存储过程的行为与常规查询不同

Question

我遇到了一个问题，我执行了一个存储过程，该过程对数据库中的几个表进行更新。错误的发生使得每次只有一个字段（表示相同的记录id）更新为新的id。这是spoc:

DROP PROCEDURE IF EXISTS feeds_transfer;
DELIMITER $$
CREATE PROCEDURE feeds_transfer(IN original_owner INT, IN new_owner INT, IN feed_id INT)
    BEGIN
        UPDATE `events` SET `user_id` = new_owner WHERE `user_id` = original_owner AND `feed_id` = feed_id;
        UPDATE `feeds` SET `partner_id` = new_owner WHERE `partner_id` = original_owner AND `id` = feed_id;
    UPDATE `ics_uploads` SET `user_id` = new_owner WHERE `user_id` = original_owner AND `context_type` = 'feed' AND `context_id` = feed_id;
    UPDATE `images` SET `user_id` = new_owner WHERE `user_id` = original_owner AND `context` = 'feed' AND `context_id` = feed_id;
    UPDATE `private_feed_invitees` SET `user_id` = new_owner WHERE `user_id` = original_owner AND `feed_id` = feed_id;
        UPDATE `subscribed_feeds` SET `user_id` = new_owner WHERE `user_id` = original_owner AND `feed_id` = feed_id;
    END $$
DELIMITER ;

结果是subscribed_feeds表将更新user_id=original_owner的每条记录。它就像它忽略feed_id=feed_id 一样

Answer 1

好吧，这里发生的事情并不是因为feed_id=feed_id子句被忽略，而是因为它总是正确的，所以它总是触发比预期更多的更新。这个故事的寓意是，您永远不应该将存储过程的输入变量命名为与您计划对其进行逻辑检查的任何字段相同的名称。