combine.py

#!/usr/bin/env python3
"""
Combines multiple source files into a single Markdown document.

Usage:
    python combine.py path/to/file1.js path/to/file2.ts path/to/file3.prisma
    python combine.py $(cat filelist.txt)

Output:
    Creates 'combined_output.md' in the current directory.
"""

import sys
import os

# Map file extensions to markdown language identifiers
EXTENSION_MAP = {
    '.js': 'js',
    '.jsx': 'jsx',
    '.ts': 'ts',
    '.tsx': 'tsx',
    '.py': 'python',
    '.prisma': 'prisma',
    '.json': 'json',
    '.sql': 'sql',
    '.html': 'html',
    '.css': 'css',
    '.yml': 'yaml',
    '.yaml': 'yaml',
    '.md': 'markdown',
    '.env': 'env',
    '.sh': 'bash',
    '.bash': 'bash',
    '.txt': 'text',
    '.xml': 'xml',
    '.toml': 'toml',
    '.graphql': 'graphql',
    '.dockerfile': 'dockerfile',
}


def get_language(filepath: str) -> str:
    """Determine the markdown code fence language from file extension."""
    _, ext = os.path.splitext(filepath)
    ext = ext.lower()

    # Handle special filenames without typical extensions
    basename = os.path.basename(filepath).lower()
    if basename == 'dockerfile':
        return 'dockerfile'
    if basename == '.env' or basename.startswith('.env'):
        return 'env'

    return EXTENSION_MAP.get(ext, ext.lstrip('.') or 'text')


def combine_files(filepaths: list[str], output_path: str = 'combined_output.md'):
    """Read all files and write them into a single markdown document."""
    found = []
    missing = []

    for fp in filepaths:
        if os.path.isfile(fp):
            found.append(fp)
        else:
            missing.append(fp)

    if missing:
        print(f"\n⚠️  {len(missing)} file(s) not found:")
        for m in missing:
            print(f"   ✗ {m}")

    if not found:
        print("\n❌ No valid files to process. Exiting.")
        sys.exit(1)

    lines = []
    for fp in found:
        lang = get_language(fp)
        # Normalize path separators for display
        display_path = fp.replace('\\', '/')

        try:
            with open(fp, 'r', encoding='utf-8', errors='replace') as f:
                content = f.read()
        except Exception as e:
            print(f"   ⚠️  Could not read {fp}: {e}")
            content = f"# ERROR: Could not read file — {e}"

        lines.append(f"## `{display_path}`")
        lines.append(f"```{lang}")
        lines.append(content.rstrip())
        lines.append("```")
        lines.append("")  # blank line between files

    output = '\n'.join(lines)

    with open(output_path, 'w', encoding='utf-8') as f:
        f.write(output)

    print(f"\n✅ Combined {len(found)} file(s) → {output_path}")
    print(f"   Size: {len(output):,} characters")

    if missing:
        print(f"   ⚠️  Skipped {len(missing)} missing file(s)")


def main():
    if len(sys.argv) < 2:
        print("Usage: python combine.py <file1> <file2> <file3> ...")
        print("")
        print("Example:")
        print("  python combine.py backend/prisma/schema.prisma backend/server.js")
        print("")
        print("Tip: You can also pipe a file list:")
        print("  python combine.py $(cat files_to_send.txt)")
        sys.exit(1)

    filepaths = sys.argv[1:]
    combine_files(filepaths)


if __name__ == '__main__':
    main()