This script automatically mirrors GitHub starred repositories to a Gitea instance.

mirror_github_starred_repos.sh

#!/bin/bash

#########################################################################
# GitHub to Gitea Starred Repository Mirror
#
# This script automatically mirrors your GitHub starred repositories to
# a Gitea instance.
#
# - Fetches all repositories starred on GitHub
# - Recreate organization / owner (as org) to mirror repos into
# - Creates mirror repositories on Gitea for each starred GitHub repo
# - Handles pagination for users with many starred repositories
#
# Requirements:
# - GitHub Personal Access Token with permissions to read starred repos
# - Gitea Access Token with organization and repository write access
# - curl, jq installed on your system
#########################################################################

# Read variables from .env file.
if ! [ -f ".env" ]; then
  echo "No .env file found to source variables from."
  echo "Please create a .env file with the variables:"
  echo 'GITHUB_USER="jane_doe"'
  echo 'GITHUB_TOKEN="github_pat_..."'
  echo 'GITEA_URL="https://gitea.example.com"'
  echo 'GITEA_USER="jane_doe"'
  echo 'GITEA_TOKEN="..."'
  exit 1
fi

source '.env'

# Make sure all variables are set.
required_variables=("GITHUB_USER" "GITHUB_TOKEN" "GITEA_URL" "GITEA_USER" "GITEA_TOKEN")
for variable_name in "${required_variables[@]}"; do
  if ! [ -v $variable_name ]; then
    echo "$variable_name is not set."
    exit 1
  fi
done

# $1: method (e.g. POST)
# $2: endpoint (e.g. users/some_user/starred?page=1)
# $3: data (e.g. '{"key": "value"}')
function github_query() {
  method=$1
  endpoint=$2
  if [ -v 3 ]; then
    data=$3
    curl -s -X "$method" -H "Authorization: token $GITHUB_TOKEN" -H "Accept: application/vnd.github.v3+json" -d "$data" "https://api.github.com/$endpoint"
  else
    curl -s -X "$method" -H "Authorization: token $GITHUB_TOKEN" -H "Accept: application/vnd.github.v3+json" "https://api.github.com/$endpoint"
  fi
}

function gitea_query() {
  method=$1
  endpoint=$2
  if [ -v 3 ]; then
    data=$3
    curl -s -X "$method" -H "Authorization: token $GITEA_TOKEN" -H "Content-Type: application/json" -d "$data" "$GITEA_URL/api/v1/$endpoint"
  else
    curl -s -X "$method" -H "Authorization: token $GITEA_TOKEN" -H "Content-Type: application/json" "$GITEA_URL/api/v1/$endpoint"
  fi
}


echo "Fetching Gitea organisations..."
page=1
existing_orgs=()

while true; do
  orgs_page=$(gitea_query "GET" "orgs?page=$page")

  if [ "$(echo "$orgs_page" | jq '. | length')" = "0" ]; then
    break
  fi

  page=$((page + 1))

  while read -r org; do
    name=$(echo "$org" | jq -r '.name')
    existing_orgs+=("$name")
  done < <(echo "$orgs_page" | jq -c '.[]')
done

echo "Found ${#existing_orgs[@]} organisations on Gitea."

# Flatten for regex-check below.
existing_orgs=$(printf "%s\n" "${existing_orgs[@]}")


echo "Fetching Gitea repositories..."
page=1
existing_repos=()

while true; do
  repos_page=$(gitea_query "GET" "repos/search?page=$page")

  if [ "$(echo "$repos_page" | jq '.data | length')" = "0" ]; then
    break
  fi

  page=$((page + 1))

  while read -r repo; do
      name=$(echo "$repo" | jq -r '.full_name')
      existing_repos+=("$name")
  done < <(echo "$repos_page" | jq -c '.data[]')
done

echo "Found ${#existing_repos[@]} repositories on Gitea."

# Flatten for regex-check below.
existing_repos=$(printf "%s\n" "${existing_repos[@]}")


echo "Fetching starred repositories from GitHub..."
page=1

while true; do
  starred_page=$(github_query "GET" "users/$GITHUB_USER/starred?per_page=100&page=$page")

  if [ "$(echo "$starred_page" | jq '. | length')" = "0" ]; then
    break
  fi

  page=$((page + 1))

  while read -r repo; do
    # Read values in one query for efficiency, needs readarray because of multi-line output.
    readarray -t repo_info < <(echo "$repo" | jq -r '.name, .clone_url, .description // "", .owner.login')
    name=${repo_info[0]}
    url=${repo_info[1]}
    description=${repo_info[2]}
    owner=${repo_info[3]}

    # Skip if repo already exists in Gitea.
    if echo "$existing_repos" | grep -q "^$owner/$name$"; then
      continue
    fi

    echo "Found new repository to mirror: $owner/$name..."

    if ! echo "$existing_orgs" | grep -q "^$owner$"; then
      echo "Creating new organisation: $owner..."
      query=$(jq -n --arg name "$owner" '{"username": $name, "website": "https://github.com/\($name)"}')
      result=$(gitea_query "POST" "orgs" "$query")

      id=$(echo "$result" | jq -r '.id')
      if [ -z "$id" ] || [ "$id" = "null" ]; then
        echo "Error creating Gitea organisation: $result"
        continue
      fi
    fi

    query=$(jq -n --arg name "$name" --arg url "$url" --arg description "$description" --arg owner "$owner" --arg user "$GITHUB_USER" --arg token "$GITHUB_TOKEN" '{
      "clone_addr": $url,
      "repo_owner": $owner,
      "repo_name": $name,
      "description": $description,
      "mirror": true,
      "private": false,
      "lfs": true,
      "auth_username": $user,
      "auth_token": $token
    }')
    result=$(gitea_query "POST" "repos/migrate" "$query")

    gitea_repo_url=$(echo "$result" | jq -r '.clone_url')
    if [ -z "$gitea_repo_url" ] || [ "$gitea_repo_url" = "null" ]; then
      echo "Error creating mirror: $result"
      continue
    fi

    echo "Successfully set up mirror for $owner/$name."
  done < <(echo "$starred_page" | jq -c '.[]')
done

echo "Mirroring process completed!"