]> 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 memcached stuff
[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', 120);
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 $subscribers = db_escape_string($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 $tmp_result = db_query($link, "SELECT subscribers FROM
27 ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'");
28
29 if (db_num_rows($tmp_result) == 0) {
30
31 db_query($link, "INSERT INTO ttrss_feedbrowser_cache
32 (feed_url, site_url, title, subscribers) VALUES ('$feed_url',
33 '$site_url', '$title', '$subscribers')");
34
35 ++$count;
36
37 }
38
39 }
40
41 db_query($link, "COMMIT");
42
43 return $count;
44
45 }
46
47
48 /**
49 * Update a feed batch.
50 * Used by daemons to update n feeds by run.
51 * Only update feed needing a update, and not being processed
52 * by another process.
53 *
54 * @param mixed $link Database link
55 * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT.
56 * @param boolean $from_http Set to true if you call this function from http to disable cli specific code.
57 * @param boolean $debug Set to false to disable debug output. Default to true.
58 * @return void
59 */
60 function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
61 // Process all other feeds using last_updated and interval parameters
62
63 define('PREFS_NO_CACHE', true);
64
65 // Test if the user has loggued in recently. If not, it does not update its feeds.
66 if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) {
67 if (DB_TYPE == "pgsql") {
68 $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'";
69 } else {
70 $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)";
71 }
72 } else {
73 $login_thresh_qpart = "";
74 }
75
76 // Test if the feed need a update (update interval exceded).
77 if (DB_TYPE == "pgsql") {
78 $update_limit_qpart = "AND ((
79 ttrss_feeds.update_interval = 0
80 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
81 ) OR (
82 ttrss_feeds.update_interval > 0
83 AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL)
84 ) OR ttrss_feeds.last_updated IS NULL
85 OR last_updated = '1970-01-01 00:00:00')";
86 } else {
87 $update_limit_qpart = "AND ((
88 ttrss_feeds.update_interval = 0
89 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
90 ) OR (
91 ttrss_feeds.update_interval > 0
92 AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE)
93 ) OR ttrss_feeds.last_updated IS NULL
94 OR last_updated = '1970-01-01 00:00:00')";
95 }
96
97 // Test if feed is currently being updated by another process.
98 if (DB_TYPE == "pgsql") {
99 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '5 minutes')";
100 } else {
101 $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 5 MINUTE))";
102 }
103
104 // Test if there is a limit to number of updated feeds
105 $query_limit = "";
106 if($limit) $query_limit = sprintf("LIMIT %d", $limit);
107
108 $random_qpart = sql_random_function();
109
110 // We search for feed needing update.
111 $result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id, ttrss_feeds.owner_uid,
112 ".SUBSTRING_FOR_DATE."(ttrss_feeds.last_updated,1,19) AS last_updated,
113 ttrss_feeds.update_interval
114 FROM
115 ttrss_feeds, ttrss_users, ttrss_user_prefs
116 WHERE
117 ttrss_feeds.owner_uid = ttrss_users.id
118 AND ttrss_users.id = ttrss_user_prefs.owner_uid
119 AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
120 $login_thresh_qpart $update_limit_qpart
121 $updstart_thresh_qpart
122 ORDER BY $random_qpart $query_limit");
123
124 $user_prefs_cache = array();
125
126 if($debug) _debug(sprintf("Scheduled %d feeds to update...\n", db_num_rows($result)));
127
128 // Here is a little cache magic in order to minimize risk of double feed updates.
129 $feeds_to_update = array();
130 while ($line = db_fetch_assoc($result)) {
131 $feeds_to_update[$line['id']] = $line;
132 }
133
134 // We update the feed last update started date before anything else.
135 // There is no lag due to feed contents downloads
136 // It prevent an other process to update the same feed.
137 $feed_ids = array_keys($feeds_to_update);
138 if($feed_ids) {
139 db_query($link, sprintf("UPDATE ttrss_feeds SET last_update_started = NOW()
140 WHERE id IN (%s)", implode(',', $feed_ids)));
141 }
142
143 expire_cached_files($debug);
144
145 // For each feed, we call the feed update function.
146 while ($line = array_pop($feeds_to_update)) {
147
148 if($debug) _debug("Feed: " . $line["feed_url"] . ", " . $line["last_updated"]);
149
150 update_rss_feed($link, $line["id"], true);
151
152 sleep(1); // prevent flood (FIXME make this an option?)
153 }
154
155 // Send feed digests by email if needed.
156 send_headlines_digests($link, $debug);
157
158 } // function update_daemon_common
159
160 function fetch_twitter_rss($link, $url, $owner_uid) {
161
162 require_once 'lib/tmhoauth/tmhOAuth.php';
163 require_once "lib/magpierss/rss_fetch.inc";
164 require_once 'lib/magpierss/rss_utils.inc';
165
166 $result = db_query($link, "SELECT twitter_oauth FROM ttrss_users
167 WHERE id = $owner_uid");
168
169 $access_token = json_decode(db_fetch_result($result, 0, 'twitter_oauth'), true);
170 $url_escaped = db_escape_string($url);
171
172 if ($access_token) {
173
174 $tmhOAuth = new tmhOAuth(array(
175 'consumer_key' => CONSUMER_KEY,
176 'consumer_secret' => CONSUMER_SECRET,
177 'user_token' => $access_token['oauth_token'],
178 'user_secret' => $access_token['oauth_token_secret'],
179 ));
180
181 $code = $tmhOAuth->request('GET', $url,
182 convertUrlQuery(parse_url($url, PHP_URL_QUERY)));
183
184 if ($code == 200) {
185
186 $content = $tmhOAuth->response['response'];
187
188 define('MAGPIE_CACHE_ON', false);
189
190 $rss = new MagpieRSS($content, MAGPIE_OUTPUT_ENCODING,
191 MAGPIE_INPUT_ENCODING, MAGPIE_DETECT_ENCODING );
192
193 return $rss;
