]>
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 | ||
6 | function update_feedbrowser_cache() { | |
7 | ||
8 | $result = db_query("SELECT feed_url, site_url, title, COUNT(id) AS subscribers | |
9 | FROM ttrss_feeds WHERE (SELECT COUNT(id) = 0 FROM ttrss_feeds AS tf | |
10 | WHERE tf.feed_url = ttrss_feeds.feed_url | |
11 | AND (private IS true OR auth_login != '' OR auth_pass != '' OR feed_url LIKE '%:%@%/%')) | |
12 | GROUP BY feed_url, site_url, title ORDER BY subscribers DESC LIMIT 1000"); | |
13 | ||
14 | db_query("BEGIN"); | |
15 | ||
16 | db_query("DELETE FROM ttrss_feedbrowser_cache"); | |
17 | ||
18 | $count = 0; | |
19 | ||
20 | while ($line = db_fetch_assoc($result)) { | |
21 | $subscribers = db_escape_string($line["subscribers"]); | |
22 | $feed_url = db_escape_string($line["feed_url"]); | |
23 | $title = db_escape_string($line["title"]); | |
24 | $site_url = db_escape_string($line["site_url"]); | |
25 | ||
26 | $tmp_result = db_query("SELECT subscribers FROM | |
27 | ttrss_feedbrowser_cache WHERE feed_url = '$feed_url'"); | |
28 | ||
29 | if (db_num_rows($tmp_result) == 0) { | |
30 | ||
31 | db_query("INSERT INTO ttrss_feedbrowser_cache | |
32 | (feed_url, site_url, title, subscribers) VALUES ('$feed_url', | |
33 | '$site_url', '$title', '$subscribers')"); | |
34 | ||
35 | ++$count; | |
36 | ||
37 | } | |
38 | ||
39 | } | |
40 | ||
41 | db_query("COMMIT"); | |
42 | ||
43 | return $count; | |
44 | ||
45 | } | |
46 | ||
47 | ||
48 | /** | |
49 | * Update a feed batch. | |
50 | * Used by daemons to update n feeds by run. | |
51 | * Only update feed needing a update, and not being processed | |
52 | * by another process. | |
53 | * | |
54 | * @param mixed $link Database link | |
55 | * @param integer $limit Maximum number of feeds in update batch. Default to DAEMON_FEED_LIMIT. | |
56 | * @param boolean $from_http Set to true if you call this function from http to disable cli specific code. | |
57 | * @param boolean $debug Set to false to disable debug output. Default to true. | |
58 | * @return void | |
59 | */ | |
60 | function update_daemon_common($limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) { | |
61 | // Process all other feeds using last_updated and interval parameters | |
62 | ||
63 | $schema_version = get_schema_version(); | |
64 | ||
65 | if ($schema_version != SCHEMA_VERSION) { | |
66 | die("Schema version is wrong, please upgrade the database.\n"); | |
67 | } | |
68 | ||
69 | define('PREFS_NO_CACHE', true); | |
70 | ||
71 | // Test if the user has loggued in recently. If not, it does not update its feeds. | |
72 | if (!SINGLE_USER_MODE && DAEMON_UPDATE_LOGIN_LIMIT > 0) { | |
73 | if (DB_TYPE == "pgsql") { | |
74 | $login_thresh_qpart = "AND ttrss_users.last_login >= NOW() - INTERVAL '".DAEMON_UPDATE_LOGIN_LIMIT." days'"; | |
75 | } else { | |
76 | $login_thresh_qpart = "AND ttrss_users.last_login >= DATE_SUB(NOW(), INTERVAL ".DAEMON_UPDATE_LOGIN_LIMIT." DAY)"; | |
77 | } | |
78 | } else { | |
79 | $login_thresh_qpart = ""; | |
80 | } | |
81 | ||
82 | // Test if the feed need a update (update interval exceded). | |
83 | if (DB_TYPE == "pgsql") { | |
84 | $update_limit_qpart = "AND (( | |
85 | ttrss_feeds.update_interval = 0 | |
86 | AND ttrss_user_prefs.value != '-1' | |
87 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL) | |
88 | ) OR ( | |
89 | ttrss_feeds.update_interval > 0 | |
90 | AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_feeds.update_interval || ' minutes') AS INTERVAL) | |
91 | ) OR ttrss_feeds.last_updated IS NULL | |
92 | OR last_updated = '1970-01-01 00:00:00')"; | |
93 | } else { | |
94 | $update_limit_qpart = "AND (( | |
95 | ttrss_feeds.update_interval = 0 | |
96 | AND ttrss_user_prefs.value != '-1' | |
97 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE) | |
98 | ) OR ( | |
99 | ttrss_feeds.update_interval > 0 | |
100 | AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL ttrss_feeds.update_interval MINUTE) | |
101 | ) OR ttrss_feeds.last_updated IS NULL | |
102 | OR last_updated = '1970-01-01 00:00:00')"; | |
103 | } | |
104 | ||
105 | // Test if feed is currently being updated by another process. | |
106 | if (DB_TYPE == "pgsql") { | |
107 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < NOW() - INTERVAL '10 minutes')"; | |
108 | } else { | |
109 | $updstart_thresh_qpart = "AND (ttrss_feeds.last_update_started IS NULL OR ttrss_feeds.last_update_started < DATE_SUB(NOW(), INTERVAL 10 MINUTE))"; | |
110 | } | |
111 | ||
112 | // Test if there is a limit to number of updated feeds | |
113 | $query_limit = ""; | |
114 | if($limit) $query_limit = sprintf("LIMIT %d", $limit); | |
115 | ||
116 | $query = "SELECT DISTINCT ttrss_feeds.feed_url, ttrss_feeds.last_updated | |
117 | FROM | |
118 | ttrss_feeds, ttrss_users, ttrss_user_prefs | |
119 | WHERE | |
120 | ttrss_feeds.owner_uid = ttrss_users.id | |
121 | AND ttrss_users.id = ttrss_user_prefs.owner_uid | |
122 | AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' | |
123 | $login_thresh_qpart $update_limit_qpart | |
124 | $updstart_thresh_qpart | |
125 | ORDER BY last_updated $query_limit"; | |
126 | ||
127 | // We search for feed needing update. | |
128 | $result = db_query($query); | |
129 | ||
130 | if($debug) _debug(sprintf("Scheduled %d feeds to update...", db_num_rows($result))); | |
131 | ||
132 | // Here is a little cache magic in order to minimize risk of double feed updates. | |
133 | $feeds_to_update = array(); | |
134 | while ($line = db_fetch_assoc($result)) { | |
135 | array_push($feeds_to_update, db_escape_string($line['feed_url'])); | |
136 | } | |
137 | ||
138 | // We update the feed last update started date before anything else. | |
139 | // There is no lag due to feed contents downloads | |
140 | // It prevent an other process to update the same feed. | |
141 | ||
142 | if(count($feeds_to_update) > 0) { | |
143 | $feeds_quoted = array(); | |
144 | ||
145 | foreach ($feeds_to_update as $feed) { | |
146 | array_push($feeds_quoted, "'" . db_escape_string($feed) . "'"); | |
147 | } | |
148 | ||
149 | db_query(sprintf("UPDATE ttrss_feeds SET last_update_started = NOW() | |
150 | WHERE feed_url IN (%s)", implode(',', $feeds_quoted))); | |
151 | } | |
152 | ||
153 | $nf = 0; | |
154 | ||
