Flask-SQLAlchemy 小写索引 - 跳过功能，SQLAlchemy 反射不支持 [英] Flask-SQLAlchemy Lower Case Index - skipping functional, not supported by SQLAlchemy reflection

问题描述

首先.抱歉，如果有人回答了这个问题，但我在任何地方都找不到答案.

First off. Apologies if this has been answered but I can not find the answer any where.

我需要在 Flask-SQLAlchemy 对象上定义一个小写索引.

I need to define a lowercase index on a Flask-SQLAlchemy object.

我遇到的问题是我需要将模型用户名和电子邮件字段存储为小写，以便我可以使用 User.query.filter(func.lower(username) == func.lower(username) 检查它们).first()

The problem I have is I need a models username and email fields to be stored lowercase so that I can check them with User.query.filter(func.lower(username) == func.lower(username)).first()

到目前为止，我一直通过将这些字段插入为小写来处理这个问题，但在这个特定的例子中，我需要用户名来保留它定义的大小写.

Up to now I have always dealt with this by just inserting these fields as lowercase but in this particular instance I need the username to preserve the case that it was defined with.

我认为我走在正确的轨道上，但遇到了一个我以前从未见过的问题.

I think I am on the right tracks but am running in to an issue I have not seen before.

class User(UserMixin, db.Model):
    __tablename__ = 'users'

    id = db.Column(db.Integer, primary_key=True)
    # Indexes for username and email are defined below User
    # They are functional lowercase indexes
    username = db.Column(db.String(32))
    email = db.Column(db.String(255))
    password_hash = db.Column(db.String(255))

    ...

db.Index('ix_users_username', func.lower(User.username), unique=True)
db.Index('ix_users_email', func.lower(User.email), unique=True)

我看不出这有什么问题.我猜 func.lower() 需要将索引定义为与我的其他搜索可以分辨的不同.

I can't see a problem with this. I am guessing that the func.lower() requires the Index to be defined separate from what I can tell with my other searches.

现在问题来了，当我运行迁移时，我收到以下错误:

Now the problem comes when I run the migration I get the following errors:

INFO  [alembic.runtime.migration] Context impl PostgresqlImpl.
INFO  [alembic.runtime.migration] Will assume transactional DDL.
INFO  [alembic.autogenerate.compare] Detected added table 'users'
.../lib/python3.4/site-packages/alembic/util/messaging.py:69: UserWarning: autogenerate skipping functional index ix_users_email; not supported by SQLAlchemy reflection
  warnings.warn(msg)
.../lib/python3.4/site-packages/alembic/util/messaging.py:69: UserWarning: autogenerate skipping functional index ix_users_username; not supported by SQLAlchemy reflection

我不是 100% 确定为什么这不受反射支持.我期望在我的迁移中它会像以前一样添加它们，但将其包裹在较低的位置.

I'm not 100% sure why this isn't supported by reflection. I was expecting in my migration that it would add them as it did before but with it wrapped in lower.

我并不反对手动编写迁移(因为我假设它是可能的，而不是 100% 确定如何)，但谁能指出为什么这不能立即使用?

I am not adverse to writing the migration manually (as I'm presuming its possible, not 100% sure how) but can anyone point out why this doesn't work out of the box?

提前致谢乔

我通过在迁移中添加以下行解决了这个问题.

I solved this by adding the following line in to the migration.

op.create_index('ix_users_username', 'users', [sa.text('lower(username)')])
op.create_index('ix_users_email', 'users', [sa.text('lower(email)')])

推荐答案

您需要使用 alembic 的 execute 方法手动添加 UNIQUE INDEX.将类似的内容添加到您的迁移脚本中.

You'd need to add the UNIQUE INDEX manually using alembic's execute method. Add something like this to your migration script.

from alembic import op
# ...

def upgrade():
    # ...
    op.execute(
        """ CREATE UNIQUE INDEX users_normalized_username
            ON users
            (lower(username))
        """
    )

def downgrade():
    # ...
    op.execute("DROP INDEX users_normalized_username")

您还可以添加 ColumnProperty 以便您可以访问 username 的规范化表示.

You could also add a ColumnProperty so you have access the normalized representation of the username.

from sqlalchemy import Column, String, func, orm

class User(Base):
    __tablename__ = 'users'

    username = Column(String(32))
    normalized_username = orm.column_property(func.lower(username))

    # ...

这不会在您的数据库中创建额外的列，但您可以对其进行查询和比较.

This won't create an extra column in your database but you'll be able to query and compare against it.

session.query(User).filter(User.normalized_username == func.lower(username)).first()

这篇关于Flask-SQLAlchemy 小写索引 - 跳过功能，SQLAlchemy 反射不支持的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！