194
195 } else {
196
197 db_query($link, "UPDATE ttrss_feeds
198 SET last_error = 'OAuth authorization failed ($code).'
199 WHERE feed_url = '$url_escaped' AND owner_uid = $owner_uid");
200 }
201
202 } else {
203
204 db_query($link, "UPDATE ttrss_feeds
205 SET last_error = 'OAuth information not found.'
206 WHERE feed_url = '$url_escaped' AND owner_uid = $owner_uid");
207
208 return false;
209 }
210 }
211
212 function update_rss_feed($link, $feed, $ignore_daemon = false, $no_cache = false,
213 $override_url = false) {
214
215 require_once "lib/simplepie/simplepie.inc";
216 require_once "lib/magpierss/rss_fetch.inc";
217 require_once 'lib/magpierss/rss_utils.inc';
218
219 $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug'];
220
221 if (!$_REQUEST["daemon"] && !$ignore_daemon) {
222 return false;
223 }
224
225 if ($debug_enabled) {
226 _debug("update_rss_feed: start");
227 }
228
229 if (!$ignore_daemon) {
230
231 if (DB_TYPE == "pgsql") {
232 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '120 seconds')";
233 } else {
234 $updstart_thresh_qpart = "(ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 120 SECOND))";
235 }
236
237 $result = db_query($link, "SELECT id,update_interval,auth_login,
238 auth_pass,cache_images,update_method,last_updated
239 FROM ttrss_feeds WHERE id = '$feed' AND $updstart_thresh_qpart");
240
241 } else {
242
243 $result = db_query($link, "SELECT id,update_interval,auth_login,
244 feed_url,auth_pass,cache_images,update_method,last_updated,
245 mark_unread_on_update, owner_uid, update_on_checksum_change,
246 pubsub_state
247 FROM ttrss_feeds WHERE id = '$feed'");
248
249 }
250
251 if (db_num_rows($result) == 0) {
252 if ($debug_enabled) {
253 _debug("update_rss_feed: feed $feed NOT FOUND/SKIPPED");
254 }
255 return false;
256 }
257
258 $update_method = db_fetch_result($result, 0, "update_method");
259 $last_updated = db_fetch_result($result, 0, "last_updated");
260 $owner_uid = db_fetch_result($result, 0, "owner_uid");
261 $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
262 0, "mark_unread_on_update"));
263 $update_on_checksum_change = sql_bool_to_bool(db_fetch_result($result,
264 0, "update_on_checksum_change"));
265 $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
266
267 db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
268 WHERE id = '$feed'");
269
270 $auth_login = db_fetch_result($result, 0, "auth_login");
271 $auth_pass = db_fetch_result($result, 0, "auth_pass");
272
273 if ($update_method == 0)
274 $update_method = DEFAULT_UPDATE_METHOD + 1;
275
276 // 1 - Magpie
277 // 2 - SimplePie
278 // 3 - Twitter OAuth
279
280 if ($update_method == 2)
281 $use_simplepie = true;
282 else
283 $use_simplepie = false;
284
285 if ($debug_enabled) {
286 _debug("update method: $update_method (feed setting: $update_method) (use simplepie: $use_simplepie)\n");
287 }
288
289 if ($update_method == 1) {
290 $auth_login = urlencode($auth_login);
291 $auth_pass = urlencode($auth_pass);
292 }
293
294 $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
295 $fetch_url = db_fetch_result($result, 0, "feed_url");
296
297 $feed = db_escape_string($feed);
298
299 if ($auth_login && $auth_pass ){
300 $url_parts = array();
301 preg_match("/(^[^:]*):\/\/(.*)/", $fetch_url, $url_parts);
302
303 if ($url_parts[1] && $url_parts[2]) {
304 $fetch_url = $url_parts[1] . "://$auth_login:$auth_pass@" . $url_parts[2];
305 }
306
307 }
308
309 if ($override_url)
310 $fetch_url = $override_url;
311
312 if ($debug_enabled) {
313 _debug("update_rss_feed: fetching [$fetch_url]...");
314 }
315
316 // Ignore cache if new feed or manual update.
317 $cache_age = (is_null($last_updated) || $last_updated == '1970-01-01 00:00:00') ?
318 -1 : get_feed_update_interval($link, $feed) * 60;
319
320 if ($update_method == 3) {
321 $rss = fetch_twitter_rss($link, $fetch_url, $owner_uid);
322 } else if ($update_method == 1) {
323
324 define('MAGPIE_CACHE_AGE', $cache_age);
325 define('MAGPIE_CACHE_ON', !$no_cache);
326 define('MAGPIE_FETCH_TIME_OUT', 60);
327 define('MAGPIE_CACHE_DIR', CACHE_DIR . "/magpie");
328
329 $rss = @fetch_rss($fetch_url);
330 } else {
331 $simplepie_cache_dir = CACHE_DIR . "/simplepie";
332
333 if (!is_dir($simplepie_cache_dir)) {
334 mkdir($simplepie_cache_dir);
335 }
336
337 $rss = new SimplePie();
338 $rss->set_useragent(SELF_USER_AGENT);
339 # $rss->set_timeout(10);
340 $rss->set_feed_url($fetch_url);
341 $rss->set_output_encoding('UTF-8');
342 //$rss->force_feed(true);
343
344 if ($debug_enabled) {
345 _debug("feed update interval (sec): " .
