以下是我的情况：
• I have a table of users <br><br>
• I have a table of movies <br><br>
• I have a table of user_movies that associates users to movies that stores a frequently changing "score" column on each record.<br><br>
• I need to be able to bulk update the user_movies table for speed and to do so I believe I need a unique index to be able to pull off bulk updates like this.<br><br>
• Considering each user_movie combination is unique, would it be possible to do
将(User_id--->) uid_mid (<----movie_id)作为唯一标识符？如果是这样，怎么会将现有的表转换成这个呢？
答案 0 :(得分：0)
我想认为有一种方法可以为您的问题使用双列索引，但是......
我想回答原来的问题并假设你真的想将这些列合并为一个，我建议考虑以下内容：
ALTER TABLE user_movies ADD COLUMN new_id VARCHAR(250);
UPDATE user_moves SET new_id=CONCAT(CAST(user_id as CHAR(125)), LPAD(CAST(move_id as CHAR(125)),125,'0'));
CREATE UNIQUE INDEX myindex ON user_movies(new_id);
数字250和125当然是任意的，但是你想要它们足够大以适应你的实际ID。我填充第二个ID以确保它们是真正独特的（因为我们不希望user_id = 55 / movie_id = 66获得与user_id = 5 / movie_id = 566相同的ID）。