python-require (repyrer)

Provides nodejs-style requiring of modules for python

This module allows imports of python modules using syntax similar to the node.js require function. In particular, it takes string arguments comprised of (optionally) a UNIX-style directory path and a module name, and returns a reference to the loaded module. Specifying no directory path will have the module import attempted first in the current working directory, then on the python path variable (sys.path), allowing both local-to-the-folder imports and standard library imports (as well as anything else you've configured into sys.path). If an (absolute) directory path is included (i.e. a leading / in the directory path), the module will first be attempted to be loaded from that directory first, then the rest of the sys.path as a fallback. If a relative directory path is provided (like lib/module_name or ../other_lib/module_name) the current working directory is used to resolve the relative path.

NOTE: the leading ./ characters on a directory path (as in ./lib/module_name) indicate that the directory path should be resolved relative to the current file NOT the current working directory. For example, suppose you have the following file structure:

top_folder
|-----main.py
|-----lib
  |-----A.py
  |-----sub_lib
    |-----B.py

Suppose main.py imports A.py as a module, and A.py in tern imports B.py as a module. Inside of A.py, it should use B = repyre('./sublib/B'), NOT B = repyre('sublib/B') since the latter will fail when main.py imports A.py (because the current working directory will be top_folder NOT lib). However, main.py is free to import A.py as A = repyre('lib/A') or A = repyre('./lib/A') provided main.py will only ever be run directly as a script from top_folder.

Example Usage

You can use repyrer exactly as you would import to access the standard library, and anything else you might have pip installed:

>>> from repyrer import repyre
>>> os = repyre('os') # import standard library
>>> np = repyre('numpy') # import other libraries installed with pip, etc.
>>> plt = repyre('matplotlib.pyplot') # import sub-modules using '.' notation

You can also specify a directory path (relative to the current working directory, or absolute) preceding the module name to access local libraries, suppose you have a file structure like:

folder
|----CWD
  |----main.py
  |----adjacent_file.py
  |----my_dev_lib
    |-----module_A.py
|----my_other_lib
  |-----module_B.py

Then, in the python shell

>>> adjacent = repyre('adjacent_file') # import other files relative to the CWD
>>> A = repyre('my_dev_lib/module_A') # import modules from a relative directory path (relative to CWD)
>>> B = repyre('../my_other_lib/module_B') # can use '..' in directory paths

And in a script like main.py:

from repyrer import repyre

adjacent = repyre('adjacent_file') # only works if script is run in this folder
adjacent = repyre('./adjacent_file') # safe to run from anywhere