Question

在MySQL中是否存在着某种污点?

请允许我说,我们有一个桌子。

user | action  | target
-----------------------
1    | likes   | 14
2    | follows | 190

I don t want to store "action" as text, because it takes much space and is slow to index. Actions are likely to be limited (up to 50 actions, I guess) but can be added/removed in the future. So I would like to avoid storing all actions by numbers in PHP. I would like to have a table that handles this transparently.

For example, table above would be stored as (1,1,14), (2,2,190) internally, and keys would be stored in another table (1 = likes, 2 = follows).

INSERT INTO table (41, "likes", 153)

这里的“类似”已解决到1个。

INSERT INTO table (23, "dislikes", 1245)

在这里,我们对于在内部增加和储存“类似”没有任何关键。

可能吗?

Answer 1

如果你有一套固定(或合理固定)的价值观,那么你就可以使用大田。它是在内部实施的,因此需要少量磁盘空间。这里是一个例子:

CREATE TABLE enum_test (
    myEnum enum( enabled ,  disabled ,  unknown )
);

Answer 2

是的,它有这样的顺序:

INSERT INTO table (23, (SELECT id FROM actions WHERE action="dislikes") , 1245)

这样,就能够不从PHP一侧了解该数据库,而只了解行动名称,并继续将它输入数据库,作为ID。

假设你采取行动

id | action
-----------
1  | like
2  | dislike

Answer 3

You want a table called "actions", and a foreign key called "action_id". That is how database normalization works:

用户——行动:

user | action_id  | target
-----------------------
1    | 1          | 14
2    | 2          | 190

行动:

id  | name    
--------------
1   | likes   
2   | follows

在将<代码>插入用户_actions(1,类似于47)的工作中: 你们应该给予护理。努力使你坐下来,是毫无道理的;你绝不应在你的申请书中写字。数据库互动应当由一层模型/商业物体处理,其内部实施应当由您处理。

为使<代码>sert成为用户_actions(1,不相类似,47)自动在<代码>中产生新的记录<>actions。表: 这再次阻碍了数据库的工作。你们的榜样应当处理这个问题。

友情链接