SQLAlchemy – Mapping Table Columns

In this article, we will see how to map table columns using SQLAlchemy in Python.

You will need a database (MySQL, PostgreSQL, SQLite, etc) to work with. Since we are going to use MySQL in this post, we will also install a SQL connector for MySQL in Python. However, none of the code implementations changes with change in the database except for the SQL connectors.

pip install pymysql

We will use the sample sakila database from MySQL. All the examples covered in this article will make use of the actor table within the sakila database. If you do not have the sakila database and want to follow along with this article without installing it then use the SQL script present in the link mentioned below to create the required schema and actor table along with the records.

Databased Used: Sakila Actor Table Script

We will be referring to the same SQL query in each of the examples mentioned below –

SELECT first_name FROM sakila.actor LIMIT 1;

The different ways in which we can map the columns in SQLAlchemy are –

• Mapping columns directly to the attribute names
• Mapping columns distinctly from attribute names
• Mapping columns using reflection
• Mapping columns using a prefix

Mapping columns directly to the attribute names

In the below example, the column mapping is done by mapping each of the table columns as class attributes. Each of the attributes is provided with the same name as the corresponding table columns that it represents. We then establish the SQLAlchemy engine connected to the sakila database in MySQL. Then a session object is created to query the database. Using this session object, we will query the first record in the `actor` table. We get the value of the column `first_name` for the first record by accessing the `first_name` attribute of the `result` object. This shows that the column of the actor table is mapped against the attributes of the Actor class.

Python3

import sqlalchemy as db
from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

# MAPPING TABLE ACTOR
class Actor(Base):

    __tablename__ = 'actor'

    actor_id    = db.Column(db.SmallInteger, autoincrement=True, primary_key=True)
    first_name  = db.Column(db.String(45), nullable=False)
    last_name   = db.Column(db.String(45), nullable=False)
    last_update = db.Column(db.TIMESTAMP, nullable=False)

# DEFINE THE ENGINE (CONNECTION OBJECT)
engine = db.create_engine("mysql+pymysql://root:password@localhost/sakila")

# CREATE A SESSION OBJECT TO INITIATE QUERY IN DATABASE
from sqlalchemy.orm import sessionmaker
Session = sessionmaker(bind = engine)
session = Session()

# SELECT * FROM sakila.actor LIMIT 1;
result = session.query(Actor).first()

# DISPLAY FIRST NAME OF FIRST RECORD IN ACTOR TABLE
print("First Name (Record 1):", result.first_name)

Output:

First Name (Record 1): PENELOPE

Mapping columns distinctly from attribute names

This is similar to the first example with a small change. The attribute names mentioned in this example are different than the column names. This is possible by providing an additional parameter inside the `Column()` method. The method’s first argument takes in the actual column name which allows using different attribute names for referring to these columns. If we look at the final `print()` method, the first name of the first record in the actor table is referenced using the `fname` attribute as opposed to the `first_name` or actual column name seen in the first example.

Python3

import sqlalchemy as db
from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

# MAPPING TABLE ACTOR
class Actor(Base):

    __tablename__ = 'actor'

    id        = db.Column('actor_id', db.SmallInteger, autoincrement=True,
                          primary_key=True)
    fname     = db.Column('first_name', db.String(45), nullable=False)
    lname     = db.Column('last_name', db.String(45), nullable=False)
    update_on = db.Column('last_update', db.TIMESTAMP, nullable=False)

# DEFINE THE ENGINE (CONNECTION OBJECT)
engine = db.create_engine("mysql+pymysql://root:password@localhost/sakila")

# CREATE A SESSION OBJECT TO INITIATE QUERY IN DATABASE
from sqlalchemy.orm import sessionmaker
Session = sessionmaker(bind = engine)
session = Session()

# SELECT * FROM sakila.actor LIMIT 1;
result = session.query(Actor).first()

# DISPLAY FIRST NAME OF FIRST RECORD IN ACTOR TABLE
print("First Name (Record 1):", result.fname)

Output:

First Name (Record 1): PENELOPE

Mapping columns using reflection

In the previous two examples, we needed to explicitly map each column with the table using class and its attributes. In this method, we do not need to provide this explicit mapping of each table column separately. Using reflection, this task is automatically done by providing the metadata object and the SQLAlchemy engine connection. In the `__table__` attribute. We can then use the engine and session objects to query the actor table to fetch the first name as done in earlier examples using the column name `first_name` as the attribute name itself.

Python3

import sqlalchemy as db
from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

# DEFINE THE ENGINE (CONNECTION OBJECT)
engine = db.create_engine("mysql+pymysql://root:password@localhost/sakila")

class Actor(Base):
    __table__ = db.Table("actor", Base.metadata, autoload_with=engine)

# CREATE A SESSION OBJECT TO INITIATE QUERY IN DATABASE
from sqlalchemy.orm import sessionmaker
Session = sessionmaker(bind = engine)
session = Session()

# SELECT COUNT(*) FROM Table LIMIT 1;
result = session.query(Actor).first()

# DISPLAY FIRST NAME OF FIRST RECORD IN ACTOR TABLE
print("First Name (Record 1):", result.first_name)

Output:

First Name (Record 1): PENELOPE

Mapping columns using a prefix

The usage of prefixes is rare but still, it can be found in some use cases. In this example, it can be seen that we have used an additional attribute `__mapper_args__` which is a python dictionary. It is provided with a key as `column_prefix` and a value of `_`. This means that we want to prefix all the column names or attribute names with an underscore. For this reason, we used `_first_name` instead of `first_name` as the attribute for the respective column.

Python3

import sqlalchemy as db
from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

# DEFINE THE ENGINE (CONNECTION OBJECT)
engine = db.create_engine("mysql+pymysql://root:password@localhost/sakila")

class Actor(Base):
    __table__ = db.Table("actor", Base.metadata, autoload_with=engine)
    __mapper_args__ = {'column_prefix': '_'}

# DEFINE THE ENGINE (CONNECTION OBJECT)
engine = db.create_engine("mysql+pymysql://root:password@localhost/sakila")

# CREATE A SESSION OBJECT TO INITIATE QUERY IN DATABASE
from sqlalchemy.orm import sessionmaker
Session = sessionmaker(bind = engine)
session = Session()

# SELECT COUNT(*) FROM Table LIMIT 1;
result = session.query(Actor).first()

# DISPLAY FIRST NAME OF FIRST RECORD IN ACTOR TABLE
print("First Name (Record 1):", result._first_name)

Output:

First Name (Record 1): PENELOPE