This is designed to add and process annotations to python objects, a-la Java Annotations.

Annotations are just metadata, something that is not directly used by code itself but can be queried by other parts of the program.

Python itself doesn't need special low-level implementations in order to support metadata, this library's target is just to provide a standard way of setting and querying metadata from python objects.

Python 3 actually contains a basic version of metadata, but it's limited to the arguments and return values of functions; this library is designed to extend such support to other objects, so you can annotate a class, a function or any object with any kind of data.

Current status

Currently contains a PEP-3107 compatible signature annotation implementation for Python 2.x.

Example code

Basic function annotation:

>>> from annopyte.annotations.signature import annotate_f
>>> @annotate_f("return_value_annotation", param1="asd", param2="fgh")
... def myfunc(param1, param2=None):
...     pass
>>> print myfunc.__annotations__
{'return': 'return_value_annotation', 'param2': 'fgh', 'param1': 'asd'}

Prospective code ( to be done )

That's the basic idea I'd like to implement for metadata usage:

class Author(Annotation)
    name = "default"

@Author(name="John Doe")
class MyClass(object):

>>> query_for_metadata("mypackage", Author, name="John Doe")
[<class 'mypackage.subpackage.MyClass'>]

That's it.

Contact me

Alan Franzoni <username at> (please note: write LITERALLY username in the email address!)