346 get_feed_update_interval($link, $feed)*60);
347 }
348
349 $rss->enable_cache(!$no_cache);
350
351 if (!$no_cache) {
352 $rss->set_cache_location($simplepie_cache_dir);
353 $rss->set_cache_duration($cache_age);
354 }
355
356 $rss->init();
357 }
358
359 // print_r($rss);
360
361 if ($debug_enabled) {
362 _debug("update_rss_feed: fetch done, parsing...");
363 }
364
365 $feed = db_escape_string($feed);
366
367 if ($update_method == 2) {
368 $fetch_ok = !$rss->error();
369 } else {
370 $fetch_ok = !!$rss;
371 }
372
373 if ($fetch_ok) {
374
375 if ($debug_enabled) {
376 _debug("update_rss_feed: processing feed data...");
377 }
378
379 // db_query($link, "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 $result = db_query($link, "SELECT title,icon_url,site_url,owner_uid,
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 $orig_icon_url = db_fetch_result($result, 0, "icon_url");
394 $orig_site_url = db_fetch_result($result, 0, "site_url");
395 $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0,
396 "favicon_needs_check"));
397
398 $owner_uid = db_fetch_result($result, 0, "owner_uid");
399
400 if ($use_simplepie) {
401 $site_url = db_escape_string(trim($rss->get_link()));
402 } else {
403 $site_url = db_escape_string(trim($rss->channel["link"]));
404 }
405
406 // weird, weird Magpie
407 if (!$use_simplepie) {
408 if (!$site_url) $site_url = db_escape_string($rss->channel["link_"]);
409 }
410
411 $site_url = rewrite_relative_url($fetch_url, $site_url);
412 $site_url = substr($site_url, 0, 250);
413
414 if ($debug_enabled) {
415 _debug("update_rss_feed: checking favicon...");
416 }
417
418 if ($favicon_needs_check) {
419 check_feed_favicon($site_url, $feed, $link);
420
421 db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
422 WHERE id = '$feed'");
423 }
424
425 if (!$registered_title || $registered_title == "[Unknown]") {
426
427 if ($use_simplepie) {
428 $feed_title = db_escape_string($rss->get_title());
429 } else {
430 $feed_title = db_escape_string($rss->channel["title"]);
431 }
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 // print "I: " . $rss->channel["image"]["url"];
447
448 if (!$use_simplepie) {
449 $icon_url = db_escape_string(trim($rss->image["url"]));
450 } else {
451 $icon_url = db_escape_string(trim($rss->get_image_url()));
452 }
453
454 $icon_url = rewrite_relative_url($fetch_url, $icon_url);
455 $icon_url = substr($icon_url, 0, 250);
456
457 if ($icon_url && $orig_icon_url != $icon_url) {
458 db_query($link, "UPDATE ttrss_feeds SET icon_url = '$icon_url' WHERE id = '$feed'");
459 }
460
461 if ($debug_enabled) {
462 _debug("update_rss_feed: loading filters...");
463 }
464
465 $filters = load_filters($link, $feed, $owner_uid);
466
467 // if ($debug_enabled) {
468 // print_r($filters);
469 // }
470
471 if ($use_simplepie) {
472 $iterator = $rss->get_items();
473 } else {
474 $iterator = $rss->items;
475 if (!$iterator || !is_array($iterator)) $iterator = $rss->entries;
476 if (!$iterator || !is_array($iterator)) $iterator = $rss;
477 }
478
479 if (!is_array($iterator)) {
480 /* db_query($link, "UPDATE ttrss_feeds
481 SET last_error = 'Parse error: can\'t find any articles.'
482 WHERE id = '$feed'"); */
483
484 // clear any errors and mark feed as updated if fetched okay
485 // even if it's blank
486
487 if ($debug_enabled) {
488 _debug("update_rss_feed: entry iterator is not an array, no articles?");
489 }
490
491 db_query($link, "UPDATE ttrss_feeds
492 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
493
494 return; // no articles
495 }
496
497 if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) {
498
499 if ($debug_enabled) _debug("update_rss_feed: checking for PUSH hub...");
500
501 $feed_hub_url = false;
502 if ($use_simplepie) {
503 $links = $rss->get_links('hub');
504
505 if ($links && is_array($links)) {
506 foreach ($links as $l) {
507 $feed_hub_url = $l;
508 break;
509 }
510 }
511
512 } else {
513 $atom = $rss->channel['atom'];
514
515 if ($atom) {
516 if ($atom['link@rel'] == 'hub') {
517 $feed_hub_url = $atom['link@href'];
518 }
519
520 if (!$feed_hub_url && $atom['link#'] > 1) {
521 for ($i = 2; $i <= $atom['link#']; $i++) {
522 if ($atom["link#$i@rel"] == 'hub') {
523 $feed_hub_url = $atom["link#$i@href"];
524 break;
525 }
526 }
527 }
528 } else {
529 $feed_hub_url = $rss->channel['link_hub'];
530 }
531 }
532
533 if ($debug_enabled) _debug("update_rss_feed: feed hub url: $feed_hub_url");
534
535 if ($feed_hub_url && function_exists('curl_init') &&
536 !ini_get("open_basedir")) {
537
538 require_once 'lib/pubsubhubbub/subscriber.php';
539
540 $callback_url = get_self_url_prefix() .
