]>
Commit | Line | Data |
---|---|---|
1 | <?php | |
2 | define_default('DAEMON_UPDATE_LOGIN_LIMIT', 30); | |
3 | define_default('DAEMON_FEED_LIMIT', 500); | |
4 | define_default('DAEMON_SLEEP_INTERVAL', 120); | |
5 | define_default('_MIN_CACHE_IMAGE_SIZE', 1024); | |
6 | ||
7 | function update_feedbrowser_cache() { | |
8 | ||
9 | $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers | |
10 | FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf | |
11 | WHERE tf.feed_url = ttrss_feeds.feed_url | |
12 | AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')) | |
13 | GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000"); | |
14 | ||
15 | db_query("BEGIN"); | |
16 | ||
17 | db_query("DELETE FROM ttrss_feedbrowser_cache"); | |
18 | ||
19 | $count = 0; | |
20 | ||
21 | while ($line = db_fetch_assoc($result)) { | |
22 | $subscribers = db_escape_string($line["subscribers"]); | |
23 | $feed_url = db_escape_string($line["feed_url"]); | |
24 | $title = db_escape_string($line["title"]); | |
25 | $site_url = db_escape_string($line["site_url"]); | |
26 | ||
27 | $tmp_result = db_query("SELECT subscribers FROM | |
28 | ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'"); | |
29 | ||
30 | if (db_num_rows($tmp_result) == 0) { | |
31 | ||
32 | db_query("INSERT INTO ttrss_feedbrowser_cache | |
33 | (feed_url, site_url, title, subscribers) VALUES ('$feed_url', | |
34 | '$site_url', '$title', '$subscribers')"); | |
35 | ||
36 | ++$count; | |
37 | ||
38 | } | |
39 | ||
40 | } | |
41 | ||
42 | db_query("COMMIT"); | |
43 | ||
44 | return $count; | |
45 | ||
46 | } | |
47 | ||
48 | ||
49 | /** | |
50 | * Update a feed batch. | |
51 | * Used by daemons to update n feeds by run. | |
52 | * Only update feed needing a update, and not being processed | |
53 | * by another process. | |
54 | * | |
55 | * @param mixed $link Database link | |
56 | * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT. | |
57 | * @param boolean $from_http Set to true if you call this function from http to disable cli specific code. | |
58 | * @param boolean $debug Set to false to disable debug output. Default to true. | |
59 | * @return void | |
60 | */ | |
61 | function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) { | |
62 | // Process all other feeds using last_updated and interval parameters | |
63 | ||
64 | $schema_version = get_schema_version(); | |
65 | ||
66 | if ($schema_version != SCHEMA_VERSION) { | |
67 | die("Schema version is wrong, please upgrade the database.\n"); | |
68 | } | |
69 | ||
70 | define('PREFS_NO_CACHE', true); | |
71 | ||
72 | // Test if the user has loggued in recently. If not, it does not update its feeds. | |
73 | if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) { | |
74 | if (DB_TYPE == "pgsql") { | |
75 | $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'"; | |
76 | } else { | |
77 | $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)"; | |
78 | } | |
79 | } else { | |
80 | $login_thresh_qpart = ""; | |
81 | } | |
82 | ||
83 | // Test if the feed need a update (update interval exceeded). | |
84 | if (DB_TYPE == "pgsql") { | |
85 | $update_limit_qpart = "AND (( | |
86 | ttrss_feeds.update_interval = 0 | |
87 | AND ttrss_user_prefs.value != '-1' | |
88 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL) | |
89 | ) OR ( | |
90 | ttrss_feeds.update_interval > 0 | |
91 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL) | |
92 | ) OR (ttrss_feeds.last_updated IS NULL | |
93 | AND ttrss_user_prefs.value != '-1') | |
94 | OR (last_updated = '1970-01-01 00:00:00' | |
95 | AND ttrss_user_prefs.value != '-1'))"; | |
96 | } else { | |
97 | $update_limit_qpart = "AND (( | |
98 | ttrss_feeds.update_interval = 0 | |
99 | AND ttrss_user_prefs.value != '-1' | |
100 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE) | |
101 | ) OR ( | |
102 | ttrss_feeds.update_interval > 0 | |
103 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE) | |
104 | ) OR (ttrss_feeds.last_updated IS NULL | |
105 | AND ttrss_user_prefs.value != '-1') | |
106 | OR (last_updated = '1970-01-01 00:00:00' | |
107 | AND ttrss_user_prefs.value != '-1'))"; | |
108 | } | |
109 | ||
110 | // Test if feed is currently being updated by another process. | |
111 | if (DB_TYPE == "pgsql") { | |
112 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')"; | |
113 | } else { | |
114 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))"; | |
115 | } | |
116 | ||
117 | // Test if there is a limit to number of updated feeds | |
118 | $query_limit = ""; | |
119 | if($limit) $query_limit = sprintf("LIMIT %d", $limit); | |
120 | ||
121 | $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated | |
122 | FROM | |
123 | ttrss_feeds, ttrss_users, ttrss_user_prefs | |
124 | WHERE | |
125 | ttrss_feeds.owner_uid = ttrss_users.id | |
126 | AND ttrss_user_prefs.profile IS NULL | |
127 | AND ttrss_users.id = ttrss_user_prefs.owner_uid | |
128 | AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' | |
129 | $login_thresh_qpart $update_limit_qpart | |
130 | $updstart_thresh_qpart | |
131 | ORDER BY last_updated $query_limit"; | |
132 | ||
133 | // We search for feed needing update. | |
134 | $result = db_query($query); | |
135 | ||
136 | if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result))); | |
137 | ||
138 | // Here is a little cache magic in order to minimize risk of double feed updates. | |
139 | $feeds_to_update = array(); | |
140 | while ($line = db_fetch_assoc($result)) { | |
141 | array_push($feeds_to_update, db_escape_string($line['feed_url'])); | |
142 | } | |
143 | ||
144 | // We update the feed last update started date before anything else. | |
145 | // There is no lag due to feed contents downloads | |
146 | // It prevent an other process to update the same feed. | |
147 | ||
148 | if(count($feeds_to_update) > 0) { | |
149 | $feeds_quoted = array(); | |
150 | ||
151 | foreach ($feeds_to_update as $feed) { | |
152 | array_push($feeds_quoted, "'" . db_escape_string($feed) . "'"); | |
153 | } | |
154 | ||
155 | db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW() | |
156 | WHERE feed_url IN (%s)", implode(',', $feeds_quoted))); | |
157 | } | |
158 | ||
159 | $nf = 0; | |
