我有一个存储3k记录的表,
我是Entity Framework和ORM的新手,但我能理解出错了。
当我运行这个linq查询时:
repo.GetQuery<Article>().Where(foo=>foo.Expires_date>=date
&& foo.Istoparticle==true
&& foo.Isfrontpage==false)
.OrderByDescending(foo=>foo.Date)
.Take(4);
我在mysql端得到了这个查询:
SELECT `Project1`.`Id`, `Project1`.`User_id`, `Project1`.`Category_id`, `Project1`.`Title`, `Project1`.`Value`, `Project1`.`Keywords`, `Project1`.`Description`, `Project1`.`Images`, `Project1`.`Votes`, `Project1`.`Views`, `Project1`.`Isvisible`, `Project1`.`Isfrontpage`, `Project1`.`Istoparticle`, `Project1`.`Date`, `Project1`.`Expires_date`, `Project1`.`Votes_sum` FROM (SELECT `Extent1`.`Id`, `Extent1`.`User_id`, `Extent1`.`Category_id`, `Extent1`.`Title`, `Extent1`.`Value`, `Extent1`.`Keywords`, `Extent1`.`Description`, `Extent1`.`Images`, `Extent1`.`Votes`, `Extent1`.`Votes_sum`, `Extent1`.`Views`, `Extent1`.`Isvisible`, `Extent1`.`Isfrontpage`, `Extent1`.`Istoparticle`, `Extent1`.`Expires_date`, `Extent1`.`Date` FROM `tcms_articles` AS `Extent1` WHERE `Extent1`.`Expires_date` >= '2012-06-24 13:41:47.816') AS `Project1` ORDER BY `Project1`.`Date` DESC LIMIT 4
执行此查询大约需要3.50秒。
解释此查询:
+----+-------------+------------+-------+---------------+--------------+---------+------+------+----------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+------------+-------+---------------+--------------+---------+------+------+----------------+ | 1 | PRIMARY | | ALL | NULL | NULL | NULL | NULL | 4054 | Using filesort | | 2 | DERIVED | Extent1 | range | expires_date | expires_date | 8 | NULL | 4053 | Using where | +----+-------------+------------+-------+---------------+--------------+---------+------+------+----------------+
当我查询时:
SELECT *
FROM tcms_articles
WHERE expires_date >= '2012-06-24 13:41:47.816'
ORDER BY date DESC
limit 4
我得到0.01秒......
再次运行解释我得到:
+----+-------------+---------------+-------+---------------+------+---------+------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+---------------+-------+---------------+------+---------+------+------+-------------+ | 1 | SIMPLE | tcms_articles | index | expires_date | date | 8 | NULL | 11 | Using where | +----+-------------+---------------+-------+---------------+------+---------+------+------+-------------+
我不明白为什么会这样。
实体框架4.3 MySQL Connector Net 6.5.4.0
tcms_articles:
CREATE TABLE `tcms_articles` ( `Id` int(10) unsigned NOT NULL AUTO_INCREMENT, `User_id` int(10) unsigned DEFAULT NULL, `Category_id` int(10) unsigned DEFAULT NULL, `Title` varchar(255) DEFAULT NULL, `Value` longtext, `Keywords` varchar(255) NOT NULL DEFAULT '', `Description` varchar(255) NOT NULL DEFAULT '', `Images` longtext NOT NULL, `Votes` int(10) unsigned NOT NULL DEFAULT '1', `Votes_sum` int(10) unsigned NOT NULL DEFAULT '5', `Views` int(10) unsigned NOT NULL DEFAULT '0', `Isvisible` tinyint(1) unsigned NOT NULL DEFAULT '1', `Isfrontpage` tinyint(1) unsigned NOT NULL DEFAULT '0', `Istoparticle` tinyint(1) unsigned NOT NULL DEFAULT '1', `Expires_date` datetime NOT NULL DEFAULT '2099-12-31 00:00:00', `Date` datetime NOT NULL, PRIMARY KEY (`Id`), KEY `article_users` (`User_id`) USING BTREE, KEY `article_section` (`Category_id`) USING BTREE, KEY `Isvisible_index2` (`Isvisible`) USING BTREE, KEY `Istoparticle_index2` (`Istoparticle`) USING BTREE, KEY `Expires_date_index2` (`Expires_date`) USING BTREE, KEY `isfrontpage2` (`Isfrontpage`) USING BTREE, KEY `Date_index2` (`Date`) USING BTREE, CONSTRAINT `tcms_articles_ibfk_1` FOREIGN KEY (`Category_id`) REFERENCES `tcms_categories` (`Id`) ON DELETE CASCADE ON UPDATE CASCADE, CONSTRAINT `tcms_articles_ibfk_2` FOREIGN KEY (`User_id`) REFERENCES `tcms_users` (`Id`) ON DELETE CASCADE ON UPDATE CASCADE ) ENGINE=InnoDB AUTO_INCREMENT=80 DEFAULT CHARSET=utf8;
那么为什么Linq产生这个查询以及如何解决这个问题?
答案 0 :(得分:0)
repo.GetQuery<Article>().Where(foo=>foo.Expires_date>=date -- Note 1
&& foo.Istoparticle==true
&& foo.Isfrontpage==false)
.OrderByDescending(foo=>foo.Date) -- Note 2
.Take(4);
在这两个地方使用foo.Expires_date
。