]> 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
Added average color calculation of feeds' favicons for banded display.
[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 if (file_exists($cache_filename) &&
263 is_readable($cache_filename) &&
264 !$auth_login && !$auth_pass &&
265 filemtime($cache_filename) > time() - $cache_age) {
266
267 if ($debug_enabled) {
268 _debug("update_rss_feed: using local cache.");
269 }
270
271 if ($cache_timestamp > $last_updated_timestamp) {
272 @$rss_data = file_get_contents($cache_filename);
273
274 if ($rss_data) {
275 $rss_hash = sha1($rss_data);
276 @$rss = unserialize($rss_data);
277 }
278 } else {
279 if ($debug_enabled) {
280 _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
281 }
282 return;
283 }
284 }
285
286 if (!$rss) {
287
288 if (!$feed_data) {
289 if ($debug_enabled) {
290 _debug("update_rss_feed: fetching [$fetch_url] (ts: $cache_timestamp/$last_updated_timestamp)");
291 }
292
293 $force_refetch = isset($_REQUEST["force_refetch"]);
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, $rss->get_link()), 0, 245));
405
406 if ($debug_enabled) {
407 _debug("update_rss_feed: checking favicon...");
408 }
409
410 if ($favicon_needs_check) {
411 $favicon_file = check_feed_favicon($site_url, $feed, $link);
412 if ($favicon_file) {
413 $favicon_color = calculate_avg_color($favicon_file);
414 if (is_array($favicon_color))
415 $favicon_colorstring = ",favicon_avg_color = '" . implode("|", array_slice($favicon_color, 0, 3)) . "'";
416 }
417
418 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW() $favicon_colorstring
419 WHERE id = '$feed'");
420 }
421
422 if (!$registered_title || $registered_title == "[Unknown]") {
423
424 $feed_title = db_escape_string($link, $rss->get_title());
425
426 if ($debug_enabled) {
427 _debug("update_rss_feed: registering title: $feed_title");
428 }
429
430 db_query($link, "UPDATE ttrss_feeds SET
431 title = '$feed_title' WHERE id = '$feed'");
432 }
433
434 if ($site_url && $orig_site_url != $site_url) {
435 db_query($link, "UPDATE ttrss_feeds SET
436 site_url = '$site_url' WHERE id = '$feed'");
437 }
438
439 if ($debug_enabled) {
440 _debug("update_rss_feed: loading filters & labels...");
441 }
442
443 $filters = load_filters($link, $feed, $owner_uid);
444 $labels = get_all_labels($link, $owner_uid);
445
446 if ($debug_enabled) {
447 //print_r($filters);
448 _debug("update_rss_feed: " . count($filters) . " filters loaded.");
449 }
450
451 $items = $rss->get_items();
452
453 if (!is_array($items)) {
454 if ($debug_enabled) {
455 _debug("update_rss_feed: no articles found.");
456 }
457
458 db_query($link, "UPDATE ttrss_feeds
459 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
460
461 return; // no articles
462 }
463
464 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
465
466 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
467
468 $feed_hub_url = false;
469
470 $links = $rss->get_links('hub');
471
472 if ($links && is_array($links)) {
473 foreach ($links as $l) {
474 $feed_hub_url = $l;
475 break;
476 }
477 }
478
479 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
480
481 if ($feed_hub_url && function_exists('curl_init') &&
482 !ini_get("open_basedir")) {
483
484 require_once 'lib/pubsubhubbub/subscriber.php';
485
486 $callback_url = get_self_url_prefix() .