160 | ||
161 | // For each feed, we call the feed update function. | |
162 | foreach ($feeds_to_update as $feed) { | |
163 | if($debug) _debug("Base feed: $feed"); | |
164 | ||
165 | //update_rss_feed($line["id"], true); | |
166 | ||
167 | // since we have the data cached, we can deal with other feeds with the same url | |
168 | ||
169 | $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated,ttrss_feeds.owner_uid | |
170 | FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE | |
171 | ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND | |
172 | ttrss_users.id = ttrss_user_prefs.owner_uid AND | |
173 | ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND | |
174 | ttrss_user_prefs.profile IS NULL AND | |
175 | feed_url = '".db_escape_string($feed)."' AND | |
176 | (ttrss_feeds.update_interval > 0 OR | |
177 | ttrss_user_prefs.value != '-1') | |
178 | $login_thresh_qpart | |
179 | ORDER BY ttrss_feeds.id $query_limit"); | |
180 | ||
181 | if (db_num_rows($tmp_result) > 0) { | |
182 | $rss = false; | |
183 | ||
184 | while ($tline = db_fetch_assoc($tmp_result)) { | |
185 | if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"] . " " . $tline["owner_uid"]); | |
186 | ||
187 | $rss = update_rss_feed($tline["id"], true, false); | |
188 | _debug_suppress(false); | |
189 | ++$nf; | |
190 | } | |
191 | } | |
192 | } | |
193 | ||
194 | require_once "digest.php"; | |
195 | ||
196 | // Send feed digests by email if needed. | |
197 | send_headlines_digests($debug); | |
198 | ||
199 | return $nf; | |
200 | ||
201 | } // function update_daemon_common | |
202 | ||
203 | // ignore_daemon is not used | |
204 | function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, $rss = false) { | |
205 | ||
206 | $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug']; | |
207 | ||
208 | _debug_suppress(!$debug_enabled); | |
209 | _debug("start", $debug_enabled); | |
210 | ||
211 | $result = db_query("SELECT id,update_interval,auth_login, | |
212 | feed_url,auth_pass,cache_images, | |
213 | mark_unread_on_update, owner_uid, | |
214 | pubsub_state, auth_pass_encrypted, | |
215 | (SELECT max(date_entered) FROM | |
216 | ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp | |
217 | FROM ttrss_feeds WHERE id = '$feed'"); | |
218 | ||
219 | if (db_num_rows($result) == 0) { | |
220 | _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled); | |
221 | return false; | |
222 | } | |
223 | ||
224 | $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp")); | |
225 | ||
226 | if (defined('_DISABLE_HTTP_304')) | |
227 | $last_article_timestamp = 0; | |
228 | ||
229 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
230 | $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result, | |
231 | 0, "mark_unread_on_update")); | |
232 | $pubsub_state = db_fetch_result($result, 0, "pubsub_state"); | |
233 | $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result, | |
234 | 0, "auth_pass_encrypted")); | |
235 | ||
236 | db_query("UPDATE ttrss_feeds SET last_update_started = NOW() | |
237 | WHERE id = '$feed'"); | |
238 | ||
239 | $auth_login = db_fetch_result($result, 0, "auth_login"); | |
240 | $auth_pass = db_fetch_result($result, 0, "auth_pass"); | |
241 | ||
242 | if ($auth_pass_encrypted) { | |
243 | require_once "crypt.php"; | |
244 | $auth_pass = decrypt_string($auth_pass); | |
245 | } | |
246 | ||
247 | $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images")); | |
248 | $fetch_url = db_fetch_result($result, 0, "feed_url"); | |
249 | ||
250 | $feed = db_escape_string($feed); | |
251 | ||
252 | $date_feed_processed = date('Y-m-d H:i'); | |
253 | ||
254 | $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".xml"; | |
255 | ||
256 | $pluginhost = new PluginHost(); | |
257 | $pluginhost->set_debug($debug_enabled); | |
258 | $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid); | |
259 | ||
260 | $pluginhost->load(PLUGINS, PluginHost::KIND_ALL); | |
261 | $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid); | |
262 | $pluginhost->load_data(); | |
263 | ||
264 | if ($rss && is_object($rss) && get_class($rss) == "FeedParser") { | |
265 | _debug("using previously initialized parser object"); | |
266 | } else { | |
267 | $rss_hash = false; | |
268 | ||
269 | $force_refetch = isset($_REQUEST["force_refetch"]); | |
270 | ||
271 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FETCH_FEED) as $plugin) { | |
272 | $feed_data = $plugin->hook_fetch_feed($feed_data, $fetch_url, $owner_uid, $feed, $last_article_timestamp, $auth_login, $auth_pass); | |
273 | } | |
274 | ||
275 | // try cache | |
276 | if (!$feed_data && | |
277 | file_exists($cache_filename) && | |
278 | is_readable($cache_filename) && | |
279 | !$auth_login && !$auth_pass && | |
280 | filemtime($cache_filename) > time() - 30) { | |
281 | ||
282 | _debug("using local cache [$cache_filename].", $debug_enabled); | |
283 | ||
284 | @$feed_data = file_get_contents($cache_filename); | |
285 | ||
286 | if ($feed_data) { | |
287 | $rss_hash = sha1($feed_data); | |
288 | } | |
289 | ||
290 | } else { | |
291 | _debug("local cache will not be used for this feed", $debug_enabled); | |
292 | } | |
293 | ||
294 | // fetch feed from source | |
295 | if (!$feed_data) { | |
296 | _debug("fetching [$fetch_url]...", $debug_enabled); | |
297 | _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled); | |
298 | ||
299 | $feed_data = fetch_file_contents($fetch_url, false, | |
300 | $auth_login, $auth_pass, false, | |
301 | $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT, | |
302 | $force_refetch ? 0 : $last_article_timestamp); | |
303 | ||
304 | global $fetch_curl_used; | |
305 | ||
306 | if (!$fetch_curl_used) { | |
307 | $tmp = @gzdecode($feed_data); | |
308 | ||
309 | if ($tmp) $feed_data = $tmp; | |
310 | } | |
311 | ||
312 | $feed_data = trim($feed_data); | |
313 | ||
314 | _debug("fetch done.", $debug_enabled); | |
315 | ||
316 | /* if ($feed_data) { | |
317 | $error = verify_feed_xml($feed_data); | |
318 | ||
319 | if ($error) { | |
320 | _debug("error verifying XML, code: " . $error->code, $debug_enabled); | |
321 | ||
322 | if ($error->code == 26) { | |
323 | _debug("got error 26, trying to decode entities...", $debug_enabled); | |
