 30);
3 define('DAEMON_FEED_LIMIT', 100);
4 define('DAEMON_SLEEP_INTERVAL', 60);
5
6 function update_feedbrowser_cache($link) {
7
8 $result = db_query($link, "SELECT feed_url, site_url, title, COUNT(id) AS subscribers
9 FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf
10 WHERE tf.feed_url = ttrss_feeds.feed_url
11 AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%'))
12 GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000");
13
14 db_query($link, "BEGIN");
15
16 db_query($link, "DELETE FROM ttrss_feedbrowser_cache");
17
18 $count = 0;
19
20 while ($line = db_fetch_assoc($result)) {
21 $subscribers = db_escape_string($link, $line["subscribers"]);
22 $feed_url = db_escape_string($link, $line["feed_url"]);
23 $title = db_escape_string($link, $line["title"]);
24 $site_url = db_escape_string($link, $line["site_url"]);
25
26 $tmp_result = db_query($link, "SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query($link, "COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 define('PREFS_NO_CACHE', true);
64
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
67 if (DB_TYPE == "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
69 } else {
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
71 }
72 } else {
73 $login_thresh_qpart = "";
74 }
75
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE == "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_user_prefs.value != '-1'
81 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
82 ) OR (
83 ttrss_feeds.update_interval > 0
84 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
85 ) OR ttrss_feeds.last_updated IS NULL
86 OR last_updated = '1970-01-01 00:00:00')";
87 } else {
88 $update_limit_qpart = "AND ((
89 ttrss_feeds.update_interval = 0
90 AND ttrss_user_prefs.value != '-1'
91 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
92 ) OR (
93 ttrss_feeds.update_interval > 0
94 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
95 ) OR ttrss_feeds.last_updated IS NULL
96 OR last_updated = '1970-01-01 00:00:00')";
97 }
98
99 // Test if feed is currently being updated by another process.
100 if (DB_TYPE == "pgsql") {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
102 } else {
103 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
104 }
105
106 // Test if there is a limit to number of updated feeds
107 $query_limit = "";
108 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
109
110 $random_qpart = sql_random_function();
111
112 // We search for feed needing update.
113 $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
114 FROM
115 ttrss_feeds, ttrss_users, ttrss_user_prefs
116 WHERE
117 ttrss_feeds.owner_uid = ttrss_users.id
118 AND ttrss_users.id = ttrss_user_prefs.owner_uid
119 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
120 $login_thresh_qpart $update_limit_qpart
121 $updstart_thresh_qpart
122 ORDER BY $random_qpart $query_limit");
123
124 $user_prefs_cache = array();
125
126 if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result)));
127
128 // Here is a little cache magic in order to minimize risk of double feed updates.
129 $feeds_to_update = array();
130 while ($line = db_fetch_assoc($result)) {
131 array_push($feeds_to_update, db_escape_string($link, $line['feed_url']));
132 }
133
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
137
138 if(count($feeds_to_update) > 0) {
139 $feeds_quoted = array();
140
141 foreach ($feeds_to_update as $feed) {
142 array_push($feeds_quoted, "'" . db_escape_string($link, $feed) . "'");
143 }
144
145 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
146 WHERE feed_url IN (%s)", implode(',', $feeds_quoted)));
147 }
148
149 expire_cached_files($debug);
150 expire_lock_files($debug);
151
152 $nf = 0;
153
154 // For each feed, we call the feed update function.
155 foreach ($feeds_to_update as $feed) {
156 if($debug) _debug("Base feed: $feed");
157
158 //update_rss_feed($link, $line["id"], true);
159
160 // since we have the data cached, we can deal with other feeds with the same url
161
162 $tmp_result = db_query($link, "SELECT DISTINCT ttrss_feeds.id,last_updated
163 FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
164 ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
165 ttrss_users.id = ttrss_user_prefs.owner_uid AND
166 ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
167 feed_url = '".db_escape_string($link, $feed)."' AND
168 (ttrss_feeds.update_interval > 0 OR
169 ttrss_user_prefs.value != '-1')
170 $login_thresh_qpart
171 ORDER BY ttrss_feeds.id $query_limit");
172
173 if (db_num_rows($tmp_result) > 0) {
174 while ($tline = db_fetch_assoc($tmp_result)) {
175 if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
176 update_rss_feed($link, $tline["id"], true);
177 ++$nf;
178 }
179 }
180 }
181
182 require_once "digest.php";
183
184 // Send feed digests by email if needed.