541 "/public.php?op=pubsub&id=$feed";
542
543 $s = new Subscriber($feed_hub_url, $callback_url);
544
545 $rc = $s->subscribe($fetch_url);
546
547 if ($debug_enabled)
548 _debug("update_rss_feed: feed hub url found, subscribe request sent.");
549
550 db_query($link, "UPDATE ttrss_feeds SET pubsub_state = 1
551 WHERE id = '$feed'");
552 }
553 }
554
555 if ($debug_enabled) {
556 _debug("update_rss_feed: processing articles...");
557 }
558
559 foreach ($iterator as $item) {
560
561 if ($_REQUEST['xdebug'] == 2) {
562 print_r($item);
563 }
564
565 if ($use_simplepie) {
566 $entry_guid = $item->get_id();
567 if (!$entry_guid) $entry_guid = $item->get_link();
568 if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
569
570 } else {
571
572 $entry_guid = $item["id"];
573
574 if (!$entry_guid) $entry_guid = $item["guid"];
575 if (!$entry_guid) $entry_guid = $item["about"];
576 if (!$entry_guid) $entry_guid = $item["link"];
577 if (!$entry_guid) $entry_guid = make_guid_from_title($item["title"]);
578 }
579
580 if ($debug_enabled) {
581 _debug("update_rss_feed: guid $entry_guid");
582 }
583
584 if (!$entry_guid) continue;
585
586 $entry_timestamp = "";
587
588 if ($use_simplepie) {
589 $entry_timestamp = strtotime($item->get_date());
590 } else {
591 $rss_2_date = $item['pubdate'];
592 $rss_1_date = $item['dc']['date'];
593 $atom_date = $item['issued'];
594 if (!$atom_date) $atom_date = $item['updated'];
595
596 if ($atom_date != "") $entry_timestamp = parse_w3cdtf($atom_date);
597 if ($rss_1_date != "") $entry_timestamp = parse_w3cdtf($rss_1_date);
598 if ($rss_2_date != "") $entry_timestamp = strtotime($rss_2_date);
599
600 }
601
602 if ($entry_timestamp == "" || $entry_timestamp == -1 || !$entry_timestamp) {
603 $entry_timestamp = time();
604 $no_orig_date = 'true';
605 } else {
606 $no_orig_date = 'false';
607 }
608
609 $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp);
610
611 if ($debug_enabled) {
612 _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
613 }
614
615 if ($use_simplepie) {
616 $entry_title = $item->get_title();
617 } else {
618 $entry_title = trim(strip_tags($item["title"]));
619 }
620
621 if ($use_simplepie) {
622 $entry_link = $item->get_link();
623 } else {
624 // strange Magpie workaround
625 $entry_link = $item["link_"];
626 if (!$entry_link) $entry_link = $item["link"];
627 }
628
629 $entry_link = rewrite_relative_url($site_url, $entry_link);
630
631 if ($debug_enabled) {
632 _debug("update_rss_feed: title $entry_title");
633 _debug("update_rss_feed: link $entry_link");
634 }
635
636 if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);;
637
638 $entry_link = strip_tags($entry_link);
639
640 if ($use_simplepie) {
641 $entry_content = $item->get_content();
642 if (!$entry_content) $entry_content = $item->get_description();
643 } else {
644 $entry_content = $item["content:escaped"];
645
646 if (!$entry_content) $entry_content = $item["content:encoded"];
647 if (!$entry_content && is_array($entry_content)) $entry_content = $item["content"]["encoded"];
648 if (!$entry_content) $entry_content = $item["content"];
649
650 if (is_array($entry_content)) $entry_content = $entry_content[0];
651
652 // Magpie bugs are getting ridiculous
653 if (trim($entry_content) == "Array") $entry_content = false;
654
655 if (!$entry_content) $entry_content = $item["atom_content"];
656 if (!$entry_content) $entry_content = $item["summary"];
657
658 if (!$entry_content ||
659 strlen($entry_content) < strlen($item["description"])) {
660 $entry_content = $item["description"];
661 };
662
663 // WTF
664 if (is_array($entry_content)) {
665 $entry_content = $entry_content["encoded"];
666 if (!$entry_content) $entry_content = $entry_content["escaped"];
667 }
668 }
669
670 if ($cache_images && is_writable(CACHE_DIR . '/images'))
671 $entry_content = cache_images($entry_content, $site_url, $debug_enabled);
672
673 if ($_REQUEST["xdebug"] == 2) {
674 print "update_rss_feed: content: ";
675 print $entry_content;
676 print "\n";
677 }
678
679 $entry_content_unescaped = $entry_content;
680
681 if ($use_simplepie) {
682 $entry_comments = strip_tags($item->data["comments"]);
683 if ($item->get_author()) {
684 $entry_author_item = $item->get_author();
685 $entry_author = $entry_author_item->get_name();
686 if (!$entry_author) $entry_author = $entry_author_item->get_email();
687
688 $entry_author = db_escape_string($entry_author);
689 }
690 } else {
691 $entry_comments = strip_tags($item["comments"]);
692
693 $entry_author = db_escape_string(strip_tags($item['dc']['creator']));
