slurpfiles

slurpfiles

#!/usr/bin/env python3
"""
This script concatenates files in a directory and prints the content.

The filename and relative path is atop each text blob.

$ slurpfiles /optional/path
$ slurpfiles -r /path/with/nested/dirs
$ slurpfiles --ignore *.css *.txt /path
$ slurpfiles -o path/to/foo.js path/to/another/thing.js
"""

import os
import argparse
import fnmatch
import subprocess

def is_binary(file_path):
    with open(file_path, 'rb') as f:
        chunk = f.read(1024)
    return b'\0' in chunk

def is_file_modified(file_path):
    cmd = f'git status --porcelain {file_path}'
    result = subprocess.run(cmd, shell=True, capture_output=True, text=True)
    return result.stdout.strip() != ""

def concatenate_files(files, ignore_patterns=[], modified_only=False):
    for file_path in files:
        if os.path.isfile(file_path):
            if any(fnmatch.fnmatch(os.path.basename(file_path), pattern) for pattern in ignore_patterns):
                continue

            if modified_only and not is_file_modified(file_path):
                continue

            if not is_binary(file_path):
                print(f'# {file_path}\n')
                with open(file_path, 'r', encoding='utf-8') as f:
                    print(f.read())

def get_files(target_dir, recursive=False):
    files = []
    if recursive:
        for root, _, filenames in os.walk(target_dir):
            files.extend(os.path.join(root, filename) for filename in filenames)
    else:
        files = [os.path.join(target_dir, f) for f in os.listdir(target_dir) if os.path.isfile(os.path.join(target_dir, f))]
    return files

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Concatenate files in a directory")
    parser.add_argument("target_dir", help="Target directory (default: current directory)", nargs="?", default=".")
    parser.add_argument("-r", "--recursive", help="Recursively process subdirectories", action="store_true")
    parser.add_argument("-m", "--modified", help="Only include modified files (in git)", action="store_true")
    parser.add_argument("--ignore", help="Ignore files matching patterns (e.g., *.css)", nargs="*", default=[])
    parser.add_argument("-o", "--only", help="Only process specified files", nargs="+")
    args = parser.parse_args()

    if args.only:
        files = args.only
    else:
        files = get_files(args.target_dir, args.recursive)

    concatenate_files(files, args.ignore, args.modified)