Semanticore Release Bot 🤖 🦁 🐉

About

Your friendly Semanticore Release bot helps maintaining the changelog for a project and automates the related tagging process.

How to use it

Semanticore runs along every pipeline in the main branch, and will analyze the commit messages.

It maintains an open Merge Request for the project with all the required Changelog adjustments.

It detects the current version and suggests the next version based on the changes made.

Once a release commit is detected, it will automatically create the related Git tag on the next pipeline run.

Conventions

• Commit messages should follow the Conventional Commits so semanticore can decide whether a minor or patch level release is required.
• Releases are indicated with a commit with a commit messages which should match: Release vX.Y.Z

Supported Commit Types

Currently Semanticore supports the following commit types:

Type	Prefixes	Meaning
🆕 Feature	feat	New Feature, creates a minor commit
🚨 Security Fix	sec	Security relevant change/fix
👾 Bugfix	fix, bug	Bugfix
🛡 Test	test	(Unit-)Tests
🔁 Refactor	refactor, rework	Refactorings or reworking
🤖 Devops/CI	ops, ci, cd, build	Operations, Build, CI/CD, Pipelines
📚 Documentation	doc	Documentation
⚡️ Performance	perf	Performance improvements
🧹 Chore	chore, update	Chores, (Dependency-)Updates
📝 Other	everything else	Everything not matched by another prefix

Major versions

To enable support for major releases (breaking APIs), use the -major flag.

Configuration

The SEMANTICORE_TOKEN is required - that's a Gitlab or Github Token which has basic contributor rights and allows to perform the related Git and API operations.

Set Author and committer

Semanticore respects Git Environment variables

• GIT_AUTHOR_NAME
• GIT_AUTHOR_EMAIL
• GIT_COMMITTER_NAME
• GIT_COMMITTER_EMAIL

The values can also be overridden by adding the appropriate flags. Run with -help to get the details.

If none of these is set, Semanticore will use Semanticore Bot as name and [email protected] as E-Mail for Author and Committer.

Using Semanticore

To test Semanticore locally you can run it without an API token to create an example Changelog:

go run github.com/aoepeople/semanticore@v0 <optional path to repository>

Example Configurations

Github Action

.github/workflows/semanticore.yml

name: Semanticore

on:
  push:
    branches:
      - main
jobs:
  semanticore:
    runs-on: ubuntu-latest
    name: Semanticore
    steps:
      - uses: actions/checkout@v3
        with:
          fetch-depth: 0
      - name: Setup Go
        uses: actions/setup-go@v3
        with:
          go-version: '1.*'
      - name: Semanticore
        run: go run github.com/aoepeople/semanticore@v0
        env:
          SEMANTICORE_TOKEN: ${{secrets.GITHUB_TOKEN}}

Gitlab CI

Create a secret SEMANTICORE_TOKEN containing an API token with api and write_repository scope.

.gitlab-ci.yml

stages:
  - semanticore

semanticore:
  image: golang:1
  stage: semanticore
  script:
    - go run github.com/aoepeople/semanticore@v0
  only:
    - main

Make sure you set the repositories clone depth too a large enough value, the default of 50 might be too low.