Develop Reference

inspect.getcode on lambda in class decorator returns entire class in PY3

We are finally migrating from Python 2.7.13 to Python 3.6.6 and I have found some strange behaviour of inspect.getcode that behaves very differently in Python 3.6.6 compared to Python 2.7.13.
We have a number of classes that have a decorator that defines if an instance of this class is allowed to do a certain operation. This check is done by a separate module that manages these decorators rather than by the class itself.
To avoid unintentional changes, we use unit testing to record the code inside the decorator and compare it against a benchmark. The code is recorded with inspect.getcode().
If the decorator uses an external method, then result captured by PY2 and PY3 is identical - the code of the method.
However, if the decorator uses a lambda expression, the result is quite different:
PY2 returns the entire decorator, not just the lambda expression, but that's not really a problem for the unit test.
PY3, however, returns the decorator and the entire class - and this is a problem because now any change in the class would break this unit test.
I could not find any documented difference in inspect.getsource() that would explain this. Am I missing something here or is this a bug? I know that there are easy workarounds, e.g. I could just check the returned string and clip everything below the lambda - but I'd rather understand why this is happening.
Here is some example code that you can directly run in PY2 and PY3 (needs six).
This code is written from scratch to demonstrate the problem - please do not worry about the actual implementation of the decorator or other ways of implementing this.
checks = dict() # a mapping of class to check function
# the class decorator
def check(check_function):
def wrap(wrapped):
checks[wrapped] = check_function
return wrapped
return wrap
# example 1: check with lambda. This demonstrates the problem
#check(lambda cls: True)
class checked_with_lambda(object):
def run(self):
pass
# example 2: check with external helper. This works as expected
def my_checker(obj):
return True
#check(my_checker)
class checked_with_function(object):
def run(self):
pass
# this is how the check would be used. This works fine
c = checked_with_lambda()
if checks[type(c)](c):
c.run()
# this is what a unit test would do
# (I have added the looks-like line to show that the check function is actually what I am expecting)
import six, inspect
for cls, check in six.iteritems(checks):
print('Check for "{}":\n'
'looks like: "{}"\n'
'code: "{}"\n'.format(cls.__name__,
repr(check),
inspect.getsource(check)))
If I run this code with Python 2.7.13, this result is as expected:
Check for "checked_with_lambda":
looks like: "<function <lambda> at 0x0000025967FB2A58>"
code: "#check(lambda cls: True)
"
Check for "checked_with_function":
looks like: "<function my_checker at 0x0000025967FB2AC8>"
code: "def my_checker(obj):
return True
"
In PY3, this looks quite different - note how the check-function code for the first class includes the entire class code:
Check for "checked_with_lambda":
looks like: "<function <lambda> at 0x000002A2B295BC80>"
code: "#check(lambda cls: True)
class checked_with_lambda(object):
def run(self):
pass
"
Check for "checked_with_function":
looks like: "<function my_checker at 0x000002A2B295BD08>"
code: "def my_checker(obj):
return True
"

AFAICT this is a deliberate change in Python 3. inspect.getsource is line based, and in python 3 the inspect.BlockFinder helper class detects the line the lambda definition is in is a decorator and returns the whole decorated object. Python 2 did not do this.
You will just have to work around this.

cleaning up nested function calls

I have written several functions that run sequentially, each one taking as its input the output of the previous function so in order to run it, I have to run this line of code
make_list(cleanup(get_text(get_page(URL))))
and I just find that ugly and inefficient, is there a better way to do sequential function calls?