185 send_headlines_digests($link, $debug);
186
187 return $nf;
188
189 } // function update_daemon_common
190
191 // ignore_daemon is not used
192 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
193 $override_url = false) {
194
195 require_once "lib/simplepie/simplepie.inc";
196
197 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
198
199 if ($debug_enabled) {
200 _debug("update_rss_feed: start");
201 }
202
203 $result = db_query($link, "SELECT id,update_interval,auth_login,
204 feed_url,auth_pass,cache_images,last_updated,
205 mark_unread_on_update, owner_uid,
206 pubsub_state, auth_pass_encrypted
207 FROM ttrss_feeds WHERE id = '$feed'");
208
209 if (db_num_rows($result) == 0) {
210 if ($debug_enabled) {
211 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
212 }
213 return false;
214 }
215
216 $last_updated = db_fetch_result($result, 0, "last_updated");
217 $owner_uid = db_fetch_result($result, 0, "owner_uid");
218 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
219 0, "mark_unread_on_update"));
220 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
221 $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
222 0, "auth_pass_encrypted"));
223
224 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
225 WHERE id = '$feed'");
226
227 $auth_login = db_fetch_result($result, 0, "auth_login");
228 $auth_pass = db_fetch_result($result, 0, "auth_pass");
229
230 if ($auth_pass_encrypted) {
231 require_once "crypt.php";
232 $auth_pass = decrypt_string($auth_pass);
233 }
234
235 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
236 $fetch_url = db_fetch_result($result, 0, "feed_url");
237
238 $feed = db_escape_string($link, $feed);
239
240 if ($override_url) $fetch_url = $override_url;
241
242 $date_feed_processed = date('Y-m-d H:i');
243
244 $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed";
245
246 // Ignore cache if new feed or manual update.
247 $cache_age = ($no_cache || is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
248 30 : get_feed_update_interval($link, $feed) * 60;
249
250 if ($debug_enabled) {
251 _debug("update_rss_feed: cache filename: $cache_filename exists: " . file_exists($cache_filename));
252 _debug("update_rss_feed: cache age: $cache_age; no cache: $no_cache");
253 }
254
255 $cached_feed_data_hash = false;
256
257 $rss = false;
258 $rss_hash = false;
259 $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
260 $last_updated_timestamp = strtotime($last_updated);
261
262 $force_refetch = isset($_REQUEST["force_refetch"]);
263
264 if (file_exists($cache_filename) &&
265 is_readable($cache_filename) &&
266 !$auth_login && !$auth_pass &&
267 filemtime($cache_filename) > time() - $cache_age) {
268
269 if ($debug_enabled) {
270 _debug("update_rss_feed: using local cache.");
271 }
272
273 if ($cache_timestamp > $last_updated_timestamp) {
274 @$rss_data = file_get_contents($cache_filename);
275
276 if ($rss_data) {
277 $rss_hash = sha1($rss_data);
278 @$rss = unserialize($rss_data);
279 }
280 } else if (!$force_refetch) {
281 if ($debug_enabled) {
282 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
283 }
284 return;
285 }
286 }
287
288 if (!$rss) {
289
290 if (!$feed_data) {
291 if ($debug_enabled) {
292 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
293 }
294
295 $feed_data = fetch_file_contents($fetch_url, false,
296 $auth_login, $auth_pass, false,
297 $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
298 $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
299
300 if ($debug_enabled) {
301 _debug("update_rss_feed: fetch done.");
302 }
303
304 }
305
306 if (!$feed_data) {
307 global $fetch_last_error;
308 global $fetch_last_error_code;
309
310 if ($debug_enabled) {
311 _debug("update_rss_feed: unable to fetch: $fetch_last_error [$fetch_last_error_code]");
312 }
313
314 $error_escaped = '';
315
316 // If-Modified-Since
317 if ($fetch_last_error_code != 304) {
318 $error_escaped = db_escape_string($link, $fetch_last_error);
319 } else {
320 if ($debug_enabled) {
321 _debug("update_rss_feed: source claims data not modified, nothing to do.");
322 }
323 }
324
325 db_query($link,
326 "UPDATE ttrss_feeds SET last_error = '$error_escaped',
327 last_updated = NOW() WHERE id = '$feed'");
328
329 return;
330 }
331 }
332
333 $pluginhost = new PluginHost($link);
334 $pluginhost->set_debug($debug_enabled);
335 $user_plugins = get_pref($link, "_ENABLED_PLUGINS", $owner_uid);
336
337 $pluginhost->load(PLUGINS, $pluginhost::KIND_ALL);
338 $pluginhost->load($user_plugins, $pluginhost::KIND_USER, $owner_uid);
339 $pluginhost->load_data();
340
341 foreach ($pluginhost->get_hooks($pluginhost::HOOK_FEED_FETCHED) as $plugin) {
342 $feed_data = $plugin->hook_feed_fetched($feed_data);
343 }
344
345 if (!$rss) {
346 $rss = new SimplePie();
347 $rss->set_sanitize_class("SanitizeDummy");
348 // simplepie ignores the above and creates default sanitizer anyway,
349 // so let's override it...
