Coverage for nltk.decorators : 58%

""" 

Decorator module by Michele Simionato <michelesimionato@libero.it> 

Copyright Michele Simionato, distributed under the terms of the BSD License (see below). 

http://www.phyast.pitt.edu/~micheles/python/documentation.html 

Included in NLTK for its support of a nice memoization decorator. 

""" 

from __future__ import print_function 

__docformat__ = 'restructuredtext en' 

## The basic trick is to generate the source code for the decorated function 

## with the right signature and to evaluate it. 

## Uncomment the statement 'print >> sys.stderr, func_src'  in _decorator 

## to understand what is going on. 

__all__ = ["decorator", "new_wrapper", "getinfo"] 

import sys 

# Hack to keep NLTK's "tokenize" module from colliding with the "tokenize" in 

# the Python standard library. 

old_sys_path = sys.path[:] 

sys.path = [p for p in sys.path if "nltk" not in p] 

import inspect 

sys.path = old_sys_path 

try: 

    set 

except NameError: 

    from sets import Set as set 

def getinfo(func): 

    """ 

    Returns an info dictionary containing: 

    - name (the name of the function : str) 

    - argnames (the names of the arguments : list) 

    - defaults (the values of the default arguments : tuple) 

    - signature (the signature : str) 

    - doc (the docstring : str) 

    - module (the module name : str) 

    - dict (the function __dict__ : str) 

    >>> def f(self, x=1, y=2, *args, **kw): pass 

    >>> info = getinfo(f) 

    >>> info["name"] 

    'f' 

    >>> info["argnames"] 

    ['self', 'x', 'y', 'args', 'kw'] 

    >>> info["defaults"] 

    (1, 2) 

    >>> info["signature"] 

    'self, x, y, *args, **kw' 

    """ 

    assert inspect.ismethod(func) or inspect.isfunction(func) 

    regargs, varargs, varkwargs, defaults = inspect.getargspec(func) 

    argnames = list(regargs) 

    if varargs: 

        argnames.append(varargs) 

    if varkwargs: 

        argnames.append(varkwargs) 

    signature = inspect.formatargspec(regargs, varargs, varkwargs, defaults, 

                                      formatvalue=lambda value: "")[1:-1] 

    # pypy compatibility 

    if hasattr(func, '__closure__'): 

        _closure = func.__closure__ 

        _globals = func.__globals__ 

    else: 

        _closure = func.func_closure 

        _globals = func.func_globals 

    return dict(name=func.__name__, argnames=argnames, signature=signature, 

                defaults = func.__defaults__, doc=func.__doc__, 

                module=func.__module__, dict=func.__dict__, 

                globals=_globals, closure=_closure) 

# akin to functools.update_wrapper 

def update_wrapper(wrapper, model, infodict=None): 

    infodict = infodict or getinfo(model) 

    try: 

        wrapper.__name__ = infodict['name'] 

    except: # Python version < 2.4 

        pass 

    wrapper.__doc__ = infodict['doc'] 

    wrapper.__module__ = infodict['module'] 

    wrapper.__dict__.update(infodict['dict']) 

    wrapper.__defaults__ = infodict['defaults'] 

    wrapper.undecorated = model 

    return wrapper 

def new_wrapper(wrapper, model): 

    """ 

    An improvement over functools.update_wrapper. The wrapper is a generic 

    callable object. It works by generating a copy of the wrapper with the 

    right signature and by updating the copy, not the original. 

    Moreovoer, 'model' can be a dictionary with keys 'name', 'doc', 'module', 

    'dict', 'defaults'. 

    """ 

    if isinstance(model, dict): 

        infodict = model 

    else: # assume model is a function 

        infodict = getinfo(model) 

    assert not '_wrapper_' in infodict["argnames"], ( 

        '"_wrapper_" is a reserved argument name!') 

    src = "lambda %(signature)s: _wrapper_(%(signature)s)" % infodict 

    funcopy = eval(src, dict(_wrapper_=wrapper)) 

    return update_wrapper(funcopy, model, infodict) 

# helper used in decorator_factory 

def __call__(self, func): 

    return new_wrapper(lambda *a, **k : self.call(func, *a, **k), func) 

def decorator_factory(cls): 

    """ 

    Take a class with a ``.caller`` method and return a callable decorator 

    object. It works by adding a suitable __call__ method to the class; 

    it raises a TypeError if the class already has a nontrivial __call__ 

    method. 

    """ 

    attrs = set(dir(cls)) 

    if '__call__' in attrs: 

        raise TypeError('You cannot decorate a class with a nontrivial ' 

                        '__call__ method') 

    if 'call' not in attrs: 

        raise TypeError('You cannot decorate a class without a ' 

                        '.call method') 

    cls.__call__ = __call__ 

    return cls 

def decorator(caller): 

    """ 

    General purpose decorator factory: takes a caller function as 

    input and returns a decorator with the same attributes. 

    A caller function is any function like this:: 

     def caller(func, *args, **kw): 

         # do something 

         return func(*args, **kw) 

    Here is an example of usage: 

    >>> @decorator 

    ... def chatty(f, *args, **kw): 

    ...     print("Calling %r" % f.__name__) 

    ...     return f(*args, **kw) 

    >>> chatty.__name__ 

    'chatty' 

    >>> @chatty 

    ... def f(): pass 

    ... 

    >>> f() 

    Calling 'f' 

    decorator can also take in input a class with a .caller method; in this 

    case it converts the class into a factory of callable decorator objects. 

    See the documentation for an example. 

    """ 

    if inspect.isclass(caller): 

        return decorator_factory(caller) 

    def _decorator(func): # the real meat is here 

        infodict = getinfo(func) 

        argnames = infodict['argnames'] 

        assert not ('_call_' in argnames or '_func_' in argnames), ( 

            'You cannot use _call_ or _func_ as argument names!') 

        src = "lambda %(signature)s: _call_(_func_, %(signature)s)" % infodict 

        # import sys; print >> sys.stderr, src # for debugging purposes 

        dec_func = eval(src, dict(_func_=func, _call_=caller)) 

        return update_wrapper(dec_func, func, infodict) 

    return update_wrapper(_decorator, caller) 

def getattr_(obj, name, default_thunk): 

    "Similar to .setdefault in dictionaries." 

    try: 

        return getattr(obj, name) 

    except AttributeError: 

        default = default_thunk() 

        setattr(obj, name, default) 

        return default 

@decorator 

def memoize(func, *args): 

    dic = getattr_(func, "memoize_dic", dict) 

    # memoize_dic is created at the first call 

    if args in dic: 

        return dic[args] 

    else: 

        result = func(*args) 

        dic[args] = result 

        return result 

if __name__ == "__main__": 

    import doctest; doctest.testmod() 

##########################     LEGALESE    ############################### 

##   Redistributions of source code must retain the above copyright 

##   notice, this list of conditions and the following disclaimer. 

##   Redistributions in bytecode form must reproduce the above copyright 

##   notice, this list of conditions and the following disclaimer in 

##   the documentation and/or other materials provided with the 

##   distribution. 

##   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 

##   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 

##   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 

##   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 

##   HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 

##   INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 

##   BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS 

##   OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND 

##   ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR 

##   TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE 

##   USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH 

##   DAMAGE.