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