github-to-gitlab-backup

A shell script to retrieve the public repos from a user in GitHub and sync them to a GitLab instance

It will retrieve and sync the following information for each public repo:

• all the branches
• all the tags
• latest 10 releases & assets (asset retention of latest 10 releases, can be disabled in delete_asset function)
• wiki if available

Prerequisites

• GITHUB_TOKEN: the user's GitHub token
• GITLAB_URL: the target GitLab instance
• GITLAB_TOKEN: the user's token in the target GitLab instance
• STORAGE_URL: the dufs instance url

All urls should NOT end with a slash (/).

Install packages:

sudo apt install curl jq

How to run

Make file executable:

chmod a+x script.sh

Execute:

./script.sh --github-username "BitWarden" --gitlab-id "418"

or clone specific repositories only:

# Include specific repositories only
./script.sh --github-username "BitWarden" --gitlab-id "418" --include-only "server clients"
# Exlude specific repositories
./script.sh --github-username "BitWarden" --gitlab-id "418" --exclude "server clients"
# Skip first 10 repositories (ordered by name)
./script.sh --github-username "BitWarden" --gitlab-id "418" --skip 10

The option --exclude has a higher priority than --include-only. If a repository is mentioned in both options, it will be excluded.

The option --skip will be executed after filtering with the other two options. For example for the repostitories repo1, repo2, repo3 and repo4, by calling the command:

./script.sh --exclude "repo2" --skip 2

we will fetch only repo4 (repo1 and repo3 will be skipped from option --skip).