/doculog

Automated Changelog/release note generation

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

doculog

README generated with Documatic.

Quickly generate changelogs and release notes by analysing your git history. A tool written in python, but works on any language.

Getting started

Requirements

  • python >= 3.8
  • git
  • "good" commit messages
  • Git version tags

Minimum python 3.8. Project actively supports python 3.8, 3.9, 3.10. To install, clone the repository and run pip install -e . to package locally OR pip install doculog.

Doculog works by reading git commit messages and inferring what changes are being made. It assumes that you are writing your commit messages as actions: e.g. "Add some feature", "Fix a particular bug". While it's good practice to have the action in the present, imperitive tense, doculog accepts past verbs. See git best practices for more information on this git commit writing style. Standard doculog looks through a list of expected verbs (open an issue/contribute a PR if there are some missing!), but the extended version includes additional logic for classifying commit message, which allows you to be more lax with your commit messages.

API key

To generate a changelog with a full feature-set, doculog requires a (free) API key. Join the waitlist for an API key by signing up here. Someone will be in touch with your API key. In the meantime, doculog works without an API key (you just won't have access to advanced features).

doculog uses python-dotenv to load environment variables stored in a .env file. To use your API key, create a .env file in your project root directory with the following fields:

DOCUMATIC_API_KEY = <your-api-key>

IMPORTANT: DO NOT ADD .env TO VERSION CONTROL. YOUR API KEY MUST BE KEPT SECRET.

Generate a Changelog

In a terminal, run doculog to create a CHANGELOG.md from your git commit history, or update an existing changelog. The "Unreleased" section corresponds to updates not attached to a version. Each changelog update version may contain the following sections: "Added", "Removed", "Deprecated", "Fixed", "Changed". Each section header will only appear in the version if it has at least one update. Note: doculog will overwrite changes made to the "Unreleased" section every time it is run, however tagged versions are not overwritten. Therefore, you can manually edit and add updates to a version release.

To get the best out of the changelog, read the concepts below for information on configuration, git commits and version tags.

Concepts

Git commit parsing

The initial logic for generating a changelog comes from reading your git commit messages. doculog expects commit messages to begin with an imperitive verb, and to written passively. doculog parses the message for signalling words and phrases.

E.g. Rename 'my_func' to 'my_awesome_func' will get interpreted as a "Changed" feature. Whereas 'my_func' -> 'my_awesome_func' will not.

Version tags

Changelogs break down your project's featureset by each release. Currently, doculog infers a release has been made by reading the git tags of your project. If you don't have any git tags, your changelog will only have an "Unreleased" section. To make a git tag, run git tag -a v<MAJOR>-<MINOR>-<PATCH> (and git push --tags to push to your remote); This assumes you're using semver versioning system.

Note: not using semver or git tags to release your project? Open an issue on the doculog repo detailing your method to get it supported by doculog.

Configuration

You can configure how doculog runs by adding a tool.doculog section to pyproject.toml.

Field Purpose Required Default value
changelog Name of changelog file generated. ".md" suffix added if not present. No CHANGELOG.md
project The name of your project. Used to title the changelog No The name of your root project folder
local If true, use a local sever for advanced features. Only used for project development No false

For example, your pyproject.toml file might be:

[tool.doculog]
changelog = "CHANGELOG"
project = "My Cool Project"

Developers

Read the contributing guide for information on coding styles and workflow.

Run pip install -r requirements-dev.txt to get developer requirements.

CI file Purpose
test.yml Linting and unit testing. Runs on every pull request

FAQ

I want more intelligent featureset generation. What can I do?

Request access to the free Documatic API to generate a changelog driven by machine learning. Follow Documatic on GitHub and socials to stay up to date with the latest features and releases.

How do I get my API key?

Once you've joined the waitlist, we will be in touch shortly with your API key.

The changelog is great, but I want more!

Get in touch - info@documatic.com.

I'm not getting a complete changelog. What's gone wrong?

Check that you have appropriate version tags and commit messages. If you have the advanced featureset (i.e. have an API key) then you will get better changelog updates which don't require you to follow the commit process so strictly. If you're still not getting good results, please open a bug report.

Can I contribute to doculog?

Absolutely: feature requests, bug fixes, bug reports and PRs of all shapes and sizes are welcome. See the developers section.

License

Licensed under GNU GPL3. Please see the [LICENSE] for terms in full.

Generated by Documatic.