ClusterFuzz tools

Status: Early prototype phase

The tools supports various tasks (e.g. reproduce a crash locally) needed by ClusterFuzz's users.

Installation

pip install clusterfuzz

Usage

See clusterfuzz --help.

Currently, it supports reproducing a crash locally. In the future, it will support uploading a fuzzer, tailing fuzzer log, and uploading a testcase.

Develop

1. Install virtualenv.
2. Setting up virtualenv by running virtualenv ENV and source ENV/bin/activate.
3. Install dependencies by running pip install -r requirements.txt.
4. Build "the development mode" binary by running pip install -e ..
5. Run the binary by running clusterfuzz --help.
6. Run the tests python test.py --help.
7. Run the tests with coverage coverage run test.py.
8. Generate the report with coverage html -d /tmp/coverage and see /tmp/coverage/index.html.

Please note that we need to run pip install -e . before running clusterfuzz if the code has been changed.

Publish

1. Create and merge a pull request to increase the version number.
2. Clear previously built artifacts with rm -rf dist.
3. Build artifact by running python setup.py sdist bdist_wheel.
4. Upload to pypi.python.org by running twine upload dist/*.
5. Tag the current version with git tag -a <version> -m "Version <version>".
6. Push the tag git push --tags.