350 $rss->sanitize = new SanitizeDummy();
351 $rss->set_output_encoding('UTF-8');
352 $rss->set_raw_data($feed_data);
353 $rss->enable_cache(false);
354
355 @$rss->init();
356 }
357
358 // print_r($rss);
359
360 $feed = db_escape_string($link, $feed);
361
362 if (!$rss->error()) {
363
364 // cache data for later
365 if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) {
366 $rss_data = serialize($rss);
367 $new_rss_hash = sha1($rss_data);
368
369 if ($new_rss_hash != $rss_hash) {
370 if ($debug_enabled) {
371 _debug("update_rss_feed: saving $cache_filename");
372 }
373 @file_put_contents($cache_filename, serialize($rss));
374 }
375 }
376
377 // We use local pluginhost here because we need to load different per-user feed plugins
378 $pluginhost->run_hooks($pluginhost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss);
379
380 if ($debug_enabled) {
381 _debug("update_rss_feed: processing feed data...");
382 }
383
384 // db_query($link, "BEGIN");
385
386 if (DB_TYPE == "pgsql") {
387 $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'";
388 } else {
389 $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)";
390 }
391
392 $result = db_query($link, "SELECT title,site_url,owner_uid,
393 (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS
394 favicon_needs_check
395 FROM ttrss_feeds WHERE id = '$feed'");
396
397 $registered_title = db_fetch_result($result, 0, "title");
398 $orig_site_url = db_fetch_result($result, 0, "site_url");
399 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
400 "favicon_needs_check"));
401
402 $owner_uid = db_fetch_result($result, 0, "owner_uid");
403
404 $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
405
406 if ($favicon_needs_check || $force_refetch) {
407 if ($debug_enabled) {
408 _debug("update_rss_feed: checking favicon...");
409 }
410
411 check_feed_favicon($site_url, $feed, $link);
412 $favicon_file = ICONS_DIR . "/$feed.ico";
413
414 if (file_exists($favicon_file)) {
415 require_once "colors.php";
416
417 $favicon_color = db_escape_string($link,
418 calculate_avg_color($favicon_file));
419
420 if ($debug_enabled) _debug("color: $favicon_color");
421
422 $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
423 }
424
425 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW() $favicon_colorstring
426 WHERE id = '$feed'");
427 }
428
429 if (!$registered_title || $registered_title == "[Unknown]") {
430
431 $feed_title = db_escape_string($link, $rss->get_title());
432
433 if ($debug_enabled) {
434 _debug("update_rss_feed: registering title: $feed_title");
435 }
436
437 db_query($link, "UPDATE ttrss_feeds SET
438 title = '$feed_title' WHERE id = '$feed'");
439 }
440
441 if ($site_url && $orig_site_url != $site_url) {
442 db_query($link, "UPDATE ttrss_feeds SET
443 site_url = '$site_url' WHERE id = '$feed'");
444 }
445
446 if ($debug_enabled) {
447 _debug("update_rss_feed: loading filters & labels...");
448 }
449
450 $filters = load_filters($link, $feed, $owner_uid);
451 $labels = get_all_labels($link, $owner_uid);
452
453 if ($debug_enabled) {
454 //print_r($filters);
455 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
456 }
457
458 $items = $rss->get_items();
459
460 if (!is_array($items)) {
461 if ($debug_enabled) {
462 _debug("update_rss_feed: no articles found.");
463 }
464
465 db_query($link, "UPDATE ttrss_feeds
466 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
467
468 return; // no articles
469 }
470
471 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
472
473 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
474
475 $feed_hub_url = false;
476
477 $links = $rss->get_links('hub');
478
479 if ($links && is_array($links)) {
480 foreach ($links as $l) {
481 $feed_hub_url = $l;
482 break;
483 }
484 }
485
486 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
487
488 if ($feed_hub_url && function_exists('curl_init') &&
489 !ini_get("open_basedir")) {
490
491 require_once 'lib/pubsubhubbub/subscriber.php';
492
493 $callback_url = get_self_url_prefix() .
