# INSERT

# INSERT, ON DUPLICATE KEY UPDATE

INSERT INTO `table_name` 
    (`index_field`, `other_field_1`, `other_field_2`) 
    VALUES 
    ('index_value', 'insert_value', 'other_value') 
    ON DUPLICATE KEY UPDATE 
        `other_field_1` = 'update_value',
        `other_field_2` = VALUES(`other_field_2`);

This will INSERT into table_name the specified values, but if the unique key already exists, it will update the other_field_1 to have a new value.
Sometimes, when updating on duplicate key it comes in handy to use VALUES() (opens new window) in order to access the original value that was passed to the INSERT instead of setting the value directly. This way, you can set different values by using INSERT and UPDATE. See the example above where other_field_1 is set to insert_value on INSERT or to update_value on UPDATE while other_field_2 is always set to other_value.

Crucial for the Insert on Duplicate Key Update (IODKU) to work is the schema containing a unique key that will signal a duplicate clash. This unique key can be a Primary Key or not. It can be a unique key on a single column, or a multi-column (composite key).

# Inserting multiple rows

INSERT INTO `my_table` (`field_1`, `field_2`) VALUES 
    ('data_1', 'data_2'),
    ('data_1', 'data_3'),
    ('data_4', 'data_5');

This is an easy way to add several rows at once with one INSERT statement.

This kind of 'batch' insert is much faster than inserting rows one by one. Typically, inserting 100 rows in a single batch insert this way is 10 times as fast as inserting them all individually.

# Ignoring existing rows

When importing large datasets, it may be preferable under certain circumstances to skip rows that would usually cause the query to fail due to a column restraint e.g. duplicate primary keys. This can be done using INSERT IGNORE.

Consider following example database:

SELECT * FROM `people`;
--- Produces:
+----+------+
| id | name |
+----+------+
|  1 | john |
|  2 | anna |
+----+------+


INSERT IGNORE INTO `people` (`id`, `name`) VALUES
    ('2', 'anna'), --- Without the IGNORE keyword, this record would produce an error
    ('3', 'mike');

SELECT * FROM `people`;
--- Produces:
+----+--------+
| id |  name  |
+----+--------+
|  1 |  john  |
|  2 |  anna  |
|  3 |  mike  |
+----+--------+

The important thing to remember is that INSERT IGNORE will also silently skip other errors too, here is what Mysql official documentations says:

Data conversions that would trigger errors abort the statement if IGNORE is not > specified. With IGNORE, invalid values are adjusted to the closest values and >inserted; warnings are produced but the statement does not abort.

Note :- The section below is added for the sake of completeness, but is not considered best practice (this would fail, for example, if another column was added into the table).

If you specify the value of the corresponding column for all columns in the table, you can ignore the column list in the INSERT statement as follows:

INSERT INTO `my_table` VALUES 
        ('data_1', 'data_2'),
        ('data_1', 'data_3'),
        ('data_4', 'data_5');

# Basic Insert

INSERT INTO `table_name` (`field_one`, `field_two`) VALUES ('value_one', 'value_two');

In this trivial example, table_name is where the data are to be added, field_one and field_two are fields to set data against, and value_one and value_two are the data to do against field_one and field_two respectively.

It's good practice to list the fields you are inserting data into within your code, as if the table changes and new columns are added, your insert would break should they not be there

# INSERT with AUTO_INCREMENT + LAST_INSERT_ID()

When a table has an AUTO_INCREMENT PRIMARY KEY, normally one does not insert into that column. Instead, specify all the other columns, then ask what the new id was.

CREATE TABLE t (
    id SMALLINT UNSIGNED AUTO_INCREMENT NOT NULL,
    this ...,
    that ...,
    PRIMARY KEY(id) );

INSERT INTO t (this, that) VALUES (..., ...);
SELECT LAST_INSERT_ID() INTO @id;
INSERT INTO another_table (..., t_id, ...) VALUES (..., @id, ...);

Note that LAST_INSERT_ID() is tied to the session, so even if multiple connections are inserting into the same table, each with get its own id.