694
695 if ($item['author']) {
696
697 if (is_array($item['author'])) {
698
699 if (!$entry_author) {
700 $entry_author = db_escape_string(strip_tags($item['author']['name']));
701 }
702
703 if (!$entry_author) {
704 $entry_author = db_escape_string(strip_tags($item['author']['email']));
705 }
706 }
707
708 if (!$entry_author) {
709 $entry_author = db_escape_string(strip_tags($item['author']));
710 }
711 }
712 }
713
714 if (preg_match('/^[\t\n\r ]*$/', $entry_author)) $entry_author = '';
715
716 $entry_guid = db_escape_string(strip_tags($entry_guid));
717 $entry_guid = mb_substr($entry_guid, 0, 250);
718
719 $result = db_query($link, "SELECT id FROM ttrss_entries
720 WHERE guid = '$entry_guid'");
721
722 $entry_content = db_escape_string($entry_content, false);
723
724 $content_hash = "SHA1:" . sha1(strip_tags($entry_content));
725
726 $entry_title = db_escape_string($entry_title);
727 $entry_link = db_escape_string($entry_link);
728 $entry_comments = mb_substr(db_escape_string($entry_comments), 0, 250);
729 $entry_author = mb_substr($entry_author, 0, 250);
730
731 if ($use_simplepie) {
732 $num_comments = 0; #FIXME#
733 } else {
734 $num_comments = db_escape_string($item["slash"]["comments"]);
735 }
736
737 if (!$num_comments) $num_comments = 0;
738
739 if ($debug_enabled) {
740 _debug("update_rss_feed: looking for tags [1]...");
741 }
742
743 // parse <category> entries into tags
744
745 $additional_tags = array();
746
747 if ($use_simplepie) {
748
749 $additional_tags_src = $item->get_categories();
750
751 if (is_array($additional_tags_src)) {
752 foreach ($additional_tags_src as $tobj) {
753 array_push($additional_tags, $tobj->get_term());
754 }
755 }
756
757 if ($debug_enabled) {
758 _debug("update_rss_feed: category tags:");
759 print_r($additional_tags);
760 }
761
762 } else {
763
764 $t_ctr = $item['category#'];
765
766 if ($t_ctr == 0) {
767 $additional_tags = array();
768 } else if ($t_ctr > 0) {
769 $additional_tags = array($item['category']);
770
771 if ($item['category@term']) {
772 array_push($additional_tags, $item['category@term']);
773 }
774
775 for ($i = 0; $i <= $t_ctr; $i++ ) {
776 if ($item["category#$i"]) {
777 array_push($additional_tags, $item["category#$i"]);
778 }
779
780 if ($item["category#$i@term"]) {
781 array_push($additional_tags, $item["category#$i@term"]);
782 }
783 }
784 }
785
786 // parse <dc:subject> elements
787
788 $t_ctr = $item['dc']['subject#'];
789
790 if ($t_ctr > 0) {
791 array_push($additional_tags, $item['dc']['subject']);
792
793 for ($i = 0; $i <= $t_ctr; $i++ ) {
794 if ($item['dc']["subject#$i"]) {
795 array_push($additional_tags, $item['dc']["subject#$i"]);
796 }
797 }
798 }
799 }
800
801 if ($debug_enabled) {
802 _debug("update_rss_feed: looking for tags [2]...");
803 }
804
805 /* taaaags */
806 // <a href="..." rel="tag">Xorg</a>, //
807
808 $entry_tags = null;
809
810 preg_match_all("/<a.*?rel=['\"]tag['\"].*?\>([^<]+)<\/a>/i",
811 $entry_content_unescaped, $entry_tags);
812
813 $entry_tags = $entry_tags[1];
814
815 $entry_tags = array_merge($entry_tags, $additional_tags);
816 $entry_tags = array_unique($entry_tags);
817
818 for ($i = 0; $i < count($entry_tags); $i++)
819 $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8');
820
821 if ($debug_enabled) {
822 _debug("update_rss_feed: unfiltered tags found:");
823 print_r($entry_tags);
824 }
825
826 # sanitize content
827
828 $entry_content = sanitize_article_content($entry_content);
829 $entry_title = sanitize_article_content($entry_title);
830
831 if ($debug_enabled) {
832 _debug("update_rss_feed: done collecting data [TITLE:$entry_title]");
833 }
834
835 db_query($link, "BEGIN");
836
837 if (db_num_rows($result) == 0) {
838
839 if ($debug_enabled) {
840 _debug("update_rss_feed: base guid not found");
841 }
842
843 // base post entry does not exist, create it
844
845 $result = db_query($link,
846 "INSERT INTO ttrss_entries
847 (title,
848 guid,
849 link,
850 updated,
851 content,
852 content_hash,
853 no_orig_date,
854 date_updated,
855 date_entered,
856 comments,
857 num_comments,
858 author)
859 VALUES
860 ('$entry_title',
861 '$entry_guid',
862 '$entry_link',
863 '$entry_timestamp_fmt',
864 '$entry_content',
865 '$content_hash',
866 $no_orig_date,
867 NOW(),
868 NOW(),
869 '$entry_comments',
870 '$num_comments',
871 '$entry_author')");
872 } else {
873 // we keep encountering the entry in feeds, so we need to
874 // update date_updated column so that we don't get horrible
875 // dupes when the entry gets purged and reinserted again e.g.
