SQLAlchemyを使用したPostgresの列挙型の配列

SqlAlchemyのソースコードで素晴らしい回避策を見つけました：

import re

from sqlalchemy import TypeDecorator, cast
from sqlalchemy.dialects.postgresql import ARRAY


class ArrayOfEnum(TypeDecorator):

    impl = ARRAY

    def bind_expression(self, bindvalue):
        return cast(bindvalue, self)

    def result_processor(self, dialect, coltype):
        super_rp = super(ArrayOfEnum, self).result_processor(dialect, coltype)

        def handle_raw_string(value):
            inner = re.match(r"^{(.*)}$", value).group(1)

            return inner.split(",") if inner else []

        def process(value):
            if value is None:
                return None

            return super_rp(handle_raw_string(value))

        return process

そして今：

achievements = Column(ArrayOfEnum(Enum(AchievementsType)))

そして：

career.achievements = [AchievementsType.world, AchievementsType.local]