155 | // For each feed, we call the feed update function. | |
156 | foreach ($feeds_to_update as $feed) { | |
157 | if($debug) _debug("Base feed: $feed"); | |
158 | ||
159 | //update_rss_feed($line["id"], true); | |
160 | ||
161 | // since we have the data cached, we can deal with other feeds with the same url | |
162 | ||
163 | $tmp_result = db_query("SELECT DISTINCT ttrss_feeds.id,last_updated | |
164 | FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE | |
165 | ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND | |
166 | ttrss_users.id = ttrss_user_prefs.owner_uid AND | |
167 | ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND | |
168 | feed_url = '".db_escape_string($feed)."' AND | |
169 | (ttrss_feeds.update_interval > 0 OR | |
170 | ttrss_user_prefs.value != '-1') | |
171 | $login_thresh_qpart | |
172 | ORDER BY ttrss_feeds.id $query_limit"); | |
173 | ||
174 | if (db_num_rows($tmp_result) > 0) { | |
175 | while ($tline = db_fetch_assoc($tmp_result)) { | |
176 | if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]); | |
177 | update_rss_feed($tline["id"], true); | |
178 | ++$nf; | |
179 | } | |
180 | } | |
181 | } | |
182 | ||
183 | require_once "digest.php"; | |
184 | ||
185 | // Send feed digests by email if needed. | |
186 | send_headlines_digests($debug); | |
187 | ||
188 | return $nf; | |
189 | ||
190 | } // function update_daemon_common | |
191 | ||
192 | // ignore_daemon is not used | |
193 | function update_rss_feed($feed, $ignore_daemon = false, $no_cache = false, | |
194 | $override_url = false) { | |
195 | ||
196 | require_once "lib/simplepie/simplepie.inc"; | |
197 | ||
198 | $debug_enabled = defined('DAEMON_EXTENDED_DEBUG') || $_REQUEST['xdebug']; | |
199 | ||
200 | _debug("start", $debug_enabled); | |
201 | ||
202 | $result = db_query("SELECT id,update_interval,auth_login, | |
203 | feed_url,auth_pass,cache_images,last_updated, | |
204 | mark_unread_on_update, owner_uid, | |
205 | pubsub_state, auth_pass_encrypted, | |
206 | (SELECT max(date_entered) FROM | |
207 | ttrss_entries, ttrss_user_entries where ref_id = id AND feed_id = '$feed') AS last_article_timestamp | |
208 | FROM ttrss_feeds WHERE id = '$feed'"); | |
209 | ||
210 | if (db_num_rows($result) == 0) { | |
211 | _debug("feed $feed NOT FOUND/SKIPPED", $debug_enabled); | |
212 | return false; | |
213 | } | |
214 | ||
215 | // set last update to now so if anything *simplepie* crashes later we won't be | |
216 | // continuously failing on the same feed | |
217 | db_query("UPDATE ttrss_feeds SET last_updated = NOW() WHERE id = '$feed'"); | |
218 | ||
219 | $last_updated = db_fetch_result($result, 0, "last_updated"); | |
220 | $last_article_timestamp = @strtotime(db_fetch_result($result, 0, "last_article_timestamp")); | |
221 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
222 | $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result, | |
223 | 0, "mark_unread_on_update")); | |
224 | $pubsub_state = db_fetch_result($result, 0, "pubsub_state"); | |
225 | $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result, | |
226 | 0, "auth_pass_encrypted")); | |
227 | ||
228 | db_query("UPDATE ttrss_feeds SET last_update_started = NOW() | |
229 | WHERE id = '$feed'"); | |
230 | ||
231 | $auth_login = db_fetch_result($result, 0, "auth_login"); | |
232 | $auth_pass = db_fetch_result($result, 0, "auth_pass"); | |
233 | ||
234 | if ($auth_pass_encrypted) { | |
235 | require_once "crypt.php"; | |
236 | $auth_pass = decrypt_string($auth_pass); | |
237 | } | |
238 | ||
239 | $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images")); | |
240 | $fetch_url = db_fetch_result($result, 0, "feed_url"); | |
241 | ||
242 | $feed = db_escape_string($feed); | |
243 | ||
244 | if ($override_url) $fetch_url = $override_url; | |
245 | ||
246 | $date_feed_processed = date('Y-m-d H:i'); | |
247 | ||
248 | $cache_filename = CACHE_DIR . "/simplepie/" . sha1($fetch_url) . ".feed"; | |
249 | ||
250 | // Ignore cache if new feed or manual update. | |
251 | $cache_age = ($no_cache || is_null($last_updated) || strpos($last_updated, '1970-01-01') === 0) ? 30 : get_feed_update_interval($feed) * 60; | |
252 | ||
253 | _debug("cache filename: $cache_filename exists: " . file_exists($cache_filename), $debug_enabled); | |
254 | _debug("cache age: $cache_age; no cache: $no_cache", $debug_enabled); | |
255 | ||
256 | $cached_feed_data_hash = false; | |
257 | ||
258 | $rss = false; | |
259 | $rss_hash = false; | |
260 | $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0; | |
261 | ||
262 | $force_refetch = isset($_REQUEST["force_refetch"]); | |
263 | ||
264 | if (file_exists($cache_filename) && | |
265 | is_readable($cache_filename) && | |
266 | !$auth_login && !$auth_pass && | |
267 | filemtime($cache_filename) > time() - $cache_age) { | |
268 | ||
269 | _debug("using local cache.", $debug_enabled); | |
270 | ||
271 | if ($cache_timestamp > $last_article_timestamp) { | |
272 | @$rss_data = file_get_contents($cache_filename); | |
273 | ||
274 | if ($rss_data) { | |
275 | $rss_hash = sha1($rss_data); | |
276 | @$rss = unserialize($rss_data); | |
277 | } | |
278 | } else if (!$force_refetch) { | |
279 | _debug("local cache valid and older than last_updated, nothing to do.", $debug_enabled); | |
280 | return; | |
281 | } | |
282 | } else { | |
283 | _debug("local cache will not be used for this feed", $debug_enabled); | |
284 | } | |
285 | ||
286 | if (!$rss) { | |
287 | ||
288 | if (!$feed_data) { | |
289 | _debug("fetching [$fetch_url]...", $debug_enabled); | |
290 | _debug("If-Modified-Since: ".gmdate('D, d M Y H:i:s \G\M\T', $last_article_timestamp), $debug_enabled); | |
291 | ||
292 | $feed_data = fetch_file_contents($fetch_url, false, | |
293 | $auth_login, $auth_pass, false, | |
294 | $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT, | |
295 | $force_refetch ? 0 : $last_article_timestamp); | |
296 | ||
297 | global $fetch_curl_used; | |
298 | ||
299 | if (!$fetch_curl_used) { | |
300 | $tmp = @gzdecode($feed_data); | |
301 | ||
302 | if ($tmp) $feed_data = $tmp; | |
303 | } | |
304 | ||
305 | $feed_data = trim($feed_data); | |
306 | ||
307 | _debug("fetch done.", $debug_enabled); | |
308 | ||
309 | if ($feed_data) { | |
310 | $error = verify_feed_xml($feed_data); | |
311 | ||
312 | if ($error) { | |
