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