fixture.loadable.sqlalchemy_loadable

Components for loading and unloading data using SQLAlchemy.

See Using LoadableFixture for examples.

class fixture.loadable.sqlalchemy_loadable.SQLAlchemyFixture(engine=None, connection=None, session=None, scoped_session=None, **kw)

Bases: fixture.loadable.loadable.DBLoadableFixture

A fixture that knows how to load DataSet objects into SQLAlchemy objects.

>>> from fixture import SQLAlchemyFixture

The recommended way to deal with connections is to either pass in your own engine object or let implicit binding govern how connections are made. This is because SQLAlchemyFixture will use an internally scoped session to avoid conflicts with that of the Application Under Test. If you need to bypass this behavior then pass in your own session or scoped_session.

For examples of usage see Using LoadableFixture

Keyword Arguments:

style

A Style object to translate names with

env

A dict or module that contains either mapped classes or Table objects, or both. This will be searched when Style translates DataSet names into storage media. See EnvLoadableFixture.attach_storage_medium for details on how env works.

engine

A specific connectable/engine object to use when one is not bound. engine.connect() will be called.

session

A session from sqlalchemy.create_session(). See Contextual/Thread-local Sessions for more info. This will override the ScopedSession and SessionContext approaches. Only declare a session if you have to. The preferred way is to let fixture use its own session in a private scope.

scoped_session

A class-like Session object created by scoped_session(sessionmaker()). Only declare a custom Session if you have to. The preferred way is to let fixture use its own Session which defines a private scope to avoid conflicts with that of the Application Under Test.

connection

A specific connection / engine to use when one is not bound.

dataclass

SuperSet class to represent loaded data with

medium

A custom StorageMediumAdapter to instantiate when storing a DataSet. By default, a medium adapter will be negotiated based on the type of SQLAlchemy object so you should only set this if you know what you doing.

begin(unloading=False)

Begin loading data

• creates and stores a connection with engine.connect() if an engine was passed
  • binds the connection or engine to fixture’s internal session
• uses an unbound internal session if no engine or connection was passed in

commit()

Commit the load transaction and flush the session

create_transaction()

Create a session transaction or a connection transaction

• if a custom connection was used, calls connection.begin
• otherwise calls session.begin()

dispose()

Dispose of this fixture instance entirely

Closes all connection, session, and transaction objects and calls engine.dispose()

After calling fixture.dispose() you cannot use the fixture instance. Instead you have to create a new instance like:

fixture = SQLAlchemyFixture(...)

rollback()

Rollback load transaction

class fixture.loadable.sqlalchemy_loadable.MappedClassMedium(*a, **kw)

Bases: fixture.loadable.loadable.StorageMediumAdapter

Adapter for SQLAlchemy mapped classes.

For example, in mapper(TheClass, the_table) TheClass is a mapped class. If using Elixir then any class descending from elixir.Entity is treated like a mapped class.

clear(obj)

Delete this object from the session

save(row, column_vals)

Save a new object to the session if it doesn’t already exist in the session.

visit_loader(loader)

Visits the SQLAlchemyFixture loader and stores a reference to its session

class fixture.loadable.sqlalchemy_loadable.TableMedium(*a, **kw)

Bases: fixture.loadable.loadable.StorageMediumAdapter

Adapter for SQLAlchemy Table objects

If no connection or engine is configured in the SQLAlchemyFixture then statements will be executed directly on the Table object itself which adheres to implicit connection rules. Otherwise, the respective connection or engine will be used to execute statements.

clear(obj)

Constructs a delete statement per each primary key and executes it either explicitly or implicitly

save(row, column_vals)

Constructs an insert statement with the given values and executes it either explicitly or implicitly

visit_loader(loader)

Visits the SQLAlchemyFixture loader and stores a reference to its connection if there is one.