oh right, you can filter them out. Use include_object:
http://alembic.readthedocs.org/en/latest/api.html#alembic.environment.EnvironmentContext.configure.params.include_object
if type == ‘foreign_key_constraint’: return False
dewey <de...@pathoz.com> wrote:
> Thanks for looking at it.
>
> Is there a simple way I can tell Alembic to ignore foreign keys altogether
> and i can manage those manually??
> I really prefer that to hand-coding all of my field names.
>
> Thanks for any suggestions.
> Dewey
>
>
>
> On Friday, January 2, 2015 3:13:07 PM UTC-6, Michael Bayer wrote:
> dewey <de...@pathoz.com <javascript:>> wrote:
>
>> Ok, I've confirmed the column rename event is called PRIOR to the Alembic
>> analysis of the models.
>>
>> I'm not sure how to interpret all this, but I suspect the output below will
>> tell us what's going on.
>>
>> I added this line at the end of your "class
>> _fk_constraint_sig(_constraint_sig):" constructor:
>
> first problem is that the render is wrong. the name we set is not present in
> the constraints.
>
> starting with:
>
> Table(
> 't1', target_metadata,
> Column('id', Integer, primary_key=True),
> Column('foo', String)
> )
>
> Table(
> 't2', target_metadata,
> Column('id', Integer, primary_key=True),
> Column('t1id', Integer, ForeignKey('t1.id <http://t1.id/>'))
> )
>
> it auto generates:
>
> op.create_table('t2',
> sa.Column(u't2_id', sa.Integer(), nullable=False),
> sa.Column(u't2_t1id', sa.Integer(), nullable=True),
> sa.ForeignKeyConstraint(['t1id'], ['t1.id <http://t1.id/>'], ),
> sa.PrimaryKeyConstraint('id')
> )
>
> that’s wrong. It isn’t putting our .key in the Column objects, so the PK
> and FK’s point to nothing, because these are rendered as pointing to .key.
> The bug here is that if we aren’t rendering .key in Column (which I think we
> shouldn’t, for reasons below), we need to render the constraints as referring
> to the columns by name, not key, which can be different. So this is issue
> https://bitbucket.org/zzzeek/alembic/issue/259/autogen-erroneously-gens-constraints-using
>
> <https://bitbucket.org/zzzeek/alembic/issue/259/autogen-erroneously-gens-constraints-using>.
>
> If alembic rendered the .key in the migration file like this:
>
> op.create_table('t1',
> sa.Column(u't1_id', sa.Integer(), key='id', nullable=False),
> sa.Column(u't1_foo', sa.String(), key='foo', nullable=True),
> sa.PrimaryKeyConstraint('id')
> )
>
> op.create_table('t2',
> sa.Column(u't2_id', sa.Integer(), key='id', nullable=False),
> sa.Column(u't2_t1id', sa.Integer(), key='t1id', nullable=True),
> sa.ForeignKeyConstraint(['t1id'], ['t1.id <http://t1.id/>'], ),
> sa.PrimaryKeyConstraint('id')
> )
>
> that breaks; because there is no MetaData object shared between t1 and t2 in
> the above context, Alembic has no way to locate ‘t1.id <http://t1.id/>’ when
> it is onto ‘t2’, so that raises. Hence I think it’s best that migration
> scripts be generated fully on column names without keys.
>
> So we can do that in our migration:
>
> op.create_table('t1',
> sa.Column(u't1_id', sa.Integer(), nullable=False),
> sa.Column(u't1_foo', sa.String(), nullable=True),
> sa.PrimaryKeyConstraint('t1_id')
> )
> op.create_table('t2',
> sa.Column(u't2_id', sa.Integer(), nullable=False),
> sa.Column(u't2_t1id', sa.Integer(), nullable=True),
> sa.ForeignKeyConstraint(['t2_t1id'], ['t1.t1_id'], ),
> sa.PrimaryKeyConstraint('t2_id')
> )
>
> somehow, you got past all of this part of it.
>
> The second issue is that the FK comparison is done using column.key from the
> reflected table. But column.key here has the table name added to it which
> does not match our local metadata. that is a bug, which is
> https://bitbucket.org/zzzeek/alembic/issue/260/autogenerate-fk-comparison-and-uniques
>
> <https://bitbucket.org/zzzeek/alembic/issue/260/autogenerate-fk-comparison-and-uniques>.
>
> Some attempts to work around the second issue haven’t succeeded.
>
>
> So short answer is, your use case here is not supported right now; even if
> you name the columns up front with the naming convention, assuming you still
> want the .key on the old name, the second issue here is still going to come
> up.