487 "/public.php?op=pubsub&id=$feed";
488
489 $s = new Subscriber($feed_hub_url, $callback_url);
490
491 $rc = $s->subscribe($fetch_url);
492
493 if ($debug_enabled)
494 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
495
496 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
497 WHERE id = '$feed'");
498 }
499 }
500
501 if ($debug_enabled) {
502 _debug("update_rss_feed: processing articles...");
503 }
504
505 foreach ($items as $item) {
506 if ($_REQUEST['xdebug'] == 3) {
507 print_r($item);
508 }
509
510 $entry_guid = $item->get_id();
511 if (!$entry_guid) $entry_guid = $item->get_link();
512 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
513
514 if ($debug_enabled) {
515 _debug("update_rss_feed: guid $entry_guid");
516 }
517
518 if (!$entry_guid) continue;
519
520 $entry_guid = "$owner_uid,$entry_guid";
521
522 $entry_timestamp = "";
523
524 $entry_timestamp = strtotime($item->get_date());
525
526 if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) {
527 $entry_timestamp = time();
528 $no_orig_date = 'true';
529 } else {
530 $no_orig_date = 'false';
531 }
532
533 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
534
535 if ($debug_enabled) {
536 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
537 }
538
539 $entry_title = html_entity_decode($item->get_title());
540
541 $entry_link = rewrite_relative_url($site_url, $item->get_link());
542
543 if ($debug_enabled) {
544 _debug("update_rss_feed: title $entry_title");
545 _debug("update_rss_feed: link $entry_link");
546 }
547
548 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
549
550 $entry_content = $item->get_content();
551 if (!$entry_content) $entry_content = $item->get_description();
552
553 if ($_REQUEST["xdebug"] == 2) {
554 print "update_rss_feed: content: ";
555 print $entry_content;
556 print "\n";
557 }
558
559 $entry_comments = $item->data["comments"];
560
561 if ($item->get_author()) {
562 $entry_author_item = $item->get_author();
563 $entry_author = $entry_author_item->get_name();
564 if (!$entry_author) $entry_author = $entry_author_item->get_email();
565
566 $entry_author = db_escape_string($link, $entry_author);
567 }
568
569 $entry_guid = db_escape_string($link, mb_substr($entry_guid, 0, 245));
570
571 $entry_comments = db_escape_string($link, mb_substr($entry_comments, 0, 245));
572 $entry_author = db_escape_string($link, mb_substr($entry_author, 0, 245));
573
574 $num_comments = $item->get_item_tags('http://purl.org/rss/1.0/modules/slash/', 'comments');
575
576 if (is_array($num_comments) && is_array($num_comments[0])) {
577 $num_comments = (int) $num_comments[0]["data"];
578 } else {
579 $num_comments = 0;
580 }
581
582 if ($debug_enabled) {
583 _debug("update_rss_feed: num_comments: $num_comments");
584 _debug("update_rss_feed: looking for tags [1]...");
585 }
586
587 // parse <category> entries into tags
588
589 $additional_tags = array();
590
591 $additional_tags_src = $item->get_categories();
592
593 if (is_array($additional_tags_src)) {
594 foreach ($additional_tags_src as $tobj) {
595 array_push($additional_tags, $tobj->get_term());
596 }
597 }
598
599 if ($debug_enabled) {
600 _debug("update_rss_feed: category tags:");
601 print_r($additional_tags);
602 }
603
604 if ($debug_enabled) {
605 _debug("update_rss_feed: looking for tags [2]...");
606 }
607
608 $entry_tags = array_unique($additional_tags);
609
610 for ($i = 0; $i < count($entry_tags); $i++)
611 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
612
613 if ($debug_enabled) {
614 //_debug("update_rss_feed: unfiltered tags found:");
615 //print_r($entry_tags);
616 }
617
618 if ($debug_enabled) {
619 _debug("update_rss_feed: done collecting data.");
620 }
621
622 // TODO: less memory-hungry implementation
623
624 if ($debug_enabled) {
625 _debug("update_rss_feed: applying plugin filters..");
626 }
627
628 // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
629 $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
630 WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
631
632 if (db_num_rows($result) != 0) {
633 $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
634 $stored_article = array("title" => db_fetch_result($result, 0, "title"),
