This actions synchronizes a monolithic repository to standalone repositories by using splitsh-lite.
Specify using
withkeyword
config-path- (Required) Location of the subtree split mapping configurationbatch-size- How many subtreesplits should be processed in parallel. It is recommended to keep this amount low (or even at 1) for large repositories.
Create a configuration file with the mapping of the different subtree splits. Each subtree split item contains a name (must be unique), directory in the current repository and a target git repository.
{
"subtree-splits": [
{
"name": "core",
"directory": "src/Core",
"target": "git@github.com:example/core-package.git"
}
]
}
Example workflow to sync commits and tags.
on:
push:
# Only trigger for specific branches or changes in specific paths.
branches:
- '*'
paths:
- src/**
# Tag push events should be ignored, they will be handled with the create event below.
tags-ignore:
- '*'
create:
tags:
- '*'
delete:
tags:
- '*'
jobs:
sync_commits:
runs-on: ubuntu-latest
name: Sync commits
if: github.repository == 'your-org/your-repository' # Execute this workflow job only on the main repository.
steps:
- uses: actions/checkout@v2
with:
persist-credentials: false
fetch-depth: 0
# Add a personal access token to the repository secrets. This will allow the splitter action to push the new commits
- uses: frankdejonge/use-github-token@1.0.2
with:
authentication: 'username:${{ secrets.PERSONAL_GITHUB_TOKEN }}'
user_name: 'Committer name'
user_email: 'Committer email'
# Cache the splitsh executable to speedup future runs
- name: Cache splitsh-lite
uses: actions/cache@v2
with:
path: './splitsh'
key: '${{ runner.os }}-splitsh-v101'
# Sync commits and tags for the configured subtree splits
- name: subtree split
uses: acrobat/subtree-splitter@v1.1.3
with:
config-path: .github/subtree-splitter-config.json # Reference the location where you saved your config file