313 | _debug("error verifying XML, code: " . $error->code, $debug_enabled); | |
314 | ||
315 | if ($error->code == 26) { | |
316 | _debug("got error 26, trying to decode entities...", $debug_enabled); | |
317 | ||
318 | $feed_data = html_entity_decode($feed_data, ENT_COMPAT, 'UTF-8'); | |
319 | ||
320 | $error = verify_feed_xml($feed_data); | |
321 | ||
322 | if ($error) $feed_data = ''; | |
323 | } | |
324 | } | |
325 | } | |
326 | } | |
327 | ||
328 | if (!$feed_data) { | |
329 | global $fetch_last_error; | |
330 | global $fetch_last_error_code; | |
331 | ||
332 | _debug("unable to fetch: $fetch_last_error [$fetch_last_error_code]", $debug_enabled); | |
333 | ||
334 | $error_escaped = ''; | |
335 | ||
336 | // If-Modified-Since | |
337 | if ($fetch_last_error_code != 304) { | |
338 | $error_escaped = db_escape_string($fetch_last_error); | |
339 | } else { | |
340 | _debug("source claims data not modified, nothing to do.", $debug_enabled); | |
341 | } | |
342 | ||
343 | db_query( | |
344 | "UPDATE ttrss_feeds SET last_error = '$error_escaped', | |
345 | last_updated = NOW() WHERE id = '$feed'"); | |
346 | ||
347 | return; | |
348 | } | |
349 | } | |
350 | ||
351 | $pluginhost = new PluginHost(); | |
352 | $pluginhost->set_debug($debug_enabled, $debug_enabled); | |
353 | $user_plugins = get_pref("_ENABLED_PLUGINS", $owner_uid); | |
354 | ||
355 | $pluginhost->load(PLUGINS, PluginHost::KIND_ALL); | |
356 | $pluginhost->load($user_plugins, PluginHost::KIND_USER, $owner_uid); | |
357 | $pluginhost->load_data(); | |
358 | ||
359 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_FEED_FETCHED) as $plugin) { | |
360 | $feed_data = $plugin->hook_feed_fetched($feed_data); | |
361 | } | |
362 | ||
363 | if (!$rss) { | |
364 | /* $rss = new SimplePie(); | |
365 | $rss->set_sanitize_class("SanitizeDummy"); | |
366 | // simplepie ignores the above and creates default sanitizer anyway, | |
367 | // so let's override it... | |
368 | $rss->sanitize = new SanitizeDummy(); | |
369 | $rss->set_output_encoding('UTF-8'); | |
370 | $rss->set_raw_data($feed_data); | |
371 | $rss->enable_cache(false); */ | |
372 | ||
373 | $rss = new FeedParser($feed_data); | |
374 | $rss->init(); | |
375 | } | |
376 | ||
377 | // print_r($rss); | |
378 | ||
379 | $feed = db_escape_string($feed); | |
380 | ||
381 | if (!$rss->error()) { | |
382 | ||
383 | // cache data for later | |
384 | if (!$auth_pass && !$auth_login && is_writable(CACHE_DIR . "/simplepie")) { | |
385 | $rss_data = serialize($rss); | |
386 | $new_rss_hash = sha1($rss_data); | |
387 | ||
388 | if ($new_rss_hash != $rss_hash) { | |
389 | _debug("saving $cache_filename", $debug_enabled); | |
390 | //@file_put_contents($cache_filename, serialize($rss)); NOT YET | |
391 | } | |
392 | } | |
393 | ||
394 | // We use local pluginhost here because we need to load different per-user feed plugins | |
395 | $pluginhost->run_hooks(PluginHost::HOOK_FEED_PARSED, "hook_feed_parsed", $rss); | |
396 | ||
397 | _debug("processing feed data...", $debug_enabled); | |
398 | ||
399 | // db_query("BEGIN"); | |
400 | ||
401 | if (DB_TYPE == "pgsql") { | |
402 | $favicon_interval_qpart = "favicon_last_checked < NOW() - INTERVAL '12 hour'"; | |
403 | } else { | |
404 | $favicon_interval_qpart = "favicon_last_checked < DATE_SUB(NOW(), INTERVAL 12 HOUR)"; | |
405 | } | |
406 | ||
407 | $result = db_query("SELECT title,site_url,owner_uid,favicon_avg_color, | |
408 | (favicon_last_checked IS NULL OR $favicon_interval_qpart) AS | |
409 | favicon_needs_check | |
410 | FROM ttrss_feeds WHERE id = '$feed'"); | |
411 | ||
412 | $registered_title = db_fetch_result($result, 0, "title"); | |
413 | $orig_site_url = db_fetch_result($result, 0, "site_url"); | |
414 | $favicon_needs_check = sql_bool_to_bool(db_fetch_result($result, 0, | |
415 | "favicon_needs_check")); | |
416 | $favicon_avg_color = db_fetch_result($result, 0, "favicon_avg_color"); | |
417 | ||
418 | $owner_uid = db_fetch_result($result, 0, "owner_uid"); | |
419 | ||
420 | $site_url = db_escape_string(mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245)); | |
421 | ||
422 | _debug("site_url: $site_url", $debug_enabled); | |
423 | _debug("feed_title: " . $rss->get_title(), $debug_enabled); | |
424 | ||
425 | if ($favicon_needs_check || $force_refetch) { | |
426 | ||
427 | /* terrible hack: if we crash on floicon shit here, we won't check | |
428 | * the icon avgcolor again (unless the icon got updated) */ | |
429 | ||
430 | $favicon_file = ICONS_DIR . "/$feed.ico"; | |
431 | $favicon_modified = @filemtime($favicon_file); | |
432 | ||
433 | _debug("checking favicon...", $debug_enabled); | |
434 | ||
435 | check_feed_favicon($site_url, $feed, $link); | |
436 | $favicon_modified_new = @filemtime($favicon_file); | |
437 | ||
438 | if ($favicon_modified_new > $favicon_modified) | |
439 | $favicon_avg_color = ''; | |
440 | ||
441 | if (file_exists($favicon_file) && function_exists("imagecreatefromstring") && $favicon_avg_color == '') { | |
442 | require_once "colors.php"; | |
443 | ||
444 | db_query("UPDATE ttrss_feeds SET favicon_avg_color = 'fail' WHERE | |
445 | id = '$feed'"); | |
446 | ||
447 | $favicon_color = db_escape_string( | |
448 | calculate_avg_color($favicon_file)); | |
449 | ||
450 | $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'"; | |
451 | } else if ($favicon_avg_color == 'fail') { | |
452 | _debug("floicon failed on this file, not trying to recalculate avg color", $debug_enabled); | |
453 | } | |
454 | ||
455 | db_query("UPDATE ttrss_feeds SET favicon_last_checked = NOW() | |
456 | $favicon_colorstring | |
457 | WHERE id = '$feed'"); | |
458 | } | |
459 | ||
460 | if (!$registered_title || $registered_title == "[Unknown]") { | |
461 | ||
462 | $feed_title = db_escape_string($rss->get_title()); | |
463 | ||
464 | _debug("registering title: $feed_title", $debug_enabled); | |
465 | ||
466 | db_query("UPDATE ttrss_feeds SET | |
467 | title = '$feed_title' WHERE id = '$feed'"); | |
468 | } | |
469 | ||
470 | if ($site_url && $orig_site_url != $site_url) { | |
471 | db_query("UPDATE ttrss_feeds SET | |
472 | site_url = '$site_url' WHERE id = '$feed'"); | |
473 | } | |
474 | ||
475 | _debug("loading filters & labels...", $debug_enabled); | |
476 | ||
477 | $filters = load_filters($feed, $owner_uid); | |
