pganti / micheles Goto Github PK

While it is easy for me to see that this is under a BSD license on this site, I 
would prefer that you had a LICENSE.txt file in the project filled in with your 
name and the appropriate copyright year(s). That just makes it that much easier 
to get things right when including it in another project.

$ PYTHONPATH="build/lib" python3.3 documentation3.py
**********************************************************************
File "documentation3.py", line 133, in __main__
Failed example:
    f1(0, 1)
Expected:
    Traceback (most recent call last):
       ...
    TypeError: f1() takes exactly 1 positional argument (2 given)
Got:
    Traceback (most recent call last):
      File "/usr/lib64/python3.3/doctest.py", line 1287, in __run
        compileflags, 1), test.globs)
      File "<doctest __main__[3]>", line 1, in <module>
        f1(0, 1)
      File "documentation3.py", line 908, in memoize
        cache[key] = result = func(*args, **kw)
    TypeError: f1() takes 1 positional argument but 2 were given
**********************************************************************
File "documentation3.py", line 1063, in __main__.test_kwonlyargs
Failed example:
    func('a', 'b', 'c', 'd', 'e', y='y', z='z', cat='dog')
Expected:
    calling func with args ('a', 'b', 'c', 'd', 'e'), {'y': 'y', 'z': 'z', 'cat': 'dog'}
    ('y', 'z')
Got:
    calling func with args ('a', 'b', 'c', 'd', 'e'), {'cat': 'dog', 'z': 'z', 'y': 'y'}
    ('y', 'z')
**********************************************************************
2 items had failures:
   1 of  63 in __main__
   1 of   2 in __main__.test_kwonlyargs
***Test Failed*** 2 failures.

What steps will reproduce the problem?

$ pip install decorator==2.3.2
Downloading/unpacking decorator==2.3.2
  Could not find a version that satisfies the requirement decorator==2.3.2 (from versions: )
No distributions matching the version for decorator==2.3.2
Storing complete log in /home/alfborge/.pip/pip.log

What is the expected output? What do you see instead?

$ pip install decorator==3.3.1
Downloading/unpacking decorator==3.3.1
  Downloading decorator-3.3.1.tar.gz
  Running setup.py egg_info for package decorator

    warning: no previously-included files found matching 'Makefile'
Installing collected packages: decorator
  Running setup.py install for decorator

    warning: no previously-included files found matching 'Makefile'
Successfully installed decorator
Cleaning up...

Since we're using a product that has a requirement on decorator <= 2.3.2, we 
can no longer rely on pip to pull in our requirements.  If at all possible, 
please put back the old versions.

Thank you,
Alf Lervåg

[deleted issue]

I'm trying to decorate a fabric `WrappedCallableTask` object. The comment on 
line 73 of `decorator.py` says `func` can be a class or callable, but not an 
instance method. `WrappedCallableTask` objects are callable.

This seems to be because `self.signature` is only assigned for actual functions 
(not callables). If I set it to an empty string instead, it works as expected.

Here's an interactive shell session demonstrating the problem.

In [1]: from fabric.api import task

In [2]: import decorator

In [3]: @decorator.decorator
   ...: def foo(func, *args, **kwargs):
   ...:         print 'foo'
   ...:         return func(*args, **kwargs)
   ...: 

In [4]: def x():
   ...:         print 'X'
   ...:     

In [5]: x
Out[5]: <function __main__.x>

In [6]: x()
X

In [7]: foo(x)
Out[7]: <function __main__.x>

In [8]: foo(x)()
foo
X

In [9]: y = task(x)

In [10]: y
Out[10]: <fabric.tasks.WrappedCallableTask at 0x107171350>

# `y` is callable, just like `x`.
In [11]: y()
X

# But I can't decorate it with a decorator that has been created by 
`decorator.decorator`.
In [12]: foo(y)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-12-c34da9d57b44> in <module>()
----> 1 foo(y)

<string> in foo(func)

/Users/tailee/.virtualenvs/ixcmembers/lib/python2.7/site-packages/decorator.pyc 
in decorator(caller, func)
    198         return FunctionMaker.create(
    199             func, "return _call_(_func_, %(shortsignature)s)",
--> 200             evaldict, undecorated=func, __wrapped__=func)
    201     else: # returns a decorator
    202         if inspect.isclass(caller):

/Users/tailee/.virtualenvs/ixcmembers/lib/python2.7/site-packages/decorator.pyc 
in create(cls, obj, body, evaldict, defaults, doc, module, addsource, **attrs)
    182             signature = None
    183             func = obj
--> 184         self = cls(func, name, signature, defaults, doc, module)
    185         ibody = '\n'.join('    ' + line for line in body.splitlines())
    186         return self.make('def %(name)s(%(signature)s):\n' + ibody,

