This script inspects a SQL database to find foreign keys referring to columns that do not have a unique key. For example:
CREATE TABLE parent (
id INT
);
CREATE TABLE child (
parent_id INT,
FOREIGN KEY(parent_id) REFERENCES parent(id)
);
Here, the key declared on the parent_id
column refers to a non-unique column.
This, according to MySQL documentation, may cause undefined behavior:
The handling of foreign key references to nonunique keys or keys that contain NULL values is not well defined for operations such as UPDATE or DELETE CASCADE. You are advised to use foreign keys that reference only UNIQUE (including PRIMARY) and NOT NULL keys.
Note: this particular example is rather obvious and somewhat contrived, but there may be cases where the problematic key is harder to find (for example, when the unique key on the referred column evolved into a multi-column key, but analogous change was not applied to the child table).
Clone this repository and make sure you have Python 3.x and SQLAlchemy installed:
pip3 install -r requirements.txt
You will also need to install a SQLAlchemy dialect suitable for your database:
pip3 install mysqlclient
~/path/sqlnufk.py mysql://bar_user:barP455word@host.bar/bar_database
For MySQL databases you can also use an option group name instead of the
database URL. This works similar to the --defaults-group-suffix
parameter of
the mysql
command line utility. For example, if you have the following
options saved in your ~/.my.cnf
file:
[client_foo]
host=host.bar
database=bar_database
user=bar_user
password=barP455word
then you can run:
~/path/sqlnufk.py foo
The script will exit with code 0 if no problematic foreign keys are found. If problematic keys are found, they will be printed on standard output and the script will exit with code 64. Any other exit code means that the script could not check the keys (connection problem, driver / dialect not installed, invalid URL etc.)