PROGRESS,CTIME,ID,HASH,COLOR,MODE,FONTSIZE,TEXT
0.0,2023-08-11 19:43:01,1413322056778563840,e997dd79,#84C0CB,5,25,梦结束的地方

import requests
# import pandas as pd
import xml.etree.ElementTree as ET

from datetime import datetime
from tqdm import tqdm
from sys import argv

import warnings
warnings.filterwarnings('ignore')

class Converter:
    def __init__(self):
        self.base58 = 'fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF'
        self.digit_map = [11, 10, 3, 8, 4, 6]
        self.xor_magic = 177451812
        self.add_magic = 8728348608
        self.template = list('BV1  4 1 7  ')
    def decode(self, cipher):
        plain = 0
        for i, digit in enumerate(self.digit_map):
            plain += self.base58.index(cipher[digit]) * 58 ** i
        return 'av%s' % ((plain - self.add_magic) ^ self.xor_magic)

if __name__ == '__main__':

    URL = argv[1]
    if URL.startswith('http'):
        URL = URL.replace('/?', '?').split('?')[0].split('/')[-1]
    if URL.startswith('BV'):
        assert len(URL) == 12, 'Invalid BV number'
        URL = Converter().decode(URL)
    if not URL.startswith('av'):
        raise ValueError('Invalid input')

    cid_api = 'https://api.bilibili.com/x/player/pagelist?aid=%s' % URL[2:]
    ret = requests.get(cid_api).json()['data'][0]
    cid, title = ret['cid'], ret['part']

    dmk_api = 'https://api.bilibili.com/x/v1/dm/list.so?oid=%d' % cid
    dmk = requests.get(dmk_api).content

    # processed = pd.DataFrame(columns=['PROGRESS', 'CTIME', 'ID', 'HASH', 'COLOR', 'MODE', 'FONTSIZE', 'TEXT'])
    processed = []

    for elt in tqdm(ET.fromstring(dmk).findall('d')):

        progress, mode, fontsize, color, ctime, _, hash, id, *_ = elt.attrib['p'].split(',')
        if color == 'undefined':
            color = '16777215'

        # text = elt.text.replace('\n', ' ').replace('\r', ' ').replace('\t', ' ')
        text = elt.text.replace('"', '""').replace('\n', ' ').replace('\r', ' ').replace('\t', ' ')
        if '""' in text or ',' in text:
            text = '"%s"' % text

        # processed = pd.concat([processed, pd.DataFrame({
        #     'PROGRESS': '%08.3f' % float(progress),
        #     'CTIME': datetime.fromtimestamp(int(ctime)).strftime('%Y-%m-%d %H:%M:%S'),
        #     'ID': '%19d' % int(id),
        #     'HASH': '%08x' % int(hash, 16),
        #     'COLOR': '#%06X' % int(color),
        #     'MODE': mode,
        #     'FONTSIZE': fontsize,
        #     'TEXT': text
        # }, index=[0])], ignore_index=True)
        processed.append((
            '%08.3f' % float(progress),
            datetime.fromtimestamp(int(ctime)).strftime('%Y-%m-%d %H:%M:%S'),
            '%19d' % int(id),
            '%08x' % int(hash, 16),
            '#%06X' % int(color),
            mode, fontsize, text
        ))

    FILENAME_OUT = 'danmaku_%s_%s_%d.csv' % (URL, title, int(datetime.now().timestamp()))
    # processed = processed.drop_duplicates('ID')
    # processed = processed.sort_values(['PROGRESS', 'CTIME'])
    # processed.to_csv(FILENAME_OUT, index=False, encoding='utf-8')
    with open(FILENAME_OUT, 'w', encoding='utf-8') as fout:
        fout.write('PROGRESS,CTIME,ID,HASH,COLOR,MODE,FONTSIZE,TEXT\n')
        processed = sorted(processed, key=lambda x: x[1])
        processed = sorted(processed, key=lambda x: x[0])
        for line in processed:
            fout.write(','.join(line) + '\n')