Find status of all repos in a given directory

find_git_repos.sh

#!/usr/bin/env bash
set -euo pipefail

usage() {
  cat <<'EOF'
Usage: find_git_repos.sh [--output <file>] <directory>

Scans immediate subdirectories of <directory>:
 - If a subdirectory is a git repo, prints its repo name and remotes
 - If not, prints a warning that no repo is present

Outputs lines like:
Directory: /path/project | Repo: project | Remotes: origin, upstream | GitHub URLs: https://github.com/org/repo.git
Directory: /path/notes   | WARNING: no git repo present

Options:
  -o, --output <file>   Write a copy of the output to <file> (also prints to stdout). Default: git_repo_report_YYYYmmdd-HHMMSS.txt
  -h, --help            Show this help and exit
EOF
}

if [[ ${1-} == "-h" || ${1-} == "--help" ]]; then
  usage
  exit 0
fi

# Parse args
output_file=""
args=()
while [[ $# -gt 0 ]]; do
  case "$1" in
    -o|--output)
      if [[ $# -lt 2 ]]; then
        echo "Error: missing value for $1" >&2
        exit 1
      fi
      output_file="$2"
      shift 2
      ;;
    -h|--help)
      usage
      exit 0
      ;;
    --)
      shift; break ;;
    -*)
      echo "Error: unknown option: $1" >&2
      echo >&2
      usage >&2
      exit 1
      ;;
    *)
      args+=("$1"); shift ;;
  esac
done

if [[ ${#args[@]} -ne 1 ]]; then
  echo "Error: exactly one directory path is required" >&2
  echo >&2
  usage >&2
  exit 1
fi

base_dir=${args[0]}
if [[ ! -d "$base_dir" ]]; then
  echo "Error: '$base_dir' is not a directory" >&2
  exit 1
fi

# Prepare output capture
ts=$(date '+%Y%m%d-%H%M%S')
default_out="git_repo_report_${ts}.txt"
outfile=${output_file:-$default_out}

# Mirror stdout/stderr to both console and file
exec > >(tee "$outfile") 2>&1

# Helper: normalize and detect GitHub URLs
is_github_url() {
  local url=$1
  # Match common GitHub URL forms: https, http, git, ssh, and scp-like
  if [[ $url =~ (^https?://([^/]*\.)?github\.com/) || \
        $url =~ (^git://([^/]*\.)?github\.com/) || \
        $url =~ (^ssh://git@([^/]*\.)?github\.com/) || \
        $url =~ (^git@([^:]*\.)?github\.com:) ]]; then
    return 0
  fi
  return 1
}

# Summarize repository cleanliness and upstream push status
repo_status_summary() {
  local repo_dir=$1
  local status_parts=()

  local porcelain
  porcelain=$(git -C "$repo_dir" status --porcelain 2>/dev/null || true)
  if [[ -n "$porcelain" ]]; then
    status_parts+=("dirty")

    local untracked_count=0
    local tracked_changes=0
    while IFS= read -r entry; do
      [[ -z "$entry" ]] && continue
      if [[ ${entry:0:2} == "??" ]]; then
        ((untracked_count++))
      elif [[ ${entry:0:1} == "!" ]]; then
        continue
      else
        ((tracked_changes++))
      fi
    done <<< "$porcelain"

    if (( untracked_count > 0 )); then
      if (( untracked_count == 1 )); then
        status_parts+=("1 untracked file")
      else
        status_parts+=("${untracked_count} untracked files")
      fi
    fi

    if (( tracked_changes > 0 )); then
      if (( tracked_changes == 1 )); then
        status_parts+=("1 file with changes")
      else
        status_parts+=("${tracked_changes} files with changes")
      fi
    fi
  fi

  local status_branch
  status_branch=$(git -C "$repo_dir" status --porcelain=2 --branch 2>/dev/null || true)
  local upstream=""
  while IFS= read -r line; do
    case "$line" in
      '# branch.upstream '*)
        upstream=${line#'# branch.upstream '}
        ;;
      '# branch.ab '*)
        local trimmed=${line#'# branch.ab '}
        local ahead_token=${trimmed%% -*}
        local ahead=${ahead_token#+}
        if [[ $ahead =~ ^[0-9]+$ ]] && (( ahead > 0 )); then
          if [[ -n "$upstream" ]]; then
            status_parts+=("ahead of $upstream by $ahead")
          else
            status_parts+=("ahead of upstream by $ahead")
          fi
        fi
        ;;
    esac
  done <<< "$status_branch"

  if [[ ${#status_parts[@]} -eq 0 ]]; then
    echo "clean"
  else
    local joined="${status_parts[0]}"
    for ((i=1; i<${#status_parts[@]}; i++)); do
      joined+=", ${status_parts[$i]}"
    done
    echo "$joined"
  fi
}

# Iterate immediate subdirectories (including those with spaces)
while IFS= read -r -d '' dir; do
  # Skip nested VCS metadata directories like .git
  name=$(basename "$dir")
  if [[ "$name" == ".git" ]]; then
    continue
  fi

  # Detect if this directory is inside a git work tree
  if git -C "$dir" rev-parse --is-inside-work-tree >/dev/null 2>&1; then
    # Determine repo toplevel and a human-friendly repo name
    toplevel=$(git -C "$dir" rev-parse --show-toplevel 2>/dev/null || true)
    if [[ -n "${toplevel}" && -d "${toplevel}" ]]; then
      repo_name=$(basename "${toplevel}")
    else
      # Fallback to directory name
      repo_name="$name"
    fi

    repo_root=${toplevel:-$dir}
    repo_status=$(repo_status_summary "$repo_root")

    # Collect remote names (if any)
    # Use mapfile to safely capture lines; fallback if not available
    if remotes=$(git -C "$dir" remote 2>/dev/null); then
      if [[ -z "$remotes" ]]; then
        echo "Directory: $dir | Repo: $repo_name | WARNING: no remote present | Status: $repo_status"
      else
        # Join remote names with ', '
        remote_list=$(printf '%s' "$remotes" | paste -sd ', ' -)

        # Collect GitHub remote URLs (all remotes/all URLs)
        gh_urls=()
        while IFS= read -r remote; do
          # get-url --all prints multiple lines for multiple push/fetch URLs
          while IFS= read -r url; do
            if is_github_url "$url"; then
              gh_urls+=("$url")
            fi
          done < <(git -C "$dir" remote get-url --all "$remote" 2>/dev/null || true)
        done <<< "$remotes"

        if [[ ${#gh_urls[@]} -gt 0 ]]; then
          # Deduplicate while preserving order
          uniq_urls=()
          declare -A seen=()
          for u in "${gh_urls[@]}"; do
            if [[ -z "${seen[$u]:-}" ]]; then
              uniq_urls+=("$u")
              seen[$u]=1
            fi
          done
          github_list=$(printf '%s' "${uniq_urls[0]}"; for ((i=1;i<${#uniq_urls[@]};i++)); do printf ', %s' "${uniq_urls[$i]}"; done)
          echo "Directory: $dir | Repo: $repo_name | Remotes: $remote_list | GitHub URLs: $github_list | Status: $repo_status"
        else
          echo "Directory: $dir | Repo: $repo_name | Remotes: $remote_list | GitHub URLs: none | Status: $repo_status"
        fi
      fi
    else
      echo "Directory: $dir | Repo: $repo_name | Remotes: error reading remotes | Status: $repo_status"
    fi
  else
    echo "Directory: $dir | WARNING: no git repo present"
  fi
done < <(find "$base_dir" -mindepth 1 -maxdepth 1 -type d -print0 | sort -z)