876 // in the case of SLOW SLOW OMG SLOW updating feeds
877
878 $base_entry_id = db_fetch_result($result, 0, "id");
879
880 db_query($link, "UPDATE ttrss_entries SET date_updated = NOW()
881 WHERE id = '$base_entry_id'");
882 }
883
884 // now it should exist, if not - bad luck then
885
886 $result = db_query($link, "SELECT
887 id,content_hash,no_orig_date,title,
888 ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
889 ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
890 num_comments
891 FROM
892 ttrss_entries
893 WHERE guid = '$entry_guid'");
894
895 $entry_ref_id = 0;
896 $entry_int_id = 0;
897
898 if (db_num_rows($result) == 1) {
899
900 if ($debug_enabled) {
901 _debug("update_rss_feed: base guid found, checking for user record");
902 }
903
904 // this will be used below in update handler
905 $orig_content_hash = db_fetch_result($result, 0, "content_hash");
906 $orig_title = db_fetch_result($result, 0, "title");
907 $orig_num_comments = db_fetch_result($result, 0, "num_comments");
908 $orig_date_updated = strtotime(db_fetch_result($result,
909 0, "date_updated"));
910
911 $ref_id = db_fetch_result($result, 0, "id");
912 $entry_ref_id = $ref_id;
913
914 // check for user post link to main table
915
916 // do we allow duplicate posts with same GUID in different feeds?
917 if (get_pref($link, "ALLOW_DUPLICATE_POSTS", $owner_uid, false)) {
918 $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)";
919 } else {
920 $dupcheck_qpart = "";
921 }
922
923 /* Collect article tags here so we could filter by them: */
924
925 $article_filters = get_article_filters($filters, $entry_title,
926 $entry_content, $entry_link, $entry_timestamp, $entry_author,
927 $entry_tags);
928
929 if ($debug_enabled) {
930 _debug("update_rss_feed: article filters: ");
931 if (count($article_filters) != 0) {
932 print_r($article_filters);
933 }
934 }
935
936 if (find_article_filter($article_filters, "filter")) {
937 db_query($link, "COMMIT"); // close transaction in progress
938 continue;
939 }
940
941 $score = calculate_article_score($article_filters);
942
943 if ($debug_enabled) {
944 _debug("update_rss_feed: initial score: $score");
945 }
946
947 $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE
948 ref_id = '$ref_id' AND owner_uid = '$owner_uid'
949 $dupcheck_qpart";
950
951 // if ($_REQUEST["xdebug"]) print "$query\n";
952
953 $result = db_query($link, $query);
954
955 // okay it doesn't exist - create user entry
956 if (db_num_rows($result) == 0) {
957
958 if ($debug_enabled) {
959 _debug("update_rss_feed: user record not found, creating...");
960 }
961
962 if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) {
963 $unread = 'true';
964 $last_read_qpart = 'NULL';
965 } else {
966 $unread = 'false';
967 $last_read_qpart = 'NOW()';
968 }
969
970 if (find_article_filter($article_filters, 'mark') || $score > 1000) {
971 $marked = 'true';
972 } else {
973 $marked = 'false';
974 }
975
976 if (find_article_filter($article_filters, 'publish')) {
977 $published = 'true';
978 } else {
979 $published = 'false';
980 }
981
982 $result = db_query($link,
983 "INSERT INTO ttrss_user_entries
984 (ref_id, owner_uid, feed_id, unread, last_read, marked,
985 published, score, tag_cache, label_cache, uuid)
986 VALUES ('$ref_id', '$owner_uid', '$feed', $unread,
987 $last_read_qpart, $marked, $published, '$score', '', '', '')");
988
989 if (PUBSUBHUBBUB_HUB && $published == 'true') {
990 $rss_link = get_self_url_prefix() .
991 "/public.php?op=rss&id=-2&key=" .
992 get_feed_access_key($link, -2, false, $owner_uid);
993
994 $p = new Publisher(PUBSUBHUBBUB_HUB);
995
996 $pubsub_result = $p->publish_update($rss_link);
997 }
998
999 $result = db_query($link,
1000 "SELECT int_id FROM ttrss_user_entries WHERE
1001 ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND
1002 feed_id = '$feed' LIMIT 1");
1003
1004 if (db_num_rows($result) == 1) {
1005 $entry_int_id = db_fetch_result($result, 0, "int_id");
1006 }
1007 } else {
1008 if ($debug_enabled) {
1009 _debug("update_rss_feed: user record FOUND");
1010 }
1011
1012 $entry_ref_id = db_fetch_result($result, 0, "ref_id");
1013 $entry_int_id = db_fetch_result($result, 0, "int_id");
1014 }
1015
1016 if ($debug_enabled) {
1017 _debug("update_rss_feed: RID: $entry_ref_id, IID: $entry_int_id");
1018 }
1019
1020 $post_needs_update = false;
1021 $update_insignificant = false;
1022
1023 if ($orig_num_comments != $num_comments) {
1024 $post_needs_update = true;
1025 $update_insignificant = true;
1026 }
1027
1028 if ($content_hash != $orig_content_hash) {
1029 $post_needs_update = true;
1030 $update_insignificant = false;
1031 }
