]> git.wh0rd.org - tt-rss.git/blobdiff - include/rssfuncs.php
add schema checking to feed updating and update script
[tt-rss.git] / include / rssfuncs.php
index 49d9e6009c448a8ab7d7e33281602b37df9a701c..9ea2662fccd0ea07cef1b0e3f6a3344469dc0f02 100644 (file)
        function update_daemon_common($link, $limit = DAEMON_FEED_LIMIT, $from_http = false, $debug = true) {
                // Process all other feeds using last_updated and interval parameters
 
+               $schema_version = get_schema_version($link);
+
+               if ($schema_version != SCHEMA_VERSION) {
+                       die("Schema version is wrong, please upgrade the database.\n");
+               }
+
                define('PREFS_NO_CACHE', true);
 
                // Test if the user has loggued in recently. If not, it does not update its feeds.
@@ -77,6 +83,7 @@
                if (DB_TYPE == "pgsql") {
                        $update_limit_qpart = "AND ((
                                        ttrss_feeds.update_interval = 0
+                                       AND ttrss_user_prefs.value != '-1'
                                        AND ttrss_feeds.last_updated < NOW() - CAST((ttrss_user_prefs.value || ' minutes') AS INTERVAL)
                                ) OR (
                                        ttrss_feeds.update_interval > 0
@@ -86,6 +93,7 @@
                } else {
                        $update_limit_qpart = "AND ((
                                        ttrss_feeds.update_interval = 0
+                                       AND ttrss_user_prefs.value != '-1'
                                        AND ttrss_feeds.last_updated < DATE_SUB(NOW(), INTERVAL CONVERT(ttrss_user_prefs.value, SIGNED INTEGER) MINUTE)
                                ) OR (
                                        ttrss_feeds.update_interval > 0
                $random_qpart = sql_random_function();
 
                // We search for feed needing update.
-               $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url
+               $result = db_query($link, "SELECT DISTINCT ttrss_feeds.feed_url,$random_qpart
                        FROM
                                ttrss_feeds, ttrss_users, ttrss_user_prefs
                        WHERE
                                AND ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL'
                                $login_thresh_qpart $update_limit_qpart
                                $updstart_thresh_qpart
-                       ORDER BY feed_url $query_limit");
+                       ORDER BY $random_qpart $query_limit");
 
                $user_prefs_cache = array();
 
 
                expire_cached_files($debug);
                expire_lock_files($debug);
+               expire_error_log($link, $debug);
+
+               $nf = 0;
 
                // For each feed, we call the feed update function.
                foreach ($feeds_to_update as $feed) {
 
                        // since we have the data cached, we can deal with other feeds with the same url
 
-                       $tmp_result = db_query($link, "SELECT ttrss_feeds.feed_url,ttrss_feeds.id,last_updated
-                       FROM ttrss_feeds, ttrss_users WHERE
-                               ttrss_users.id = ttrss_feeds.owner_uid AND
+                       $tmp_result = db_query($link, "SELECT DISTINCT ttrss_feeds.id,last_updated
+                       FROM ttrss_feeds, ttrss_users, ttrss_user_prefs WHERE
+                               ttrss_user_prefs.owner_uid = ttrss_feeds.owner_uid AND
+                               ttrss_users.id = ttrss_user_prefs.owner_uid AND
+                               ttrss_user_prefs.pref_name = 'DEFAULT_UPDATE_INTERVAL' AND
                                feed_url = '".db_escape_string($link, $feed)."' AND
-                               ttrss_feeds.update_interval != -1
+                               (ttrss_feeds.update_interval > 0 OR
+                                       ttrss_user_prefs.value != '-1')
                                $login_thresh_qpart
-                       ORDER BY feed_url $query_limit");
+                       ORDER BY ttrss_feeds.id $query_limit");
 
                        if (db_num_rows($tmp_result) > 0) {
                                while ($tline = db_fetch_assoc($tmp_result)) {
                                        if($debug) _debug(" => " . $tline["last_updated"] . ", " . $tline["id"]);
                                        update_rss_feed($link, $tline["id"], true);
+                                       ++$nf;
                                }
                        }
                }
                // Send feed digests by email if needed.
                send_headlines_digests($link, $debug);
 
