Skip to content

README.md

Latest commit

 

History

History
105 lines (66 loc) · 3.26 KB

README.md

File metadata and controls

105 lines (66 loc) · 3.26 KB

# module-inject

module-inject is a Python library designed to inject configuration values into modules before they are imported, enabling centralized and declarative control over runtime behavior across dynamic or plugin-based systems.

This is particularly useful in applications where modules need to be configured externally, but standard import mechanisms do not allow mutation of module-level state prior to execution.

Features

  • Inject values into modules before import-time execution
  • Fine-grained control with per-module and per-package scopes
  • Lazy and dynamic module loading after configuration
  • Optional type enforcement and stack-based namespace lookup

Installation

pip install module-inject

Usage

Consider a plugin system where plugins are python modules under a plugins/ directory need to be initialized with context-specific parameters.

Example

main.py

from pathlib import Path
from module_inject import Module

# Create a controller for the package
package = Module.from_import_path("plugins")

# Set a package-scoped variable
package.set(role="slave")

# Set a variable for the current script
Module.from_path(Path(__file__)).set(role="master")

# Iterate over all module files and inject values before importing
for path in Path("plugins").glob("*.py"):
    module = Module.from_path(path)
    module.set(variable=f"value for {module.import_path}")
    module.load().init()

plugins/interesting_plugin.py

import module_inject as module

def init():
    role = module.get("role", str, package=True)
    print(role)

# Module-level config access
print(module.get("variable"))
print(module.get_namespace())

Output

value for plugins.interesting_plugin
{'variable': 'value for plugins.interesting_plugin'}
slave

API

Module.from_import_path(path: str) -> Module

Creates a Module controller for the given import path (e.g., "plugins.interesting_plugin").

Module.from_path(path: Path) -> Module

Creates a Module controller from a filesystem path to a .py file. Searches for the file in sys.path under the hood to resolve import path.

module.get(key: str, type_: type[T] = None, package: bool = False, recursive: bool = False) -> T

Retrieves the value associated with key. If type_ is set - checks if the value in namespace is instance of the provided type.

If package=True, it will attempt to retrieve the value from the package scope rather than the module-specific scope.

If recursive=True, it will walk over the call stack to find the name.

module.get_namespace() -> dict

Returns a dictionary of all key-value pairs injected into the current module.

Module.set(**kwargs) -> None

Sets key-value pairs to be injected into the associated module upon import.

Module.load() -> types.ModuleType

Loads (imports) the target module with the injected configuration. Must be called to finalize injection.

Design Philosophy

This library provides a low-level interface to orchestrate dynamic configuration of Python modules, particularly in systems requiring deferred or context-aware initialization. Unlike environment variables or global config objects, this approach allows values to be injected into module namespace dictionaries before execution.