324 | ||
325 | $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8'); | |
326 | ||
327 | $error = verify_feed_xml($feed_data); | |
328 | ||
329 | if ($error) $feed_data = ''; | |
330 | } | |
331 | } | |
332 | } */ | |
333 | ||
334 | // cache vanilla feed data for re-use | |
335 | if ($feed_data && !$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) { | |
336 | $new_rss_hash = sha1($feed_data); | |
337 | ||
338 | if ($new_rss_hash != $rss_hash) { | |
339 | _debug("saving $cache_filename", $debug_enabled); | |
340 | @file_put_contents($cache_filename, $feed_data); | |
341 | } | |
342 | } | |
343 | } | |
344 | ||
345 | if (!$feed_data) { | |
346 | global $fetch_last_error; | |
347 | global $fetch_last_error_code; | |
348 | ||
349 | _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled); | |
350 | ||
351 | $error_escaped = ''; | |
352 | ||
353 | // If-Modified-Since | |
354 | if ($fetch_last_error_code != 304) { | |
355 | $error_escaped = db_escape_string($fetch_last_error); | |
356 | } else { | |
357 | _debug("source claims data not modified, nothing to do.", $debug_enabled); | |
358 | } | |
359 | ||
360 | db_query( | |
361 | "UPDATE ttrss_feeds SET last_error = '$error_escaped', | |
362 | last_updated = NOW() WHERE id = '$feed'"); | |
363 | ||
364 | return; | |
365 | } | |
366 | } | |
367 | ||
368 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) { | |
369 | $feed_data = $plugin->hook_feed_fetched($feed_data, $fetch_url, $owner_uid, $feed); | |
370 | } | |
371 | ||
372 | // set last update to now so if anything *simplepie* crashes later we won't be | |
373 | // continuously failing on the same feed | |
374 | //db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'"); | |
375 | ||
376 | if (!$rss) { | |
377 | $rss = new FeedParser($feed_data); | |
378 | $rss->init(); | |
379 | } | |
380 | ||
381 | if (DETECT_ARTICLE_LANGUAGE) { | |
382 | require_once "lib/languagedetect/LanguageDetect.php"; | |
383 | ||
384 | $lang = new Text_LanguageDetect(); | |
385 | $lang->setNameMode(2); | |
386 | } | |
387 | ||
388 | // print_r($rss); | |
389 | ||
390 | $feed = db_escape_string($feed); | |
391 | ||
392 | if (!$rss->error()) { | |
393 | ||
394 | // We use local pluginhost here because we need to load different per-user feed plugins | |
395 | $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss); | |
396 | ||
397 | _debug("processing feed data...", $debug_enabled); | |
398 | ||
399 | // db_query("BEGIN"); | |
400 | ||
401 | if (DB_TYPE == "pgsql") { | |
402 | $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'"; | |
403 | } else { | |
404 | $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)"; | |
405 | } | |
406 | ||
407 | $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color, | |
408 | (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS | |
409 | favicon_needs_check | |
410 | FROM ttrss_feeds WHERE id = '$feed'"); | |
411 | ||
412 | $registered_title = db_fetch_result($result, 0, "title"); | |
413 | $orig_site_url = db_fetch_result($result, 0, "site_url"); | |
414 | $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0, | |
415 | "favicon_needs_check")); | |
416 | $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color"); | |
417 | ||
418 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
419 | ||
420 | $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); | |
421 | ||
422 | _debug("site_url: $site_url", $debug_enabled); | |
423 | _debug("feed_title: " . $rss->get_title(), $debug_enabled); | |
424 | ||
425 | if ($favicon_needs_check || $force_refetch) { | |
426 | ||
427 | /* terrible hack: if we crash on floicon shit here, we won't check | |
428 | * the icon avgcolor again (unless the icon got updated) */ | |
429 | ||
430 | $favicon_file = ICONS_DIR . "/$feed.ico"; | |
431 | $favicon_modified = @filemtime($favicon_file); | |
432 | ||
433 | _debug("checking favicon...", $debug_enabled); | |
434 | ||
435 | check_feed_favicon($site_url, $feed); | |
436 | $favicon_modified_new = @filemtime($favicon_file); | |
437 | ||
438 | if ($favicon_modified_new > $favicon_modified) | |
439 | $favicon_avg_color = ''; | |
440 | ||
441 | if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') { | |
442 | require_once "colors.php"; | |
443 | ||
444 | db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE | |
445 | id = '$feed'"); | |
446 | ||
447 | $favicon_color = db_escape_string( | |
448 | calculate_avg_color($favicon_file)); | |
449 | ||
450 | $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'"; | |
451 | } else if ($favicon_avg_color == 'fail') { | |
452 | _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled); | |
453 | } | |
454 | ||
455 | db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW() | |
456 | $favicon_colorstring | |
457 | WHERE id = '$feed'"); | |
458 | } | |
459 | ||
460 | if (!$registered_title || $registered_title == "[Unknown]") { | |
461 | ||
462 | $feed_title = db_escape_string($rss->get_title()); | |
463 | ||
464 | if ($feed_title) { | |
465 | _debug("registering title: $feed_title", $debug_enabled); | |
466 | ||
467 | db_query("UPDATE ttrss_feeds SET | |
468 | title = '$feed_title' WHERE id = '$feed'"); | |
469 | } | |
470 | } | |
471 | ||
472 | if ($site_url && $orig_site_url != $site_url) { | |
473 | db_query("UPDATE ttrss_feeds SET | |
474 | site_url = '$site_url' WHERE id = '$feed'"); | |
475 | } | |
476 | ||
477 | _debug("loading filters & labels...", $debug_enabled); | |
478 | ||
479 | $filters = load_filters($feed, $owner_uid); | |
480 | $labels = get_all_labels($owner_uid); | |
481 | ||
482 | _debug("" . count($filters) . " filters loaded.", $debug_enabled); | |
483 | ||
484 | $items = $rss->get_items(); | |
485 | ||
486 | if (!is_array($items)) { | |
487 | _debug("no articles found.", $debug_enabled); | |
488 | ||
489 | db_query("UPDATE ttrss_feeds | |
490 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
491 | ||
492 | return; // no articles | |
493 | } | |
494 | ||
495 | if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) { | |
496 | ||
497 | _debug("checking for PUSH hub...", $debug_enabled); | |
498 | ||
499 | $feed_hub_url = false; | |
500 | ||
501 | $links = $rss->get_links('hub'); | |