635 "content" => db_fetch_result($result, 0, "content"),
636 "link" => db_fetch_result($result, 0, "link"),
637 "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")),
638 "author" => db_fetch_result($result, 0, "author"));
639 } else {
640 $entry_plugin_data = "";
641 $stored_article = array();
642 }
643
644 $article = array("owner_uid" => $owner_uid, // read only
645 "guid" => $entry_guid, // read only
646 "title" => $entry_title,
647 "content" => $entry_content,
648 "link" => $entry_link,
649 "tags" => $entry_tags,
650 "plugin_data" => $entry_plugin_data,
651 "author" => $entry_author,
652 "stored" => $stored_article);
653
654 foreach ($pluginhost->get_hooks($pluginhost::HOOK_ARTICLE_FILTER) as $plugin) {
655 $article = $plugin->hook_article_filter($article);
656 }
657
658 $entry_tags = $article["tags"];
659 $entry_guid = db_escape_string($link, $entry_guid);
660 $entry_title = db_escape_string($link, $article["title"]);
661 $entry_author = db_escape_string($link, $article["author"]);
662 $entry_link = db_escape_string($link, $article["link"]);
663 $entry_plugin_data = db_escape_string($link, $article["plugin_data"]);
664 $entry_content = $article["content"]; // escaped below
665
666
667 if ($debug_enabled) {
668 _debug("update_rss_feed: plugin data: $entry_plugin_data");
669 }
670
671 if ($cache_images && is_writable(CACHE_DIR . '/images'))
672 cache_images($entry_content, $site_url, $debug_enabled);
673
674 $entry_content = db_escape_string($link, $entry_content, false);
675
676 $content_hash = "SHA1:" . sha1($entry_content);
677
678 db_query($link, "BEGIN");
679
680 $result = db_query($link, "SELECT id FROM ttrss_entries
681 WHERE guid = '$entry_guid'");
682
683 if (db_num_rows($result) == 0) {
684
685 if ($debug_enabled) {
686 _debug("update_rss_feed: base guid [$entry_guid] not found");
687 }
688
689 // base post entry does not exist, create it
690
691 $result = db_query($link,
692 "INSERT INTO ttrss_entries
693 (title,
694 guid,
695 link,
696 updated,
697 content,
698 content_hash,
699 cached_content,
700 no_orig_date,
701 date_updated,
702 date_entered,
703 comments,
704 num_comments,
705 plugin_data,
706 author)
707 VALUES
708 ('$entry_title',
709 '$entry_guid',
710 '$entry_link',
711 '$entry_timestamp_fmt',
712 '$entry_content',
713 '$content_hash',
714 '',
715 $no_orig_date,
716 NOW(),
717 '$date_feed_processed',
718 '$entry_comments',
719 '$num_comments',
720 '$entry_plugin_data',
721 '$entry_author')");
722
723 $article_labels = array();
724
725 } else {
726 // we keep encountering the entry in feeds, so we need to
727 // update date_updated column so that we don't get horrible
728 // dupes when the entry gets purged and reinserted again e.g.
729 // in the case of SLOW SLOW OMG SLOW updating feeds
730
731 $base_entry_id = db_fetch_result($result, 0, "id");
732
733 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
734 WHERE id = '$base_entry_id'");
735
736 $article_labels = get_article_labels($link, $base_entry_id, $owner_uid);
737 }
738
739 // now it should exist, if not - bad luck then
740
741 $result = db_query($link, "SELECT
742 id,content_hash,no_orig_date,title,plugin_data,
743 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
744 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
745 num_comments
746 FROM
747 ttrss_entries
748 WHERE guid = '$entry_guid'");
749
750 $entry_ref_id = 0;
751 $entry_int_id = 0;
752
753 if (db_num_rows($result) == 1) {
754
755 if ($debug_enabled) {
756 _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
757 }
758
759 // this will be used below in update handler
760 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
761 $orig_title = db_fetch_result($result, 0, "title");
762 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
763 $orig_date_updated = strtotime(db_fetch_result($result,
764 0, "date_updated"));
765 $orig_plugin_data = db_fetch_result($result, 0, "plugin_data");
766
767 $ref_id = db_fetch_result($result, 0, "id");
768 $entry_ref_id = $ref_id;
769
770 // check for user post link to main table
771
772 // do we allow duplicate posts with same GUID in different feeds?