+               return $nf;
+
        } // function update_daemon_common
 
        // ignore_daemon is not used
                $result = db_query($link, "SELECT id,update_interval,auth_login,
                        feed_url,auth_pass,cache_images,last_updated,
                        mark_unread_on_update, owner_uid,
-                       pubsub_state
+                       pubsub_state, auth_pass_encrypted
                        FROM ttrss_feeds WHERE id = '$feed'");
 
                if (db_num_rows($result) == 0) {
                $mark_unread_on_update = sql_bool_to_bool(db_fetch_result($result,
                        0, "mark_unread_on_update"));
                $pubsub_state = db_fetch_result($result, 0, "pubsub_state");
+               $auth_pass_encrypted = sql_bool_to_bool(db_fetch_result($result,
+                       0, "auth_pass_encrypted"));
 
                db_query($link, "UPDATE ttrss_feeds SET last_update_started = NOW()
                        WHERE id = '$feed'");
                $auth_login = db_fetch_result($result, 0, "auth_login");
                $auth_pass = db_fetch_result($result, 0, "auth_pass");
 
+               if ($auth_pass_encrypted) {
+                       require_once "crypt.php";
+                       $auth_pass = decrypt_string($auth_pass);
+               }
+
                $cache_images = sql_bool_to_bool(db_fetch_result($result, 0, "cache_images"));
                $fetch_url = db_fetch_result($result, 0, "feed_url");
 
                $cache_timestamp = file_exists($cache_filename) ? filemtime($cache_filename) : 0;
                $last_updated_timestamp = strtotime($last_updated);
 
+               $force_refetch = isset($_REQUEST["force_refetch"]);
+
                if (file_exists($cache_filename) &&
                        is_readable($cache_filename) &&
                        !$auth_login && !$auth_pass &&
                                                $rss_hash = sha1($rss_data);
                                                @$rss = unserialize($rss_data);
                                        }
-                               } else {
+                               } else if (!$force_refetch) {
                                        if ($debug_enabled) {
                                                _debug("update_rss_feed: local cache valid and older than last_updated, nothing to do.");
                                        }
                                }
 
                                $feed_data = fetch_file_contents($fetch_url, false,
-                                       $auth_login, $auth_pass, false, $no_cache ? 15 : 45,
-                                       max($last_updated_timestamp, $cache_timestamp));
+                                       $auth_login, $auth_pass, false,
+                                       $no_cache ? FEED_FETCH_NO_CACHE_TIMEOUT : FEED_FETCH_TIMEOUT,
+                                       $force_refetch ? 0 : max($last_updated_timestamp, $cache_timestamp));
 
                                if ($debug_enabled) {
                                        _debug("update_rss_feed: fetch done.");
 
                        $site_url = db_escape_string($link, mb_substr(rewrite_relative_url($fetch_url, $rss->get_link()), 0, 245));
 
-                       if ($debug_enabled) {
-                               _debug("update_rss_feed: checking favicon...");
-                       }
+                       if ($favicon_needs_check || $force_refetch) {
+                               if ($debug_enabled) {
+                                       _debug("update_rss_feed: checking favicon...");
+                               }
 
-                       if ($favicon_needs_check) {
                                check_feed_favicon($site_url, $feed, $link);
+                               $favicon_file = ICONS_DIR . "/$feed.ico";
+
+                               if (file_exists($favicon_file)) {
+                                               require_once "colors.php";
+
+                                               $favicon_color = db_escape_string($link,
+                                                       calculate_avg_color($favicon_file));
+
+                                               $favicon_colorstring = ",favicon_avg_color = '".$favicon_color."'";
+                               }
 
                                db_query($link, "UPDATE ttrss_feeds SET favicon_last_checked = NOW()
+                                       $favicon_colorstring
                                        WHERE id = '$feed'");
                        }
 
                                if (!$entry_guid) $entry_guid = $item->get_link();
                                if (!$entry_guid) $entry_guid = make_guid_from_title($item->get_title());
 
-                               if ($debug_enabled) {
-                                       _debug("update_rss_feed: guid $entry_guid");
-                               }
-
                                if (!$entry_guid) continue;
 
                                $entry_guid = "$owner_uid,$entry_guid";
 
+                               $entry_guid_hashed = db_escape_string($link, 'SHA1:' . sha1($entry_guid));
+
+                               if ($debug_enabled) {
+                                       _debug("update_rss_feed: guid $entry_guid / $entry_guid_hashed");
+                               }
+
                                $entry_timestamp = "";
 
                                $entry_timestamp = strtotime($item->get_date());
                                        _debug("update_rss_feed: date $entry_timestamp [$entry_timestamp_fmt]");
                                }
 
-                               $entry_title = $item->get_title();
+                               $entry_title = html_entity_decode($item->get_title());
 
                                $entry_link = rewrite_relative_url($site_url, $item->get_link());
 
