svn to git migration

svn_to_git.sh

#!/bin/bash

# SVN to Git Migration Script
# ==========================
#
# This script performs a complete migration from SVN to Git while preserving
# history, usernames, branches, and tags.
#
# Usage:
#   ./svn-to-git-migrate.sh <SVN_REPO_URL> <GIT_REPO_NAME>
#
# Examples:
#   ./svn-to-git-migrate.sh http://svn.example.com/repo my-project
#   ./svn-to-git-migrate.sh svn://svn.example.com/repo/trunk project-name
#   ./svn-to-git-migrate.sh file:///path/to/local/svn/repo local-project
#
# Prerequisites:
#   - git
#   - git-svn
#   - svn
#
# The script will:
#   1. Create an authors mapping file (authors.txt)
#   2. Convert SVN history to Git format
#   3. Convert SVN branches and tags to Git format
#   4. Create a server-ready Git repository
#   5. Provide instructions for pushing to GitHub/GitLab
#
# Important Notes:
#   - Review authors.txt before proceeding with migration
#   - Ensure sufficient disk space (at least 2x the SVN repository size)
#   - Large repositories may take several hours to migrate
#   - The script assumes standard SVN layout (trunk/branches/tags)
#   - Have your GitHub/GitLab repository URL ready for the final push
#
# Output:
#   - authors.txt: SVN to Git username mapping
#   - migration.log: Detailed migration log
#   - temp_migration/: Temporary working directory
#   - <GIT_REPO_NAME>.git/: Final server-ready Git repository with complete history
#
# Post-Migration:
#   The script will provide instructions for pushing to GitHub/GitLab:
#   1. Create an empty repository on GitHub/GitLab (DO NOT initialize with README)
#   2. Follow the provided commands to push all branches and tags
#
# Example authors.txt format:
#   svnuser = Git User <[email protected]>
#   john = John Doe <[email protected]>
#

set -e

SVN_REPO_URL=""
GIT_REPO_NAME=""
AUTHORS_FILE="authors.txt"
TEMP_DIR="temp_migration"
LOG_FILE="migration.log"

# Function to show usage
show_usage() {
    grep '^#' "$0" | tail -n +2 | cut -c 3-
    exit 1
}

# Function to log messages
log_message() {
    echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1" | tee -a "$LOG_FILE"
}

# Function to verify prerequisites
check_prerequisites() {
    log_message "Checking prerequisites..."
    
    # Check for required commands
    commands=("git" "git-svn" "svn")
    for cmd in "${commands[@]}"; do
        if ! command -v "$cmd" >/dev/null 2>&1; then
            log_message "ERROR: $cmd is not installed"
            exit 1
        fi
    done
}

# Function to create authors file
create_authors_file() {
    log_message "Creating authors mapping file..."
    
    if [ -f "$AUTHORS_FILE" ]; then
        log_message "Authors file already exists, skipping..."
        return
    }

    svn log -q "$SVN_REPO_URL" | \
        awk -F '|' '/^r/ {sub("^ ", "", $2); sub(" $", "", $2); print $2}' | \
        sort -u | \
        while read -r author; do
            echo "$author = $author <[email protected]>"
        done > "$AUTHORS_FILE"
    
    log_message "AUTHORS_FILE created. Please review and update email addresses!"
    exit 1
}

# Function to display post-migration instructions
show_post_migration_instructions() {
    local repo_name="$1"
    
    echo
    echo "=== Post-Migration Instructions ==="
    echo
    echo "Your repository is ready! Follow these steps to push to GitHub or GitLab:"
    echo
    echo "1. Create a new empty repository on GitHub or GitLab"
    echo "   - DO NOT initialize it with a README, license, or .gitignore"
    echo "   - GitHub: Click 'New' at github.com/yourname?tab=repositories"
    echo "   - GitLab: Click 'New project' at gitlab.com/projects/new"
    echo
    echo "2. Push the repository (replace REMOTE_URL with your GitHub/GitLab URL):"
    echo
    echo "   cd ${repo_name}.git"
    echo "   # For GitHub:"
    echo "   git remote add origin https://github.com/YOUR_USERNAME/${repo_name}.git"
    echo "   # OR for GitLab:"
    echo "   git remote add origin https://gitlab.com/YOUR_USERNAME/${repo_name}.git"
    echo
    echo "   # Push all branches and tags:"
    echo "   git push origin --all"
    echo "   git push origin --tags"
    echo
    echo "3. Verify the migration:"
    echo "   - Check the repository on GitHub/GitLab"
    echo "   - Verify branches and tags are present"
    echo "   - Review commit history"
    echo "   - Clone a fresh copy and check the contents:"
    echo "     git clone https://github.com/YOUR_USERNAME/${repo_name}.git"
    echo
    echo "4. Clean up local migration files (optional):"
    echo "   cd .."
    echo "   rm -rf temp_migration"
    echo "   # Keep ${repo_name}.git if you want a local backup"
    echo
    echo "Migration log is available in: migration.log"
    echo
}

# Function to perform the migration
perform_migration() {
    log_message "Starting migration process..."
    
    # Create and enter temporary directory
    mkdir -p "$TEMP_DIR"
    cd "$TEMP_DIR"

    # Initialize git-svn
    log_message "Initializing git-svn..."
    git svn init "$SVN_REPO_URL" --no-metadata \
        --authors-file="../$AUTHORS_FILE" \
        --trunk=trunk \
        --tags=tags \
        --branches=branches \
        --prefix=origin/

    # Fetch SVN repository
    log_message "Fetching SVN repository (this may take a while)..."
    git svn fetch

    # Fix tags
    log_message "Converting SVN tags to Git tags..."
    git for-each-ref refs/remotes/origin/tags | cut -d / -f 5- | \
    while read -r tag; do
        git tag "$tag" "origin/tags/$tag^" || true
        git branch -r -d "origin/tags/$tag" || true
    done

    # Fix branches
    log_message "Converting SVN branches to Git branches..."
    git for-each-ref refs/remotes/origin | cut -d / -f 4- | \
    while read -r branch; do
        if [ "$branch" != "trunk" ]; then
            git branch "$branch" "refs/remotes/origin/$branch" || true
            git branch -r -d "origin/$branch" || true
        fi
    done

    # Create master from trunk
    log_message "Creating master branch from trunk..."
    git branch -m trunk master

    # Clean up
    log_message "Cleaning up..."
    git gc --aggressive
    git prune

    # Create server-ready repository
    log_message "Creating server-ready repository with complete history..."
    cd ..
    git clone --bare "$TEMP_DIR" "$GIT_REPO_NAME.git"
    
    # Verify repository integrity
    cd "$GIT_REPO_NAME.git"
    log_message "Verifying repository integrity..."
    git fsck --full
    
    log_message "Migration completed successfully! Repository is ready for GitHub/GitLab."
    
    # Return to original directory
    cd ..
    
    # Show instructions for pushing to GitHub/GitLab
    show_post_migration_instructions "$GIT_REPO_NAME"
}

# Main execution
main() {
    # Show usage if --help flag is passed or no arguments provided
    if [ "$1" = "--help" ] || [ "$1" = "-h" ] || [ -z "$1" ] || [ -z "$2" ]; then
        show_usage
    fi

    SVN_REPO_URL="$1"
    GIT_REPO_NAME="$2"

    # Initialize log file
    echo "=== Migration Log Started $(date) ===" > "$LOG_FILE"

    check_prerequisites
    create_authors_file
    perform_migration
}

main "$@"