502 | ||
503 | if ($links && is_array($links)) { | |
504 | foreach ($links as $l) { | |
505 | $feed_hub_url = $l; | |
506 | break; | |
507 | } | |
508 | } | |
509 | ||
510 | _debug("feed hub url: $feed_hub_url", $debug_enabled); | |
511 | ||
512 | $feed_self_url = $fetch_url; | |
513 | ||
514 | $links = $rss->get_links('self'); | |
515 | ||
516 | if ($links && is_array($links)) { | |
517 | foreach ($links as $l) { | |
518 | $feed_self_url = $l; | |
519 | break; | |
520 | } | |
521 | } | |
522 | ||
523 | _debug("feed self url = $feed_self_url"); | |
524 | ||
525 | if ($feed_hub_url && $feed_self_url && function_exists('curl_init') && | |
526 | !ini_get("open_basedir")) { | |
527 | ||
528 | require_once 'lib/pubsubhubbub/subscriber.php'; | |
529 | ||
530 | $callback_url = get_self_url_prefix() . | |
531 | "/public.php?op=pubsub&id=$feed"; | |
532 | ||
533 | $s = new Subscriber($feed_hub_url, $callback_url); | |
534 | ||
535 | $rc = $s->subscribe($feed_self_url); | |
536 | ||
537 | _debug("feed hub url found, subscribe request sent. [rc=$rc]", $debug_enabled); | |
538 | ||
539 | db_query("UPDATE ttrss_feeds SET pubsub_state = 1 | |
540 | WHERE id = '$feed'"); | |
541 | } | |
542 | } | |
543 | ||
544 | _debug("processing articles...", $debug_enabled); | |
545 | ||
546 | foreach ($items as $item) { | |
547 | if ($_REQUEST['xdebug'] == 3) { | |
548 | print_r($item); | |
549 | } | |
550 | ||
551 | $entry_guid = $item->get_id(); | |
552 | if (!$entry_guid) $entry_guid = $item->get_link(); | |
553 | if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title()); | |
554 | ||
555 | _debug("f_guid $entry_guid", $debug_enabled); | |
556 | ||
557 | if (!$entry_guid) continue; | |
558 | ||
559 | $entry_guid = "$owner_uid,$entry_guid"; | |
560 | ||
561 | $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid)); | |
562 | ||
563 | _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled); | |
564 | ||
565 | $entry_timestamp = ""; | |
566 | ||
567 | $entry_timestamp = $item->get_date(); | |
568 | ||
569 | _debug("orig date: " . $item->get_date(), $debug_enabled); | |
570 | ||
571 | if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) { | |
572 | $entry_timestamp = time(); | |
573 | } | |
574 | ||
575 | $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp); | |
576 | ||
577 | _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled); | |
578 | ||
579 | // $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8'); | |
580 | // $entry_title = decode_numeric_entities($entry_title); | |
581 | $entry_title = $item->get_title(); | |
582 | ||
583 | $entry_link = rewrite_relative_url($site_url, $item->get_link()); | |
584 | ||
585 | _debug("title $entry_title", $debug_enabled); | |
586 | _debug("link $entry_link", $debug_enabled); | |
587 | ||
588 | if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);; | |
589 | ||
590 | $entry_content = $item->get_content(); | |
591 | if (!$entry_content) $entry_content = $item->get_description(); | |
592 | ||
593 | if ($_REQUEST["xdebug"] == 2) { | |
594 | print "content: "; | |
595 | print $entry_content; | |
596 | print "\n"; | |
597 | } | |
598 | ||
599 | $entry_language = ""; | |
600 | ||
601 | if (DETECT_ARTICLE_LANGUAGE) { | |
602 | $entry_language = $lang->detect($entry_title . " " . $entry_content, 1); | |
603 | ||
604 | if (count($entry_language) > 0) { | |
605 | $entry_language = array_keys($entry_language); | |
606 | ||
607 | // the fuck? | |
608 | if (is_array($entry_language)) | |
609 | $entry_language = ""; | |
610 | else | |
611 | $entry_language = db_escape_string(substr($entry_language[0], 0, 2)); | |
612 | ||
613 | _debug("detected language: $entry_language", $debug_enabled); | |
614 | } else { | |
615 | $entry_language = ""; | |
616 | } | |
617 | } | |
618 | ||
619 | $entry_comments = $item->get_comments_url(); | |
620 | $entry_author = $item->get_author(); | |
621 | ||
622 | $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245)); | |
623 | ||
624 | $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245)); | |
625 | $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245)); | |
626 | ||
627 | $num_comments = (int) $item->get_comments_count(); | |
628 | ||
629 | _debug("author $entry_author", $debug_enabled); | |
630 | _debug("num_comments: $num_comments", $debug_enabled); | |
631 | _debug("looking for tags...", $debug_enabled); | |
632 | ||
633 | // parse <category> entries into tags | |
634 | ||
635 | $additional_tags = array(); | |
636 | ||
637 | $additional_tags_src = $item->get_categories(); | |
638 | ||
639 | if (is_array($additional_tags_src)) { | |
640 | foreach ($additional_tags_src as $tobj) { | |
641 | array_push($additional_tags, $tobj); | |
642 | } | |
643 | } | |
644 | ||
645 | $entry_tags = array_unique($additional_tags); | |
646 | ||
647 | for ($i = 0; $i < count($entry_tags); $i++) | |
648 | $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8'); | |
649 | ||
650 | _debug("tags found: " . join(",", $entry_tags), $debug_enabled); | |
651 | ||
652 | _debug("done collecting data.", $debug_enabled); | |
653 | ||
654 | // TODO: less memory-hungry implementation | |
655 | ||
656 | _debug("applying plugin filters..", $debug_enabled); | |
657 | ||
658 | // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?) | |
659 | $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries | |
660 | WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid"); | |
661 | ||
662 | if (db_num_rows($result) != 0) { | |
663 | $entry_plugin_data = db_fetch_result($result, 0, "plugin_data"); | |
664 | $stored_article = array("title" => db_fetch_result($result, 0, "title"), | |
665 | "content" => db_fetch_result($result, 0, "content"), | |
666 | "link" => db_fetch_result($result, 0, "link"), | |
667 | "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")), | |
668 | "author" => db_fetch_result($result, 0, "author")); | |
669 | } else { | |
670 | $entry_plugin_data = ""; | |
671 | $stored_article = array(); | |
672 | } | |
673 | ||
674 | $article = array("owner_uid" => $owner_uid, // read only | |
675 | "guid" => $entry_guid, // read only | |
676 | "title" => $entry_title, | |
677 | "content" => $entry_content, | |
678 | "link" => $entry_link, | |
