如何处理数据库中的重复？

Question

在一个程序中，我们是否应该使用try catch来检查表中是否存在重复值，或者我们应该检查表中是否已存在该值并避免插入？

Answer 1

This is easy enough to enforce with a UNIQUE constraint on the database side so that's my recommendation. I try to put as much of the data integrity into the database so that I can avoid having bad data (although sometimes unavoidable).

If this is how you already have it you might as well just catch the mysql exception for duplicate value insertion on such a table as doing the check then the insertion is more costly then having the database do one simple lookup (and possibly an insert).

Answer 2

Depends upon whether you are inserting one, or a million, as well as whether the duplicate is the primary key.

If its the primary key, read: http://database-programmer.blogspot.com/2009/06/approaches-to-upsert.html

An UPSERT or ON DUPLICATE KEY... The idea behind an UPSERT is simple. The client issues an INSERT command. If a row already exists with the given primary key, then instead of throwing a key violation error, it takes the non-key values and updates the row.

This is one of those strange (and very unusual) cases where MySQL actually supports something you will not find in all of the other more mature databases. So if you are using MySQL, you do not need to do anything special to make an UPSERT. You just add the term "ON DUPLICATE KEY UPDATE" to the INSERT statement:

If it's not the primary key, and you are inserting just one row, then you can still make sure this doesn't cause a failure.

For your actual question, I don't really like the idea of using try/catch for program flow, but really, you have to evaluate readability and user experience (in this case performance), and pick what you think is the best of mix of the two.

Answer 3

You can add a UNIQUE constraint to your table.. Something like

CREATE TABLE IF NOT EXISTS login
(
    loginid SMALLINT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY,
    loginname CHAR(20) NOT NULL,
    UNIQUE (loginname) 
);

This will ensure no two login names are the same.

Answer 4

您可以创建唯一的复合键

ALTER TABLE `TableName` ADD UNIQUE KEY (KeyOne, KeyTwo, ...);

Answer 5

您只需在表中创建一个唯一键，以便它不允许再次添加相同的值。

Answer 6

You should try inserting the value and catch the exception. In a busy system, if you check for the existience of a value it might get inserted between the time you check and the time you insert it.

Let the database do it's job, let the database check for the duplicate entry.

Answer 7

A database is a computerized representation of a set of business rules and a DBMS is used to enforce these business rules as constraints. Neither can verify a proposition in the database is true in the real world. For example, if the model in question is the employees of an enterprise and the Employees table contains two people named 'Jimmy Barnes' DBMS (nor the database) cannot know whether one is a duplicate, whether either are real people, etc. A trusted source is required to determine existence and identity. In the above example, the enterprise's personnel department is responsible for checking public records, perusing references, ensuring the person is not already on the payroll, etc then allocating an unique employee reference number that can be used as a key. This is why we look for industry-standard identifiers with a trusted source: ISBN for books, VIN for cars, ISO 4217 for currencies, ISO 3166 for countries, etc.

Answer 8

I think it is better to check if the value already exists and avoid the insertion. The check for duplicate values can be done in the procedure that saves the data (using exists if your database is an SQL database).

If a duplicate exists you avoid the insertion and can return a value to your app indicating so and then show a message accordingly.

For example, a piece of SQL code could be something like this:

    select @ret_val = 0
   If exists (select * from employee where last_name = @param_ln and first_name = @param_fn)
       select @ret_val = -1
    Else
       -- your insert statement here

   Select @ret_val

Your condition for duplicate values will depend on what you define as a duplicate record. In your application you would use the return value to know if the data was a duplicate. Good luck!