494 "/public.php?op=pubsub&id=$feed";
495
496 $s = new Subscriber($feed_hub_url, $callback_url);
497
498 $rc = $s->subscribe($fetch_url);
499
500 if ($debug_enabled)
501 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
502
503 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
504 WHERE id = '$feed'");
505 }
506 }
507
508 if ($debug_enabled) {
509 _debug("update_rss_feed: processing articles...");
510 }
511
512 foreach ($items as $item) {
513 if ($_REQUEST['xdebug'] == 3) {
514 print_r($item);
515 }
516
517 $entry_guid = $item->get_id();
518 if (!$entry_guid) $entry_guid = $item->get_link();
519 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
520
521 if ($debug_enabled) {
522 _debug("update_rss_feed: guid $entry_guid");
523 }
524
525 if (!$entry_guid) continue;
526
527 $entry_guid = "$owner_uid,$entry_guid";
528
529 $entry_timestamp = "";
530
531 $entry_timestamp = strtotime($item->get_date());
532
533 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
534 $entry_timestamp = time();
535 $no_orig_date = 'true';
536 } else {
537 $no_orig_date = 'false';
538 }
539
540 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
541
542 if ($debug_enabled) {
543 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
544 }
545
546 $entry_title = html_entity_decode($item->get_title());
547
548 $entry_link = rewrite_relative_url($site_url, $item->get_link());
549
550 if ($debug_enabled) {
551 _debug("update_rss_feed: title $entry_title");
552 _debug("update_rss_feed: link $entry_link");
553 }
554
555 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
556
557 $entry_content = $item->get_content();
558 if (!$entry_content) $entry_content = $item->get_description();
559
560 if ($_REQUEST["xdebug"] == 2) {
561 print "update_rss_feed: content: ";
562 print $entry_content;
563 print "\n";
564 }
565
566 $entry_comments = $item->data["comments"];
567
568 if ($item->get_author()) {
569 $entry_author_item = $item->get_author();
570 $entry_author = $entry_author_item->get_name();
571 if (!$entry_author) $entry_author = $entry_author_item->get_email();
572
573 $entry_author = db_escape_string($link, $entry_author);
574 }
575
576 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
577
578 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
579 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
580
581 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
582
583 if (is_array($num_comments) && is_array($num_comments[0])) {
584 $num_comments = (int) $num_comments[0]["data"];
585 } else {
586 $num_comments = 0;
587 }
588
589 if ($debug_enabled) {
590 _debug("update_rss_feed: num_comments: $num_comments");
591 _debug("update_rss_feed: looking for tags [1]...");
592 }
593
594 // parse <category> entries into tags
595
596 $additional_tags = array();
597
598 $additional_tags_src = $item->get_categories();
599
600 if (is_array($additional_tags_src)) {
601 foreach ($additional_tags_src as $tobj) {
602 array_push($additional_tags, $tobj->get_term());
603 }
604 }
605
606 if ($debug_enabled) {
607 _debug("update_rss_feed: category tags:");
608 print_r($additional_tags);
609 }
610
611 if ($debug_enabled) {
612 _debug("update_rss_feed: looking for tags [2]...");
613 }
614
615 $entry_tags = array_unique($additional_tags);
616
617 for ($i = 0; $i < count($entry_tags); $i++)
618 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
619
620 if ($debug_enabled) {
621 //_debug("update_rss_feed: unfiltered tags found:");
622 //print_r($entry_tags);
623 }
624
625 if ($debug_enabled) {
626 _debug("update_rss_feed: done collecting data.");
627 }
628
629 // TODO: less memory-hungry implementation
630
631 if ($debug_enabled) {
632 _debug("update_rss_feed: applying plugin filters..");
633 }
634
635 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
636 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
637 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
638
639 if (db_num_rows($result) != 0) {
640 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
641 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
642 "content" => db_fetch_result($result, 0, "content"),
643 "link" => db_fetch_result($result, 0, "link"),
644 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
645 "author" => db_fetch_result($result, 0, "author"));
646 } else {
647 $entry_plugin_data = "";
648 $stored_article = array();
649 }
650
651 $article = array("owner_uid" => $owner_uid, // read only
652 "guid" => $entry_guid, // read only
653 "title" => $entry_title,
654 "content" => $entry_content,
655 "link" => $entry_link,
656 "tags" => $entry_tags,
657 "plugin_data" => $entry_plugin_data,
658 "author" => $entry_author,
659 "stored" => $stored_article);
660
661 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
662 $article = $plugin->hook_article_filter($article);
663 }
664
665 $entry_tags = $article["tags"];
666 $entry_guid = db_escape_string($link, $entry_guid);
667 $entry_title = db_escape_string($link, $article["title"]);
668 $entry_author = db_escape_string($link, $article["author"]);
669 $entry_link = db_escape_string($link, $article["link"]);
670 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
671 $entry_content = $article["content"]; // escaped below
672
673
674 if ($debug_enabled) {
675 _debug("update_rss_feed: plugin data: $entry_plugin_data");
676 }
677
678 if ($cache_images && is_writable(CACHE_DIR . '/images'))
679 cache_images($entry_content, $site_url, $debug_enabled);
680
681 $entry_content = db_escape_string($link, $entry_content, false);
682
683 $content_hash = "SHA1:" . sha1($entry_content);
684
685 db_query($link, "BEGIN");
686
687 $result = db_query($link, "SELECT id FROM ttrss_entries
688 WHERE guid = '$entry_guid'");
689
690 if (db_num_rows($result) == 0) {
691
692 if ($debug_enabled) {
693 _debug("update_rss_feed: base guid [$entry_guid] not found");
694 }
695
696 // base post entry does not exist, create it
697
698 $result = db_query($link,
699 "INSERT INTO ttrss_entries
700 (title,
701 guid,
702 link,
703 updated,
704 content,
705 content_hash,
706 cached_content,
707 no_orig_date,
708 date_updated,
709 date_entered,
710 comments,
711 num_comments,
712 plugin_data,
713 author)
714 VALUES
715 ('$entry_title',
716 '$entry_guid',
717 '$entry_link',
718 '$entry_timestamp_fmt',
719 '$entry_content',
720 '$content_hash',
721 '',
722 $no_orig_date,
723 NOW(),
724 '$date_feed_processed',
725 '$entry_comments',
726 '$num_comments',
727 '$entry_plugin_data',
728 '$entry_author')");
729
730 $article_labels = array();
731
732 } else {
733 // we keep encountering the entry in feeds, so we need to
734 // update date_updated column so that we don't get horrible
735 // dupes when the entry gets purged and reinserted again e.g.