478 | $labels = get_all_labels($owner_uid); | |
479 | ||
480 | _debug("" . count($filters) . " filters loaded.", $debug_enabled); | |
481 | ||
482 | $items = $rss->get_items(); | |
483 | ||
484 | if (!is_array($items)) { | |
485 | _debug("no articles found.", $debug_enabled); | |
486 | ||
487 | db_query("UPDATE ttrss_feeds | |
488 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
489 | ||
490 | return; // no articles | |
491 | } | |
492 | ||
493 | if ($pubsub_state != 2 && PUBSUBHUBBUB_ENABLED) { | |
494 | ||
495 | _debug("checking for PUSH hub...", $debug_enabled); | |
496 | ||
497 | $feed_hub_url = false; | |
498 | ||
499 | $links = $rss->get_links('hub'); | |
500 | ||
501 | if ($links && is_array($links)) { | |
502 | foreach ($links as $l) { | |
503 | $feed_hub_url = $l; | |
504 | break; | |
505 | } | |
506 | } | |
507 | ||
508 | _debug("feed hub url: $feed_hub_url", $debug_enabled); | |
509 | ||
510 | if ($feed_hub_url && function_exists('curl_init') && | |
511 | !ini_get("open_basedir")) { | |
512 | ||
513 | require_once 'lib/pubsubhubbub/subscriber.php'; | |
514 | ||
515 | $callback_url = get_self_url_prefix() . | |
516 | "/public.php?op=pubsub&id=$feed"; | |
517 | ||
518 | $s = new Subscriber($feed_hub_url, $callback_url); | |
519 | ||
520 | $rc = $s->subscribe($fetch_url); | |
521 | ||
522 | _debug("feed hub url found, subscribe request sent.", $debug_enabled); | |
523 | ||
524 | db_query("UPDATE ttrss_feeds SET pubsub_state = 1 | |
525 | WHERE id = '$feed'"); | |
526 | } | |
527 | } | |
528 | ||
529 | _debug("processing articles...", $debug_enabled); | |
530 | ||
531 | foreach ($items as $item) { | |
532 | if ($_REQUEST['xdebug'] == 3) { | |
533 | print_r($item); | |
534 | } | |
535 | ||
536 | $entry_guid = $item->get_id(); | |
537 | if (!$entry_guid) $entry_guid = $item->get_link(); | |
538 | if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title()); | |
539 | ||
540 | _debug("f_guid $entry_guid", $debug_enabled); | |
541 | ||
542 | if (!$entry_guid) continue; | |
543 | ||
544 | $entry_guid = "$owner_uid,$entry_guid"; | |
545 | ||
546 | $entry_guid_hashed = db_escape_string('SHA1:' . sha1($entry_guid)); | |
547 | ||
548 | _debug("guid $entry_guid / $entry_guid_hashed", $debug_enabled); | |
549 | ||
550 | $entry_timestamp = ""; | |
551 | ||
552 | $entry_timestamp = strtotime($item->get_date()); | |
553 | ||
554 | if ($entry_timestamp == -1 || !$entry_timestamp || $entry_timestamp > time()) { | |
555 | $entry_timestamp = time(); | |
556 | $no_orig_date = 'true'; | |
557 | } else { | |
558 | $no_orig_date = 'false'; | |
559 | } | |
560 | ||
561 | $entry_timestamp_fmt = strftime("%Y/%m/%d %H:%M:%S", $entry_timestamp); | |
562 | ||
563 | _debug("date $entry_timestamp [$entry_timestamp_fmt]", $debug_enabled); | |
564 | ||
565 | $entry_title = html_entity_decode($item->get_title(), ENT_COMPAT, 'UTF-8'); | |
566 | $entry_title = decode_numeric_entities($entry_title); | |
567 | ||
568 | $entry_link = rewrite_relative_url($site_url, $item->get_link()); | |
569 | ||
570 | _debug("title $entry_title", $debug_enabled); | |
571 | _debug("link $entry_link", $debug_enabled); | |
572 | ||
573 | if (!$entry_title) $entry_title = date("Y-m-d H:i:s", $entry_timestamp);; | |
574 | ||
575 | $entry_content = $item->get_content(); | |
576 | if (!$entry_content) $entry_content = $item->get_description(); | |
577 | ||
578 | if ($_REQUEST["xdebug"] == 2) { | |
579 | print "content: "; | |
580 | print $entry_content; | |
581 | print "\n"; | |
582 | } | |
583 | ||
584 | $entry_comments = $item->get_comments_url(); | |
585 | $entry_author = $item->get_author(); | |
586 | ||
587 | $entry_guid = db_escape_string(mb_substr($entry_guid, 0, 245)); | |
588 | ||
589 | $entry_comments = db_escape_string(mb_substr(trim($entry_comments), 0, 245)); | |
590 | $entry_author = db_escape_string(mb_substr(trim($entry_author), 0, 245)); | |
591 | ||
592 | $num_comments = $item->get_comments_count(); | |
593 | ||
594 | if (is_array($num_comments) && is_array($num_comments[0])) { | |
595 | $num_comments = (int) $num_comments[0]["data"]; | |
596 | } else { | |
597 | $num_comments = 0; | |
598 | } | |
599 | ||
600 | _debug("author $entry_author", $debug_enabled); | |
601 | _debug("num_comments: $num_comments", $debug_enabled); | |
602 | _debug("looking for tags [1]...", $debug_enabled); | |
603 | ||
604 | // parse <category> entries into tags | |
605 | ||
606 | $additional_tags = array(); | |
607 | ||
608 | $additional_tags_src = $item->get_categories(); | |
609 | ||
610 | if (is_array($additional_tags_src)) { | |
611 | foreach ($additional_tags_src as $tobj) { | |
612 | array_push($additional_tags, $tobj); | |
613 | } | |
614 | } | |
615 | ||
616 | _debug("category tags:", $debug_enabled); | |
617 | _debug("looking for tags [2]...", $debug_enabled); | |
618 | ||
619 | $entry_tags = array_unique($additional_tags); | |
620 | ||
621 | for ($i = 0; $i < count($entry_tags); $i++) | |
622 | $entry_tags[$i] = mb_strtolower($entry_tags[$i], 'utf-8'); | |
623 | ||
624 | _debug("done collecting data.", $debug_enabled); | |
625 | ||
626 | // TODO: less memory-hungry implementation | |
627 | ||
628 | _debug("applying plugin filters..", $debug_enabled); | |
629 | ||
630 | // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?) | |
631 | $result = db_query("SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries | |
632 | WHERE ref_id = id AND (guid = '".db_escape_string($entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid"); | |
633 | ||
634 | if (db_num_rows($result) != 0) { | |
635 | $entry_plugin_data = db_fetch_result($result, 0, "plugin_data"); | |
636 | $stored_article = array("title" => db_fetch_result($result, 0, "title"), | |
637 | "content" => db_fetch_result($result, 0, "content"), | |
638 | "link" => db_fetch_result($result, 0, "link"), | |
639 | "tags" => explode(",", db_fetch_result($result, 0, "tag_cache")), | |
640 | "author" => db_fetch_result($result, 0, "author")); | |
641 | } else { | |
642 | $entry_plugin_data = ""; | |
643 | $stored_article = array(); | |
644 | } | |
645 | ||
646 | $article = array("owner_uid" => $owner_uid, // read only | |
647 | "guid" => $entry_guid, // read only | |
648 | "title" => $entry_title, | |
649 | "content" => $entry_content, | |
