process_audio_chunks.sh

#!/bin/bash

# ANSI color codes
GREEN='\033[0;32m'
RED='\033[0;31m'
NC='\033[0m' # No Color

# Function to print colored output
print_success() {
    echo -e "${GREEN}$1${NC}"
}

print_error() {
    echo -e "${RED}$1${NC}" >&2
}

print_info() {
    echo "$1"
}

# Function to display help message
show_help() {
    echo "Usage: $0 [options] <directory> <workers>"
    echo
    echo "Process multiple audio chunks in parallel using generate_tiktok.sh"
    echo
    echo "Arguments:"
    echo "  directory    Directory containing chunk_* subdirectories"
    echo "  workers      Number of parallel workers (default: 1)"
    echo
    echo "Options:"
    echo "  -h, --help        Show this help message"
    echo "  --debug           Enable debug mode"
    echo "  --regenerate      Force regeneration of outputs"
}

# Parse command line arguments
DEBUG=""
REGENERATE=""
DIRECTORY=""
WORKERS=1

while [[ $# -gt 0 ]]; do
    case $1 in
        -h|--help)
            show_help
            exit 0
            ;;
        --debug)
            DEBUG="--debug"
            shift
            ;;
        --regenerate)
            REGENERATE="--regenerate"
            shift
            ;;
        *)
            if [[ -z "$DIRECTORY" ]]; then
                DIRECTORY="$1"
            elif [[ -z "$WORKERS" || "$WORKERS" -eq 1 ]]; then
                if [[ "$1" =~ ^[0-9]+$ ]]; then
                    WORKERS="$1"
                else
                    print_error "Error: Workers must be a number"
                    show_help
                    exit 1
                fi
            else
                print_error "Error: Unexpected argument '$1'"
                show_help
                exit 1
            fi
            shift
            ;;
    esac
done

# Validate arguments
if [[ -z "$DIRECTORY" ]]; then
    print_error "Error: Directory is required"
    show_help
    exit 1
fi

if [[ ! -d "$DIRECTORY" ]]; then
    print_error "Error: Directory '$DIRECTORY' does not exist"
    exit 1
fi

# Function to process a chunk directory
process_chunk() {
    local chunk_dir="$1"
    print_info "Processing $chunk_dir..."
    if ./generate_tiktok.sh $DEBUG $REGENERATE "$chunk_dir"; then
        print_success "Completed processing $chunk_dir"
        return 0
    else
        print_error "Failed processing $chunk_dir"
        return 1
    fi
}

# Find all chunk directories and sort them numerically
chunk_dirs=($(find "$DIRECTORY" -type d -name "chunk_*" | sort -t_ -k2 -n))

if [ ${#chunk_dirs[@]} -eq 0 ]; then
    print_error "No chunk_* directories found in $DIRECTORY"
    exit 1
fi

print_info "Found ${#chunk_dirs[@]} chunks to process"
print_info "Using $WORKERS parallel workers"

# Process chunks in parallel with maximum number of workers
pids=()
for chunk_dir in "${chunk_dirs[@]}"; do
    # Wait if we've reached maximum number of workers
    while [ ${#pids[@]} -ge $WORKERS ]; do
        for i in "${!pids[@]}"; do
            if ! kill -0 ${pids[$i]} 2>/dev/null; then
                wait ${pids[$i]}
                exit_code=$?
                unset pids[$i]
                if [ $exit_code -ne 0 ]; then
                    print_error "A worker process failed"
                    exit 1
                fi
            fi
        done
        sleep 1
    done

    # Start new process
    process_chunk "$chunk_dir" &
    pids+=($!)
done

# Wait for remaining processes to complete
for pid in "${pids[@]}"; do
    wait $pid
    if [ $? -ne 0 ]; then
        print_error "A worker process failed"
        exit 1
    fi
done

print_success "All chunks processed successfully!"