773 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
774 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
775 } else {
776 $dupcheck_qpart = "";
777 }
778
779 /* Collect article tags here so we could filter by them: */
780
781 $article_filters = get_article_filters($filters, $entry_title,
782 $entry_content, $entry_link, $entry_timestamp, $entry_author,
783 $entry_tags);
784
785 if ($debug_enabled) {
786 _debug("update_rss_feed: article filters: ");
787 if (count($article_filters) != 0) {
788 print_r($article_filters);
789 }
790 }
791
792 if (find_article_filter($article_filters, "filter")) {
793 db_query($link, "COMMIT"); // close transaction in progress
794 continue;
795 }
796
797 $score = calculate_article_score($article_filters);
798
799 if ($debug_enabled) {
800 _debug("update_rss_feed: initial score: $score");
801 }
802
803 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
804 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
805 $dupcheck_qpart";
806
807 // if ($_REQUEST["xdebug"]) print "$query\n";
808
809 $result = db_query($link, $query);
810
811 // okay it doesn't exist - create user entry
812 if (db_num_rows($result) == 0) {
813
814 if ($debug_enabled) {
815 _debug("update_rss_feed: user record not found, creating...");
816 }
817
818 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
819 $unread = 'true';
820 $last_read_qpart = 'NULL';
821 } else {
822 $unread = 'false';
823 $last_read_qpart = 'NOW()';
824 }
825
826 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
827 $marked = 'true';
828 } else {
829 $marked = 'false';
830 }
831
832 if (find_article_filter($article_filters, 'publish')) {
833 $published = 'true';
834 } else {
835 $published = 'false';
836 }
837
838 // N-grams
839
840 if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) {
841
842 $result = db_query($link, "SELECT COUNT(*) AS similar FROM
843 ttrss_entries,ttrss_user_entries
844 WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day'
845 AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD."
846 AND owner_uid = $owner_uid");
847
848 $ngram_similar = db_fetch_result($result, 0, "similar");
849
850 if ($debug_enabled) {
851 _debug("update_rss_feed: N-gram similar results: $ngram_similar");
852 }
853
854 if ($ngram_similar > 0) {
855 $unread = 'false';
856 }
857 }
858
859 $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL';
860 $last_published = ($published == 'true') ? 'NOW()' : 'NULL';
861
862 $result = db_query($link,
863 "INSERT INTO ttrss_user_entries
864 (ref_id, owner_uid, feed_id, unread, last_read, marked,
865 published, score, tag_cache, label_cache, uuid,
866 last_marked, last_published)
867 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
868 $last_read_qpart, $marked, $published, '$score', '', '',
869 '', $last_marked, $last_published)");
870
871 if (PUBSUBHUBBUB_HUB && $published == 'true') {
872 $rss_link = get_self_url_prefix() .
873 "/public.php?op=rss&id=-2&key=" .
874 get_feed_access_key($link, -2, false, $owner_uid);
875
876 $p = new Publisher(PUBSUBHUBBUB_HUB);
877
878 $pubsub_result = $p->publish_update($rss_link);
879 }
880
881 $result = db_query($link,
882 "SELECT int_id FROM ttrss_user_entries WHERE
883 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
884 feed_id = '$feed' LIMIT 1");
885
886 if (db_num_rows($result) == 1) {
887 $entry_int_id = db_fetch_result($result, 0, "int_id");
888 }
889 } else {
890 if ($debug_enabled) {
891 _debug("update_rss_feed: user record FOUND");
892 }
893
894 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
895 $entry_int_id = db_fetch_result($result, 0, "int_id");
896 }
897
898 if ($debug_enabled) {
899 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
900 }
901
902 $post_needs_update = false;
903 $update_insignificant = false;
904
905 if ($orig_num_comments != $num_comments) {
906 $post_needs_update = true;
907 $update_insignificant = true;
908 }
909
910 if ($entry_plugin_data != $orig_plugin_data) {