650 | "link" => $entry_link, | |
651 | "tags" => $entry_tags, | |
652 | "plugin_data" => $entry_plugin_data, | |
653 | "author" => $entry_author, | |
654 | "stored" => $stored_article); | |
655 | ||
656 | foreach ($pluginhost->get_hooks(PluginHost::HOOK_ARTICLE_FILTER) as $plugin) { | |
657 | $article = $plugin->hook_article_filter($article); | |
658 | } | |
659 | ||
660 | $entry_tags = $article["tags"]; | |
661 | $entry_guid = db_escape_string($entry_guid); | |
662 | $entry_title = db_escape_string($article["title"]); | |
663 | $entry_author = db_escape_string($article["author"]); | |
664 | $entry_link = db_escape_string($article["link"]); | |
665 | $entry_plugin_data = db_escape_string($article["plugin_data"]); | |
666 | $entry_content = $article["content"]; // escaped below | |
667 | ||
668 | ||
669 | _debug("plugin data: $entry_plugin_data", $debug_enabled); | |
670 | ||
671 | if ($cache_images && is_writable(CACHE_DIR . '/images')) | |
672 | cache_images($entry_content, $site_url, $debug_enabled); | |
673 | ||
674 | $entry_content = db_escape_string($entry_content, false); | |
675 | ||
676 | $content_hash = "SHA1:" . sha1($entry_content); | |
677 | ||
678 | db_query("BEGIN"); | |
679 | ||
680 | $result = db_query("SELECT id FROM ttrss_entries | |
681 | WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')"); | |
682 | ||
683 | if (db_num_rows($result) == 0) { | |
684 | ||
685 | _debug("base guid [$entry_guid] not found", $debug_enabled); | |
686 | ||
687 | // base post entry does not exist, create it | |
688 | ||
689 | $result = db_query( | |
690 | "INSERT INTO ttrss_entries | |
691 | (title, | |
692 | guid, | |
693 | link, | |
694 | updated, | |
695 | content, | |
696 | content_hash, | |
697 | cached_content, | |
698 | no_orig_date, | |
699 | date_updated, | |
700 | date_entered, | |
701 | comments, | |
702 | num_comments, | |
703 | plugin_data, | |
704 | author) | |
705 | VALUES | |
706 | ('$entry_title', | |
707 | '$entry_guid_hashed', | |
708 | '$entry_link', | |
709 | '$entry_timestamp_fmt', | |
710 | '$entry_content', | |
711 | '$content_hash', | |
712 | '', | |
713 | $no_orig_date, | |
714 | NOW(), | |
715 | '$date_feed_processed', | |
716 | '$entry_comments', | |
717 | '$num_comments', | |
718 | '$entry_plugin_data', | |
719 | '$entry_author')"); | |
720 | ||
721 | $article_labels = array(); | |
722 | ||
723 | } else { | |
724 | // we keep encountering the entry in feeds, so we need to | |
725 | // update date_updated column so that we don't get horrible | |
726 | // dupes when the entry gets purged and reinserted again e.g. | |
727 | // in the case of SLOW SLOW OMG SLOW updating feeds | |
728 | ||
729 | $base_entry_id = db_fetch_result($result, 0, "id"); | |
730 | ||
731 | db_query("UPDATE ttrss_entries SET date_updated = NOW() | |
732 | WHERE id = '$base_entry_id'"); | |
733 | ||
734 | $article_labels = get_article_labels($base_entry_id, $owner_uid); | |
735 | } | |
736 | ||
737 | // now it should exist, if not - bad luck then | |
738 | ||
739 | $result = db_query("SELECT | |
740 | id,content_hash,no_orig_date,title,plugin_data,guid, | |
741 | ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated, | |
742 | ".SUBSTRING_FOR_DATE."(updated,1,19) as updated, | |
743 | num_comments | |
744 | FROM | |
745 | ttrss_entries | |
746 | WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'"); | |
747 | ||
748 | $entry_ref_id = 0; | |
749 | $entry_int_id = 0; | |
750 | ||
751 | if (db_num_rows($result) == 1) { | |
752 | ||
753 | _debug("base guid found, checking for user record", $debug_enabled); | |
754 | ||
755 | // this will be used below in update handler | |
756 | $orig_content_hash = db_fetch_result($result, 0, "content_hash"); | |
757 | $orig_title = db_fetch_result($result, 0, "title"); | |
758 | $orig_num_comments = db_fetch_result($result, 0, "num_comments"); | |
759 | $orig_date_updated = strtotime(db_fetch_result($result, | |
760 | 0, "date_updated")); | |
761 | $orig_plugin_data = db_fetch_result($result, 0, "plugin_data"); | |
762 | ||
763 | $ref_id = db_fetch_result($result, 0, "id"); | |
764 | $entry_ref_id = $ref_id; | |
765 | ||
766 | /* $stored_guid = db_fetch_result($result, 0, "guid"); | |
767 | if ($stored_guid != $entry_guid_hashed) { | |
768 | if ($debug_enabled) _debug("upgrading compat guid to hashed one", $debug_enabled); | |
769 | ||
770 | db_query("UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE | |
771 | id = '$ref_id'"); | |
772 | } */ | |
773 | ||
774 | // check for user post link to main table | |
775 | ||
776 | // do we allow duplicate posts with same GUID in different feeds? | |
777 | if (get_pref("ALLOW_DUPLICATE_POSTS", $owner_uid, false)) { | |
778 | $dupcheck_qpart = "AND (feed_id = '$feed' OR feed_id IS NULL)"; | |
779 | } else { | |
780 | $dupcheck_qpart = ""; | |
781 | } | |
782 | ||
783 | /* Collect article tags here so we could filter by them: */ | |
784 | ||
785 | $article_filters = get_article_filters($filters, $entry_title, | |
786 | $entry_content, $entry_link, $entry_timestamp, $entry_author, | |
787 | $entry_tags); | |
788 | ||
789 | if ($debug_enabled) { | |
790 | _debug("article filters: ", $debug_enabled); | |
791 | if (count($article_filters) != 0) { | |
792 | print_r($article_filters); | |
793 | } | |
794 | } | |
795 | ||
796 | if (find_article_filter($article_filters, "filter")) { | |
797 | db_query("COMMIT"); // close transaction in progress | |
798 | continue; | |
799 | } | |
800 | ||
801 | $score = calculate_article_score($article_filters); | |
802 | ||
803 | _debug("initial score: $score", $debug_enabled); | |
804 | ||
805 | $query = "SELECT ref_id, int_id FROM ttrss_user_entries WHERE | |
806 | ref_id = '$ref_id' AND owner_uid = '$owner_uid' | |
807 | $dupcheck_qpart"; | |
808 | ||
809 | // if ($_REQUEST["xdebug"]) print "$query\n"; | |
810 | ||
811 | $result = db_query($query); | |
812 | ||
813 | // okay it doesn't exist - create user entry | |
814 | if (db_num_rows($result) == 0) { | |
815 | ||
816 | _debug("user record not found, creating...", $debug_enabled); | |
817 | ||
818 | if ($score >= -500 && !find_article_filter($article_filters, 'catchup')) { | |
819 | $unread = 'true'; | |
820 | $last_read_qpart = 'NULL'; | |
821 | } else { | |
822 | $unread = 'false'; | |
