This package can be used to extract functions, import statements (TODO: class) in a module file, and their call dependencies in a source repository.

Example:

Consider a simple repository structure

simple_repo
  |
  |_ file1.py
  |_ folder1
        |_ file2.py     

The content in each file is

# file1.py
def print_function(str):
  print(str)

# folder1/file2.py
from file1 import print_function

def start_print():
  print_function("Author: Nam Le Hai")

def print_hello():
  print_function("hello")

The tool can use to find the dependency of print_hello() function in file2.py which is the print_function() function imported from file1.py.

Currently, our package only supports extracting dependency for functions and import statements in a module file.

Python >= 3.10

Using Anaconda (feel free to use other env)

conda create -n YOUR_ENV_NAME python=3.10
conda activate YOUR_ENV_NAME
pip install pydepcall

from pydepcall import Extractor

If you want to extract all module files in the repository

reposrc = YOUR_LOCAL_PATH_OF_REPO
extractor = Extractor(reposrc)
output = extractor.extract()

If you want to extract a specific module file in the repository

reposrc = YOUR_LOCAL_PATH_OF_REPO
module_file = YOUR_LOCAL_PATH_OF_FILE_IN_REPO
extractor = Extractor(reposrc, module_file)
output = extractor.extract()

Example: See more in example

>>> from pydepcall import Extractor
>>> reposrc = "simple_repo"
>>> extractor = Extractor(reposrc)
>>> output = extractor.extract()
>>> output
{'simple_repo/file1.py': <pydepcall.Node.ModuleNode object at 0x7faeb6d84580>, 'simple_repo/folder1/file2.py': <pydepcall.Node.ModuleNode object at 0x7faeb7822050>}

>>> output["simple_repo/folder1/file2.py"].function_list
[<pydepcall.Node.FunctionNode object at 0x7faeb6bc2740>, <pydepcall.Node.FunctionNode object at 0x7faeb6bc2530>]

>>> output["simple_repo/folder1/file2.py"].function_list[0].children
[<pydepcall.Node.ImportNode object at 0x7fc5fade6320>]

>>> output["simple_repo/folder1/file2.py"].function_list[0].children[0].children
[<pydepcall.Node.FunctionNode object at 0x7faeb6bc22c0>]

>>> output["simple_repo/folder1/file2.py"].function_list[0].children[0].children[0].content
'def print_function(str):\n  print(str)'

For extracting a specific module, the output will be a ModuleNode of the input file.

For extracting the whole repository, the output will be the dictionary of ModuleNode with the keys are all module files in the repository.

The package has 5 main nodes:

• ModuleNode: contains all FunctionNode and ImportNode in a module file
• ImportNode: a node represents an import statement
• FunctionNode: a node represents a python function
• ClassNode: a node represents a class
• BlockNode: a node represents a codeblock in the module file

Every node except ModuleNode has the following attributes:

• path: the module file's local path contains that node
• content: the text content of the node (function, import, class or codeblock)
• position_in_file: the position of the node in the module file

For FunctionNode and ImportNode, we can acquire their dependencies through their children (node.children) attribute.

Please see Node.py for more details.