911 $post_needs_update = true;
912 $update_insignificant = true;
913 }
914
915 if ($content_hash != $orig_content_hash) {
916 $post_needs_update = true;
917 $update_insignificant = false;
918 }
919
920 if (db_escape_string($link, $orig_title) != $entry_title) {
921 $post_needs_update = true;
922 $update_insignificant = false;
923 }
924
925 // if post needs update, update it and mark all user entries
926 // linking to this post as updated
927 if ($post_needs_update) {
928
929 if (defined('DAEMON_EXTENDED_DEBUG')) {
930 _debug("update_rss_feed: post $entry_guid needs update...");
931 }
932
933 // print "<!-- post $orig_title needs update : $post_needs_update -->";
934
935 db_query($link, "UPDATE ttrss_entries
936 SET title = '$entry_title', content = '$entry_content',
937 content_hash = '$content_hash',
938 updated = '$entry_timestamp_fmt',
939 num_comments = '$num_comments',
940 plugin_data = '$entry_plugin_data'
941 WHERE id = '$ref_id'");
942
943 if (!$update_insignificant) {
944 if ($mark_unread_on_update) {
945 db_query($link, "UPDATE ttrss_user_entries
946 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
947 }
948 }
949 }
950 }
951
952 db_query($link, "COMMIT");
953
954 if ($debug_enabled) {
955 _debug("update_rss_feed: assigning labels...");
956 }
957
958 assign_article_to_label_filters($link, $entry_ref_id, $article_filters,
959 $owner_uid, $article_labels);
960
961 if ($debug_enabled) {
962 _debug("update_rss_feed: looking for enclosures...");
963 }
964
965 // enclosures
966
967 $enclosures = array();
968
969 $encs = $item->get_enclosures();
970
971 if (is_array($encs)) {
972 foreach ($encs as $e) {
973 $e_item = array(
974 $e->link, $e->type, $e->length);
975 array_push($enclosures, $e_item);
976 }
977 }
978
979 if ($debug_enabled) {
980 _debug("update_rss_feed: article enclosures:");
981 print_r($enclosures);
982 }
983
984 db_query($link, "BEGIN");
985
986 foreach ($enclosures as $enc) {
987 $enc_url = db_escape_string($link, $enc[0]);
988 $enc_type = db_escape_string($link, $enc[1]);
989 $enc_dur = db_escape_string($link, $enc[2]);
990
991 $result = db_query($link, "SELECT id FROM ttrss_enclosures
992 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
993
994 if (db_num_rows($result) == 0) {
995 db_query($link, "INSERT INTO ttrss_enclosures
996 (content_url, content_type, title, duration, post_id) VALUES
997 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
998 }
999 }
1000
1001 db_query($link, "COMMIT");
1002
1003 // check for manual tags (we have to do it here since they're loaded from filters)
1004
1005 foreach ($article_filters as $f) {
1006 if ($f["type"] == "tag") {
1007
1008 $manual_tags = trim_array(explode(",", $f["param"]));
1009
1010 foreach ($manual_tags as $tag) {
1011 if (tag_is_valid($tag)) {
1012 array_push($entry_tags, $tag);
1013 }
1014 }
1015 }
1016 }
1017
1018 // Skip boring tags
1019
1020 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1021 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1022
1023 $filtered_tags = array();
1024 $tags_to_cache = array();
1025
1026 if ($entry_tags && is_array($entry_tags)) {
1027 foreach ($entry_tags as $tag) {
1028 if (array_search($tag, $boring_tags) === false) {
1029 array_push($filtered_tags, $tag);
1030 }
1031 }
1032 }
1033
1034 $filtered_tags = array_unique($filtered_tags);
1035
1036 if ($debug_enabled) {
1037 _debug("update_rss_feed: filtered article tags:");
1038 print_r($filtered_tags);
1039 }
1040
1041 // Save article tags in the database
1042
1043 if (count($filtered_tags) > 0) {
1044
1045 db_query($link, "BEGIN");
1046
1047 foreach ($filtered_tags as $tag) {
1048
1049 $tag = sanitize_tag($tag);
1050 $tag = db_escape_string($link, $tag);
1051
1052 if (!tag_is_valid($tag)) continue;
1053
1054 $result = db_query($link, "SELECT id FROM ttrss_tags
1055 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1056 owner_uid = '$owner_uid' LIMIT 1");
1057
1058 if ($result && db_num_rows($result) == 0) {
1059
1060 db_query($link, "INSERT INTO ttrss_tags