823 | $last_read_qpart = 'NOW()'; | |
824 | } | |
825 | ||
826 | if (find_article_filter($article_filters, 'mark') || $score > 1000) { | |
827 | $marked = 'true'; | |
828 | } else { | |
829 | $marked = 'false'; | |
830 | } | |
831 | ||
832 | if (find_article_filter($article_filters, 'publish')) { | |
833 | $published = 'true'; | |
834 | } else { | |
835 | $published = 'false'; | |
836 | } | |
837 | ||
838 | // N-grams | |
839 | ||
840 | if (DB_TYPE == "pgsql" and defined('_NGRAM_TITLE_DUPLICATE_THRESHOLD')) { | |
841 | ||
842 | $result = db_query("SELECT COUNT(*) AS similar FROM | |
843 | ttrss_entries,ttrss_user_entries | |
844 | WHERE ref_id = id AND updated >= NOW() - INTERVAL '7 day' | |
845 | AND similarity(title, '$entry_title') >= "._NGRAM_TITLE_DUPLICATE_THRESHOLD." | |
846 | AND owner_uid = $owner_uid"); | |
847 | ||
848 | $ngram_similar = db_fetch_result($result, 0, "similar"); | |
849 | ||
850 | _debug("N-gram similar results: $ngram_similar", $debug_enabled); | |
851 | ||
852 | if ($ngram_similar > 0) { | |
853 | $unread = 'false'; | |
854 | } | |
855 | } | |
856 | ||
857 | $last_marked = ($marked == 'true') ? 'NOW()' : 'NULL'; | |
858 | $last_published = ($published == 'true') ? 'NOW()' : 'NULL'; | |
859 | ||
860 | $result = db_query( | |
861 | "INSERT INTO ttrss_user_entries | |
862 | (ref_id, owner_uid, feed_id, unread, last_read, marked, | |
863 | published, score, tag_cache, label_cache, uuid, | |
864 | last_marked, last_published) | |
865 | VALUES ('$ref_id', '$owner_uid', '$feed', $unread, | |
866 | $last_read_qpart, $marked, $published, '$score', '', '', | |
867 | '', $last_marked, $last_published)"); | |
868 | ||
869 | if (PUBSUBHUBBUB_HUB && $published == 'true') { | |
870 | $rss_link = get_self_url_prefix() . | |
871 | "/public.php?op=rss&id=-2&key=" . | |
872 | get_feed_access_key(-2, false, $owner_uid); | |
873 | ||
874 | $p = new Publisher(PUBSUBHUBBUB_HUB); | |
875 | ||
876 | $pubsub_result = $p->publish_update($rss_link); | |
877 | } | |
878 | ||
879 | $result = db_query( | |
880 | "SELECT int_id FROM ttrss_user_entries WHERE | |
881 | ref_id = '$ref_id' AND owner_uid = '$owner_uid' AND | |
882 | feed_id = '$feed' LIMIT 1"); | |
883 | ||
884 | if (db_num_rows($result) == 1) { | |
885 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
886 | } | |
887 | } else { | |
888 | _debug("user record FOUND", $debug_enabled); | |
889 | ||
890 | $entry_ref_id = db_fetch_result($result, 0, "ref_id"); | |
891 | $entry_int_id = db_fetch_result($result, 0, "int_id"); | |
892 | } | |
893 | ||
894 | _debug("RID: $entry_ref_id, IID: $entry_int_id", $debug_enabled); | |
895 | ||
896 | $post_needs_update = false; | |
897 | $update_insignificant = false; | |
898 | ||
899 | if ($orig_num_comments != $num_comments) { | |
900 | $post_needs_update = true; | |
901 | $update_insignificant = true; | |
902 | } | |
903 | ||
904 | if ($entry_plugin_data != $orig_plugin_data) { | |
905 | $post_needs_update = true; | |
906 | $update_insignificant = true; | |
907 | } | |
908 | ||
909 | if ($content_hash != $orig_content_hash) { | |
910 | $post_needs_update = true; | |
911 | $update_insignificant = false; | |
912 | } | |
913 | ||
914 | if (db_escape_string($orig_title) != $entry_title) { | |
915 | $post_needs_update = true; | |
916 | $update_insignificant = false; | |
917 | } | |
918 | ||
919 | // if post needs update, update it and mark all user entries | |
920 | // linking to this post as updated | |
921 | if ($post_needs_update) { | |
922 | ||
923 | if (defined('DAEMON_EXTENDED_DEBUG')) { | |
924 | _debug("post $entry_guid_hashed needs update...", $debug_enabled); | |
925 | } | |
926 | ||
927 | // print "<!-- post $orig_title needs update : $post_needs_update -->"; | |
928 | ||
929 | db_query("UPDATE ttrss_entries | |
930 | SET title = '$entry_title', content = '$entry_content', | |
931 | content_hash = '$content_hash', | |
932 | updated = '$entry_timestamp_fmt', | |
933 | num_comments = '$num_comments', | |
934 | plugin_data = '$entry_plugin_data' | |
935 | WHERE id = '$ref_id'"); | |
936 | ||
937 | if (!$update_insignificant) { | |
938 | if ($mark_unread_on_update) { | |
939 | db_query("UPDATE ttrss_user_entries | |
940 | SET last_read = null, unread = true WHERE ref_id = '$ref_id'"); | |
941 | } | |
942 | } | |
943 | } | |
944 | } | |
945 | ||
946 | db_query("COMMIT"); | |
947 | ||
948 | _debug("assigning labels...", $debug_enabled); | |
949 | ||
950 | assign_article_to_label_filters($entry_ref_id, $article_filters, | |
951 | $owner_uid, $article_labels); | |
952 | ||
953 | _debug("looking for enclosures...", $debug_enabled); | |
954 | ||
955 | // enclosures | |
956 | ||
957 | $enclosures = array(); | |
958 | ||
959 | $encs = $item->get_enclosures(); | |
960 | ||
961 | if (is_array($encs)) { | |
962 | foreach ($encs as $e) { | |
963 | $e_item = array( | |
964 | $e->link, $e->type, $e->length); | |
965 | array_push($enclosures, $e_item); | |
966 | } | |
967 | } | |
968 | ||
969 | if ($debug_enabled) { | |
970 | _debug("article enclosures:", $debug_enabled); | |
971 | print_r($enclosures); | |
972 | } | |
973 | ||
974 | db_query("BEGIN"); | |
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 | ||
981 | $result = db_query("SELECT id FROM ttrss_enclosures | |
982 | WHERE content_url = '$enc_url' AND post_id = '$entry_ref_id'"); | |
983 | ||
984 | if (db_num_rows($result) == 0) { | |
985 | db_query("INSERT INTO ttrss_enclosures | |
986 | (content_url, content_type, title, duration, post_id) VALUES | |
987 | ('$enc_url', '$enc_type', '', '$enc_dur', '$entry_ref_id')"); | |
988 | } | |
989 | } | |
990 | ||
991 | db_query("COMMIT"); | |
992 | ||
993 | // check for manual tags (we have to do it here since they're loaded from filters) | |
994 | ||
995 | foreach ($article_filters as $f) { | |
996 | if ($f["type"] == "tag") { | |
997 | ||
998 | $manual_tags = trim_array(explode(",", $f["param"])); | |
999 | ||
1000 | foreach ($manual_tags as $tag) { | |
1001 | if (tag_is_valid($tag)) { | |
1002 | array_push($entry_tags, $tag); | |
1003 | } | |
1004 | } | |
1005 | } | |
1006 | } | |
1007 | ||
1008 | // Skip boring tags | |
1009 | ||
1010 | $boring_tags = trim_array(explode(",", mb_strtolower(get_pref( | |
1011 | 'BLACKLISTED_TAGS', $owner_uid, ''), 'utf-8'))); | |
1012 | ||