/Users/tailee/.virtualenvs/ixcmembers/lib/python2.7/site-packages/decorator.pyc 
in __init__(self, func, name, signature, defaults, doc, module, funcdict)
    122         assert hasattr(self, 'name')
    123         if not hasattr(self, 'signature'):
--> 124             raise TypeError('You are decorating a non function: %s' % 
func)
    125 
    126     def update(self, func, **kw):

TypeError: You are decorating a non function: <fabric.tasks.WrappedCallableTask 
object at 0x107171350>

Hi Michele and thanks for your work on decorator.py! Sending attached a quick 
patch for allowing methods to be decorated. Tested on 2.7, I haven't jumped on 
Python 3 yet but shouldn't be too hard to adapt it.

Hi Michele. Thanks for your great job!

Wrote a small decorator(django by default caches everything in 'default' cache):

def filesystem_cache(key_prefix, cache_time=None):
    """
    Caches function based on key_prefix and function args/kwargs.
    Stores function result in filesystem cache for a certain cache_time.
    """
    if cache_time:
        FILESYSTEM_CACHE_TIME = cache_time
    else:
        FILESYSTEM_CACHE_TIME = settings.CACHES['filesystem'].get('TIMEOUT')

    @decorator
    def wrapfunc(func, *args, **kwargs):
        ignore_cache = kwargs.pop('_ignore_cache', False)
        full_args = list()
        full_args.extend(args)
        for k, v in kwargs.items():
            full_args.append('%s:%s' % (str(k), str(v)))
        md5_args = md5_constructor(u':'.join([urlquote(var) for var in full_args]))
        cache_key = 'template.cache.%s.%s' % (key_prefix, md5_args.hexdigest())
        filesystem_cache = get_cache(FILESYSTEM_CACHE_NAME)
        cached_value = filesystem_cache.get(cache_key)
        if cached_value and not ignore_cache:
            # if cached value exists - return it
            return cached_value
        result = func(*args, **kwargs)
        filesystem_cache.set(cache_key, result, FILESYSTEM_CACHE_TIME)
        return result
    return wrapfunc

Everything works ok except that all args and kwargs for function fall into args 
section : 
In [2]: @filesystem_cache('bla')
   ...: def hello(a,b,c, d=10,e=12):
   ...:     print a
   ...:     print b
   ...:     print c
   ...:     print d
   ...:     return e

In [3]: hello(5,6,7, d=11,e=18)
> /home/dev/imax/mws/src/imax_main/utils.py(29)wrapfunc()
-> md5_args = md5_constructor(u':'.join([urlquote(var) for var in full_args]))
(Pdb) l
 24             full_args = list()
 25             full_args.extend(args)
 26             for k, v in kwargs.items():
 27                 full_args.append('%s:%s' % (str(k), str(v)))
 28             import pdb; pdb.set_trace()
 29  ->         md5_args = md5_constructor(u':'.join([urlquote(var) for var in full_args]))
 30             cache_key = 'template.cache.%s.%s' % (key_prefix, md5_args.hexdigest())
 31             filesystem_cache = get_cache(FILESYSTEM_CACHE_NAME)
 32             cached_value = filesystem_cache.get(cache_key)
 33             if cached_value and not ignore_cache:
 34                 # if cached value exists - return it
(Pdb) args
func = <function hello at 0xa63f454>
args = (5, 6, 7, 11, 18)
kwargs = {}
(Pdb) kwargs
{}

Why d and e kwargs are in args array ? Thanks !

An example to explain the issue (for python2.6):

@decorator
def example_decorator(func, *ar, **kw):
    print '*ar: ', ar
    print '**kw: ', kw
    func(*ar, **kw)

@example_decorator
def print_pars(z, x, c, v):
    pass

Output:

>>> print_pars(1, 2, 3, 4)
*ar: (1, 2, 3, 4)
**kw: {}

Thus, as you see, **kw dictionary is empty.

documentation*.py scripts return a successful return code even when the tests 
fail. This makes it unsuitable for automated testing.

This happens because the result of doctest.testmod() is unused. It can be 
easily fixed through using it like:

  sys.exit(0 if doctest.testmod()[0] == 0 else 1)

What steps will reproduce the problem?
1. pip install decorator==3.3.0

What is the expected output? What do you see instead?
No distributions matching the version for decorator==3.3.0

What version of the product are you using? On what operating system?
3.3.0

Please provide any additional information below.
Where did it go? My build just broke.
I've tried to use 'pip install -e' but gave up (ignorance and out of time).
Thanks!

Thank you for a nice lib. Good job.  

I have a small issue when using default function or method values. Decorated 
function just fails.

