How do I set the schema in SQLAlchemy for MSSQL?

Question

I currently do this:

#!/usr/bin/env python

# 3rd party modules
from sqlalchemy import create_engine  # requires pymssql

# local modules
from config import cfg

connection_string = 'mssql+pymssql://{user}:{password}@{host}:{port}/{db}'

engine = create_engine(connection_string
                       .format(host=cfg['db']['host'],
                               db=cfg['db']['database'],
                               user=cfg['db']['user'],
                               password=cfg['db']['password'],
                               port=cfg['db']['port'],
                               schema=cfg['db']['schema']))

with engine.begin() as conn:
    sql = ('SELECT foo FROM bar;')
    rows = conn.execute(sql)
    print(rows)

But I get

  File "/usr/local/lib/python3.5/dist-packages/sqlalchemy/engine/default.py", line 470, in do_execute
    cursor.execute(statement, parameters)
  File "pymssql.pyx", line 464, in pymssql.Cursor.execute (pymssql.c:7491)
sqlalchemy.exc.ProgrammingError: (pymssql.ProgrammingError) (208, b"Invalid object name 'bar'.DB-Lib error message 20018, severity 16:\nGeneral SQL Server error: Check messages from the SQL Server\n") [SQL: 'SELECT foo FROM bar;']

I think the problem is that I have to use the schema exampleschema. I can access the table bar with the column foo in the schema exampleschema with the same credentials via DBeaver.

But when I add /{schema} to the connection string, I get

sqlalchemy.exc.OperationalError: (pymssql.OperationalError)
    (18456,
     b"Login failed for user 'exampleuser'.DB-Lib error message 20018,
     severity 14:\nGeneral SQL Server error: Check messages from the SQL Server\nDB-Lib error message 20002,
     severity 9:\nAdaptive Server connection failed (192.168.123.456:1433)\n")

How do I set the schema?

Solution 1

Reading this SO Question Possible to set default schema from connection string? it seems not possible to select schema using connection string.

try a full qualified name :

sql = ('SELECT foo FROM exampleschema.bar;')

Solution 2

You can also specify the schema name in the class definition (is not your specific case but I think it's a common situation).

For example, if you have a table "dog" into the "animal" schema:

from sqlalchemy import Column, Integer, String
from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

class Notification(Base):
    __tablename__ = "dog"
    __table_args__ = {"schema": "animal"}
    id = Column(Integer, primary_key=True)
    name = Column(String)

Solution 3

If you load metadata from the database and load table from the metadata, you can set the schema during table load.

Then create a session and query the table

from sqlalchemy import MetaData, Table
from sqlalchemy.orm import sessionmaker
table = Table({Table Name}, {metadata}, autoload=True, autoload_with={engine}, schema={Schema name})

Session = sessionmaker()
Session.configure(bind={engine})
session = Session()
query = session.query(table)

Comments

[matteogll]

You can use raw query this way: from sqlalchemy.orm import Session session.execute("SELECT * FROM animal.dog")