1013 | $filtered_tags = array(); | |
1014 | $tags_to_cache = array(); | |
1015 | ||
1016 | if ($entry_tags && is_array($entry_tags)) { | |
1017 | foreach ($entry_tags as $tag) { | |
1018 | if (array_search($tag, $boring_tags) === false) { | |
1019 | array_push($filtered_tags, $tag); | |
1020 | } | |
1021 | } | |
1022 | } | |
1023 | ||
1024 | $filtered_tags = array_unique($filtered_tags); | |
1025 | ||
1026 | if ($debug_enabled) { | |
1027 | _debug("filtered article tags:", $debug_enabled); | |
1028 | print_r($filtered_tags); | |
1029 | } | |
1030 | ||
1031 | // Save article tags in the database | |
1032 | ||
1033 | if (count($filtered_tags) > 0) { | |
1034 | ||
1035 | db_query("BEGIN"); | |
1036 | ||
1037 | foreach ($filtered_tags as $tag) { | |
1038 | ||
1039 | $tag = sanitize_tag($tag); | |
1040 | $tag = db_escape_string($tag); | |
1041 | ||
1042 | if (!tag_is_valid($tag)) continue; | |
1043 | ||
1044 | $result = db_query("SELECT id FROM ttrss_tags | |
1045 | WHERE tag_name = '$tag' AND post_int_id = '$entry_int_id' AND | |
1046 | owner_uid = '$owner_uid' LIMIT 1"); | |
1047 | ||
1048 | if ($result && db_num_rows($result) == 0) { | |
1049 | ||
1050 | db_query("INSERT INTO ttrss_tags | |
1051 | (owner_uid,tag_name,post_int_id) | |
1052 | VALUES ('$owner_uid','$tag', '$entry_int_id')"); | |
1053 | } | |
1054 | ||
1055 | array_push($tags_to_cache, $tag); | |
1056 | } | |
1057 | ||
1058 | /* update the cache */ | |
1059 | ||
1060 | $tags_to_cache = array_unique($tags_to_cache); | |
1061 | ||
1062 | $tags_str = db_escape_string(join(",", $tags_to_cache)); | |
1063 | ||
1064 | db_query("UPDATE ttrss_user_entries | |
1065 | SET tag_cache = '$tags_str' WHERE ref_id = '$entry_ref_id' | |
1066 | AND owner_uid = $owner_uid"); | |
1067 | ||
1068 | db_query("COMMIT"); | |
1069 | } | |
1070 | ||
1071 | if (get_pref("AUTO_ASSIGN_LABELS", $owner_uid, false)) { | |
1072 | _debug("auto-assigning labels...", $debug_enabled); | |
1073 | ||
1074 | foreach ($labels as $label) { | |
1075 | $caption = preg_quote($label["caption"]); | |
1076 | ||
1077 | if ($caption && preg_match("/\b$caption\b/i", "$tags_str " . strip_tags($entry_content) . " $entry_title")) { | |
1078 | if (!labels_contains_caption($article_labels, $caption)) { | |
1079 | label_add_article($entry_ref_id, $caption, $owner_uid); | |
1080 | } | |
1081 | } | |
1082 | } | |
1083 | } | |
1084 | ||
1085 | _debug("article processed", $debug_enabled); | |
1086 | } | |
1087 | ||
1088 | if (!$last_updated) { | |
1089 | _debug("new feed, catching it up...", $debug_enabled); | |
1090 | catchup_feed($feed, false, $owner_uid); | |
1091 | } | |
1092 | ||
1093 | _debug("purging feed...", $debug_enabled); | |
1094 | ||
1095 | purge_feed($feed, 0, $debug_enabled); | |
1096 | ||
1097 | db_query("UPDATE ttrss_feeds | |
1098 | SET last_updated = NOW(), last_error = '' WHERE id = '$feed'"); | |
1099 | ||
1100 | // db_query("COMMIT"); | |
1101 | ||
1102 | } else { | |
1103 | ||
1104 | $error_msg = db_escape_string(mb_substr($rss->error(), 0, 245)); | |
1105 | ||
1106 | _debug("error fetching feed: $error_msg", $debug_enabled); | |
1107 | ||
1108 | db_query( | |
1109 | "UPDATE ttrss_feeds SET last_error = '$error_msg', | |
1110 | last_updated = NOW() WHERE id = '$feed'"); | |
1111 | } | |
1112 | ||
1113 | unset($rss); | |
1114 | ||
1115 | _debug("done", $debug_enabled); | |
1116 | } | |
1117 | ||
1118 | function cache_images($html, $site_url, $debug) { | |
1119 | $cache_dir = CACHE_DIR . "/images"; | |
1120 | ||
1121 | libxml_use_internal_errors(true); | |
1122 | ||
1123 | $charset_hack = '<head> | |
1124 | <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/> | |
1125 | </head>'; | |
1126 | ||
1127 | $doc = new DOMDocument(); | |
1128 | $doc->loadHTML($charset_hack . $html); | |
1129 | $xpath = new DOMXPath($doc); | |
1130 | ||
1131 | $entries = $xpath->query('(//img[@src])'); | |
1132 | ||
1133 | foreach ($entries as $entry) { | |
1134 | if ($entry->hasAttribute('src')) { | |
1135 | $src = rewrite_relative_url($site_url, $entry->getAttribute('src')); | |
1136 | ||
1137 | $local_filename = CACHE_DIR . "/images/" . sha1($src) . ".png"; | |
1138 | ||
1139 | if ($debug) _debug("cache_images: downloading: $src to $local_filename"); | |
1140 | ||
1141 | if (!file_exists($local_filename)) { | |
1142 | $file_content = fetch_file_contents($src); | |
1143 | ||
1144 | if ($file_content && strlen($file_content) > 1024) { | |
1145 | file_put_contents($local_filename, $file_content); | |
1146 | } | |
1147 | } | |
1148 | ||
1149 | if (file_exists($local_filename)) { | |
1150 | $entry->setAttribute('src', SELF_URL_PATH . '/image.php?url=' . | |
1151 | base64_encode($src)); | |
1152 | } | |
1153 | } | |
1154 | } | |
1155 | ||
1156 | $node = $doc->getElementsByTagName('body')->item(0); | |
1157 | ||
1158 | return $doc->saveXML($node); | |
1159 | } | |
1160 | ||
1161 | function expire_error_log($debug) { | |
1162 | if ($debug) _debug("Removing old error log entries..."); | |
1163 | ||
1164 | if (DB_TYPE == "pgsql") { | |
1165 | db_query("DELETE FROM ttrss_error_log | |
1166 | WHERE created_at < NOW() - INTERVAL '7 days'"); | |
1167 | } else { | |
1168 | db_query("DELETE FROM ttrss_error_log | |
1169 | WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)"); | |
1170 | } | |
1171 | ||
1172 | } | |
1173 | ||
1174 | function expire_lock_files($debug) { | |
1175 | //if ($debug) _debug("Removing old lock files..."); | |
1176 | ||
1177 | $num_deleted = 0; | |
1178 | ||
1179 | if (is_writable(LOCK_DIRECTORY)) { | |
1180 | $files = glob(LOCK_DIRECTORY . "/*.lock"); | |
1181 | ||
1182 | if ($files) { | |
1183 | foreach ($files as $file) { | |
1184 | if (!file_is_locked(basename($file)) && time() - filemtime($file) > 86400*2) { | |
1185 | unlink($file); | |
1186 | ++$num_deleted; | |
1187 | } | |
1188 | } | |
1189 | } | |
1190 | } | |
1191 | ||
1192 | if ($debug) _debug("Removed $num_deleted old lock files."); | |
1193 | } | |
1194 | ||
1195 | function expire_cached_files($debug) { | |
1196 | foreach (array("simplepie", "images", "export", "upload") as $dir) { | |
1197 | $cache_dir = CACHE_DIR . "/$dir"; | |
1198 | ||
1199 | // if ($debug) _debug("Expiring $cache_dir"); | |
1200 | ||
1201 | $num_deleted = 0; | |
1202 | ||
1203 | if (is_writable($cache_dir)) { | |
