Flask Migrate执行upgrade失败后无法执行downgrade的问题求助

大家好！我最近跟着Miguel Grinberg的Flask Mega Tutorial开发Flask应用，修改数据库模型的时候踩了大坑，现在flask db upgrade执行失败，连回滚用的flask db downgrade也跑不了，实在没辙了来求助各位！

问题经过 #

最开始我的Post模型是这样的：

class Post(db.Model):
    id: so.Mapped[int] = so.mapped_column(primary_key=True)
    body: so.Mapped[str] = so.mapped_column(sa.String(140))
    timestamp: so.Mapped[datetime] = so.mapped_column(
        index=True, default=lambda: datetime.now(timezone.utc))
    user_id: so.Mapped[int] = so.mapped_column(sa.ForeignKey(User.id), index=True)
    school_id: so.Mapped[Optional[int]] = so.mapped_column(sa.ForeignKey(School.id))
    
    author: so.Mapped['User'] = so.relationship(back_populates='posts')
    school: so.Mapped['School'] = so.relationship(back_populates='posts')
    
    def __repr__(self):
        return '<Post {}>'.format(self.body)

我想实现帖子可以@用户或学校的功能，所以打算新增at_user_id字段记录@的用户，同时把原来的school_id改名为at_school_id记录@的学校，修改后的模型代码如下：

class Post(db.Model):
    id: so.Mapped[int] = so.mapped_column(primary_key=True)
    body: so.Mapped[str] = so.mapped_column(sa.String(140))
    timestamp: so.Mapped[datetime] = so.mapped_column(
        index=True, default=lambda: datetime.now(timezone.utc))
    user_id: so.Mapped[int] = so.mapped_column(sa.ForeignKey(User.id), index=True)
    at_user_id: so.Mapped[int] = so.mapped_column(sa.ForeignKey(User.id), index=True)
    at_school_id: so.Mapped[Optional[int]] = so.mapped_column(sa.ForeignKey(School.id))
    
    author: so.Mapped['User'] = so.relationship(back_populates='posts')
    school: so.Mapped['School'] = so.relationship(back_populates='posts')
    
    def __repr__(self):
        return '<Post {}>'.format(self.body)

改完后我先跑了flask db migrate，看起来一切正常，结果执行flask db upgrade的时候直接报错，栈追踪的最后一行是：

ValueError: Constraint must have a name

我意识到修改肯定有问题，赶紧把模型代码改回了原来的版本，但现在连flask db downgrade也执行失败，还是报同样的错误。

目前的迁移状态 #

我执行flask db history得到的输出是：

5894df3de404 -> c99a4bb66db5 (head), change post school reference column name
d536cbcefe42 -> 5894df3de404, remove unique constraint on coder.name
81bde43194e8 -> d536cbcefe42, add errors table
da1665d06269 -> 81bde43194e8, aligned db with refined workflow
84c4df60bcf5 -> da1665d06269, update to working model, add admin privilages column to user table
d366444514a2 -> 84c4df60bcf5, followers
aa7ad342e8fa -> d366444514a2, new fields in user model

再执行flask db check，输出如下：

INFO [alembic.runtime.migration] Context impl SQLiteImpl.
INFO [alembic.runtime.migration] Will assume non-transactional DDL.
INFO [alembic.autogenerate.compare] Detected added index ''ix_coder_name'' on '('name',)'
ERROR [flask_migrate] Error: New upgrade operations detected:
[('add_index', Index('ix_coder_name', Column('name', String(length=64), table=<coder>, nullable=False)))]

现在我完全卡壳了：模型已经改回原样，但迁移状态好像被卡住了，既不能升级也不能回滚，有没有大佬能指点一下该怎么解决这个问题呀？

内容来源于stack exchange