sqlalchemy - reflecting tables and columns with spaces

Question

How can I use sqlalchemy on a database where the column names (and table names) have spaces in them?

db.auth_stuff.filter(\"db.auth_stuff.first name\"==\'Joe\'

Answer 1

you can do this using a reflection event to give the columns a .key, however the full recipe has a bug when primary key columns are involved, which was fixed in the still-unreleased 0.8.3 version (as well as master). If you check out 0.8.3 at https://bitbucket.org/zzzeek/sqlalchemy/get/rel_0_8.zip this recipe will work even with primary key cols:

from sqlalchemy import *
from sqlalchemy.orm import *
from sqlalchemy.ext.declarative import declarative_base, DeferredReflection

Base = declarative_base(cls=DeferredReflection)


e = create_engine("sqlite://", echo=True)
e.execute("""
    create table "user table" (
            "id col" integer primary key,
            "data col" varchar(30)
    )
""")

from sqlalchemy import event

@event.listens_for(Table, "column_reflect")
def reflect_col(inspector, table, column_info):
    column_info['key'] = column_info['name'].replace(' ', '_')

class User(Base):
    __tablename__ = "user table"

Base.prepare(e)

s = Session(e)
print s.query(User).filter(User.data_col == "some data")

DeferredReflection is an optional helper to use with declarative + reflection.