1204 | $files = glob("$cache_dir/*"); | |
1205 | ||
1206 | if ($files) { | |
1207 | foreach ($files as $file) { | |
1208 | if (time() - filemtime($file) > 86400*7) { | |
1209 | unlink($file); | |
1210 | ||
1211 | ++$num_deleted; | |
1212 | } | |
1213 | } | |
1214 | } | |
1215 | } | |
1216 | ||
1217 | if ($debug) _debug("$cache_dir: removed $num_deleted files."); | |
1218 | } | |
1219 | } | |
1220 | ||
1221 | /** | |
1222 | * Source: http://www.php.net/manual/en/function.parse-url.php#104527 | |
1223 | * Returns the url query as associative array | |
1224 | * | |
1225 | * @param string query | |
1226 | * @return array params | |
1227 | */ | |
1228 | function convertUrlQuery($query) { | |
1229 | $queryParts = explode('&', $query); | |
1230 | ||
1231 | $params = array(); | |
1232 | ||
1233 | foreach ($queryParts as $param) { | |
1234 | $item = explode('=', $param); | |
1235 | $params[$item[0]] = $item[1]; | |
1236 | } | |
1237 | ||
1238 | return $params; | |
1239 | } | |
1240 | ||
1241 | function get_article_filters($filters, $title, $content, $link, $timestamp, $author, $tags) { | |
1242 | $matches = array(); | |
1243 | ||
1244 | foreach ($filters as $filter) { | |
1245 | $match_any_rule = $filter["match_any_rule"]; | |
1246 | $inverse = $filter["inverse"]; | |
1247 | $filter_match = false; | |
1248 | ||
1249 | foreach ($filter["rules"] as $rule) { | |
1250 | $match = false; | |
1251 | $reg_exp = str_replace('/', '\/', $rule["reg_exp"]); | |
1252 | $rule_inverse = $rule["inverse"]; | |
1253 | ||
1254 | if (!$reg_exp) | |
1255 | continue; | |
1256 | ||
1257 | switch ($rule["type"]) { | |
1258 | case "title": | |
1259 | $match = @preg_match("/$reg_exp/i", $title); | |
1260 | break; | |
1261 | case "content": | |
1262 | // we don't need to deal with multiline regexps | |
1263 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1264 | ||
1265 | $match = @preg_match("/$reg_exp/i", $content); | |
1266 | break; | |
1267 | case "both": | |
1268 | // we don't need to deal with multiline regexps | |
1269 | $content = preg_replace("/[\r\n\t]/", "", $content); | |
1270 | ||
1271 | $match = (@preg_match("/$reg_exp/i", $title) || @preg_match("/$reg_exp/i", $content)); | |
1272 | break; | |
1273 | case "link": | |
1274 | $match = @preg_match("/$reg_exp/i", $link); | |
1275 | break; | |
1276 | case "author": | |
1277 | $match = @preg_match("/$reg_exp/i", $author); | |
1278 | break; | |
1279 | case "tag": | |
1280 | foreach ($tags as $tag) { | |
1281 | if (@preg_match("/$reg_exp/i", $tag)) { | |
1282 | $match = true; | |
1283 | break; | |
1284 | } | |
1285 | } | |
1286 | break; | |
1287 | } | |
1288 | ||
1289 | if ($rule_inverse) $match = !$match; | |
1290 | ||
1291 | if ($match_any_rule) { | |
1292 | if ($match) { | |
1293 | $filter_match = true; | |
1294 | break; | |
1295 | } | |
1296 | } else { | |
1297 | $filter_match = $match; | |
1298 | if (!$match) { | |
1299 | break; | |
1300 | } | |
1301 | } | |
1302 | } | |
1303 | ||
1304 | if ($inverse) $filter_match = !$filter_match; | |
1305 | ||
1306 | if ($filter_match) { | |
1307 | foreach ($filter["actions"] AS $action) { | |
1308 | array_push($matches, $action); | |
1309 | ||
1310 | // if Stop action encountered, perform no further processing | |
1311 | if ($action["type"] == "stop") return $matches; | |
1312 | } | |
1313 | } | |
1314 | } | |
1315 | ||
1316 | return $matches; | |
1317 | } | |
1318 | ||
1319 | function find_article_filter($filters, $filter_name) { | |
1320 | foreach ($filters as $f) { | |
1321 | if ($f["type"] == $filter_name) { | |
1322 | return $f; | |
1323 | }; | |
1324 | } | |
1325 | return false; | |
1326 | } | |
1327 | ||
1328 | function find_article_filters($filters, $filter_name) { | |
1329 | $results = array(); | |
1330 | ||
1331 | foreach ($filters as $f) { | |
1332 | if ($f["type"] == $filter_name) { | |
1333 | array_push($results, $f); | |
1334 | }; | |
1335 | } | |
1336 | return $results; | |
1337 | } | |
1338 | ||
1339 | function calculate_article_score($filters) { | |
1340 | $score = 0; | |
1341 | ||
1342 | foreach ($filters as $f) { | |
1343 | if ($f["type"] == "score") { | |
1344 | $score += $f["param"]; | |
1345 | }; | |
1346 | } | |
1347 | return $score; | |
1348 | } | |
1349 | ||
1350 | function labels_contains_caption($labels, $caption) { | |
1351 | foreach ($labels as $label) { | |
1352 | if ($label[1] == $caption) { | |
1353 | return true; | |
1354 | } | |
1355 | } | |
1356 | ||
1357 | return false; | |
1358 | } | |
1359 | ||
1360 | function assign_article_to_label_filters($id, $filters, $owner_uid, $article_labels) { | |
1361 | foreach ($filters as $f) { | |
1362 | if ($f["type"] == "label") { | |
1363 | if (!labels_contains_caption($article_labels, $f["param"])) { | |
1364 | label_add_article($id, $f["param"], $owner_uid); | |
1365 | } | |
1366 | } | |
1367 | } | |
1368 | } | |
1369 | ||
1370 | function make_guid_from_title($title) { | |
1371 | return preg_replace("/[ \"\',.:;]/", "-", | |
1372 | mb_strtolower(strip_tags($title), 'utf-8')); | |
1373 | } | |
1374 | ||
1375 | function verify_feed_xml($feed_data) { | |
1376 | libxml_use_internal_errors(true); | |
1377 | $doc = new DOMDocument(); | |
1378 | $doc->loadXML($feed_data); | |
1379 | $error = libxml_get_last_error(); | |
1380 | libxml_clear_errors(); | |
1381 | return $error; | |
1382 | } | |
1383 | ||
1384 | function housekeeping_common($debug) { | |
1385 | expire_cached_files($debug); | |
1386 | expire_lock_files($debug); | |
1387 | expire_error_log($debug); | |
1388 | ||
1389 | $count = update_feedbrowser_cache(); | |
1390 | _debug("Feedbrowser updated, $count feeds processed."); | |
1391 | ||
1392 | purge_orphans( true); | |
1393 | $rc = cleanup_tags( 14, 50000); | |
1394 | ||
1395 | _debug("Cleaned $rc cached tags."); | |
1396 | } | |
1397 | ||
1398 | function utf8_entity_decode($entity){ | |
1399 | $convmap = array(0x0, 0x10000, 0, 0xfffff); | |
1400 | return mb_decode_numericentity($entity, $convmap, 'UTF-8'); | |
1401 | } | |
1402 | ||
1403 | function decode_numeric_entities($body) { | |
1404 | $body = preg_replace('/&#\d{2,5};/ue', "utf8_entity_decode('$0')", $body ); | |
1405 | $body = preg_replace('/&#x([a-fA-F0-7]{2,8});/ue', "utf8_entity_decode('&#'.hexdec('$1').';')", $body ); | |
1406 | return $body; | |
1407 | } | |
1408 | ?> |