SQL ALCHEMY – How to return query(model).all() as a tuple of values instead of model instance

Question

I’m using SQL Alchemy for my project but I’m having a hard time finding out how to simply query a model and return a list of tuples as a result.

I have a simple model:

class Security(Base):

    __tablename__   = 'Securities'
    id              = Column(Integer, primary_key=True, autoincrement=True)
    ticker          = Column(String, index= True)
    company         = Column(String)
    sector          = Column(String)
    industry        = Column(String)
    country         = Column(String)

When I query this model within a session, I get a model in return:

s = Session()
q = s.query(Security).all()

If I simply print the query, I get a list of the model class.

[<models.Security at 0x7ff6eaf75a90>,
 <models.Security at 0x7ff6eaf759a0>,
 <models.Security at 0x7ff6eaf75940>,
 <models.Security at 0x7ff6eaf758e0>,
 <models.Security at 0x7ff6eaf75040>,
 <models.Security at 0x7ff6eaf750a0>,
 <models.Security at 0x7ff6eaf75100>,
 <models.Security at 0x7ff6eaf75070>,
 <models.Security at 0x7ff6eaf751c0>,..

So I’ve found a solution for now:

[(i.id, i.ticker, i.company, i. sector, i.industry, i.country) for i in q]

Is there a more pythonic way to do this?
When I was querying my database with mysql–connector-python (or pymysql, can’t remember), the results would be printed out directly as a list of tuple.

Asked By: Paul Marmagne

||

Source

Answer 1

If you want to get tuples rather than model instances you can query the underlying Table object rather than the model class, for example

# 1.x style
rows = session.query(Security.__table__).all()

# 2.0 style
rows = session.scalars(sa.select(Security.__table__)).all()

Note that these approaches may not work for relationships.

If the application is designed to only handle raw tuples then there is little point using SQLAlchemy’s ORM layer – use the core layer instead where you can work with Table objects directly.

Answered By: snakecharmerb

Answer 2

Another approach is to take advantage of the fact that for select(Thing)

if we use Session.execute() we get ORM objects back, but
if we use Connection.execute() we get tuple-like LegacyRow objects back

class Thing(Base):
    __tablename__ = "thing"
    id = Column(Integer, primary_key=True, autoincrement=False)
    txt = Column(String(50))


with Session(engine) as sess:
    query = select(Thing)

    results = sess.execute(query).all()
    pprint(results)
    """
    [(<__main__.Thing object at 0x000000FD8375C0D0>,),
     (<__main__.Thing object at 0x000000FD8375C160>,)]
    """

    with sess.connection() as conn:
        results = conn.execute(query).all()
        pprint(results)
        """
        [(1, 'foo'), (2, 'bar')]
        """

Answered By: Gord Thompson

SQL ALCHEMY – How to return query(model).all() as a tuple of values instead of model instance

Question:

Answers: