This is the Python version of the Codemodder Framework.

Codemodder is sponsored by pixee.ai.

As of v0.80.0 the codemod API is relatively stable. However, backwards compatibility between releases will not be guaranteed until version 1.0.0.

See CHANGELOG.md for detailed release notes.

The codemodder package is available on PyPI. To install, run:

$ pip install codemodder

WARNING: You should also install the semgrep optional dependency for now as it isn't deprecated yet:

$ pip install codemodder[semgrep]

To install the package from source, use pip:

$ pip install /path/to/codemodder-python

Codemodder can be run as a CLI. The codemodder package provides an executable called codemodder. This should be available on your path by default after installation.

For basic usage, run the codemodder command with a target directory path:

$ codemodder /path/to/my-project

Note that by default codemodder will make changes to files in your target directory. To run codemodder without making updates on disk, use the --dry-run flag:

$ codemodder --dry-run /path/to/my-project

To list all available codemods (including any that are registered with installed plugins), use the --list option:

$ codemodder --list

For a full list of options, use the --help flag:

$ codemodder --help

You can also run codemodder as a library by importing the module and running run. For basic usage, pass a target directory path and the dry_run argument:

import codemodder

output, exit_code = codemodder.run("/path/to/my-project", dry_run=True)

Unlike the CLI which has a default dry_run of False, when calling codemodder as a library you must indicate if you want codemodder to make changes to your files.

Codemods are composed of the following key components:

• Detector
• Transformer(s)
• Metadata

The Python codemodder supports a plugin infrastructure for custom codemods. For users interested in developing a custom codemod plugin, we recommend starting with the Cookiecutter template.

Coming soon!

See CONTRIBUTING.md.