736 // in the case of SLOW SLOW OMG SLOW updating feeds
737
738 $base_entry_id = db_fetch_result($result, 0, "id");
739
740 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
741 WHERE id = '$base_entry_id'");
742
743 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
744 }
745
746 // now it should exist, if not - bad luck then
747
748 $result = db_query($link, "SELECT
749 id,content_hash,no_orig_date,title,plugin_data,
750 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
751 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
752 num_comments
753 FROM
754 ttrss_entries
755 WHERE guid = '$entry_guid'");
756
757 $entry_ref_id = 0;
758 $entry_int_id = 0;
759
760 if (db_num_rows($result) == 1) {
761
762 if ($debug_enabled) {
763 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
764 }
765
766 // this will be used below in update handler
767 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
768 $orig_title = db_fetch_result($result, 0, "title");
769 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
770 $orig_date_updated = strtotime(db_fetch_result($result,
771 0, "date_updated"));
772 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
773
774 $ref_id = db_fetch_result($result, 0, "id");
775 $entry_ref_id = $ref_id;
776
777 // check for user post link to main table
778
779 // do we allow duplicate posts with same GUID in different feeds?
780 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
781 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
782 } else {
783 $dupcheck_qpart = "";
784 }
785
786 /* Collect article tags here so we could filter by them: */
787
788 $article_filters = get_article_filters($filters, $entry_title,
789 $entry_content, $entry_link, $entry_timestamp, $entry_author,
790 $entry_tags);
791
792 if ($debug_enabled) {
793 _debug("update_rss_feed: article filters: ");
794 if (count($article_filters) != 0) {
795 print_r($article_filters);
796 }
797 }
798
799 if (find_article_filter($article_filters, "filter")) {
800 db_query($link, "COMMIT"); // close transaction in progress
801 continue;
802 }
803
804 $score = calculate_article_score($article_filters);
805
806 if ($debug_enabled) {
807 _debug("update_rss_feed: initial score: $score");
808 }
809
810 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
811 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
812 $dupcheck_qpart";
813
814 // if ($_REQUEST["xdebug"]) print "$query\n";
815
816 $result = db_query($link, $query);
817
818 // okay it doesn't exist - create user entry
819 if (db_num_rows($result) == 0) {
820
821 if ($debug_enabled) {
822 _debug("update_rss_feed: user record not found, creating...");
823 }
824
825 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
826 $unread = 'true';
827 $last_read_qpart = 'NULL';
828 } else {
829 $unread = 'false';
830 $last_read_qpart = 'NOW()';
831 }
832
833 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
834 $marked = 'true';
835 } else {
836 $marked = 'false';
837 }
838
839 if (find_article_filter($article_filters, 'publish')) {
840 $published = 'true';
841 } else {
842 $published = 'false';
843 }
844
845 // N-grams
846
847 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
848
849 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
850 ttrss_entries,ttrss_user_entries
851 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
852 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
853 AND owner_uid = $owner_uid");
854
855 $ngram_similar = db_fetch_result($result, 0, "similar");
856
857 if ($debug_enabled) {
858 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
859 }
860
861 if ($ngram_similar > 0) {
862 $unread = 'false';
863 }
864 }
865
866 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
867 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
868
869 $result = db_query($link,
870 "INSERT INTO ttrss_user_entries
871 (ref_id, owner_uid, feed_id, unread, last_read, marked,
872 published, score, tag_cache, label_cache, uuid,
873 last_marked, last_published)
874 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
875 $last_read_qpart, $marked, $published, '$score', '', '',
876 '', $last_marked, $last_published)");
877
878 if (PUBSUBHUBBUB_HUB && $published == 'true') {
879 $rss_link = get_self_url_prefix() .
