How to check submodules in Python with hasattr

At runtime, the Python code gets the name of a submodule to load, which I don't know before. Now, I want to check, if this submodule exists inside an existing module. Consider this structure, where foo and bar can be specified:

master/
|
|- __init__.py
|
|- foo/
|  |
|  |- __init__.py
|
|- bar/
   |
   |- __init__.py

Now, usually I do this, which works for defs and variables:

import master

unknown_submodule = "foo"
if hasattr(master, unknown_submodule):
    pass # all's well

or I'm catching the AttributeError, which works equally.

However, with the above file structure, I'm not able to bring this approach up and working. hasattr() return开发者_运维知识库s always False (that is, there is always an AttributeError thrown).

If I look at dir(master), I see this output:

['__builtins__', '__doc__', '__file__', '__name__', '__package__', '__path__']

and even explicitly specifying __all__ in master/__init__.py doesn't help, but changes the dir() to

['__all__', '__builtins__', '__doc__', '__file__', '__name__', '__package__', '__path__']

Any idea, what I'm doing wrong, or if there is a way to achieve these kinds of tests? (Btw: Python 2.6 on Win/Cygwin, if that's of any interest)

Submodules are not attributes of their parent modules unless stated explicitly. Simply try to import the module and catch ImportError:

try:
    __import__("os.peth", fromlist=[os])
except ImportError:
    pass

I recently had to check submodule existence and since I'm using python 3.4.1, I'm giving a new answer to the question:

In python 3.4.1 you can use importlib.util.find_spec(name, package=None) (https://docs.python.org/3.4/library/importlib.html#importlib.util.find_spec)

import importlib
module_exists = importlib.util.find_spec('path.to.my.module')

That easy =)

you can do

try:
 import module.submodule

except ImportError:
  print 'failed or whatever'