Create a single file with the contents of all files of a given directory to provide as context to LLMs

#!/bin/bash
 
# Check if a directory path was provided
if [ $# -ne 1 ]; then
    echo "Usage: $0 <directory_path>"
    exit 1
fi
 
# Assign the input directory to a variable
INPUT_DIR=$1
 
# Check if the directory exists
if [ ! -d "$INPUT_DIR" ]; then
    echo "Error: Directory '$INPUT_DIR' does not exist."
    exit 1
fi
 
# Create output file name based on the directory name
OUTPUT_FILE="$(basename "$INPUT_DIR")_contents.txt"
 
# Clear the output file if it exists
> "$OUTPUT_FILE"
 
# Function to process each file
process_file() {
    local file="$1"
    local rel_path="${file#$INPUT_DIR/}"
 
    # Only process regular files (not directories, symlinks, etc.)
    if [ -f "$file" ]; then
        # Add a comment with the file name
        echo -e "\n\n# ========== FILE: $rel_path ==========" >> "$OUTPUT_FILE"
 
        # Check if the file is binary
        if file "$file" | grep -q "binary"; then
            echo "[Binary file - contents not included]" >> "$OUTPUT_FILE"
        else
            # Append the file contents
            cat "$file" >> "$OUTPUT_FILE"
        fi
    fi
}
 
# Export the function so it can be used with find
export -f process_file
export INPUT_DIR
export OUTPUT_FILE
 
# Use find to recursively process all files
find "$INPUT_DIR" -type f -not -path "*/\.*" -exec bash -c 'process_file "$0"' {} \;
 
echo "All file contents have been written to $OUTPUT_FILE"