Question

I’m getting close to my final goal, which is to generate a nice graph between modules and other imported modules.

For example if x imports from y and z, and y imports from t and v I would like to have:

x -> y, z
y -> t, v

Now I already have my import hook defined as below, but running it on a simple file I don’t get what I would expect:

python study_imports.py CollectImports simple.py

('study_imports.py', 'study_imports')

Where simple.py actually imports from study_imports.
The problem is that I want to see “simple.py” instead of “study_imports.py”, is there a way to get the path of the file actually importing the other module?

class CollectImports(object):
    """
    Import hook, adds each import request to the loaded set and dumps
    them to file
    """

    def __init__(self, output_file):
        self.loaded = set()
        self.output_file = output_file

    def __str__(self):
        return str(self.loaded)

    def cleanup(self):
        """Dump the loaded set to file
        """
        dumped_str = '\n'.join(x for x in self.loaded)
        open(self.output_file, 'w').write(dumped_str)

    def find_module(self, module_name, package=None):
        #TODO: try to find the name of the package which is actually
        #importing something else, and how it's doing it
        #use a defualtdict with empty sets as the storage for this job
        entry = (__file__, module_name)
        self.loaded.add(str(entry))

Answer 1

Maybe with the inspect module.

Module a.py

import inspect

print inspect.stack()

Module b.py

import a

when running b.py, I got :

[
   (<frame object at 0x28a9b70>, '/path/a.py', 5, '<module>', ['print inspect.stack()\n'], 0),
   (<frame object at 0x28a9660>, 'b.py', 2, '<module>', ['import to_import\n'], 0)
]

Looks like the second frame contains what you need.