880 "/public.php?op=rss&id=-2&key=" .
881 get_feed_access_key($link, -2, false, $owner_uid);
882
883 $p = new Publisher(PUBSUBHUBBUB_HUB);
884
885 $pubsub_result = $p->publish_update($rss_link);
886 }
887
888 $result = db_query($link,
889 "SELECT int_id FROM ttrss_user_entries WHERE
890 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
891 feed_id = '$feed' LIMIT 1");
892
893 if (db_num_rows($result) == 1) {
894 $entry_int_id = db_fetch_result($result, 0, "int_id");
895 }
896 } else {
897 if ($debug_enabled) {
898 _debug("update_rss_feed: user record FOUND");
899 }
900
901 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
902 $entry_int_id = db_fetch_result($result, 0, "int_id");
903 }
904
905 if ($debug_enabled) {
906 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
907 }
908
909 $post_needs_update = false;
910 $update_insignificant = false;
911
912 if ($orig_num_comments != $num_comments) {
913 $post_needs_update = true;
914 $update_insignificant = true;
915 }
916
917 if ($entry_plugin_data != $orig_plugin_data) {
918 $post_needs_update = true;
919 $update_insignificant = true;
920 }
921
922 if ($content_hash != $orig_content_hash) {
923 $post_needs_update = true;
924 $update_insignificant = false;
925 }
926
927 if (db_escape_string($link, $orig_title) != $entry_title) {
928 $post_needs_update = true;
929 $update_insignificant = false;
930 }
931
932 // if post needs update, update it and mark all user entries
933 // linking to this post as updated
934 if ($post_needs_update) {
935
936 if (defined('DAEMON_EXTENDED_DEBUG')) {
937 _debug("update_rss_feed: post $entry_guid needs update...");
938 }
939
940 // print "<!-- post $orig_title needs update : $post_needs_update -->";
941
942 db_query($link, "UPDATE ttrss_entries
943 SET title = '$entry_title', content = '$entry_content',
944 content_hash = '$content_hash',
945 updated = '$entry_timestamp_fmt',
946 num_comments = '$num_comments',
947 plugin_data = '$entry_plugin_data'
948 WHERE id = '$ref_id'");
949
950 if (!$update_insignificant) {
951 if ($mark_unread_on_update) {
952 db_query($link, "UPDATE ttrss_user_entries
953 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
954 }
955 }
956 }
957 }
958
959 db_query($link, "COMMIT");
960
961 if ($debug_enabled) {
962 _debug("update_rss_feed: assigning labels...");
963 }
964
965 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
966 $owner_uid, $article_labels);
967
968 if ($debug_enabled) {
969 _debug("update_rss_feed: looking for enclosures...");
970 }
971
972 // enclosures
973
974 $enclosures = array();
975
976 $encs = $item->get_enclosures();
977
978 if (is_array($encs)) {
979 foreach ($encs as $e) {
980 $e_item = array(
981 $e->link, $e->type, $e->length);
982 array_push($enclosures, $e_item);
983 }
984 }
985
986 if ($debug_enabled) {
987 _debug("update_rss_feed: article enclosures:");
988 print_r($enclosures);
989 }
990
991 db_query($link, "BEGIN");
992
993 foreach ($enclosures as $enc) {
994 $enc_url = db_escape_string($link, $enc[0]);
995 $enc_type = db_escape_string($link, $enc[1]);
996 $enc_dur = db_escape_string($link, $enc[2]);
997
998 $result = db_query($link, "SELECT id FROM ttrss_enclosures
999 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1000
1001 if (db_num_rows($result) == 0) {
1002 db_query($link, "INSERT INTO ttrss_enclosures
1003 (content_url, content_type, title, duration, post_id) VALUES
1004 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1005 }
1006 }
1007
1008 db_query($link, "COMMIT");
1009
1010 // check for manual tags (we have to do it here since they're loaded from filters)
1011
1012 foreach ($article_filters as $f) {
1013 if ($f["type"] == "tag") {
1014
1015 $manual_tags = trim_array(explode(",", $f["param"]));
1016
1017 foreach ($manual_tags as $tag) {
1018 if (tag_is_valid($tag)) {
1019 array_push($entry_tags, $tag);
1020 }
1021 }
1022 }
1023 }
1024
1025 // Skip boring tags
1026
1027 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1028 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1029
1030 $filtered_tags = array();
1031 $tags_to_cache = array();
1032
1033 if ($entry_tags && is_array($entry_tags)) {
1034 foreach ($entry_tags as $tag) {
1035 if (array_search($tag, $boring_tags) === false) {
1036 array_push($filtered_tags, $tag);