Your client API probably has an alternative way of getting the LAST_INSERT_ID() without actually performing a SELECT and handing the value back to the client instead of leaving it in an @variable inside MySQL. Such is usually preferable.

Longer, more detailed, example

The "normal" usage of IODKU is to trigger "duplicate key" based on some UNIQUE key, not the AUTO_INCREMENT PRIMARY KEY. The following demonstrates such. Note that it does not supply the id in the INSERT.

Setup for examples to follow:

CREATE TABLE iodku (
    id INT AUTO_INCREMENT NOT NULL,
    name VARCHAR(99) NOT NULL,
    misc INT NOT NULL,
    PRIMARY KEY(id),
    UNIQUE(name)
) ENGINE=InnoDB;

INSERT INTO iodku (name, misc)
    VALUES
    ('Leslie', 123),
    ('Sally', 456);
Query OK, 2 rows affected (0.00 sec)
Records: 2  Duplicates: 0  Warnings: 0
+----+--------+------+
| id | name   | misc |
+----+--------+------+
|  1 | Leslie |  123 |
|  2 | Sally  |  456 |
+----+--------+------+

The case of IODKU performing an "update" and LAST_INSERT_ID() retrieving the relevant id:

INSERT INTO iodku (name, misc)
    VALUES
    ('Sally', 3333)            -- should update
    ON DUPLICATE KEY UPDATE    -- `name` will trigger "duplicate key"
    id = LAST_INSERT_ID(id),
    misc = VALUES(misc);
SELECT LAST_INSERT_ID();       -- picking up existing value
+------------------+
| LAST_INSERT_ID() |
+------------------+
|                2 |
+------------------+

The case where IODKU performs an "insert" and LAST_INSERT_ID() retrieves the new id:

INSERT INTO iodku (name, misc)
    VALUES
    ('Dana', 789)          -- Should insert
    ON DUPLICATE KEY UPDATE
    id = LAST_INSERT_ID(id),
    misc = VALUES(misc);
SELECT LAST_INSERT_ID();   -- picking up new value
+------------------+
| LAST_INSERT_ID() |
+------------------+
|                3 |
+------------------+

Resulting table contents:

SELECT * FROM iodku;
+----+--------+------+
| id | name   | misc |
+----+--------+------+
|  1 | Leslie |  123 |
|  2 | Sally  | 3333 |  -- IODKU changed this
|  3 | Dana   |  789 |  -- IODKU added this
+----+--------+------+

# INSERT SELECT (Inserting data from another Table)

This is the basic way to insert data from another table with the SELECT statement.

INSERT INTO `tableA` (`field_one`, `field_two`) 
   SELECT `tableB`.`field_one`, `tableB`.`field_two` 
   FROM `tableB` 
   WHERE `tableB`.clmn <> 'someValue'
   ORDER BY `tableB`.`sorting_clmn`;

You can SELECT * FROM, but then tableA and tableB must have matching column count and corresponding datatypes.

Columns with AUTO_INCREMENT are treated as in the INSERT with VALUES clause.

This syntax makes it easy to fill (temporary) tables with data from other tables, even more so when the data is to be filtered on the insert.

# Lost AUTO_INCREMENT ids

Several 'insert' functions can "burn" ids. Here is an example, using InnoDB (other Engines may work differently):

CREATE TABLE Burn (
    id SMALLINT UNSIGNED AUTO_INCREMENT NOT NULL,
    name VARCHAR(99) NOT NULL,
    PRIMARY KEY(id),
    UNIQUE(name)
        ) ENGINE=InnoDB;

INSERT IGNORE INTO Burn (name) VALUES ('first'), ('second');
SELECT LAST_INSERT_ID();          -- 1
SELECT * FROM Burn ORDER BY id;
  +----+--------+
  |  1 | first  |
  |  2 | second |
  +----+--------+

