def drop_old_duplicate_entries_from_table(migrate_engine, table_name,
use_soft_delete, *uc_column_names):
"""Drop all old rows having the same values for columns in uc_columns.
This method drop (or mark ad `deleted` if use_soft_delete is True) old
duplicate rows form table with name `table_name`.
:param migrate_engine: Sqlalchemy engine
:param table_name: Table with duplicates
:param use_soft_delete: If True - values will be marked as `deleted`,
if False - values will be removed from table
:param uc_column_names: Unique constraint columns
"""
meta = MetaData()
meta.bind = migrate_engine
table = Table(table_name, meta, autoload=True)
columns_for_group_by = [table.c[name] for name in uc_column_names]
columns_for_select = [func.max(table.c.id)]
columns_for_select.extend(columns_for_group_by)
duplicated_rows_select = select(columns_for_select,
group_by=columns_for_group_by,
having=func.count(table.c.id) > 1)
for row in migrate_engine.execute(duplicated_rows_select):
# NOTE(boris-42): Do not remove row that has the biggest ID.
delete_condition = table.c.id != row[0]
is_none = None # workaround for pyflakes
delete_condition &= table.c.deleted_at == is_none
for name in uc_column_names:
delete_condition &= table.c[name] == row[name]
rows_to_delete_select = select([table.c.id]).where(delete_condition)
for row in migrate_engine.execute(rows_to_delete_select).fetchall():
LOG.info(_("Deleting duplicated row with id: %(id)s from table: "
"%(table)s") % dict(id=row[0], table=table_name))
if use_soft_delete:
delete_statement = table.update().\
where(delete_condition).\
values({
'deleted': literal_column('id'),
'updated_at': literal_column('updated_at'),
'deleted_at': timeutils.utcnow()
})
else:
delete_statement = table.delete().where(delete_condition)
migrate_engine.execute(delete_statement)
def soft_delete(self, session=None):
"""Mark this object as deleted."""
self.deleted = self.id
self.deleted_at = timeutils.utcnow()
self.save(session=session)
def soft_delete(self, synchronize_session='evaluate'):
return self.update({'deleted': literal_column('id'),
'updated_at': literal_column('updated_at'),
'deleted_at': timeutils.utcnow()},
synchronize_session=synchronize_session)