FreshRSS (Customized)/app/models/Entry.php, branch 0.5.0 Customized version of FreshRSS, a self-hosted RSS feed aggregator https://git.rdnlsmith.com/fresh-rss-custom/atom?h=0.5.0 2013-09-15T09:29:31+00:00 Issue #155 : correction bug prefix table SQL 2013-09-15T09:29:31+00:00 Marien Fressinaud dev@marienfressinaud.fr 2013-09-15T09:29:31+00:00 urn:sha1:35dcb5e39aee8029fee541d5649456bc385cd649 Le préfixe freshrss_ était rentré en dur dans une des requêtes Issue #155 : suppression fonctions commentées 2013-09-14T19:59:54+00:00 Marien Fressinaud dev@marienfressinaud.fr 2013-09-14T19:59:54+00:00 urn:sha1:d9975d86a2d159c0f2442b9ee02f0523b8577af3 Merge branch 'SqlOptimisation' of https://github.com/Alkarex/FreshRSS into Alkarex-SqlOptimisation 2013-09-14T17:44:53+00:00 Marien Fressinaud dev@marienfressinaud.fr 2013-09-14T17:44:53+00:00 urn:sha1:008d6a704748f04687ecf14f587b986d2fe5eacd Conflicts: app/controllers/feedController.php Fix issue #151 : marquer lus les articles du futur 2013-09-14T17:26:01+00:00 Marien Fressinaud dev@marienfressinaud.fr 2013-09-14T17:26:01+00:00 urn:sha1:73fe0eabbe4461ba8c6d24692ca05de0ac93de11 On peut désormais marquer comme lu les articles parus dans le futur Handle paging for entries with identical date 2013-09-04T07:28:27+00:00 Alexandre Alapetite alexandre@alapetite.fr 2013-09-04T07:28:27+00:00 urn:sha1:cf62bcd3d31a801a16a06608b6a953613f68fbde Paging now works even when many entries have the same date. SQL speed could probably be improved by testing first on date, and then on CONCAT. Also, having an index on date would probably help too. Work around 2013-09-04T00:20:02+00:00 Alexandre Alapetite alexandre@alapetite.fr 2013-09-04T00:20:02+00:00 urn:sha1:040e72fe4f4d6105239dc21d68e97dc8be724dc3 In the current SQL request with LIMIT, if many dates are identical, the pagination may not work properly. Added a little more tolerance, but will have to be solved better. Disable SQL LIMIT optimisation during search 2013-09-03T21:35:33+00:00 Alexandre Alapetite alexandre@alapetite.fr 2013-09-03T21:35:33+00:00 urn:sha1:f55ae730e6477f37566722961ca7e7ab9ad7f3c5 This patch is to make search work again after the new SQL optimisations, by removing some of the optimisations when searching is used. Optimisation of search is left for some future work. The whole base is indeed transfered from MySQL to PHP, which is not good. SQL and model optimisation 2013-09-02T20:06:51+00:00 Alexandre Alapetite alexandre@alapetite.fr 2013-09-02T20:06:51+00:00 urn:sha1:f855dbdca6e95ac367b7a9dae9d3a866e1f85d37 Big effect (on speed and memory), but few changes :-) Drastically reduced the number of SQL requests needed (from 233 down to 8 to load the home page with my own data set = 140 feeds in 15 categories). Drastically reduced the amount of data transferred from MySQL to PHP. Change filter of expected SQL errors 2013-08-26T21:47:59+00:00 Alexandre Alapetite alexandre@alapetite.fr 2013-08-26T21:47:59+00:00 urn:sha1:de6e1b0a86c29845a2f8179aa0ef45d48e0d7ebb Filter out "SQLSTATE Class code 23: Constraint Violation" because of expected duplicate entries. Consider adding a Minz_Log::DEBUG level Fix issue #118 : option pour garder historique 2013-08-22T09:11:22+00:00 Marien Fressinaud dev@marienfressinaud.fr 2013-08-22T09:11:22+00:00 urn:sha1:9642fbb2211368f8063f657d34e043b311df4718 *** ATTENTION, MODIFICATION DE LA BDD *** Cette option permet de garder les vieux articles d'un flux en même s'ils sont plus vieux que la limite des X mois d'historique (3 par défaut) Les modifications de la base de données : - ajout du champ "keep_history int(1) DEFAULT 0" à la table feed - suppression des champs "is_public" et "lastUpdate" de la table entry (n'étaient plus utilisés de toute façon)