cim_summary/cleanup_gcs.sh

#!/bin/bash
set -e

# Colors for output
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m' # No Color

# Configuration
PROJECT_ID="cim-summarizer"
UPLOADS_BUCKET="gs://cim-summarizer-uploads"
FIREBASE_UPLOADS_BUCKET="gs://gcf-v2-uploads-245796323861.us-central1.cloudfunctions.appspot.com"
FIREBASE_SOURCES_BUCKET="gs://gcf-v2-sources-245796323861-us-central1"

# Function to print colored output
print_status() {
    echo -e "${BLUE}[INFO]${NC} $1"
}

print_success() {
    echo -e "${GREEN}[SUCCESS]${NC} $1"
}

print_warning() {
    echo -e "${YELLOW}[WARNING]${NC} $1"
}

print_error() {
    echo -e "${RED}[ERROR]${NC} $1"
}

# Function to get storage size in human readable format
format_size() {
    local bytes=$1
    if [ $bytes -gt 1073741824 ]; then
        echo "$(echo "scale=2; $bytes/1073741824" | bc) GB"
    elif [ $bytes -gt 1048576 ]; then
        echo "$(echo "scale=2; $bytes/1048576" | bc) MB"
    elif [ $bytes -gt 1024 ]; then
        echo "$(echo "scale=2; $bytes/1024" | bc) KB"
    else
        echo "${bytes} B"
    fi
}

# Function to calculate total size of files
calculate_size() {
    local bucket=$1
    local size=$(gsutil ls -la "$bucket" | grep -E "TOTAL:" | awk '{print $2}' | head -1)
    echo "${size:-0}"
}

# Function to show usage
show_usage() {
    echo "Usage: $0 [OPTIONS]"
    echo ""
    echo "Options:"
    echo "  -d, --dry-run     Show what would be deleted without actually deleting"
    echo "  -f, --force       Skip confirmation prompts"
    echo "  -h, --help        Show this help message"
    echo ""
    echo "This script cleans up old files from GCS buckets:"
    echo "  - Removes test files from uploads bucket"
    echo "  - Removes old Firebase function deployments (keeps 2 most recent)"
    echo "  - Removes old Firebase function sources (keeps 1 most recent)"
    echo "  - Cleans up empty directories"
}

# Function to clean up test files
cleanup_test_files() {
    print_status "Cleaning up test files..."
    
    local test_files=$(gsutil ls "$UPLOADS_BUCKET/test-*.txt" 2>/dev/null || true)
    if [ -n "$test_files" ]; then
        if [ "$DRY_RUN" = true ]; then
            echo "Would remove: $test_files"
        else
            gsutil rm $test_files
            print_success "Removed test files"
        fi
    else
        print_status "No test files found"
    fi
}

# Function to clean up invalid files
cleanup_invalid_files() {
    print_status "Cleaning up invalid files..."
    
    local invalid_files=$(gsutil ls "$UPLOADS_BUCKET/uploads/staging-test-user/*.exe" 2>/dev/null || true)
    if [ -n "$invalid_files" ]; then
        if [ "$DRY_RUN" = true ]; then
            echo "Would remove: $invalid_files"
        else
            gsutil rm $invalid_files
            print_success "Removed invalid files"
        fi
    else
        print_status "No invalid files found"
    fi
}

# Function to clean up old Firebase function deployments
cleanup_firebase_deployments() {
    print_status "Cleaning up old Firebase function deployments..."
    
    local deployment_files=$(gsutil ls "$FIREBASE_UPLOADS_BUCKET/*.zip" 2>/dev/null | sort | head -n -2 || true)
    if [ -n "$deployment_files" ]; then
        if [ "$DRY_RUN" = true ]; then
            echo "Would remove old deployment files:"
            echo "$deployment_files"
        else
            echo "$deployment_files" | while read file; do
                if [ -n "$file" ]; then
                    gsutil rm "$file"
                fi
            done
            print_success "Removed old Firebase function deployments"
        fi
    else
        print_status "No old deployment files found"
    fi
}

# Function to clean up old Firebase function sources
cleanup_firebase_sources() {
    print_status "Cleaning up old Firebase function sources..."
    
    local source_files=$(gsutil ls "$FIREBASE_SOURCES_BUCKET/api/*.zip" 2>/dev/null | sort | head -n -1 || true)
    if [ -n "$source_files" ]; then
        if [ "$DRY_RUN" = true ]; then
            echo "Would remove old source files:"
            echo "$source_files"
        else
            echo "$source_files" | while read file; do
                if [ -n "$file" ]; then
                    gsutil rm "$file"
                fi
            done
            print_success "Removed old Firebase function sources"
        fi
    else
        print_status "No old source files found"
    fi
}

# Function to show storage summary
show_storage_summary() {
    print_status "Current storage usage:"
    
    local uploads_size=$(calculate_size "$UPLOADS_BUCKET")
    local firebase_uploads_size=$(calculate_size "$FIREBASE_UPLOADS_BUCKET")
    local firebase_sources_size=$(calculate_size "$FIREBASE_SOURCES_BUCKET/api")
    
    echo "📁 Main Uploads: $(format_size $uploads_size)"
    echo "📁 Firebase Function Uploads: $(format_size $firebase_uploads_size)"
    echo "📁 Firebase Function Sources: $(format_size $firebase_sources_size)"
    
    local total_size=$((uploads_size + firebase_uploads_size + firebase_sources_size))
    echo "📊 Total Storage: $(format_size $total_size)"
}

# Main script
main() {
    local dry_run=false
    local force=false
    
    # Parse command line arguments
    while [[ $# -gt 0 ]]; do
        case $1 in
            -d|--dry-run)
                dry_run=true
                shift
                ;;
            -f|--force)
                force=true
                shift
                ;;
            -h|--help)
                show_usage
                exit 0
                ;;
            *)
                print_error "Unknown option: $1"
                show_usage
                exit 1
                ;;
        esac
    done
    
    DRY_RUN=$dry_run
    
    print_status "Starting GCS cleanup process..."
    
    if [ "$dry_run" = true ]; then
        print_warning "DRY RUN MODE - No files will be deleted"
    fi
    
    # Show initial storage usage
    show_storage_summary
    
    # Perform cleanup
    cleanup_test_files
    cleanup_invalid_files
    cleanup_firebase_deployments
    cleanup_firebase_sources
    
    # Show final storage usage
    echo ""
    show_storage_summary
    
    if [ "$dry_run" = false ]; then
        print_success "GCS cleanup completed successfully!"
    else
        print_warning "Dry run completed - no files were deleted"
    fi
}

# Run main function with all arguments
main "$@"