1032
1033 if (db_escape_string($orig_title) != $entry_title) {
1034 $post_needs_update = true;
1035 $update_insignificant = false;
1036 }
1037
1038 // if post needs update, update it and mark all user entries
1039 // linking to this post as updated
1040 if ($post_needs_update) {
1041
1042 if (defined('DAEMON_EXTENDED_DEBUG')) {
1043 _debug("update_rss_feed: post $entry_guid needs update...");
1044 }
1045
1046 // print "<!-- post $orig_title needs update : $post_needs_update -->";
1047
1048 db_query($link, "UPDATE ttrss_entries
1049 SET title = '$entry_title', content = '$entry_content',
1050 content_hash = '$content_hash',
1051 updated = '$entry_timestamp_fmt',
1052 num_comments = '$num_comments'
1053 WHERE id = '$ref_id'");
1054
1055 if (!$update_insignificant) {
1056 if ($mark_unread_on_update) {
1057 db_query($link, "UPDATE ttrss_user_entries
1058 SET last_read = null, unread = true WHERE ref_id = '$ref_id'");
1059 } else if ($update_on_checksum_change) {
1060 db_query($link, "UPDATE ttrss_user_entries
1061 SET last_read = null WHERE ref_id = '$ref_id'
1062 AND unread = false");
1063 }
1064 }
1065 }
1066 }
1067
1068 db_query($link, "COMMIT");
1069
1070 if ($debug_enabled) {
1071 _debug("update_rss_feed: assigning labels...");
1072 }
1073
1074 assign_article_to_labels($link, $entry_ref_id, $article_filters,
1075 $owner_uid);
1076
1077 if ($debug_enabled) {
1078 _debug("update_rss_feed: looking for enclosures...");
1079 }
1080
1081 // enclosures
1082
1083 $enclosures = array();
1084
1085 if ($use_simplepie) {
1086 $encs = $item->get_enclosures();
1087
1088 if (is_array($encs)) {
1089 foreach ($encs as $e) {
1090 $e_item = array(
1091 $e->link, $e->type, $e->length);
1092
1093 array_push($enclosures, $e_item);
1094 }
1095 }
1096
1097 } else {
1098 // <enclosure>
1099
1100 $e_ctr = $item['enclosure#'];
1101
1102 if ($e_ctr > 0) {
1103 $e_item = array($item['enclosure@url'],
1104 $item['enclosure@type'],
1105 $item['enclosure@length']);
1106
1107 array_push($enclosures, $e_item);
1108
1109 for ($i = 0; $i <= $e_ctr; $i++ ) {
1110
1111 if ($item["enclosure#$i@url"]) {
1112 $e_item = array($item["enclosure#$i@url"],
1113 $item["enclosure#$i@type"],
1114 $item["enclosure#$i@length"]);
1115 array_push($enclosures, $e_item);
1116 }
1117 }
1118 }
1119
1120 // <media:content>
1121 // can there be many of those? yes -fox
1122
1123 $m_ctr = $item['media']['content#'];
1124
1125 if ($m_ctr > 0) {
1126 $e_item = array($item['media']['content@url'],
1127 $item['media']['content@medium'],
1128 $item['media']['content@length']);
1129
1130 array_push($enclosures, $e_item);
1131
1132 for ($i = 0; $i <= $m_ctr; $i++ ) {
1133
1134 if ($item["media"]["content#$i@url"]) {
1135 $e_item = array($item["media"]["content#$i@url"],
1136 $item["media"]["content#$i@medium"],
1137 $item["media"]["content#$i@length"]);
1138 array_push($enclosures, $e_item);
1139 }
1140 }
1141
1142 }
1143 }
1144
1145
1146 if ($debug_enabled) {
1147 _debug("update_rss_feed: article enclosures:");
1148 print_r($enclosures);
1149 }
1150
1151 db_query($link, "BEGIN");
1152
1153 foreach ($enclosures as $enc) {
1154 $enc_url = db_escape_string($enc[0]);
1155 $enc_type = db_escape_string($enc[1]);
1156 $enc_dur = db_escape_string($enc[2]);
1157
1158 $result = db_query($link, "SELECT id FROM ttrss_enclosures
1159 WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'");
1160
1161 if (db_num_rows($result) == 0) {
1162 db_query($link, "INSERT INTO ttrss_enclosures
1163 (content_url, content_type, title, duration, post_id) VALUES
1164 ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')");
1165 }
1166 }
1167
1168 db_query($link, "COMMIT");
1169
1170 // check for manual tags (we have to do it here since they're loaded from filters)
1171
1172 foreach ($article_filters as $f) {
1173 if ($f[0] == "tag") {
1174
1175 $manual_tags = trim_array(explode(",", $f[1]));
1176
1177 foreach ($manual_tags as $tag) {
1178 if (tag_is_valid($tag)) {
1179 array_push($entry_tags, $tag);
1180 }
1181 }
1182 }
1183 }
1184
1185 // Skip boring tags
1186
1187 $boring_tags = trim_array(explode(",", mb_strtolower(get_pref($link,
1188 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8')));
1189
1190 $filtered_tags = array();
1191 $tags_to_cache = array();
1192
1193 if ($entry_tags && is_array($entry_tags)) {
1194 foreach ($entry_tags as $tag) {
1195 if (array_search($tag, $boring_tags) === false) {
1196 array_push($filtered_tags, $tag);
1197 }
1198 }
1199 }
1200
1201 $filtered_tags = array_unique($filtered_tags);
1202
1203 if ($debug_enabled) {
1204 _debug("update_rss_feed: filtered article tags:");
1205 print_r($filtered_tags);