Really, this is the same as any case where you want to refactor commonly-used complex expressions or statements: just turn the expression or statement into a function. The fact that your expression happens to be a composition of function calls doesn't make any difference (but see below).
So, the obvious thing to do is to write a wrapper function that composes the functions together in one place, so everywhere else you can make a simple call to the wrapper:
def get_page_list(url):
return make_list(cleanup(get_text(get_page(url))))
things = get_page_list(url)
stuff = get_page_list(another_url)
spam = get_page_list(eggs)
If you don't always call the exact same chain of functions, you can always factor out into the pieces that you frequently call. For example:
def get_clean_text(page):
return cleanup(get_text(page))
def get_clean_page(url):
return get_clean_text(get_page(url))
This refactoring also opens the door to making the code a bit more verbose but a lot easier to debug, since it only appears once instead of multiple times:
def get_page_list(url):
page = get_page(url)
text = get_text(page)
cleantext = cleanup(text)
return make_list(cleantext)
If you find yourself needing to do exactly this kind of refactoring of composed functions very often, you can always write a helper that generates the refactored functions. For example:
def compose1(*funcs):
#wraps(funcs[0])
def composed(arg):
for func in reversed(funcs):
arg = func(arg)
return arg
return composed
get_page_list = compose1(make_list, cleanup, get_text, get_page)
If you want a more complicated compose function (that, e.g., allows passing multiple args/return values around), it can get a bit complicated to design, so you might want to look around on PyPI and ActiveState for the various existing implementations.

You could try something like this. I always like separating train wrecks(the book "Clean Code" calls those nested functions train wrecks). This is easier to read and debug. Remember you probably spend twice as long reading your code than writing it so make it easier to read. You will thank yourself later.
url = get_page(URL)
url_text = get_text(url)
make_list(cleanup(url_text))
# you can also encapsulate that into its own function
def build_page_list_from_url(url):
url = get_page(URL)
url_text = get_text(url)
return make_list(cleanup(url_text))

Options:
Refactor: implement this series of function calls as one, aptly-named method.
Look into decorators. They're syntactic sugar for 'chaining' functions in this way. E.g. implement cleanup and make_list as a decorators, then decorate get_text with them.
Compose the functions. See code in this answer.

You could shorten constructs like that with something like the following:
class ChainCalls(object):
def __init__(self, *funcs):
self.funcs = funcs
def __call__(self, *args, **kwargs):
result = self.funcs[-1](*args, **kwargs)
for func in self.funcs[-2::-1]:
result = func(result)
return result
def make_list(arg): return 'make_list(%s)' % arg
def cleanup(arg): return 'cleanup(%s)' % arg
def get_text(arg): return 'get_text(%s)' % arg
def get_page(arg): return 'get_page(%r)' % arg
mychain = ChainCalls(make_list, cleanup, get_text, get_page)
print( mychain('http://is.gd') )
Output:
make_list(cleanup(get_text(get_page('http://is.gd'))))

How to get a functions arguments and values from outside the function?

I have searched a little bit to try to figure this one out but didn't get a solution that I was exactly looking for.
This is my use case:
I would like to evaluate expressions from a functions/methods doc-string against the f/m's parameters and values, but from outside the function (when being called but outside execution of the function
I can't statically change the source code I am evaluating (cant write in new functionality) but dynamically changing (i.e. wrapping the function or adding attributes at run-time) is acceptable
I would prefer to stick with tools in the standard library but am willing to try external libraries if it will make the task a breeze
Here is a simple example of what I am trying to do:
def f1(a,b):
"""a==b"""
pass
def f2(f):
f_locals = "get f's args and values before f is executed"
return eval(f.__doc__,None,f_locals)
>>> f2(f1(2,2))

While I have no clue why you would want to do this, what you've described can be achieved with the inspect module. This example is as close to your original example that I can come up with.
from inspect import getcallargs
def f1(a,b):
"""a==b"""
pass
def f2(f, *f_args, **f_kwargs):
f_callargs = getcallargs(f, *f_args, **f_kwargs)
return eval(f.__doc__, None, f_callargs)
f2(f1, 2, 2)
This should output True.
Keep in mind that this assumes a great many things about the arguments and docstrings of any such functions passed to f2, not the least of which is that none of the examined functions are malicious or malformed. Why don't you want to call functions normally, and why don't you want to change functions?
Edit: As Pajton pointed out, getcallargs is more appropriate here, and removes the calls to both dict and zip. The above code has been updated to reflect this.

