aggregate_files.sh

#!/bin/bash

# Function to display usage information
usage() {
    echo "Usage: $0 source_directory [-o output_file] [-e exclude_dirs] [-x exclude_exts]"
    echo "  source_directory   Source directory to be processed"
    echo "  -o output_file     Optional output file name (default: output.txt)"
    echo "  -e exclude_dirs    Optional comma-separated list of directories to exclude"
    echo "  -x exclude_exts    Optional comma-separated list of file extensions to exclude"
    exit 1
}

# Check if source directory is provided
if [ $# -eq 0 ]; then
    echo "Error: Source directory is required"
    usage
fi

# Set source directory from first argument
sourcedir=$1
shift

# Default values
outputfile="output.txt"
excludedirs=()
excludeexts=()

# Parsing command-line arguments
while getopts "o:e:x:" opt; do
    case $opt in
        o)
            outputfile=$OPTARG
            ;;
        e)
            IFS=',' read -r -a excludedirs <<< "$OPTARG"
            ;;
        x)
            IFS=',' read -r -a excludeexts <<< "$OPTARG"
            ;;
        ?)
            echo "Option -$OPTARG requires an argument." >&2
            usage
            ;;
    esac
done

# Check if source directory exists
if [ ! -d "$sourcedir" ]; then
    echo "Error: Source directory '$sourcedir' does not exist"
    exit 1
fi

# Check if output file already exists
if [ -f "$outputfile" ]; then
    read -p "Output file '$outputfile' already exists. Overwrite? [Y/n]: " answer
    answer=${answer:-Y}  # Default to 'Y' if just Enter is pressed
    answer=$(echo "$answer" | tr '[:upper:]' '[:lower:]')  # Convert to lowercase
    if [ "$answer" != "y" ]; then
        echo "Operation cancelled"
        exit 1
    fi
fi

# Function to check if a directory is excluded
is_excluded_dir() {
    local dir="$1"
    for excluded in "${excludedirs[@]}"; do
        if [[ "$dir" == "$excluded" ]]; then
            return 0
        fi
    done
    return 1
}

# Function to check if a file extension is excluded
is_excluded_ext() {
    local file="$1"
    for ext in "${excludeexts[@]}"; do
        if [[ "$file" == *.$ext ]]; then
            return 0
        fi
    done
    return 1
}

# Function to check if a file is a text file
is_text_file() {
    local file="$1"
    if file "$file" | grep -qE 'text|JSON'; then
        return 0
    else
        return 1
    fi
}

# Recursive function to process directories
process_directory() {
    local dir="$1"
    if is_excluded_dir "$dir"; then
        return
    fi
    for item in "$dir"/*; do
        basename_item=$(basename "$item")
        # Skip hidden files and directories
        if [[ "$basename_item" == .* ]]; then
            continue
        fi
        if [ -d "$item" ]; then
            process_directory "$item"
        elif [ -f "$item" ]; then
            if is_excluded_ext "$item"; then
                continue
            fi
            if is_text_file "$item"; then
                echo "File '$item' is being processed"
                echo "--------------------------------------------" >> "$outputfile"
                echo "File: $item" >> "$outputfile"
                echo "--------------------------------------------" >> "$outputfile"
                echo >> "$outputfile"
                cat "$item" >> "$outputfile"
                echo >> "$outputfile"
            else
                echo "Skipping binary file: $item"
            fi
        fi
    done
}

# Start the process, truncate the file
> "$outputfile"
process_directory "$sourcedir"

echo "All text files have been concatenated into '$outputfile'"