Python Splitter

Python Splitter is a developer-focused tool designed to convert large Python scripts into modular units suitable for packaging and collaboration. It targets engineers, data scientists, and automation engineers who maintain codebases that grow beyond a single file. By producing coherent module boundaries, it reduces cognitive load during code reviews and simplifies dependency management, testing, and deployment workflows.
It analyzes Python source using the AST, identifying top-level definitions, imports, and __main__ logic. It supports splitting by function or class boundaries, by module-like regions marked with markers, and by explicit grouping rules. The process preserves imports and relative references, creates separate module files, and optionally generates __init__.py to form a package.
Compared to manual splitting, the tool offers deterministic results, handles name collisions with configurable renaming, and exports an optional package bundle ready for distribution. It is especially valuable for refactoring legacy libraries, creating micro-packages, or isolating experiment code in data projects. It also supports educational workflows by illustrating how large scripts can be decomposed into reusable components.
Core features include AST-based splitting, configurable boundaries, import and dependency preservation, optional docstring extraction, and an output structure matching common Python packaging conventions. Edge cases such as dynamic imports, eval/exec usage, and circular dependencies are flagged, with safe fallbacks to a line-based split. The tool does not execute code, nor run tests, and assumes the input conforms to Python syntax after preprocessing.