1061 (owner_uid,tag_name,post_int_id)
1062 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1063 }
1064
1065 array_push($tags_to_cache, $tag);
1066 }
1067
1068 /* update the cache */
1069
1070 $tags_to_cache = array_unique($tags_to_cache);
1071
1072 $tags_str = db_escape_string($link, join(",", $tags_to_cache));
1073
1074 db_query($link, "UPDATE ttrss_user_entries
1075 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1076 AND owner_uid = $owner_uid");
1077
1078 db_query($link, "COMMIT");
1079 }
1080
1081 if (get_pref($link, "AUTO_ASSIGN_LABELS", $owner_uid, false)) {
1082 if ($debug_enabled) {
1083 _debug("update_rss_feed: auto-assigning labels...");
1084 }
1085
1086 foreach ($labels as $label) {
1087 $caption = preg_quote($label["caption"]);
1088
1089 if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) {
1090 if (!labels_contains_caption($article_labels, $caption)) {
1091 label_add_article($link, $entry_ref_id, $caption, $owner_uid);
1092 }
1093 }
1094 }
1095 }
1096
1097 if ($debug_enabled) {
1098 _debug("update_rss_feed: article processed");
1099 }
1100 }
1101
1102 if (!$last_updated) {
1103 if ($debug_enabled) {
1104 _debug("update_rss_feed: new feed, catching it up...");
1105 }
1106 catchup_feed($link, $feed, false, $owner_uid);
1107 }
1108
1109 if ($debug_enabled) {
1110 _debug("purging feed...");
1111 }
1112
1113 purge_feed($link, $feed, 0, $debug_enabled);
1114
1115 db_query($link, "UPDATE ttrss_feeds
1116 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1117
1118 // db_query($link, "COMMIT");
1119
1120 } else {
1121
1122 $error_msg = db_escape_string($link, mb_substr($rss->error(), 0, 245));
1123
1124 if ($debug_enabled) {
1125 _debug("update_rss_feed: error fetching feed: $error_msg");
1126 }
1127
1128 db_query($link,
1129 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1130 last_updated = NOW() WHERE id = '$feed'");
1131 }
1132
1133 unset($rss);
1134
1135 if ($debug_enabled) {
1136 _debug("update_rss_feed: done");
1137 }
1138
1139 }
1140
1141 function cache_images($html, $site_url, $debug) {
1142 $cache_dir = CACHE_DIR . "/images";
1143
1144 libxml_use_internal_errors(true);
1145
1146 $charset_hack = '<head>
1147 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1148 </head>';
1149
1150 $doc = new DOMDocument();
1151 $doc->loadHTML($charset_hack . $html);
1152 $xpath = new DOMXPath($doc);
1153
1154 $entries = $xpath->query('(//img[@src])');
1155
1156 foreach ($entries as $entry) {
1157 if ($entry->hasAttribute('src')) {
1158 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1159
1160 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1161
1162 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1163
1164 if (!file_exists($local_filename)) {
1165 $file_content = fetch_file_contents($src);
1166
1167 if ($file_content && strlen($file_content) > 1024) {
1168 file_put_contents($local_filename, $file_content);
1169 }
1170 }
1171
1172 if (file_exists($local_filename)) {
1173 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1174 base64_encode($src));
1175 }
1176 }
1177 }
1178
1179 $node = $doc->getElementsByTagName('body')->item(0);
1180
1181 return $doc->saveXML($node);
1182 }
1183
1184 function expire_lock_files($debug) {
1185 if ($debug) _debug("Removing old lock files...");
1186
1187 $num_deleted = 0;
1188
1189 if (is_writable(LOCK_DIRECTORY)) {
1190 $files = glob(LOCK_DIRECTORY . "/*.lock");
1191
1192 if ($files) {
1193 foreach ($files as $file) {
1194 if (!file_is_locked($file) && time() - filemtime($file) > 86400*2) {
1195 unlink($file);
1196 ++$num_deleted;
1197 }
1198 }
1199 }
1200 }
1201
1202 if ($debug) _debug("Removed $num_deleted files.");
1203 }
1204
1205 function expire_cached_files($debug) {
1206 foreach (array("simplepie", "images", "export", "upload") as $dir) {
1207 $cache_dir = CACHE_DIR . "/$dir";
1208
1209 if ($debug) _debug("Expiring $cache_dir");
1210
1211 $num_deleted = 0;
1212
1213 if (is_writable($cache_dir)) {
