PEP 443 proposed to expose a mechanism in the functools standard library module in Python 3.4 that provides a simple form of generic programming known as single-dispatch generic functions.
This library is a backport of this functionality to Python 2.6 - 3.3.
To define a generic function, decorate it with the @singledispatch decorator. Note that the dispatch happens on the type of the first argument, create your function accordingly:
>>> from singledispatch import singledispatch >>> @singledispatch ... def fun(arg, verbose=False): ... if verbose: ... print("Let me just say,", end=" ") ... print(arg)
To add overloaded implementations to the function, use the register() attribute of the generic function. It is a decorator, taking a type parameter and decorating a function implementing the operation for that type:
>>> @fun.register(int) ... def _(arg, verbose=False): ... if verbose: ... print("Strength in numbers, eh?", end=" ") ... print(arg) ... >>> @fun.register(list) ... def _(arg, verbose=False): ... if verbose: ... print("Enumerate this:") ... for i, elem in enumerate(arg): ... print(i, elem)
To enable registering lambdas and pre-existing functions, the register() attribute can be used in a functional form:
>>> def nothing(arg, verbose=False): ... print("Nothing.") ... >>> fun.register(type(None), nothing)
The register() attribute returns the undecorated function which enables decorator stacking, pickling, as well as creating unit tests for each variant independently:
>>> @fun.register(float) ... @fun.register(Decimal) ... def fun_num(arg, verbose=False): ... if verbose: ... print("Half of your number:", end=" ") ... print(arg / 2) ... >>> fun_num is fun False
When called, the generic function dispatches on the type of the first argument:
>>> fun("Hello, world.") Hello, world. >>> fun("test.", verbose=True) Let me just say, test. >>> fun(42, verbose=True) Strength in numbers, eh? 42 >>> fun(['spam', 'spam', 'eggs', 'spam'], verbose=True) Enumerate this: 0 spam 1 spam 2 eggs 3 spam >>> fun(None) Nothing. >>> fun(1.23) 0.615
Where there is no registered implementation for a specific type, its method resolution order is used to find a more generic implementation. The original function decorated with @singledispatch is registered for the base object type, which means it is used if no better implementation is found.
To check which implementation will the generic function choose for a given type, use the dispatch() attribute:
>>> fun.dispatch(float) <function fun_num at 0x1035a2840> >>> fun.dispatch(dict) # note: default implementation <function fun at 0x103fe0000>
To access all registered implementations, use the read-only registry attribute:
>>> fun.registry.keys() dict_keys([<class 'NoneType'>, <class 'int'>, <class 'object'>, <class 'decimal.Decimal'>, <class 'list'>, <class 'float'>]) >>> fun.registry[float] <function fun_num at 0x1035a2840> >>> fun.registry[object] <function fun at 0x103fe0000>
The vanilla documentation is available at http://docs.python.org/3/library/functools.html#functools.singledispatch.
This backport is intended to keep 100% compatibility with the vanilla release in Python 3.4+. To help maintaining a version you want and expect, a versioning scheme is used where:
- the first three numbers indicate the version of Python 3.x from which the backport is done
- a backport release number is provided after the last dot
For example, 126.96.36.199 is the first release of singledispatch compatible with the library found in Python 3.4.0.
A single exception from the 100% compatibility principle is that bugs fixed before releasing another minor Python 3.x.y version will be included in the backport releases done in the mean time. This rule applies to bugs only.
This backport is maintained on BitBucket by Łukasz Langa, one of the members of the core CPython team:
Should now install flawlessly on PyPy as well. Thanks to Ryan Petrello for finding and fixing the setup.py issue.
Updated to the reference implementation as of 02-July-2013.
- more predictable dispatch order when abstract base classes are in use: abstract base classes are now inserted into the MRO of the argument's class where their functionality is introduced, i.e. issubclass(cls, abc) returns True for the class itself but returns False for all its direct base classes. Implicit ABCs for a given class (either registered or inferred from the presence of a special method like __len__) are inserted directly after the last ABC explicitly listed in the MRO of said class. This also means there are less "ambiguous dispatch" exceptions raised.
- better test coverage and improved docstrings
Updated to the reference implementation as of 31-May-2013.
- better performance
- fixed a corner case with PEP 435 enums
- calls to dispatch() also cached
- dispatching algorithm now now a module-level routine called _find_impl() with a simplified implementation and proper documentation
- dispatch() now handles all caching-related activities
- terminology more consistent: "overload" -> "implementation"
- the first public release compatible with 3.4.0
This section is technical and should bother you only if you are wondering how this backport is produced. If the implementation details of this backport are not important for you, feel free to ignore the following content.
singledispatch is converted using six so that a single codebase can be used for all compatible Python versions. Because a fully automatic conversion was not doable, I took the following branching approach:
- the upstream branch holds unchanged files synchronized from the upstream CPython repository. The synchronization is currently done by manually copying the required code parts and stating from which CPython changeset they come from. The tests should pass on Python 3.4 on this branch.
- the default branch holds the manually translated version and this is where all tests are run for all supported Python versions using Tox.