Python 2.7.5 (default, May 15 2013, 22:43:36) [MSC v.1500 32 bit (Intel)] on 
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> from decorator import decorator
>>>
>>> @decorator
... def function(value="default"):
...     print value
...
>>> function()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: function() takes exactly 1 argument (0 given)
>>>


I suppose the code above supposed tp print "default"

I am using decorator==3.4.0

What version of the product are you using? On what operating system?


Please provide any additional information below.

What steps will reproduce the problem?

    #!/usr/bin/env python
    from decorator import decorator

    @decorator
    def f(func, *args, **kwargs):
        print([args, kwargs])
        func(*args, **kwargs)

    @f
    def foo(a, b, c=3):
        pass

    foo(1, 2, c=3)



What is the expected output? What do you see instead?
  Actual: [(1, 2, 3), {}]
  Expected: [(1, 2), {'c': 3}]

What version of the product are you using? On what operating system?
  decorator-3.3.1, Python 2.7.1+, Ubuntu

What steps will reproduce the problem?
1. attempt to use decorator/_decorator.py from IronPython when not using 
FullFrames (either not using -X:FullFrames on the command line, or not 
compiling with PythonOptions.FullFrames)


What is the expected output? What do you see instead?

If you use with FullFrames, it will work. Else it will fail.

What version of the product are you using? On what operating system?

_decorator version 3.3.3 with IronPython 2.9.0.0 on Ubuntu 13.10, Linux.

Please provide any additional information below.

Patch:

https://github.com/ipython/ipython/pull/5055/files#diff-d32e06b23d48ae1c23fe66c8
601ebd97R142

What steps will reproduce the problem?
1. acquire decorator-3.4.2
2. unpack
3.

What is the expected output? What do you see instead?
all the source files

What version of the product are you using? On what operating system?
decorator-3.4.2, gentoo linux

Please provide any additional information below.
files;

documentation .html .pdf .py 3.html 3.pdf 3.py
index.html test.py.

The previous release had most of them.

Here's some example code:

from functools import partial
from operator import gt
from decorator import decorator

@decorator
def dec(func, *args, **kwargs):
    return func(*args, **kwargs)

@dec
def foo(bar: partial(gt, 5)) -> lambda x: x:
    pass


Here's the resulting traceback:

Traceback (most recent call last):
  File "foo.py", line 10, in <module>
    def foo(bar: partial(gt, 5)) -> lambda x: x:
  File "<string>", line 2, in dec
  File "/Users/kamil/.virtualenvs/covenant/lib/python3.2/site-packages/decorator.py", line 200, in decorator
    evaldict, undecorated=func, __wrapped__=func)
  File "/Users/kamil/.virtualenvs/covenant/lib/python3.2/site-packages/decorator.py", line 187, in create
    evaldict, addsource, **attrs)
  File "/Users/kamil/.virtualenvs/covenant/lib/python3.2/site-packages/decorator.py", line 154, in make
    code = compile(src, '<string>', 'single')
  File "<string>", line 1
    def foo(bar: functools.partial(<built-in function gt>, 5)) -> <function <lambda> at 0x1005cf2f8):


The BIF case could possibly be solved by doing some parsing of the annotations 
returned from getfullargspec but I'm not aware of any method of getting the 
code to define the lambda again. Perhaps it would be best to just filter out 
the annotations from the function signature, but still assign them to 
__annotations__ attribute of the resulting function?

Solved with small patch


diff -r b3270ba22de7 decorator/src/decorator.py
--- a/decorator/src/decorator.py        Wed Nov 09 15:24:46 2011 +0100
+++ b/decorator/src/decorator.py        Sat Dec 10 01:14:49 2011 +0300
@@ -88,7 +88,7 @@
                 self.name = '_lambda_' 
             self.doc = func.__doc__
             self.module = func.__module__
-            if inspect.isfunction(func):
+            if inspect.isfunction(func) or inspect.ismethod(func):
                 argspec = getfullargspec(func)
                 for a in ('args', 'varargs', 'varkw', 'defaults', 'kwonlyargs',
                           'kwonlydefaults', 'annotations'):

decorator 3.4.0 fails 1 test with Python 3.4.
(All tests pass with Python 2.6, 2.7, 3.2 and 3.3.)

$ PYTHONPATH="build/lib" python3.4 documentation3.py
**********************************************************************
File "documentation3.py", line 122, in __main__
Failed example:
    print(getargspec(f1))
Expected:
    ArgSpec(args=[], varargs='args', keywords='kw', defaults=None)
Got:
    ArgSpec(args=['x'], varargs=None, keywords=None, defaults=None)
**********************************************************************
1 items had failures:
   1 of  68 in __main__
***Test Failed*** 1 failures.