I'm not sure if this is what you are looking for, but here's an alternative without inspect module.
#!/usr/bin/python
# -*- coding: utf-8-unix -*-
"""
This is a sample implementation of Inline.pm (Perl) in Python.
Using #inline decorator, it is now possible to write any code
in any language in docstring, and let it compile down to executable
Python code at runtime.
For this specific example, it simply evals input docstring, so
code in docstring must be in Python as well.
"""
# Language compiler for MyLang
class MyLang:
#classmethod
def compile(self, docstring):
# For this example, this simply generates code that
# evals docstring.
def testfunc(*arg, **kw):
return eval(docstring, None, kw)
return testfunc
# #inline decorator
def inline(lang):
def decorate(func):
parm = func.__code__.co_varnames[0:func.__code__.co_argcount]
fgen = lang.compile(func.__doc__)
def wrap(*arg, **kw):
# turn all args into keyword-args
kw.update(dict(zip(parm, arg)))
return fgen(**kw)
return wrap
return decorate
#inline(MyLang)
def myadd(a, b):
"""a + b"""
print(myadd(1, 9))
print(myadd(b = 8, a = 2))
print(myadd(a = 3, b = 7))

Efficient way of having a function only execute once in a loop

At the moment, I'm doing stuff like the following, which is getting tedious:
run_once = 0
while 1:
if run_once == 0:
myFunction()
run_once = 1:
I'm guessing there is some more accepted way of handling this stuff?
What I'm looking for is having a function execute once, on demand. For example, at the press of a certain button. It is an interactive app which has a lot of user controlled switches. Having a junk variable for every switch, just for keeping track of whether it has been run or not, seemed kind of inefficient.

I would use a decorator on the function to handle keeping track of how many times it runs.
def run_once(f):
def wrapper(*args, **kwargs):
if not wrapper.has_run:
wrapper.has_run = True
return f(*args, **kwargs)
wrapper.has_run = False
return wrapper
#run_once
def my_function(foo, bar):
return foo+bar
Now my_function will only run once. Other calls to it will return None. Just add an else clause to the if if you want it to return something else. From your example, it doesn't need to return anything ever.
If you don't control the creation of the function, or the function needs to be used normally in other contexts, you can just apply the decorator manually as well.
action = run_once(my_function)
while 1:
if predicate:
action()
This will leave my_function available for other uses.
Finally, if you need to only run it once twice, then you can just do
action = run_once(my_function)
action() # run once the first time
action.has_run = False
action() # run once the second time

Another option is to set the func_code code object for your function to be a code object for a function that does nothing. This should be done at the end of your function body.
For example:
def run_once():
# Code for something you only want to execute once
run_once.func_code = (lambda:None).func_code
Here run_once.func_code = (lambda:None).func_code replaces your function's executable code with the code for lambda:None, so all subsequent calls to run_once() will do nothing.
This technique is less flexible than the decorator approach suggested in the accepted answer, but may be more concise if you only have one function you want to run once.

Run the function before the loop. Example:
myFunction()
while True:
# all the other code being executed in your loop
This is the obvious solution. If there's more than meets the eye, the solution may be a bit more complicated.

I'm assuming this is an action that you want to be performed at most one time, if some conditions are met. Since you won't always perform the action, you can't do it unconditionally outside the loop. Something like lazily retrieving some data (and caching it) if you get a request, but not retrieving it otherwise.
def do_something():
[x() for x in expensive_operations]
global action
action = lambda : None
action = do_something
while True:
# some sort of complex logic...
if foo:
action()

There are many ways to do what you want; however, do note that it is quite possible that —as described in the question— you don't have to call the function inside the loop.
If you insist in having the function call inside the loop, you can also do:
needs_to_run= expensive_function
while 1:
…
if needs_to_run: needs_to_run(); needs_to_run= None
…