679 | "tags" => $entry_tags, | |
680 | "plugin_data" => $entry_plugin_data, | |
681 | "author" => $entry_author, | |
682 | "stored" => $stored_article, | |
683 | "feed" => array("id" => $feed, | |
684 | "fetch_url" => $fetch_url, | |
685 | "site_url" => $site_url) | |
686 | ); | |
687 | ||
688 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) { | |
689 | $article = $plugin->hook_article_filter($article); | |
690 | } | |
691 | ||
692 | $entry_tags = $article["tags"]; | |
693 | $entry_guid = db_escape_string($entry_guid); | |
694 | $entry_title = db_escape_string($article["title"]); | |
695 | $entry_author = db_escape_string($article["author"]); | |
696 | $entry_link = db_escape_string($article["link"]); | |
697 | $entry_plugin_data = db_escape_string($article["plugin_data"]); | |
698 | $entry_content = $article["content"]; // escaped below | |
699 | ||
700 | ||
701 | _debug("plugin data: $entry_plugin_data", $debug_enabled); | |
702 | ||
703 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
704 | cache_images($entry_content, $site_url, $debug_enabled); | |
705 | ||
706 | $entry_content = db_escape_string($entry_content, false); | |
707 | ||
708 | $content_hash = "SHA1:" . sha1($entry_content); | |
709 | ||
710 | db_query("BEGIN"); | |
711 | ||
712 | $result = db_query("SELECT id FROM ttrss_entries | |
713 | WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')"); | |
714 | ||
715 | if (db_num_rows($result) == 0) { | |
716 | ||
717 | _debug("base guid [$entry_guid] not found", $debug_enabled); | |
718 | ||
719 | // base post entry does not exist, create it | |
720 | ||
721 | $result = db_query( | |
722 | "INSERT INTO ttrss_entries | |
723 | (title, | |
724 | guid, | |
725 | link, | |
726 | updated, | |
727 | content, | |
728 | content_hash, | |
729 | no_orig_date, | |
730 | date_updated, | |
731 | date_entered, | |
732 | comments, | |
733 | num_comments, | |
734 | plugin_data, | |
735 | lang, | |
736 | author) | |
737 | VALUES | |
738 | ('$entry_title', | |
739 | '$entry_guid_hashed', | |
740 | '$entry_link', | |
741 | '$entry_timestamp_fmt', | |
742 | '$entry_content', | |
743 | '$content_hash', | |
744 | false, | |
745 | NOW(), | |
746 | '$date_feed_processed', | |
747 | '$entry_comments', | |
748 | '$num_comments', | |
749 | '$entry_plugin_data', | |
750 | '$entry_language', | |
751 | '$entry_author')"); | |
752 | ||
753 | $article_labels = array(); | |
754 | ||
755 | } else { | |
756 | // we keep encountering the entry in feeds, so we need to | |
757 | // update date_updated column so that we don't get horrible | |
758 | // dupes when the entry gets purged and reinserted again e.g. | |
759 | // in the case of SLOW SLOW OMG SLOW updating feeds | |
760 | ||
761 | $base_entry_id = db_fetch_result($result, 0, "id"); | |
762 | ||
763 | db_query("UPDATE ttrss_entries SET date_updated = NOW() | |
764 | WHERE id = '$base_entry_id'"); | |
765 | ||
766 | $article_labels = get_article_labels($base_entry_id, $owner_uid); | |
767 | } | |
768 | ||
769 | // now it should exist, if not - bad luck then | |
770 | ||
771 | $result = db_query("SELECT | |
772 | id,content_hash,title,plugin_data,guid, | |
773 | num_comments | |
774 | FROM | |
775 | ttrss_entries | |
776 | WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'"); | |
777 | ||
778 | $entry_ref_id = 0; | |
779 | $entry_int_id = 0; | |
780 | ||
781 | if (db_num_rows($result) == 1) { | |
782 | ||
783 | _debug("base guid found, checking for user record", $debug_enabled); | |
784 | ||
785 | // this will be used below in update handler | |
786 | $orig_content_hash = db_fetch_result($result, 0, "content_hash"); | |
787 | $orig_title = db_fetch_result($result, 0, "title"); | |
788 | $orig_num_comments = db_fetch_result($result, 0, "num_comments"); | |
789 | $orig_plugin_data = db_fetch_result($result, 0, "plugin_data"); | |
790 | ||
791 | $ref_id = db_fetch_result($result, 0, "id"); | |
792 | $entry_ref_id = $ref_id; | |
793 | ||
794 | /* $stored_guid = db_fetch_result($result, 0, "guid"); | |
795 | if ($stored_guid != $entry_guid_hashed) { | |
796 | if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled); | |
797 | ||
798 | db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE | |
799 | id = '$ref_id'"); | |
800 | } */ | |
801 | ||
802 | // check for user post link to main table | |
803 | ||
804 | // do we allow duplicate posts with same GUID in different feeds? | |
805 | if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) { | |
806 | $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)"; | |
807 | } else { | |
808 | $dupcheck_qpart = ""; | |
809 | } | |
810 | ||
811 | /* Collect article tags here so we could filter by them: */ | |
812 | ||
813 | $article_filters = get_article_filters($filters, $entry_title, | |
814 | $entry_content, $entry_link, $entry_timestamp, $entry_author, | |
815 | $entry_tags); | |
816 | ||
817 | if ($debug_enabled) { | |
818 | _debug("article filters: ", $debug_enabled); | |
819 | if (count($article_filters) != 0) { | |
820 | print_r($article_filters); | |
821 | } | |
822 | } | |
823 | ||
824 | if (find_article_filter($article_filters, "filter")) { | |
825 | db_query("COMMIT"); // close transaction in progress | |
826 | continue; | |
827 | } | |
828 | ||
829 | $score = calculate_article_score($article_filters); | |
830 | ||
831 | _debug("initial score: $score", $debug_enabled); | |
832 | ||
833 | $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE | |
834 | ref_id = '$ref_id' AND owner_uid = '$owner_uid' | |
835 | $dupcheck_qpart"; | |
836 | ||
837 | // if ($_REQUEST["xdebug"]) print "$query\n"; | |
838 | ||
839 | $result = db_query($query); | |
840 | ||
841 | // okay it doesn't exist - create user entry | |
842 | if (db_num_rows($result) == 0) { | |
843 | ||
844 | _debug("user record not found, creating...", $debug_enabled); | |
845 | ||
846 | if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) { | |
847 | $unread = 'true'; | |
848 | $last_read_qpart = 'NULL'; | |
849 | } else { | |
850 | $unread = 'false'; | |
851 | $last_read_qpart = 'NOW()'; | |
852 | } | |
853 | ||
854 | if (find_article_filter($article_filters, 'mark') || $score > 1000) { | |
855 | $marked = 'true'; | |
