11

I am trying to add an 'id' primary key column to an already existing MySQL table using alembic. I tried the following...

op.add_column('mytable', sa.Column('id', sa.Integer(), nullable=False))
op.alter_column('mytable', 'id', autoincrement=True, existing_type=sa.Integer(), existing_server_default=False, existing_nullable=False)

but got the following error

sqlalchemy.exc.OperationalError: (OperationalError) (1075, 'Incorrect table definition; there can be only one auto column and it must be defined as a key') 'ALTER TABLE mytable CHANGE id id INTEGER NOT NULL AUTO_INCREMENT' ()

looks like the sql statement generated by alembic did not add PRIMARY KEY at the end of the alter statement. Could I have missed some settings?

Thanks in advance!

SuperShoot
  • 9,880
  • 2
  • 38
  • 55
minovsky
  • 857
  • 1
  • 15
  • 28

2 Answers2

25

I spent some time digging through the alembic source code, and this doesn't seem to be supported. You can specify primary keys when creating a table, but not when adding columns. In fact, it specifically checks and won't let you (link to source):

# from alembic.operations.toimpl.add_column, line 132
for constraint in t.constraints:
    if not isinstance(constraint, sa_schema.PrimaryKeyConstraint):
        operations.impl.add_constraint(constraint)

I looked around, and adding a primary key to an existing table may result in unspecified behavior - primary keys aren't supposed to be null, so your engine may or may not create primary keys for existing rows. See this SO discussion for more info: Insert auto increment primary key to existing table

I'd just run the alter query directly, and create primary keys if you need to.

op.execute("ALTER TABLE mytable ADD id INT PRIMARY KEY AUTO_INCREMENT;")

If you really need cross-engine compatibility, the big hammer would be to (1) create a new table identical to the old one with a primary key, (2) migrate all your data, (3)delete the old table and (4) rename the new table.

Hope that helps.

SuperShoot
  • 9,880
  • 2
  • 38
  • 55
Rachel Sanders
  • 5,734
  • 1
  • 27
  • 36
2

You have to remove the primary key that is in the table and then create a new one that includes all columns that you want as the primary key.

eg. In psql use \d <table name> to define the schema, then check the primary key constraint.

Indexes:
"enrollments_pkey" PRIMARY KEY, btree (se_crs_id, se_std_id)

then use this information in alembic

def upgrade():
# ### commands auto generated by Alembic - please adjust! ###
op.add_column('enrollments', sa.Column(
    'se_semester', sa.String(length=30), nullable=False))
op.drop_constraint('enrollments_pkey', 'enrollments', type_='primary')
op.create_primary_key('enrollments_pkey', 'enrollments', [
                      'se_std_id', 'se_crs_id', 'se_semester'])

The results after running \d enrollments should be updated to

Indexes:
"enrollments_pkey" PRIMARY KEY, btree (se_std_id, se_crs_id, se_semester)

This solution worked fine for me.

Musty
  • 41
  • 4