1037 }
1038 }
1039 }
1040
1041 $filtered_tags = array_unique($filtered_tags);
1042
1043 if ($debug_enabled) {
1044 _debug("update_rss_feed: filtered article tags:");
1045 print_r($filtered_tags);
1046 }
1047
1048 // Save article tags in the database
1049
1050 if (count($filtered_tags) > 0) {
1051
1052 db_query($link, "BEGIN");
1053
1054 foreach ($filtered_tags as $tag) {
1055
1056 $tag = sanitize_tag($tag);
1057 $tag = db_escape_string($link, $tag);
1058
1059 if (!tag_is_valid($tag)) continue;
1060
1061 $result = db_query($link, "SELECT id FROM ttrss_tags
1062 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1063 owner_uid = '$owner_uid' LIMIT 1");
1064
1065 if ($result && db_num_rows($result) == 0) {
1066
1067 db_query($link, "INSERT INTO ttrss_tags
1068 (owner_uid,tag_name,post_int_id)
1069 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1070 }
1071
1072 array_push($tags_to_cache, $tag);
1073 }
1074
1075 /* update the cache */
1076
1077 $tags_to_cache = array_unique($tags_to_cache);
1078
1079 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1080
1081 db_query($link, "UPDATE ttrss_user_entries
1082 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1083 AND owner_uid = $owner_uid");
1084
1085 db_query($link, "COMMIT");
1086 }
1087
1088 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1089 if ($debug_enabled) {
1090 _debug("update_rss_feed: auto-assigning labels...");
1091 }
1092
1093 foreach ($labels as $label) {
1094 $caption = preg_quote($label["caption"]);
1095
1096 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1097 if (!labels_contains_caption($article_labels, $caption)) {
1098 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1099 }
1100 }
1101 }
1102 }
1103
1104 if ($debug_enabled) {
1105 _debug("update_rss_feed: article processed");
1106 }
1107 }
1108
1109 if (!$last_updated) {
1110 if ($debug_enabled) {
1111 _debug("update_rss_feed: new feed, catching it up...");
1112 }
1113 catchup_feed($link, $feed, false, $owner_uid);
1114 }
1115
1116 if ($debug_enabled) {
1117 _debug("purging feed...");
1118 }
1119
1120 purge_feed($link, $feed, 0, $debug_enabled);
1121
1122 db_query($link, "UPDATE ttrss_feeds
1123 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1124
1125 // db_query($link, "COMMIT");
1126
1127 } else {
1128
1129 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1130
1131 if ($debug_enabled) {
1132 _debug("update_rss_feed: error fetching feed: $error_msg");
1133 }
1134
1135 db_query($link,
1136 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1137 last_updated = NOW() WHERE id = '$feed'");
1138 }
1139
1140 unset($rss);
1141
1142 if ($debug_enabled) {
1143 _debug("update_rss_feed: done");
1144 }
1145
1146 }
1147
1148 function cache_images($html, $site_url, $debug) {
1149 $cache_dir = CACHE_DIR . "/images";
1150
1151 libxml_use_internal_errors(true);
1152
1153 $charset_hack = '<head>
1154 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1155 </head>';
1156
1157 $doc = new DOMDocument();
1158 $doc->loadHTML($charset_hack . $html);
1159 $xpath = new DOMXPath($doc);
1160
1161 $entries = $xpath->query('(//img[@src])');
1162
1163 foreach ($entries as $entry) {
1164 if ($entry->hasAttribute('src')) {
1165 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1166
1167 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1168
1169 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1170
1171 if (!file_exists($local_filename)) {
1172 $file_content = fetch_file_contents($src);
1173
1174 if ($file_content && strlen($file_content) > 1024) {
1175 file_put_contents($local_filename, $file_content);
1176 }
1177 }
1178
1179 if (file_exists($local_filename)) {
1180 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1181 base64_encode($src));
1182 }
1183 }
1184 }
1185
1186 $node = $doc->getElementsByTagName('body')->item(0);
1187
1188 return $doc->saveXML($node);
1189 }
1190
1191 function expire_lock_files($debug) {
1192 if ($debug) _debug("Removing old lock files...");
1193
1194 $num_deleted = 0;
1195
1196 if (is_writable(LOCK_DIRECTORY)) {
1197 $files = glob(LOCK_DIRECTORY . "/*.lock");
1198
1199 if ($files) {
1200 foreach ($files as $file) {
1201 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1202 unlink($file);
1203 ++$num_deleted;
1204 }
1205 }