856 | } else { | |
857 | $marked = 'false'; | |
858 | } | |
859 | ||
860 | if (find_article_filter($article_filters, 'publish')) { | |
861 | $published = 'true'; | |
862 | } else { | |
863 | $published = 'false'; | |
864 | } | |
865 | ||
866 | // N-grams | |
867 | ||
868 | if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) { | |
869 | ||
870 | $result = db_query("SELECT COUNT(*) AS similar FROM | |
871 | ttrss_entries,ttrss_user_entries | |
872 | WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day' | |
873 | AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD." | |
874 | AND owner_uid = $owner_uid"); | |
875 | ||
876 | $ngram_similar = db_fetch_result($result, 0, "similar"); | |
877 | ||
878 | _debug("N-gram similar results: $ngram_similar", $debug_enabled); | |
879 | ||
880 | if ($ngram_similar > 0) { | |
881 | $unread = 'false'; | |
882 | } | |
883 | } | |
884 | ||
885 | $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL'; | |
886 | $last_published = ($published == 'true') ? 'NOW()' : 'NULL'; | |
887 | ||
888 | $result = db_query( | |
889 | "INSERT INTO ttrss_user_entries | |
890 | (ref_id, owner_uid, feed_id, unread, last_read, marked, | |
891 | published, score, tag_cache, label_cache, uuid, | |
892 | last_marked, last_published) | |
893 | VALUES ('$ref_id', '$owner_uid', '$feed', $unread, | |
894 | $last_read_qpart, $marked, $published, '$score', '', '', | |
895 | '', $last_marked, $last_published)"); | |
896 | ||
897 | if (PUBSUBHUBBUB_HUB && $published == 'true') { | |
898 | $rss_link = get_self_url_prefix() . | |
899 | "/public.php?op=rss&id=-2&key=" . | |
900 | get_feed_access_key(-2, false, $owner_uid); | |
901 | ||
902 | $p = new Publisher(PUBSUBHUBBUB_HUB); | |
903 | ||
904 | /* $pubsub_result = */ $p->publish_update($rss_link); | |
905 | } | |
906 | ||
907 | $result = db_query( | |
908 | "SELECT int_id FROM ttrss_user_entries WHERE | |
909 | ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND | |
910 | feed_id = '$feed' LIMIT 1"); | |
911 | ||
912 | if (db_num_rows($result) == 1) { | |
913 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
914 | } | |
915 | } else { | |
916 | _debug("user record FOUND", $debug_enabled); | |
917 | ||
918 | $entry_ref_id = db_fetch_result($result, 0, "ref_id"); | |
919 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
920 | } | |
921 | ||
922 | _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled); | |
923 | ||
924 | $post_needs_update = false; | |
925 | $update_insignificant = false; | |
926 | ||
927 | if ($orig_num_comments != $num_comments) { | |
928 | $post_needs_update = true; | |
929 | $update_insignificant = true; | |
930 | } | |
931 | ||
932 | if ($entry_plugin_data != $orig_plugin_data) { | |
933 | $post_needs_update = true; | |
934 | $update_insignificant = true; | |
935 | } | |
936 | ||
937 | if ($content_hash != $orig_content_hash) { | |
938 | $post_needs_update = true; | |
939 | $update_insignificant = false; | |
940 | } | |
941 | ||
942 | if (db_escape_string($orig_title) != $entry_title) { | |
943 | $post_needs_update = true; | |
944 | $update_insignificant = false; | |
945 | } | |
946 | ||
947 | // if post needs update, update it and mark all user entries | |
948 | // linking to this post as updated | |
949 | if ($post_needs_update) { | |
950 | ||
951 | if (defined('DAEMON_EXTENDED_DEBUG')) { | |
952 | _debug("post $entry_guid_hashed needs update...", $debug_enabled); | |
953 | } | |
954 | ||
955 | // print "<!-- post $orig_title needs update : $post_needs_update -->"; | |
956 | ||
957 | db_query("UPDATE ttrss_entries | |
958 | SET title = '$entry_title', content = '$entry_content', | |
959 | content_hash = '$content_hash', | |
960 | updated = '$entry_timestamp_fmt', | |
961 | num_comments = '$num_comments', | |
962 | plugin_data = '$entry_plugin_data' | |
963 | WHERE id = '$ref_id'"); | |
964 | ||
965 | if (!$update_insignificant) { | |
966 | if ($mark_unread_on_update) { | |
967 | db_query("UPDATE ttrss_user_entries | |
968 | SET last_read = null, unread = true WHERE ref_id = '$ref_id'"); | |
969 | } | |
970 | } | |
971 | } | |
972 | } | |
973 | ||
974 | db_query("COMMIT"); | |
975 | ||
976 | _debug("assigning labels...", $debug_enabled); | |
977 | ||
978 | assign_article_to_label_filters($entry_ref_id, $article_filters, | |
979 | $owner_uid, $article_labels); | |
980 | ||
981 | _debug("looking for enclosures...", $debug_enabled); | |
982 | ||
983 | // enclosures | |
984 | ||
985 | $enclosures = array(); | |
986 | ||
987 | $encs = $item->get_enclosures(); | |
988 | ||
989 | if (is_array($encs)) { | |
990 | foreach ($encs as $e) { | |
991 | $e_item = array( | |
992 | $e->link, $e->type, $e->length, $e->title, $e->width, $e->height); | |
993 | array_push($enclosures, $e_item); | |
994 | } | |
995 | } | |
996 | ||
997 | if ($debug_enabled) { | |
998 | _debug("article enclosures:", $debug_enabled); | |
999 | print_r($enclosures); | |
1000 | } | |
1001 | ||
1002 | db_query("BEGIN"); | |
1003 | ||
1004 | // debugging | |
1005 | // db_query("DELETE FROM ttrss_enclosures WHERE post_id = '$entry_ref_id'"); | |
1006 | ||
1007 | foreach ($enclosures as $enc) { | |
1008 | $enc_url = db_escape_string($enc[0]); | |
1009 | $enc_type = db_escape_string($enc[1]); | |
1010 | $enc_dur = db_escape_string($enc[2]); | |
1011 | $enc_title = db_escape_string($enc[3]); | |
1012 | $enc_width = intval($enc[4]); | |
1013 | $enc_height = intval($enc[5]); | |
1014 | ||
1015 | $result = db_query("SELECT id FROM ttrss_enclosures | |
1016 | WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'"); | |
1017 | ||
1018 | if (db_num_rows($result) == 0) { | |
1019 | db_query("INSERT INTO ttrss_enclosures | |
1020 | (content_url, content_type, title, duration, post_id, width, height) VALUES | |
1021 | ('$enc_url', '$enc_type', '$enc_title', '$enc_dur', '$entry_ref_id', $enc_width, $enc_height)"); | |
1022 | } | |
1023 | } | |
1024 | ||
1025 | db_query("COMMIT"); | |
1026 | ||
1027 | // check for manual tags (we have to do it here since they're loaded from filters) | |
1028 | ||
1029 | foreach ($article_filters as $f) { | |
1030 | if ($f["type"] == "tag") { | |
1031 | ||
1032 | $manual_tags = trim_array(explode(",", $f["param"])); | |