>
> These two issues are next on my list for alembic, particularly the second one
> as it is an issue in the new FK feature.
>
>
>
>
>> print 'name: %s; const: %s; cols: %s' % (self.name <http://self.name/>,
>> self.const, str(self.source_columns))
>>
>> And I got this output:
>> name: fk_typ_type_typ_stp_bits_stp_super_type; const:
>> ForeignKeyConstraint([Column(u'typ_stp_bits', Integer(),
>> ForeignKey(u'stp_super_type.bit'), table=<typ_type>, key='stp_bits')], None,
>> name='fk_typ_type_typ_stp_bits_stp_super_type', table=Table('typ_type',
>> MetaData(bind=None), Column(u'typ_mod_dttm', DateTime(), table=<typ_type>,
>> key='mod_dttm', nullable=False,
>> onupdate=ColumnDefault(<sqlalchemy.sql.functions.now at 0x1046e8310; now>),
>> default=ColumnDefault(<sqlalchemy.sql.functions.now at 0x1046dfb90; now>)),
>> Column(u'typ_encode', SMALLINT(), table=<typ_type>, key='encode',
>> nullable=False, default=ColumnDefault(0)), Column(u'typ_id', Integer(),
>> table=<typ_type>, key='id', primary_key=True, nullable=False),
>> Column(u'typ_name', String(length=50), table=<typ_type>, key='name',
>> nullable=False), Column(u'typ_stp_bits', Integer(),
>> ForeignKey(u'stp_super_type.bit'), table=<typ_type>, key='stp_bits'),
>> Column(u'typ_order', Integer(), table=<typ_type>, key='order',
>> nullable=False), schema=None)); cols: ['stp_bits']
>>
>> name: fk_typ_type_typ_stp_bits_stp_super_type; const:
>> ForeignKeyConstraint([Column('typ_stp_bits', INTEGER(display_width=11),
>> ForeignKey(u'stp_super_type.stp_bit'),
>> ForeignKey(u'stp_super_type.stp_bit'), table=<typ_type>)], None,
>> name=u'fk_typ_type_typ_stp_bits_stp_super_type', table=Table('typ_type',
>> MetaData(bind=None), Column('typ_mod_dttm', DATETIME(), table=<typ_type>,
>> nullable=False), Column('typ_encode', SMALLINT(display_width=6),
>> table=<typ_type>, nullable=False), Column('typ_id',
>> INTEGER(display_width=11), table=<typ_type>, primary_key=True,
>> nullable=False), Column('typ_name', VARCHAR(length=50), table=<typ_type>,
>> nullable=False), Column('typ_stp_bits', INTEGER(display_width=11),
>> ForeignKey(u'stp_super_type.stp_bit'),
>> ForeignKey(u'stp_super_type.stp_bit'), table=<typ_type>),
>> Column('typ_order', INTEGER(display_width=11), table=<typ_type>,
>> nullable=False), schema=None)); cols: ['typ_stp_bits']
>>
>> fkey is named fk_typ_type_typ_stp_bits_stp_super_type
>>
>> INFO [alembic.autogenerate.compare] Detected removed foreign key
>> (typ_stp_bits)(stp_bit) on table typ_type
>>
>> INFO [alembic.autogenerate.compare] Detected added foreign key
>> (stp_bits)(stp_bit) on table typ_type
>>
>>
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "sqlalchemy" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to sqlalchemy+...@googlegroups.com <javascript:>.
>> To post to this group, send email to sqlal...@googlegroups.com <javascript:>.
>> Visit this group at http://groups.google.com/group/sqlalchemy
>> <http://groups.google.com/group/sqlalchemy>.
>> For more options, visit https://groups.google.com/d/optout
>> <https://groups.google.com/d/optout>.
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "sqlalchemy" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sqlalchemy+unsubscr...@googlegroups.com
> <mailto:sqlalchemy+unsubscr...@googlegroups.com>.
> To post to this group, send email to sqlalchemy@googlegroups.com
> <mailto:sqlalchemy@googlegroups.com>.
> Visit this group at http://groups.google.com/group/sqlalchemy
> <http://groups.google.com/group/sqlalchemy>.
> For more options, visit https://groups.google.com/d/optout
> <https://groups.google.com/d/optout>.
--
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sqlalchemy+unsubscr...@googlegroups.com.
To post to this group, send email to sqlalchemy@googlegroups.com.
Visit this group at http://groups.google.com/group/sqlalchemy.
For more options, visit https://groups.google.com/d/optout.
