op.add_column() and op.drop_column() should check for SchemaType, run through the events for that type, or perhaps just create/drop

Migrated issue, originally created by Wichert Akkerman (@wichert)

I have a data model which has a simple enum column:

    size_type = schema.Column(
        types.Enum('normal', 'plus', name='size_type_type'),
        nullable=False, default='normal')

as part of an upgrade routine I need to add that column if it is missing:

    alembic.add_column('article_variant',
            ArticleVariant.__table__.c['size_type'].copy())

alembic is an operation instance here. This operation breaks due to a missing check if the enum type already exists. The resulting error is:

sqlalchemy.exc.ProgrammingError: (ProgrammingError) type "size_type_type" does not exist
LINE 1: ALTER TABLE article_variant ADD COLUMN size_type size_type_t...
                                                         ^
 'ALTER TABLE article_variant ADD COLUMN size_type size_type_type NOT NULL' {}

Start at the op.add_column() and op.drop_column() entry points and inspect how a copied SQLAlchemy Enum column is processed. Reproduce the shown migration with the size_type SchemaType, then verify that the operation handles the type's lifecycle consistently without the missing-type error.