1033 | ||
1034 | foreach ($manual_tags as $tag) { | |
1035 | if (tag_is_valid($tag)) { | |
1036 | array_push($entry_tags, $tag); | |
1037 | } | |
1038 | } | |
1039 | } | |
1040 | } | |
1041 | ||
1042 | // Skip boring tags | |
1043 | ||
1044 | $boring_tags = trim_array(explode(",", mb_strtolower(get_pref( | |
1045 | 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8'))); | |
1046 | ||
1047 | $filtered_tags = array(); | |
1048 | $tags_to_cache = array(); | |
1049 | ||
1050 | if ($entry_tags && is_array($entry_tags)) { | |
1051 | foreach ($entry_tags as $tag) { | |
1052 | if (array_search($tag, $boring_tags) === false) { | |
1053 | array_push($filtered_tags, $tag); | |
1054 | } | |
1055 | } | |
1056 | } | |
1057 | ||
1058 | $filtered_tags = array_unique($filtered_tags); | |
1059 | ||
1060 | if ($debug_enabled) { | |
1061 | _debug("filtered article tags:", $debug_enabled); | |
1062 | print_r($filtered_tags); | |
1063 | } | |
1064 | ||
1065 | // Save article tags in the database | |
1066 | ||
1067 | if (count($filtered_tags) > 0) { | |
1068 | ||
1069 | db_query("BEGIN"); | |
1070 | ||
1071 | foreach ($filtered_tags as $tag) { | |
1072 | ||
1073 | $tag = sanitize_tag($tag); | |
1074 | $tag = db_escape_string($tag); | |
1075 | ||
1076 | if (!tag_is_valid($tag)) continue; | |
1077 | ||
1078 | $result = db_query("SELECT id FROM ttrss_tags | |
1079 | WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND | |
1080 | owner_uid = '$owner_uid' LIMIT 1"); | |
1081 | ||
1082 | if ($result && db_num_rows($result) == 0) { | |
1083 | ||
1084 | db_query("INSERT INTO ttrss_tags | |
1085 | (owner_uid,tag_name,post_int_id) | |
1086 | VALUES ('$owner_uid','$tag', '$entry_int_id')"); | |
1087 | } | |
1088 | ||
1089 | array_push($tags_to_cache, $tag); | |
1090 | } | |
1091 | ||
1092 | /* update the cache */ | |
1093 | ||
1094 | $tags_to_cache = array_unique($tags_to_cache); | |
1095 | ||
1096 | $tags_str = db_escape_string(join(",", $tags_to_cache)); | |
1097 | ||
1098 | db_query("UPDATE ttrss_user_entries | |
1099 | SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id' | |
1100 | AND owner_uid = $owner_uid"); | |
1101 | ||
1102 | db_query("COMMIT"); | |
1103 | } | |
1104 | ||
1105 | if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) { | |
1106 | _debug("auto-assigning labels...", $debug_enabled); | |
1107 | ||
1108 | foreach ($labels as $label) { | |
1109 | $caption = preg_quote($label["caption"]); | |
1110 | ||
1111 | if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) { | |
1112 | if (!labels_contains_caption($article_labels, $caption)) { | |
1113 | label_add_article($entry_ref_id, $caption, $owner_uid); | |
1114 | } | |
1115 | } | |
1116 | } | |
1117 | } | |
1118 | ||
1119 | _debug("article processed", $debug_enabled); | |
1120 | } | |
1121 | ||
1122 | _debug("purging feed...", $debug_enabled); | |
1123 | ||
1124 | purge_feed($feed, 0, $debug_enabled); | |
1125 | ||
1126 | db_query("UPDATE ttrss_feeds | |
1127 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
1128 | ||
1129 | // db_query("COMMIT"); | |
1130 | ||
1131 | } else { | |
1132 | ||
1133 | $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245)); | |
1134 | ||
1135 | _debug("fetch error: $error_msg", $debug_enabled); | |
1136 | ||
1137 | if (count($rss->errors()) > 1) { | |
1138 | foreach ($rss->errors() as $error) { | |
1139 | _debug("+ $error"); | |
1140 | } | |
1141 | } | |
1142 | ||
1143 | db_query( | |
1144 | "UPDATE ttrss_feeds SET last_error = '$error_msg', | |
1145 | last_updated = NOW() WHERE id = '$feed'"); | |
1146 | ||
1147 | unset($rss); | |
1148 | } | |
1149 | ||
1150 | _debug("done", $debug_enabled); | |
1151 | ||
1152 | return $rss; | |
1153 | } | |
1154 | ||
1155 | function cache_images($html, $site_url, $debug) { | |
1156 | libxml_use_internal_errors(true); | |
1157 | ||
1158 | $charset_hack = '<head> | |
1159 | <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/> | |
1160 | </head>'; | |
1161 | ||
1162 | $doc = new DOMDocument(); | |
1163 | $doc->loadHTML($charset_hack . $html); | |
1164 | $xpath = new DOMXPath($doc); | |
1165 | ||
1166 | $entries = $xpath->query('(//img[@src])'); | |
1167 | ||
1168 | foreach ($entries as $entry) { | |
1169 | if ($entry->hasAttribute('src')) { | |
1170 | $src = rewrite_relative_url($site_url, $entry->getAttribute('src')); | |
1171 | ||
1172 | $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png"; | |
1173 | ||
1174 | if ($debug) _debug("cache_images: downloading: $src to $local_filename"); | |
1175 | ||
1176 | if (!file_exists($local_filename)) { | |
1177 | $file_content = fetch_file_contents($src); | |
1178 | ||
1179 | if ($file_content && strlen($file_content) > _MIN_CACHE_IMAGE_SIZE) { | |
1180 | file_put_contents($local_filename, $file_content); | |
1181 | } | |
1182 | } | |
1183 | ||
1184 | /* if (file_exists($local_filename)) { | |
1185 | $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' . | |
1186 | base64_encode($src)); | |
1187 | } */ | |
1188 | } | |
1189 | } | |
1190 | ||
1191 | //$node = $doc->getElementsByTagName('body')->item(0); | |
1192 | //return $doc->saveXML($node); | |
1193 | } | |
1194 | ||
1195 | function expire_error_log($debug) { | |
1196 | if ($debug) _debug("Removing old error log entries..."); | |
1197 | ||
1198 | if (DB_TYPE == "pgsql") { | |
1199 | db_query("DELETE FROM ttrss_error_log | |
1200 | WHERE created_at < NOW() - INTERVAL '7 days'"); | |
1201 | } else { | |
1202 | db_query("DELETE FROM ttrss_error_log | |
1203 | WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)"); | |
1204 | } | |
1205 | ||
1206 | } | |
1207 | ||
1208 | function expire_lock_files($debug) { | |
1209 | //if ($debug) _debug("Removing old lock files..."); | |
1210 | ||
1211 | $num_deleted = 0; | |
1212 | ||
1213 | if (is_writable(LOCK_DIRECTORY)) { | |
1214 | $files = glob(LOCK_DIRECTORY . "/*.lock"); | |
1215 | ||
1216 | if ($files) { | |
1217 | foreach ($files as $file) { | |
1218 | if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) { | |
1219 | unlink($file); | |
1220 | ++$num_deleted; | |
1221 | } | |
1222 | } | |
1223 | } | |
1224 | } | |
1225 | ||
1226 | if ($debug) _debug("Removed $num_deleted old lock files."); | |