1206 }
1207 }
1208
1209 if ($debug) _debug("Removed $num_deleted files.");
1210 }
1211
1212 function expire_cached_files($debug) {
1213 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1214 $cache_dir = CACHE_DIR . "/$dir";
1215
1216 if ($debug) _debug("Expiring $cache_dir");
1217
1218 $num_deleted = 0;
1219
1220 if (is_writable($cache_dir)) {
1221 $files = glob("$cache_dir/*");
1222
1223 if ($files) {
1224 foreach ($files as $file) {
1225 if (time() - filemtime($file) > 86400*7) {
1226 unlink($file);
1227
1228 ++$num_deleted;
1229 }
1230 }
1231 }
1232 }
1233
1234 if ($debug) _debug("Removed $num_deleted files.");
1235 }
1236 }
1237
1238 /**
1239 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1240 * Returns the url query as associative array
1241 *
1242 * @param string query
1243 * @return array params
1244 */
1245 function convertUrlQuery($query) {
1246 $queryParts = explode('&', $query);
1247
1248 $params = array();
1249
1250 foreach ($queryParts as $param) {
1251 $item = explode('=', $param);
1252 $params[$item[0]] = $item[1];
1253 }
1254
1255 return $params;
1256 }
1257
1258 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1259 $matches = array();
1260
1261 foreach ($filters as $filter) {
1262 $match_any_rule = $filter["match_any_rule"];
1263 $inverse = $filter["inverse"];
1264 $filter_match = false;
1265
1266 foreach ($filter["rules"] as $rule) {
1267 $match = false;
1268 $reg_exp = $rule["reg_exp"];
1269 $rule_inverse = $rule["inverse"];
1270
1271 if (!$reg_exp)
1272 continue;
1273
1274 switch ($rule["type"]) {
1275 case "title":
1276 $match = @preg_match("/$reg_exp/i", $title);
1277 break;
1278 case "content":
1279 // we don't need to deal with multiline regexps
1280 $content = preg_replace("/[\r\n\t]/", "", $content);
1281
1282 $match = @preg_match("/$reg_exp/i", $content);
1283 break;
1284 case "both":
1285 // we don't need to deal with multiline regexps
1286 $content = preg_replace("/[\r\n\t]/", "", $content);
1287
1288 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1289 break;
1290 case "link":
1291 $match = @preg_match("/$reg_exp/i", $link);
1292 break;
1293 case "author":
1294 $match = @preg_match("/$reg_exp/i", $author);
1295 break;
1296 case "tag":
1297 $tag_string = join(",", $tags);
1298 $match = @preg_match("/$reg_exp/i", $tag_string);
1299 break;
1300 }
1301
1302 if ($rule_inverse) $match = !$match;
1303
1304 if ($match_any_rule) {
1305 if ($match) {
1306 $filter_match = true;
1307 break;
1308 }
1309 } else {
1310 $filter_match = $match;
1311 if (!$match) {
1312 break;
1313 }
1314 }
1315 }
1316
1317 if ($inverse) $filter_match = !$filter_match;
1318
1319 if ($filter_match) {
1320 foreach ($filter["actions"] AS $action) {
1321 array_push($matches, $action);
1322
1323 // if Stop action encountered, perform no further processing
1324 if ($action["type"] == "stop") return $matches;
1325 }
1326 }
1327 }
1328
1329 return $matches;
1330 }
1331
1332 function find_article_filter($filters, $filter_name) {
1333 foreach ($filters as $f) {
1334 if ($f["type"] == $filter_name) {
1335 return $f;
1336 };
1337 }
1338 return false;
1339 }
1340
1341 function find_article_filters($filters, $filter_name) {
1342 $results = array();
1343
1344 foreach ($filters as $f) {
1345 if ($f["type"] == $filter_name) {
1346 array_push($results, $f);
1347 };
1348 }
1349 return $results;
1350 }
1351
1352 function calculate_article_score($filters) {
1353 $score = 0;
1354
1355 foreach ($filters as $f) {
1356 if ($f["type"] == "score") {
1357 $score += $f["param"];
1358 };
1359 }
1360 return $score;
1361 }
1362
1363 function labels_contains_caption($labels, $caption) {
1364 foreach ($labels as $label) {
1365 if ($label[1] == $caption) {
1366 return true;
1367 }
1368 }
1369
1370 return false;
1371 }
1372
1373 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1374 foreach ($filters as $f) {
1375 if ($f["type"] == "label") {
1376 if (!labels_contains_caption($article_labels, $f["param"])) {
1377 label_add_article($link, $id, $f["param"], $owner_uid);
1378 }
1379 }
1380 }
1381 }
1382
1383 function make_guid_from_title($title) {
1384 return preg_replace("/[ \"\',.:;]/", "-",
1385 mb_strtolower(strip_tags($title), 'utf-8'));
1386 }
1387
1388
1389 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/