How to query a SQL View from flask-SQLAlchemy

Question

I have been given access to a database (MySQL, in case it matters) to access some data that I need, but I don't have direct access to the tables, the data is exposed through a SQL View.

I want to wrap this in a small API using flask. For the database connection I intend to use SQLAlchemy (just because it's the only databse wrapper that I know in python).

The thing is, I have read a lot of documentation, and dozens of stackoverflow posts and guides through the Internet, and I can't seem to find a way to query SQL Views from flask-SQLAlchemy.

In the documentation, they always use Models. I tried to define my View as a Model, but I get could not assemble any primary key columns for mapped table 'MySQLView'.

What I tried to do:

class MyView(db.Model):
    __tablename__ = "my_view_name_in_sql"
    db.Column("id", db.Integer, primary_key=True)
    db.Column("sample_id", db.String)
    {...etc}

Is there a way to do this, or maybe I should opt-out of using flask-SQLAlchemy and use something else instead?

Answer 1

There's a few answers here I think deal with your case, especially LeoRochael's answer who suggests using the https://pypi.org/project/sqlalchemy-views/ package. I suggest checking if you can - or cannot access the database's metdata, I think that could really help you avoid having to make models manually

db_engine = sqlalchemy_package.create_engine("mysql+...")
db_metadata = db.MetaData(bind=db_engine)

you can then either try making a Table directly, or use the metadata to make views like in the answer I linked.

your_table = db.Table("my_view_name_in_sql", db_metadata, autoload=True)
view_definition = your_table.select()
your_view = CreateView(your_table, view_definition, or_replace=True)

Answer 2

I don't have experience with SQLAlchemy Flask but this is what worked for me in SQLAlchemy. I used the __mapper_args__ attribute to define the private key for a view in my database.

db_metadata = MetaData()
Base = automap_base(metadata=db_metadata)

class MyView(Base):
    __table__ = Table("View_Name", db_metadata, autoload_with=engine, schema="My_Schema")
    __mapper_args__ = {'primary_key': [__table__.c['Primary_Column_Name']]}

Base.prepare()