agate’s core featureset is designed rely on as few dependencies as possible. However, in real life you’re often going to want to interface agate with SQL, numpy or other data pipelines.
How extensions work¶
In order to support these use-cases, but not make things excessively complicated, agate support’s a simple extensibility pattern based on monkey patching. Libraries can be created that patch new methods on
TableSet. For example, agate-sql adds the ability to read and write tables from a SQL database:
import agate import agatesql agatesql.patch() # After calling patch the from_sql and to_sql methods are now part of the Table class table = agate.Table.from_sql('postgresql:///database', 'input_table') table.to_sql('postgresql:///database', 'output_table')
List of extensions¶
Here is a list of actively supported agate extensions:
Writing your own extensions¶
Writing your own extensions is straightforward. Create a class that acts as your “patch” and when you call
Table.monkeypatch() it will dynamically be added as a base class of
import agate class ExamplePatch(object): def new_method(self): print('I do something to a Table when you call me.')
Then create a function that applies your patch:
def patch() agate.Table.monkeypatch(ExamplePatch)
Table class will now have all the methods of
ExamplePatch as though they were defined as part of it.
>>> import agate >>> import myextension >>> myextension.patch() >>> table = agate.Table(rows, columns) >>> table.new_method() 'I do something to a Table when you call me.'
The same pattern also works for adding methods to
Extensions are added as base classes of
Table so you can not use them to override the implementation of an existing method. They are perfect for adding features, but if you need to actually modify how agate works, then you’ll need to use a subclass. Any shadowed method will be ignored.