INSERT IGNORE INTO Burn (name) VALUES ('second');  -- dup 'IGNOREd', but id=3 is burned
SELECT LAST_INSERT_ID();          -- Still "1" -- can't trust in this situation
SELECT * FROM Burn ORDER BY id;
  +----+--------+
  |  1 | first  |
  |  2 | second |
  +----+--------+

INSERT IGNORE INTO Burn (name) VALUES ('third');
SELECT LAST_INSERT_ID();           -- now "4"
SELECT * FROM Burn ORDER BY id;    -- note that id=3 was skipped over
  +----+--------+
  |  1 | first  |
  |  2 | second |
  |  4 | third  |    -- notice that id=3 has been 'burned'
  +----+--------+

Think of it (roughly) this way: First the insert looks to see how many rows might be inserted. Then grab that many values from the auto_increment for that table. Finally, insert the rows, using ids as needed, and burning any left overs.

The only time the leftover are recoverable is if the system is shutdown and restarted. On restart, effectively MAX(id) is performed. This may reuse ids that were burned or that were freed up by DELETEs of the highest id(s).

Essentially any flavor of INSERT (including REPLACE, which is DELETE + INSERT) can burn ids. In InnoDB, the global (not session!) variable innodb_autoinc_lock_mode (opens new window) can be used to control some of what is going on.

When "normalizing" long strings into an AUTO INCREMENT id, burning can easily happen. This could lead to overflowing the size of the INT you chose.

# Syntax

INSERT [LOW_PRIORITY | DELAYED | HIGH_PRIORITY] [IGNORE] [INTO] tbl_name [PARTITION (partition_name,...)] [(col_name,...)] {VALUES | VALUE} ({expr | DEFAULT},...),(...),... [ ON DUPLICATE KEY UPDATE col_name=expr [, col_name=expr] ... ]

INSERT [LOW_PRIORITY | DELAYED | HIGH_PRIORITY] [IGNORE] [INTO] tbl_name [PARTITION (partition_name,...)] SET col_name={expr | DEFAULT}, ... [ ON DUPLICATE KEY UPDATE col_name=expr [, col_name=expr] ... ]

INSERT [LOW_PRIORITY | HIGH_PRIORITY] [IGNORE] [INTO] tbl_name [PARTITION (partition_name,...)] [(col_name,...)] SELECT ... [ ON DUPLICATE KEY UPDATE col_name=expr [, col_name=expr] ... ]

An expression expr can refer to any column that was set earlier in a value list. For example, you can do this because the value for col2 refers to col1, which has previously been assigned:
INSERT INTO tbl_name (col1,col2) VALUES(15,col1*2);

INSERT statements that use VALUES syntax can insert multiple rows. To do this, include multiple lists of column values, each enclosed within parentheses and separated by commas. Example:
INSERT INTO tbl_name (a,b,c) VALUES(1,2,3),(4,5,6),(7,8,9);

The values list for each row must be enclosed within parentheses. The following statement is illegal because the number of values in the list does not match the number of column names:
INSERT INTO tbl_name (a,b,c) VALUES(1,2,3,4,5,6,7,8,9);

[INSERT ... SELECT Syntax](http://dev.mysql.com/doc/refman/5.7/en/insert-select.html)
INSERT [LOW_PRIORITY | HIGH_PRIORITY] [IGNORE] [INTO] tbl_name [PARTITION (partition_name,...)] [(col_name,...)] SELECT ... [ ON DUPLICATE KEY UPDATE col_name=expr, ... ]

With INSERT ... SELECT, you can quickly insert many rows into a table from one or many tables. For example:
INSERT INTO tbl_temp2 (fld_id) SELECT tbl_temp1.fld_order_id FROM tbl_temp1 WHERE tbl_temp1.fld_order_id > 100;

# Remarks

Official INSERT Syntax (opens new window)

← Comment Mysql DELETE →