#!/usr/bin/env python3
import sys
import datetime
import json
import random
import time

import requests


def get_all(video_id, from_time=datetime.datetime.fromisoformat('2020-01-01T00:00:00+00:00')):
    base_url = f'https://public.openrec.tv/external/api/v5/movies/{video_id}/chats'
    out = []
    known_ids = set()
    sess = requests.session()
    while True:
        t_gmt = from_time.astimezone(datetime.timezone.utc)
        time_str = (f'{t_gmt.year:04d}-{t_gmt.month:02d}-{t_gmt.day:02d}T'
                    f'{t_gmt.hour:02d}:{t_gmt.minute:02d}:{t_gmt.second:02d}.{t_gmt.microsecond // 1000:03d}Z')
        r = sess.get(base_url, params={'from_created_at': time_str, 'is_including_system_message': 'false'})
        r.raise_for_status()
        comments: list = r.json()

        # check dup
        for c in comments.copy():
            comment_id = c['id']
            if comment_id in known_ids:
                comments.remove(c)
            else:
                known_ids.add(comment_id)

        if len(comments) == 0:
            break
        out.extend(comments)
        new_from_time = datetime.datetime.fromisoformat(comments[-1]['posted_at'])

        print(f'{from_time.isoformat()} - {new_from_time.isoformat()}, append {len(comments)} comments')
        from_time = new_from_time

        time.sleep(random.random() * 1 + 1)

    return out


def main():
    if len(sys.argv) > 1:
        video_id = sys.argv[1]
    else:
        video_id = input('Input video id (11-character alphanumeric string): ')
    x = get_all(video_id)
    with open(f'{video_id}.json', 'w', encoding='utf-8') as f:
        json.dump(x, f, ensure_ascii=False, separators=(',', ':'))
    with open(f'{video_id}-formatted.json', 'w', encoding='utf-8') as f:
        json.dump(x, f, ensure_ascii=False, indent=1)
    print('Comments saved')


if __name__ == '__main__':
    main()