I've thought of another—slightly unusual, but very effective—way to do this that doesn't require decorator functions or classes. Instead it just uses a mutable keyword argument, which ought to work in most versions of Python. Most of the time these are something to be avoided since normally you wouldn't want a default argument value to change from call-to-call—but that ability can be leveraged in this case and used as a cheap storage mechanism. Here's how that would work:
def my_function1(_has_run=[]):
if _has_run: return
print("my_function1 doing stuff")
_has_run.append(1)
def my_function2(_has_run=[]):
if _has_run: return
print("my_function2 doing some other stuff")
_has_run.append(1)
for i in range(10):
my_function1()
my_function2()
print('----')
my_function1(_has_run=[]) # Force it to run.
Output:
my_function1 doing stuff
my_function2 doing some other stuff
----
my_function1 doing stuff
This could be simplified a little further by doing what #gnibbler suggested in his answer and using an iterator (which were introduced in Python 2.2):
from itertools import count
def my_function3(_count=count()):
if next(_count): return
print("my_function3 doing something")
for i in range(10):
my_function3()
print('----')
my_function3(_count=count()) # Force it to run.
Output:
my_function3 doing something
----
my_function3 doing something

Here's an answer that doesn't involve reassignment of functions, yet still prevents the need for that ugly "is first" check.
__missing__ is supported by Python 2.5 and above.
def do_once_varname1():
print 'performing varname1'
return 'only done once for varname1'
def do_once_varname2():
print 'performing varname2'
return 'only done once for varname2'
class cdict(dict):
def __missing__(self,key):
val=self['do_once_'+key]()
self[key]=val
return val
cache_dict=cdict(do_once_varname1=do_once_varname1,do_once_varname2=do_once_varname2)
if __name__=='__main__':
print cache_dict['varname1'] # causes 2 prints
print cache_dict['varname2'] # causes 2 prints
print cache_dict['varname1'] # just 1 print
print cache_dict['varname2'] # just 1 print
Output:
performing varname1
only done once for varname1
performing varname2
only done once for varname2
only done once for varname1
only done once for varname2

One object-oriented approach and make your function a class, aka as a "functor", whose instances automatically keep track of whether they've been run or not when each instance is created.
Since your updated question indicates you may need many of them, I've updated my answer to deal with that by using a class factory pattern. This is a bit unusual, and it may have been down-voted for that reason (although we'll never know for sure because they never left a comment). It could also be done with a metaclass, but it's not much simpler.
def RunOnceFactory():
class RunOnceBase(object): # abstract base class
_shared_state = {} # shared state of all instances (borg pattern)
has_run = False
def __init__(self, *args, **kwargs):
self.__dict__ = self._shared_state
if not self.has_run:
self.stuff_done_once(*args, **kwargs)
self.has_run = True
return RunOnceBase
if __name__ == '__main__':
class MyFunction1(RunOnceFactory()):
def stuff_done_once(self, *args, **kwargs):
print("MyFunction1.stuff_done_once() called")
class MyFunction2(RunOnceFactory()):
def stuff_done_once(self, *args, **kwargs):
print("MyFunction2.stuff_done_once() called")
for _ in range(10):
MyFunction1() # will only call its stuff_done_once() method once
MyFunction2() # ditto
Output:
MyFunction1.stuff_done_once() called
MyFunction2.stuff_done_once() called
Note: You could make a function/class able to do stuff again by adding a reset() method to its subclass that reset the shared has_run attribute. It's also possible to pass regular and keyword arguments to the stuff_done_once() method when the functor is created and the method is called, if desired.
And, yes, it would be applicable given the information you added to your question.

Assuming there is some reason why myFunction() can't be called before the loop
from itertools import count
for i in count():
if i==0:
myFunction()

Here's an explicit way to code this up, where the state of which functions have been called is kept locally (so global state is avoided). I don't much like the non-explicit forms suggested in other answers: it's too surprising to see f() and for this not to mean that f() gets called.
This works by using dict.pop which looks up a key in a dict, removes the key from the dict, and takes a default value to use in case the key isn't found.
def do_nothing(*args, *kwargs):
pass
# A list of all the functions you want to run just once.
actions = [
my_function,
other_function
]
actions = dict((action, action) for action in actions)
while True:
if some_condition:
actions.pop(my_function, do_nothing)()
if some_other_condition:
actions.pop(other_function, do_nothing)()