1206 }
1207
1208 // Save article tags in the database
1209
1210 if (count($filtered_tags) > 0) {
1211
1212 db_query($link, "BEGIN");
1213
1214 foreach ($filtered_tags as $tag) {
1215
1216 $tag = sanitize_tag($tag);
1217 $tag = db_escape_string($tag);
1218
1219 if (!tag_is_valid($tag)) continue;
1220
1221 $result = db_query($link, "SELECT id FROM ttrss_tags
1222 WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND
1223 owner_uid = '$owner_uid' LIMIT 1");
1224
1225 if ($result && db_num_rows($result) == 0) {
1226
1227 db_query($link, "INSERT INTO ttrss_tags
1228 (owner_uid,tag_name,post_int_id)
1229 VALUES ('$owner_uid','$tag', '$entry_int_id')");
1230 }
1231
1232 array_push($tags_to_cache, $tag);
1233 }
1234
1235 /* update the cache */
1236
1237 $tags_to_cache = array_unique($tags_to_cache);
1238
1239 $tags_str = db_escape_string(join(",", $tags_to_cache));
1240
1241 db_query($link, "UPDATE ttrss_user_entries
1242 SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id'
1243 AND owner_uid = $owner_uid");
1244
1245 db_query($link, "COMMIT");
1246 }
1247
1248 if ($debug_enabled) {
1249 _debug("update_rss_feed: article processed");
1250 }
1251 }
1252
1253 if (!$last_updated) {
1254 if ($debug_enabled) {
1255 _debug("update_rss_feed: new feed, catching it up...");
1256 }
1257 catchup_feed($link, $feed, false, $owner_uid);
1258 }
1259
1260 if ($debug_enabled) {
1261 _debug("purging feed...");
1262 }
1263
1264 purge_feed($link, $feed, 0, $debug_enabled);
1265
1266 db_query($link, "UPDATE ttrss_feeds
1267 SET last_updated = NOW(), last_error = '' WHERE id = '$feed'");
1268
1269 // db_query($link, "COMMIT");
1270
1271 } else {
1272
1273 if ($use_simplepie) {
1274 $error_msg = mb_substr($rss->error(), 0, 250);
1275 } else {
1276 $error_msg = mb_substr(magpie_error(), 0, 250);
1277 }
1278
1279 if ($debug_enabled) {
1280 _debug("update_rss_feed: error fetching feed: $error_msg");
1281 }
1282
1283 $error_msg = db_escape_string($error_msg);
1284
1285 db_query($link,
1286 "UPDATE ttrss_feeds SET last_error = '$error_msg',
1287 last_updated = NOW() WHERE id = '$feed'");
1288 }
1289
1290 if ($use_simplepie) {
1291 unset($rss);
1292 }
1293
1294 if ($debug_enabled) {
1295 _debug("update_rss_feed: done");
1296 }
1297
1298 }
1299
1300 function cache_images($html, $site_url, $debug) {
1301 $cache_dir = CACHE_DIR . "/images";
1302
1303 libxml_use_internal_errors(true);
1304
1305 $charset_hack = '<head>
1306 <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
1307 </head>';
1308
1309 $doc = new DOMDocument();
1310 $doc->loadHTML($charset_hack . $html);
1311 $xpath = new DOMXPath($doc);
1312
1313 $entries = $xpath->query('(//img[@src])');
1314
1315 foreach ($entries as $entry) {
1316 if ($entry->hasAttribute('src')) {
1317 $src = rewrite_relative_url($site_url, $entry->getAttribute('src'));
1318
1319 $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png";
1320
1321 if ($debug) _debug("cache_images: downloading: $src to $local_filename");
1322
1323 if (!file_exists($local_filename)) {
1324 $file_content = fetch_file_contents($src);
1325
1326 if ($file_content && strlen($file_content) > 1024) {
1327 file_put_contents($local_filename, $file_content);
1328 }
1329 }
1330
1331 if (file_exists($local_filename)) {
1332 $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' .
1333 base64_encode($src));
1334 }
1335 }
1336 }
1337
1338 $node = $doc->getElementsByTagName('body')->item(0);
1339
1340 return $doc->saveXML($node, LIBXML_NOEMPTYTAG);
1341 }
1342
1343 function expire_cached_files($debug) {
1344 foreach (array("magpie", "simplepie", "images", "export") as $dir) {
1345 $cache_dir = CACHE_DIR . "/$dir";
1346
1347 if ($debug) _debug("Expiring $cache_dir");
1348
1349 $num_deleted = 0;
1350
1351 if (is_writable($cache_dir)) {
1352 $files = glob("$cache_dir/*");
1353
1354 if ($files)
1355 foreach ($files as $file) {
1356 if (time() - filemtime($file) > 86400*7) {
1357 unlink($file);
1358
1359 ++$num_deleted;
1360 }
1361 }
1362 }
1363
1364 if ($debug) _debug("Removed $num_deleted files.");
1365 }
1366 }
1367
1368 /**
1369 * Source: http://www.php.net/manual/en/function.parse-url.php#104527
1370 * Returns the url query as associative array
1371 *
1372 * @param string query
1373 * @return array params
1374 */
1375 function convertUrlQuery($query) {
1376 $queryParts = explode('&', $query);
1377
1378 $params = array();
1379
1380 foreach ($queryParts as $param) {
1381 $item = explode('=', $param);
1382 $params[$item[0]] = $item[1];
1383 }
1384
1385 return $params;
1386 }
1387 ?>
web-based news feed (RSS/Atom) reader and aggregator; see https://tt-rss.org/