A script for concatenating all the files of a git repo into one file which you can feed into large context LLMs for understanding dependencies

concatenate_repo.sh

#!/bin/bash

# Check if the script received the correct number of arguments
if [ "$#" -ne 2 ]; then
    echo "Usage: $0 <repo_path> <output_file>"
    exit 1
fi

REPO_PATH=$1
OUTPUT_FILE=$2

# Change to the repository directory
cd "$REPO_PATH" || {
    echo "Repository path not found"
    exit 1
}

# Ensure the output file is empty
true >"$OUTPUT_FILE"

# Use git ls-files to get all tracked and not ignored files
git ls-files | while IFS= read -r file; do
    # Print the relative path of the file
    echo "\`\`\`repo/$file" >>"$OUTPUT_FILE"
    # Print the file content
    cat "$file" >>"$OUTPUT_FILE"
    # Add a new line to separate files
    echo "" >>"$OUTPUT_FILE"
    echo "\`\`\`" >>"$OUTPUT_FILE"
done

echo "All tracked and not ignored files in the repository have been concatenated into $OUTPUT_FILE"