#!/usr/bin/env python3

"""Pipe the output of `yadf --format ldjson` into this script.

Either :

yadf -f ldjson > results.ldjson
python3 keep_oldest.py results.ldjson

Or skipping the intermediate file :

yadf -f ldjson | python3 keep_oldest.py

This script is provided as an example meant to be modified and tinkered with.
"""

import fileinput
import itertools
import json
import multiprocessing
import os
import pathlib
from typing import Callable, Iterable, Sized, TypeVar

Cmp = TypeVar("Cmp", bound=Sized)
Key = Callable[[str], Cmp]
Filter = Callable[[Iterable[str]], Iterable[str]]


def main():
    cleaner = Cleaner(most_recent_modification_date, yield_all_except_first)
    sequential(fileinput.input(), cleaner)


def sequential(ldjson: Iterable[str], cleaner: "Cleaner"):
    for line in ldjson:
        cleaner(line)


def parallel(ldjson: Iterable[str], cleaner: "Cleaner"):
    with multiprocessing.Pool() as pool:
        pool.imap_unordered(cleaner, ldjson)


class Cleaner:
    def __init__(self, key: Key = None, filter: Filter = lambda f: f):
        self.key = key
        self.filter = filter

    def __call__(self, line: str):
        files: list[str] = json.loads(line)
        files.sort(key=self.key)
        # uncomment to actually delete files
        for filename in self.filter(files):
            # os.remove(filename)
            pass


def most_recent_modification_date(filename: str) -> float:
    return pathlib.Path(filename).stat().st_mtime


def yield_all_except_first(files: Iterable[str]) -> Iterable[str]:
    return itertools.islice(files, 1, None)


if __name__ == "__main__":
    main()