I use cached_property decorator from functools to run just once and save the value. Example from the official documentation https://docs.python.org/3/library/functools.html
class DataSet:
def __init__(self, sequence_of_numbers):
self._data = tuple(sequence_of_numbers)
#cached_property
def stdev(self):
return statistics.stdev(self._data)

You can also use one of the standard library functools.lru_cache or functools.cache decorators in front of the function:
from functools import lru_cache
#lru_cache
def expensive_function():
return None
https://docs.python.org/3/library/functools.html

If I understand the updated question correctly, something like this should work
def function1():
print "function1 called"
def function2():
print "function2 called"
def function3():
print "function3 called"
called_functions = set()
while True:
n = raw_input("choose a function: 1,2 or 3 ")
func = {"1": function1,
"2": function2,
"3": function3}.get(n)
if func in called_functions:
print "That function has already been called"
else:
called_functions.add(func)
func()

You have all those 'junk variables' outside of your mainline while True loop. To make the code easier to read those variables can be brought inside the loop, right next to where they are used. You can also set up a variable naming convention for these program control switches. So for example:
# # _already_done checkpoint logic
try:
ran_this_user_request_already_done
except:
this_user_request()
ran_this_user_request_already_done = 1
Note that on the first execution of this code the variable ran_this_user_request_already_done is not defined until after this_user_request() is called.

A simple function you can reuse in many places in your code (based on the other answers here):
def firstrun(keyword, _keys=[]):
"""Returns True only the first time it's called with each keyword."""
if keyword in _keys:
return False
else:
_keys.append(keyword)
return True
or equivalently (if you like to rely on other libraries):
from collections import defaultdict
from itertools import count
def firstrun(keyword, _keys=defaultdict(count)):
"""Returns True only the first time it's called with each keyword."""
return not _keys[keyword].next()
Sample usage:
for i in range(20):
if firstrun('house'):
build_house() # runs only once
if firstrun(42): # True
print 'This will print.'
if firstrun(42): # False
print 'This will never print.'

I've taken a more flexible approach inspired by functools.partial function:
DO_ONCE_MEMORY = []
def do_once(id, func, *args, **kwargs):
if id not in DO_ONCE_MEMORY:
DO_ONCE_MEMORY.append(id)
return func(*args, **kwargs)
else:
return None
With this approach you are able to have more complex and explicit interactions:
do_once('foobar', print, "first try")
do_once('foo', print, "first try")
do_once('bar', print, "second try")
# first try
# second try
The exciting part about this approach it can be used anywhere and does not require factories - it's just a small memory tracker.

Depending on the situation, an alternative to the decorator could be the following:
from itertools import chain, repeat
func_iter = chain((myFunction,), repeat(lambda *args, **kwds: None))
while True:
next(func_iter)()
The idea is based on iterators, which yield the function once (or using repeat(muFunction, n) n-times), and then endlessly the lambda doing nothing.
The main advantage is that you don't need a decorator which sometimes complicates things, here everything happens in a single (to my mind) readable line. The disadvantage is that you have an ugly next in your code.
Performance wise there seems to be not much of a difference, on my machine both approaches have an overhead of around 130 ns.

If the condition check needs to happen only once you are in the loop, having a flag signaling that you have already run the function helps. In this case you used a counter, a boolean variable would work just as fine.
signal = False
count = 0
def callme():
print "I am being called"
while count < 2:
if signal == False :
callme()
signal = True
count +=1

I'm not sure that I understood your problem, but I think you can divide loop. On the part of the function and the part without it and save the two loops.

Hashing a python function to regenerate output when the function is modified

I have a python function that has a deterministic result. It takes a long time to run and generates a large output:
def time_consuming_function():
# lots_of_computing_time to come up with the_result
return the_result
I modify time_consuming_function from time to time, but I would like to avoid having it run again while it's unchanged. [time_consuming_function only depends on functions that are immutable for the purposes considered here; i.e. it might have functions from Python libraries but not from other pieces of my code that I'd change.] The solution that suggests itself to me is to cache the output and also cache some "hash" of the function. If the hash changes, the function will have been modified, and we have to re-generate the output.
Is this possible or ridiculous?
Updated: based on the answers, it looks like what I want to do is to "memoize" time_consuming_function, except instead of (or in addition to) arguments passed into an invariant function, I want to account for a function that itself will change.

