Calculate the list of contributor that will lose their contribution after files are deleted

lost-contributors.sh

#!/usr/bin/env bash
#
# lost-contributors.sh
#
# 1. Lists files deleted in the current branch compared to a target branch
# 2. Finds contributors who ever participated in those deleted files (full history)
#    but have no lines remaining in any surviving file
#
# Usage:
#   ./lost-contributors.sh [target_branch]
#
# Defaults:
#   target_branch = main
#
# Example:
#   git checkout feature/remove-legacy-code
#   ./lost-contributors.sh main

set -euo pipefail

TARGET_BRANCH="${1:-main}"
CURRENT_BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null || echo "HEAD")

echo "=============================================="
echo " Lost Contributors Detector"
echo "=============================================="
echo ""
echo " Target branch  : $TARGET_BRANCH"
echo " Current branch : $CURRENT_BRANCH"
echo ""

# ── Step 1: Find files deleted in the current branch vs target ──

echo "[1/4] Finding deleted files..."
echo ""

deleted_files_list=$(mktemp)
git diff --name-only --diff-filter=D "$TARGET_BRANCH"..."$CURRENT_BRANCH" > "$deleted_files_list"
deleted_count=$(wc -l < "$deleted_files_list" | tr -d ' ')

if [ "$deleted_count" -eq 0 ]; then
    echo " ✅ No files were deleted. Nothing to check."
    rm -f "$deleted_files_list"
    exit 0
fi

echo "      $deleted_count file(s) deleted:"
while IFS= read -r f; do
    echo "        - $f"
done < "$deleted_files_list"
echo ""

# ── Step 2: Get ALL contributors who ever touched the deleted files (full history) ──

echo "[2/4] Collecting all historical contributors of deleted files..."
echo ""

deleted_contributors_file=$(mktemp)
while IFS= read -r file; do
    git log --all --follow --format='<%ae>' -- "$file" 2>/dev/null
done < "$deleted_files_list" | sort -u > "$deleted_contributors_file"

deleted_contrib_count=$(wc -l < "$deleted_contributors_file" | tr -d ' ')
echo "      Found $deleted_contrib_count unique contributor(s) who ever touched deleted files."
echo ""

if [ "$deleted_contrib_count" -eq 0 ]; then
    echo " ✅ No contributors found in deleted files."
    rm -f "$deleted_files_list" "$deleted_contributors_file"
    exit 0
fi

# ── Step 3: For each contributor in deleted files, check if they have ──
# ──         lines in any surviving file on the current branch          ──

echo "[3/4] Checking if those contributors have lines in surviving files..."
echo "      (this may take a while on large repos)"
echo ""

# Get the full list of files on the current branch (the surviving files)
surviving_files_list=$(mktemp)
git ls-tree -r --name-only "$CURRENT_BRANCH" > "$surviving_files_list"

# Blame all surviving files once and collect all contributor emails
surviving_contributors_file=$(mktemp)
while IFS= read -r file; do
    git blame --porcelain "$CURRENT_BRANCH" -- "$file" 2>/dev/null \
        | grep "^author-mail " \
        | sed 's/^author-mail //'
done < "$surviving_files_list" | sort -u > "$surviving_contributors_file"

surviving_contrib_count=$(wc -l < "$surviving_contributors_file" | tr -d ' ')
echo "      Found $surviving_contrib_count unique contributor(s) in surviving files."
echo ""

# ── Step 4: Find contributors only in deleted files (not in surviving) ──

echo "[4/4] Comparing..."
echo ""

lost_file=$(mktemp)
comm -23 "$deleted_contributors_file" "$surviving_contributors_file" > "$lost_file"
lost_count=$(wc -l < "$lost_file" | tr -d ' ')

echo "=============================================="
echo " Results"
echo "=============================================="
echo ""

if [ "$lost_count" -eq 0 ]; then
    echo " ✅ No contributors will be fully removed."
    echo "    All contributors in deleted files also have lines elsewhere."
else
    echo "## Lost contributors"
    echo ""
    echo "$lost_count contributor(s) participated in deleted files and have no lines remaining in the surviving codebase:"
    echo ""
    # Collect sorted usernames into an array
    usernames=()
    while IFS= read -r email; do
        clean_email=$(echo "$email" | tr -d '<>')
        if echo "$clean_email" | grep -q '@users\.noreply\.github\.com$'; then
            gh_user=$(echo "$clean_email" | sed 's/@users\.noreply\.github\.com$//' | sed 's/^[0-9]*+//')
        else
            gh_user=$(echo "$clean_email" | cut -d'@' -f1)
        fi
        usernames+=("@${gh_user}")
    done < "$lost_file"

    # Sort alphabetically (case-insensitive)
    IFS=$'\n' sorted=($(printf '%s\n' "${usernames[@]}" | sort -f)); unset IFS

    # Output as a 4-column markdown table (no header)
    echo "| | | | |"
    echo "|---|---|---|---|"
    col=0
    row=""
    for user in "${sorted[@]}"; do
        row="${row}| ${user} "
        col=$((col + 1))
        if [ "$col" -eq 4 ]; then
            echo "${row}|"
            row=""
            col=0
        fi
    done
    # Handle remaining users in the last incomplete row
    if [ "$col" -gt 0 ]; then
        while [ "$col" -lt 4 ]; do
            row="${row}| "
            col=$((col + 1))
        done
        echo "${row}|"
    fi
fi

echo ""

# Cleanup
rm -f "$deleted_files_list" "$deleted_contributors_file" \
      "$surviving_files_list" "$surviving_contributors_file" "$lost_file"