 
                                // FIXME not sure if owner_uid is a good idea here, we may have a base entry without user entry (?)
                                $result = db_query($link, "SELECT plugin_data,title,content,link,tag_cache,author FROM ttrss_entries, ttrss_user_entries
-                                       WHERE ref_id = id AND guid = '".db_escape_string($link, $entry_guid)."' AND owner_uid = $owner_uid");
+                                       WHERE ref_id = id AND (guid = '".db_escape_string($link, $entry_guid)."' OR guid = '$entry_guid_hashed') AND owner_uid = $owner_uid");
 
                                if (db_num_rows($result) != 0) {
                                        $entry_plugin_data = db_fetch_result($result, 0, "plugin_data");
                                db_query($link, "BEGIN");
 
                                $result = db_query($link, "SELECT id FROM       ttrss_entries
-                                       WHERE guid = '$entry_guid'");
+                                       WHERE (guid = '$entry_guid' OR guid = '$entry_guid_hashed')");
 
                                if (db_num_rows($result) == 0) {
 
                                                        author)
                                                VALUES
                                                        ('$entry_title',
-                                                       '$entry_guid',
+                                                       '$entry_guid_hashed',
                                                        '$entry_link',
                                                        '$entry_timestamp_fmt',
                                                        '$entry_content',
                                // now it should exist, if not - bad luck then
 
                                $result = db_query($link, "SELECT
-                                               id,content_hash,no_orig_date,title,plugin_data,
+                                               id,content_hash,no_orig_date,title,plugin_data,guid,
                                                ".SUBSTRING_FOR_DATE."(date_updated,1,19) as date_updated,
                                                ".SUBSTRING_FOR_DATE."(updated,1,19) as updated,
                                                num_comments
                                        FROM
                                                ttrss_entries
-                                       WHERE guid = '$entry_guid'");
+                                       WHERE guid = '$entry_guid' OR guid = '$entry_guid_hashed'");
 
                                $entry_ref_id = 0;
                                $entry_int_id = 0;
                                if (db_num_rows($result) == 1) {
 
                                        if ($debug_enabled) {
-                                               _debug("update_rss_feed: base guid [$entry_guid] found, checking for user record");
+                                               _debug("update_rss_feed: base guid found, checking for user record");
                                        }
 
                                        // this will be used below in update handler
                                        $ref_id = db_fetch_result($result, 0, "id");
                                        $entry_ref_id = $ref_id;
 
+                                       /* $stored_guid = db_fetch_result($result, 0, "guid");
+                                       if ($stored_guid != $entry_guid_hashed) {
+                                               if ($debug_enabled) _debug("upgrading compat guid to hashed one");
+
+                                               db_query($link, "UPDATE ttrss_entries SET guid = '$entry_guid_hashed' WHERE
+                                                       id = '$ref_id'");
+                                       } */
+
                                        // check for user post link to main table
 
                                        // do we allow duplicate posts with same GUID in different feeds?
                                        if ($post_needs_update) {
 
                                                if (defined('DAEMON_EXTENDED_DEBUG')) {
-                                                       _debug("update_rss_feed: post $entry_guid needs update...");
+                                                       _debug("update_rss_feed: post $entry_guid_hashed needs update...");
                                                }
 
 //                                             print "<!-- post $orig_title needs update : $post_needs_update -->";
                return $doc->saveXML($node);
        }
 
+       function expire_error_log($link, $debug) {
+               if ($debug) _debug("Removing old error log entries...");
+
+               if (DB_TYPE == "pgsql") {
+                       db_query($link, "DELETE FROM ttrss_error_log
+                               WHERE created_at < NOW() - INTERVAL '7 days'");
+               } else {
+                       db_query($link, "DELETE FROM ttrss_error_log
+                               WHERE created_at < DATE_SUB(NOW(), INTERVAL 7 DAY)");
+               }
+
+       }
+
        function expire_lock_files($debug) {
                if ($debug) _debug("Removing old lock files...");
 
        }
 
        function expire_cached_files($debug) {
-               foreach (array("simplepie", "images", "export") as $dir) {
+               foreach (array("simplepie", "images", "export", "upload") as $dir) {
                        $cache_dir = CACHE_DIR . "/$dir";
 
                        if ($debug) _debug("Expiring $cache_dir");
                        if ($filter_match) {
                                foreach ($filter["actions"] AS $action) {
                                        array_push($matches, $action);
+
+                                       // if Stop action encountered, perform no further processing
+                                       if ($action["type"] == "stop") return $matches;
                                }
                        }
                }