gc.zsh - Smart Git Clone with Organized Directory Structure

gc.zsh

#!/usr/bin/env zsh
#
# gc - Git Clone with Smart Directory Organization
#
# This function clones Git repositories into a well-organized directory structure
# based on the repository's domain, user/org, and name.
#
# USAGE:
#   gc <git-url>
#
# EXAMPLES:
#   gc https://github.com/kubernetes/kubernetes
#   # Creates: ~/src/github.com/kubernetes/kubernetes
#
#   gc git@github.com:torvalds/linux.git
#   # Creates: ~/src/github.com/torvalds/linux
#
#   gc https://gitlab.com/gitlab-org/gitlab
#   # Creates: ~/src/gitlab.com/gitlab-org/gitlab
#
# WHY THIS IS USEFUL:
# - Organizes all Git repos in a predictable hierarchy: ~/src/domain/user/repo
# - No more scattered repos in random locations
# - Easy to find repos: you know exactly where they'll be
# - Works with any Git hosting service (GitHub, GitLab, custom domains)
# - If the repo already exists, it just cd's into it (idempotent)
# - Domain whitelist via directory existence (only clone from trusted domains)
#
# SETUP:
# 1. Save this file to your zsh functions directory
# 2. Ensure your .zshrc has: fpath=(~/.local/share/zsh/functions $fpath)
# 3. Add: autoload -Uz gc
#
# The function expects ~/src/<domain> to already exist for the domain to be
# allowed. This acts as a simple whitelist. For example, create ~/src/github.com
# to allow cloning from GitHub.

gc () {
    url=$1

    # Get the domain, user, and repo for two different
    # format, HTTP and git, e.g.
    # https://github.com/dev-sec/ansible-ssh-hardening
    # or
    # git@github.com:dev-sec/ansible-ssh-hardening.git
    #
    if [[ "$url" =~ ^[Hh][Tt][Tt][Pp][Ss]?:// ]]
    then
        protocol=$(echo "$1" | awk -F'[/:@]' '{print $1}')
        domain=$(echo "$1" | awk -F'[/:@]' '{print $4}')
        user=$(echo "$1" | awk -F'[/:@]' '{print $5}')
        repo=$(echo "$1" | awk -F'[/:@]' '{print $6}')
        sanitized_url="$protocol://$domain/$user/$repo"
        if [ "$sanitized_url" != "$url" ]
        then
            echo "Invalid remote: $url"
        fi
    else
        if [[ "$url" =~ ^git@.*\.git ]]
        then
            gituser=$(echo "$1" | awk -F'[/:@]' '{print $1}')
            domain=$(echo "$1" | awk -F'[/:@]' '{print $2}')
            user=$(echo "$1" | awk -F'[/:@]' '{print $3}')
            repo=$(echo "$1" | awk -F'[/:@]' '{print $4}' | sed 's/\..*$//')
            sanitized_url="$gituser@$domain:$user/$repo.git"
            if [ "$sanitized_url" != "$url" ]
            then
                echo "Invalid remote: $url"
            fi
        else
            echo "Invalid remote: $url"
            return 1
        fi
    fi

    # workdir should be something like $HOME/src/github.com
    workdir="$HOME/src/$domain"

    # See if the workdir exists. In effect, certain domains
    # are whitelisted if they already exist, e.g. 'git.yale.edu'
    if [ ! -d "$workdir" ]
    then
        echo "Error: invalid domain for git checkout: $domain"
        return 1
    fi

    userdir="$HOME/src/$domain/$user"
    if [ ! -d "$userdir" ]
    then
        mkdir -p "$userdir"
    fi

    cd "$userdir" || exit
    filepath="$userdir/$repo"
    if [ ! -d "$filepath" ]
    then
        git clone "$1"
    fi
    cd "$filepath" || exit
}