DROP CASCADE in Sql Server

Question

I'm running a South migration in a Django project that uses Sql Server and pyodbc. This is backwards migration so the South is trying to delete a few of my tables.

The South executes the following method in order to drop the tables:

def delete_table(self, table_name, cascade=True):
    """
    Deletes the table 'table_name'.
    """
    params = (self.quote_name(table_name), )
    if cascade:
        self.execute('DROP TABLE %s CASCADE;' % params)
    else:
        self.execute('DROP TABLE %s;' % params)

drop_table = alias('delete_table')

The problem is that the Sql Server does not support cascade drops, so the migration fails with the following error:

pyodbc.ProgrammingError: ('42000', "[42000] [Microsoft][ODBC SQL Server Driver][SQL Server]Incorrect syntax near the keyword 'CASCADE'. (156) (SQLExecDirectW)")

I want to write to a patch for the South so that it works with the Sql Server. What would be the best way to simulate a DROP CASCADE? I think that both solutions using python or pure SQL would be valid.

Answer 1

I never found an implementation of the DROP CASCADE in python, so I ended up writing my own. Here is how it looks:

def delete_table(self, table_name, cascade=True):
    """
    Deletes the table 'table_name'.
    """
    params = (self.quote_name(table_name), )
    if cascade:
        conn = self._get_connection()

        # Get a list of related tables
        sql = "SELECT T1.TABLE_NAME \
                 FROM INFORMATION_SCHEMA.CONSTRAINT_TABLE_USAGE AS T1 \
                 JOIN SYS.FOREIGN_KEYS AS F \
                   ON (F.parent_object_id = OBJECT_ID(N'{0}') OR \
                      F.referenced_object_id = OBJECT_ID(N'{0}')) AND \
                      T1.CONSTRAINT_NAME = OBJECT_NAME(F.object_id)"

        related_tables = self.execute(sql.format(params[0]))

        # Drop all the constraints
        constraints = self.execute("SELECT CONSTRAINT_NAME \
                                    FROM INFORMATION_SCHEMA.TABLE_CONSTRAINTS \
                                    WHERE TABLE_NAME='{0}' AND \
                                    CONSTRAINT_TYPE='FOREIGN KEY';".format(table_name))

        sql = "ALTER TABLE {0} DROP CONSTRAINT {1};"
        for constraint in constraints:
            self.execute(sql.format(params[0], constraint[0]))

        for table in related_tables:
            self.delete_table(table[0], cascade)

        sql = "IF  EXISTS (SELECT * \
                           FROM sys.objects \
                           WHERE object_id = OBJECT_ID(N'{0}') AND \
                           type in (N'U')) \
               DROP TABLE {0}"

        self.execute(sql.format(params[0]))
    else:
        self.execute('DROP TABLE %s;' % params)

Answer 2

What do you mean by "simulate" a DROP CASCADE?

If it means "ignore the cascade parameter for MSSQL" then you can just check the current SQL platform/dialect in use and do whatever you like (I have no idea how/if that's possible). The sqlalchemy dialects implementation might give you some useful ideas, if you need them.

But if you want to actually implement the functionality, you'll have query the system views to build a list of tables to drop and the correct order to drop them in. The documentation for the INFORMATION_SCHEMA views or sys.foreign_keys should help. Once you have a query to get the dependent tables in the right order, you can patch the function to do the actual DROPs.