1227 | } | |
1228 | ||
1229 | function expire_cached_files($debug) { | |
1230 | foreach (array("simplepie", "images", "export", "upload") as $dir) { | |
1231 | $cache_dir = CACHE_DIR . "/$dir"; | |
1232 | ||
1233 | // if ($debug) _debug("Expiring $cache_dir"); | |
1234 | ||
1235 | $num_deleted = 0; | |
1236 | ||
1237 | if (is_writable($cache_dir)) { | |
1238 | $files = glob("$cache_dir/*"); | |
1239 | ||
1240 | if ($files) { | |
1241 | foreach ($files as $file) { | |
1242 | if (time() - filemtime($file) > 86400*7) { | |
1243 | unlink($file); | |
1244 | ||
1245 | ++$num_deleted; | |
1246 | } | |
1247 | } | |
1248 | } | |
1249 | } | |
1250 | ||
1251 | if ($debug) _debug("$cache_dir: removed $num_deleted files."); | |
1252 | } | |
1253 | } | |
1254 | ||
1255 | /** | |
1256 | * Source: http://www.php.net/manual/en/function.parse-url.php#104527 | |
1257 | * Returns the url query as associative array | |
1258 | * | |
1259 | * @param string query | |
1260 | * @return array params | |
1261 | */ | |
1262 | function convertUrlQuery($query) { | |
1263 | $queryParts = explode('&', $query); | |
1264 | ||
1265 | $params = array(); | |
1266 | ||
1267 | foreach ($queryParts as $param) { | |
1268 | $item = explode('=', $param); | |
1269 | $params[$item[0]] = $item[1]; | |
1270 | } | |
1271 | ||
1272 | return $params; | |
1273 | } | |
1274 | ||
1275 | function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) { | |
1276 | $matches = array(); | |
1277 | ||
1278 | foreach ($filters as $filter) { | |
1279 | $match_any_rule = $filter["match_any_rule"]; | |
1280 | $inverse = $filter["inverse"]; | |
1281 | $filter_match = false; | |
1282 | ||
1283 | foreach ($filter["rules"] as $rule) { | |
1284 | $match = false; | |
1285 | $reg_exp = str_replace('/', '\/', $rule["reg_exp"]); | |
1286 | $rule_inverse = $rule["inverse"]; | |
1287 | ||
1288 | if (!$reg_exp) | |
1289 | continue; | |
1290 | ||
1291 | switch ($rule["type"]) { | |
1292 | case "title": | |
1293 | $match = @preg_match("/$reg_exp/i", $title); | |
1294 | break; | |
1295 | case "content": | |
1296 | // we don't need to deal with multiline regexps | |
1297 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1298 | ||
1299 | $match = @preg_match("/$reg_exp/i", $content); | |
1300 | break; | |
1301 | case "both": | |
1302 | // we don't need to deal with multiline regexps | |
1303 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1304 | ||
1305 | $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content)); | |
1306 | break; | |
1307 | case "link": | |
1308 | $match = @preg_match("/$reg_exp/i", $link); | |
1309 | break; | |
1310 | case "author": | |
1311 | $match = @preg_match("/$reg_exp/i", $author); | |
1312 | break; | |
1313 | case "tag": | |
1314 | foreach ($tags as $tag) { | |
1315 | if (@preg_match("/$reg_exp/i", $tag)) { | |
1316 | $match = true; | |
1317 | break; | |
1318 | } | |
1319 | } | |
1320 | break; | |
1321 | } | |
1322 | ||
1323 | if ($rule_inverse) $match = !$match; | |
1324 | ||
1325 | if ($match_any_rule) { | |
1326 | if ($match) { | |
1327 | $filter_match = true; | |
1328 | break; | |
1329 | } | |
1330 | } else { | |
1331 | $filter_match = $match; | |
1332 | if (!$match) { | |
1333 | break; | |
1334 | } | |
1335 | } | |
1336 | } | |
1337 | ||
1338 | if ($inverse) $filter_match = !$filter_match; | |
1339 | ||
1340 | if ($filter_match) { | |
1341 | foreach ($filter["actions"] AS $action) { | |
1342 | array_push($matches, $action); | |
1343 | ||
1344 | // if Stop action encountered, perform no further processing | |
1345 | if ($action["type"] == "stop") return $matches; | |
1346 | } | |
1347 | } | |
1348 | } | |
1349 | ||
1350 | return $matches; | |
1351 | } | |
1352 | ||
1353 | function find_article_filter($filters, $filter_name) { | |
1354 | foreach ($filters as $f) { | |
1355 | if ($f["type"] == $filter_name) { | |
1356 | return $f; | |
1357 | }; | |
1358 | } | |
1359 | return false; | |
1360 | } | |
1361 | ||
1362 | function find_article_filters($filters, $filter_name) { | |
1363 | $results = array(); | |
1364 | ||
1365 | foreach ($filters as $f) { | |
1366 | if ($f["type"] == $filter_name) { | |
1367 | array_push($results, $f); | |
1368 | }; | |
1369 | } | |
1370 | return $results; | |
1371 | } | |
1372 | ||
1373 | function calculate_article_score($filters) { | |
1374 | $score = 0; | |
1375 | ||
1376 | foreach ($filters as $f) { | |
1377 | if ($f["type"] == "score") { | |
1378 | $score += $f["param"]; | |
1379 | }; | |
1380 | } | |
1381 | return $score; | |
1382 | } | |
1383 | ||
1384 | function labels_contains_caption($labels, $caption) { | |
1385 | foreach ($labels as $label) { | |
1386 | if ($label[1] == $caption) { | |
1387 | return true; | |
1388 | } | |
1389 | } | |
1390 | ||
1391 | return false; | |
1392 | } | |
1393 | ||
1394 | function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) { | |
1395 | foreach ($filters as $f) { | |
1396 | if ($f["type"] == "label") { | |
1397 | if (!labels_contains_caption($article_labels, $f["param"])) { | |
1398 | label_add_article($id, $f["param"], $owner_uid); | |
1399 | } | |
1400 | } | |
1401 | } | |
1402 | } | |
1403 | ||
1404 | function make_guid_from_title($title) { | |
1405 | return preg_replace("/[ \"\',.:;]/", "-", | |
1406 | mb_strtolower(strip_tags($title), 'utf-8')); | |
1407 | } | |
1408 | ||
1409 | /* function verify_feed_xml($feed_data) { | |
1410 | libxml_use_internal_errors(true); | |
1411 | $doc = new DOMDocument(); | |
1412 | $doc->loadXML($feed_data); | |
1413 | $error = libxml_get_last_error(); | |
1414 | libxml_clear_errors(); | |
1415 | return $error; | |
1416 | } */ | |
1417 | ||
1418 | function housekeeping_common($debug) { | |
1419 | expire_cached_files($debug); | |
1420 | expire_lock_files($debug); | |
1421 | expire_error_log($debug); | |
1422 | ||
1423 | $count = update_feedbrowser_cache(); | |
1424 | _debug("Feedbrowser updated, $count feeds processed."); | |
1425 | ||
1426 | purge_orphans( true); | |
1427 | $rc = cleanup_tags( 14, 50000); | |
1428 | ||
1429 | _debug("Cleaned $rc cached tags."); | |
1430 | ||
1431 | PluginHost::getInstance()->run_hooks(PluginHost::HOOK_HOUSE_KEEPING, "hook_house_keeping", ""); | |
1432 | ||
1433 | } | |
1434 | ?> |