A tool for aggregating file extensions and languages in a directory, compatible with the tokei project (https://github.com/XAMPPRocky/tokei).

lang_ext_counter.py

# The languages.json is obtained from the tokei project
# Source: https://github.com/XAMPPRocky/tokei/blob/c8e4d0703252c87b1df45382b365c6bb00769dbe/languages.json

from typing import Dict, Counter as CounterType

from collections import Counter
import json
import os
import sys


def load_ext_to_lang_name() -> Dict[str, str]:
    """
    Load a dictionary mapping file extensions to language names.
    :return: Dictionary with file extensions as keys and language names as values.
    """

    # Retrieve the configuration file
    script_dir = os.path.dirname(os.path.abspath(__file__))
    languages_json_path = os.path.join(script_dir, "languages.json")
    with open(languages_json_path, "r") as inp:
        lang_data = json.load(inp)

    # Parse the configuration file to build a dictionary from extension to language name
    tbl = lang_data["languages"]
    ext_to_name = dict()
    for lang_id, info in tbl.items():
        name = info.get("name", lang_id)
        extensions = info.get("extensions")
        if extensions:
            for ext in extensions:
                if not ext.startswith("."):
                    ext = "." + ext
                ext_to_name[ext] = name

    return ext_to_name


def count_file_extensions(directory: str)  -> CounterType[str]:
    """
    Count file extensions in a directory.
    :param directory: Path of the directory to search.
    :return: A Counter object counting files per extension.
    """

    counts = Counter()
    for _, _, files in os.walk(directory):
        for file in files:
            ext = os.path.splitext(file)[1]
            counts[ext] += 1
    return counts


__doc__ = """
lang_ext_counter.py is a tool for aggregating file extensions and languages in a directory, compatible with the tokei project (https://github.com/XAMPPRocky/tokei). 

It analyzes all files in a specified directory, counts the occurrences of each file extension, 
and maps them to their corresponding programming languages using data from tokei's languages.json. 
"""

def main() -> None:
    directory = sys.argv[1] if len(sys.argv) > 1 else os.curdir
    ext_to_lang_name = load_ext_to_lang_name()
    counts = count_file_extensions(directory)

    ext_name_count_list = [(ext, ext_to_lang_name.get(ext) or "-", count) for ext, count in counts.items()]

    print("extension\ttype\tcount")
    for ext, name, count in sorted(ext_name_count_list):
        print(f"{ext}\t{name}\t{count}")


if __name__ == "__main__":
    main()