If I understand your problem, I think I'd tackle it like this. It's a touch evil, but I think it's more reliable and on-point than the other solutions I see here.
import inspect
import functools
import json
def memoize_zeroadic_function_to_disk(memo_filename):
def decorator(f):
try:
with open(memo_filename, 'r') as fp:
cache = json.load(fp)
except IOError:
# file doesn't exist yet
cache = {}
source = inspect.getsource(f)
#functools.wraps(f)
def wrapper():
if source not in cache:
cache[source] = f()
with open(memo_filename, 'w') as fp:
json.dump(cache, fp)
return cache[source]
return wrapper
return decorator
#memoize_zeroadic_function_to_disk(...SOME PATH HERE...)
def time_consuming_function():
# lots_of_computing_time to come up with the_result
return the_result

Rather than putting the function in a string, I would put the function in its own file. Call it time_consuming.py, for example. It would look something like this:
def time_consuming_method():
# your existing method here
# Is the cached data older than this file?
if (not os.path.exists(data_file_name)
or os.stat(data_file_name).st_mtime < os.stat(__file__).st_mtime):
data = time_consuming_method()
save_data(data_file_name, data)
else:
data = load_data(data_file_name)
# redefine method
def time_consuming_method():
return data
While testing the infrastructure for this to work, I'd comment out the slow parts. Make a simple function that just returns 0, get all of the save/load stuff working to your satisfaction, then put the slow bits back in.

The first part is memoization and serialization of your lookup table. That should be straightforward enough based on some python serialization library. The second part is that you want to delete your serialized lookup table when the source code changes. Perhaps this is being overthought into some fancy solution. Presumably when you change the code you check it in somewhere? Why not add a hook to your checkin routine that deletes your serialized table? Or if this is not research data and is in production, make it part of your release process that if the revision number of your file (put this function in it's own file) has changed, your release script deletes the serialzed lookup table.

So, here is a really neat trick using decorators:
def memoize(f):
cache={};
def result(*args):
if args not in cache:
cache[args]=f(*args);
return cache[args];
return result;
With the above, you can then use:
#memoize
def myfunc(x,y,z):
# Some really long running computation
When you invoke myfunc, you will actually be invoking the memoized version of it. Pretty neat, huh? Whenever you want to redefine your function, simply use "#memoize" again, or explicitly write:
myfunc = memoize(new_definition_for_myfunc);
Edit
I didn't realize that you wanted to cache between multiple runs. In that case, you can do the following:
import os;
import os.path;
import cPickle;
class MemoizedFunction(object):
def __init__(self,f):
self.function=f;
self.filename=str(hash(f))+".cache";
self.cache={};
if os.path.exists(self.filename):
with open(filename,'rb') as file:
self.cache=cPickle.load(file);
def __call__(self,*args):
if args not in self.cache:
self.cache[args]=self.function(*args);
return self.cache[args];
def __del__(self):
with open(self.filename,'wb') as file:
cPickle.dump(self.cache,file,cPickle.HIGHEST_PROTOCOL);
def memoize(f):
return MemoizedFunction(f);

What you describe is effectively memoization. Most common functions can be memoized by defining a decorator.
A (overly simplified) example:
def memoized(f):
cache={}
def memo(*args):
if args in cache:
return cache[args]
else:
ret=f(*args)
cache[args]=ret
return ret
return memo
#memoized
def time_consuming_method():
# lots_of_computing_time to come up with the_result
return the_result
Edit:
From Mike Graham's comment and the OP's update, it is now clear that values need to be cached over different runs of the program. This can be done by using some of of persistent storage for the cache (e.g. something as simple as using Pickle or a simple text file, or maybe using a full blown database, or anything in between). The choice of which method to use depends on what the OP needs. Several other answers already give some solutions to this, so I'm not going to repeat that here.