1214 $files = glob("$cache_dir/*");
1215
1216 if ($files) {
1217 foreach ($files as $file) {
1218 if (time() - filemtime($file) > 86400*7) {
1219 unlink($file);
1220
1221 ++$num_deleted;
1222 }
1223 }
1224 }
1225 }
1226
1227 if ($debug) _debug("Removed $num_deleted files.");
1228 }
1229 }
1230
1231 /**
1232 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1233 * Returns the url query as associative array
1234 *
1235 * @param string query
1236 * @return array params
1237 */
1238 function convertUrlQuery($query) {
1239 $queryParts = explode('&', $query);
1240
1241 $params = array();
1242
1243 foreach ($queryParts as $param) {
1244 $item = explode('=', $param);
1245 $params[$item[0]] = $item[1];
1246 }
1247
1248 return $params;
1249 }
1250
1251 function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) {
1252 $matches = array();
1253
1254 foreach ($filters as $filter) {
1255 $match_any_rule = $filter["match_any_rule"];
1256 $inverse = $filter["inverse"];
1257 $filter_match = false;
1258
1259 foreach ($filter["rules"] as $rule) {
1260 $match = false;
1261 $reg_exp = $rule["reg_exp"];
1262 $rule_inverse = $rule["inverse"];
1263
1264 if (!$reg_exp)
1265 continue;
1266
1267 switch ($rule["type"]) {
1268 case "title":
1269 $match = @preg_match("/$reg_exp/i", $title);
1270 break;
1271 case "content":
1272 // we don't need to deal with multiline regexps
1273 $content = preg_replace("/[\r\n\t]/", "", $content);
1274
1275 $match = @preg_match("/$reg_exp/i", $content);
1276 break;
1277 case "both":
1278 // we don't need to deal with multiline regexps
1279 $content = preg_replace("/[\r\n\t]/", "", $content);
1280
1281 $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content));
1282 break;
1283 case "link":
1284 $match = @preg_match("/$reg_exp/i", $link);
1285 break;
1286 case "author":
1287 $match = @preg_match("/$reg_exp/i", $author);
1288 break;
1289 case "tag":
1290 $tag_string = join(",", $tags);
1291 $match = @preg_match("/$reg_exp/i", $tag_string);
1292 break;
1293 }
1294
1295 if ($rule_inverse) $match = !$match;
1296
1297 if ($match_any_rule) {
1298 if ($match) {
1299 $filter_match = true;
1300 break;
1301 }
1302 } else {
1303 $filter_match = $match;
1304 if (!$match) {
1305 break;
1306 }
1307 }
1308 }
1309
1310 if ($inverse) $filter_match = !$filter_match;
1311
1312 if ($filter_match) {
1313 foreach ($filter["actions"] AS $action) {
1314 array_push($matches, $action);
1315
1316 // if Stop action encountered, perform no further processing
1317 if ($action["type"] == "stop") return $matches;
1318 }
1319 }
1320 }
1321
1322 return $matches;
1323 }
1324
1325 function find_article_filter($filters, $filter_name) {
1326 foreach ($filters as $f) {
1327 if ($f["type"] == $filter_name) {
1328 return $f;
1329 };
1330 }
1331 return false;
1332 }
1333
1334 function find_article_filters($filters, $filter_name) {
1335 $results = array();
1336
1337 foreach ($filters as $f) {
1338 if ($f["type"] == $filter_name) {
1339 array_push($results, $f);
1340 };
1341 }
1342 return $results;
1343 }
1344
1345 function calculate_article_score($filters) {
1346 $score = 0;
1347
1348 foreach ($filters as $f) {
1349 if ($f["type"] == "score") {
1350 $score += $f["param"];
1351 };
1352 }
1353 return $score;
1354 }
1355
1356 function labels_contains_caption($labels, $caption) {
1357 foreach ($labels as $label) {
1358 if ($label[1] == $caption) {
1359 return true;
1360 }
1361 }
1362
1363 return false;
1364 }
1365
1366 function assign_article_to_label_filters($link, $id, $filters, $owner_uid, $article_labels) {
1367 foreach ($filters as $f) {
1368 if ($f["type"] == "label") {
1369 if (!labels_contains_caption($article_labels, $f["param"])) {
1370 label_add_article($link, $id, $f["param"], $owner_uid);
1371 }
1372 }
1373 }
1374 }
1375
1376 function make_guid_from_title($title) {
1377 return preg_replace("/[ \"\',.:;]/", "-",
1378 mb_strtolower(strip_tags($title), 'utf-8'));
1379 }
1380
1381
1382 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/