Before start, lets define the table such that:
Base = declarative_base()
class People(Base):
__tablename__ == 'people'
idx = Column('idx', Integer, primary_key=True)
name = column('name', char(50))
weight = column('weight', char(30))
height = column('height', char(30))
hobby = column('hobby', char(30))
With People class(or table? anyway), i know that result = session.scalars(select(People)) Are can access the values like this(Temporarily call this a 'dot access'):
for row in result:
print(f"{row.name} is the name")
print(f"weight are... {row.weight}")
print(f"let me show your {row.hobby}")
But result of result = session.scalars(select(People.name, People.weight, People.hobby)) are can not use the 'dot access'. (The type of rows are Tuple)
And i also using sqlalchemy.orm.Bundle: (select(Bundle("People", People.name, People.weight, People.hobby,)) but this are not works at all. (This also Tuple)
Finally, i really want that result of such querys are Table class! (So that i can use the 'dot access')
You should use scalars() when you are actually returning the whole objects. If you are selecting a couple of columns (as you are doing right here), what you can do is:
result = session.execute(select(People.name, People.height)).all()
# here we have an iterator of "Row" objects
for p in result:
print(p.name, p.height)
Note the execute statement, instead of scalars.
If you need to use an expression over a column, then you also should label the resulting columns:
# In case when you are using an expression, like "distinct" over a column, use "label" to preserve column names mapping
result = session.execute(select(distinct(People.name).label("name"), People.height)).all()
# here we have an iterator of "Row" objects
for p in result:
print(p.name, p.height)