boomstream.py

#!/usr/bin/env python3

import argparse
import json
import os
import re
import string
import subprocess
import shutil
import sys

from base64 import b64decode
from lxml.html import fromstring
import requests
from cryptography.hazmat.primitives.ciphers import Cipher, algorithms, modes

XOR_KEY = 'bla_bla_bla'

OUTPUT_PATH = "temp"
RESULT_OUTPUT_PATH = "output"

VALID_FILENAME_CHARS = set(f" -_.(){string.ascii_letters}{string.digits}")

headers = {
  'authority': 'play.boomstream.com',
  'pragma': 'no-cache',
  'cache-control': 'no-cache',
  'upgrade-insecure-requests': '1',
  'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/81.0.4044.92 Safari/537.36',
  'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9',
  'sec-fetch-site': 'none',
  'sec-fetch-mode': 'navigate',
  'sec-fetch-user': '?1',
  'sec-fetch-dest': 'document',
  'accept-language': 'en-US,en;q=0.9,ru;q=0.8,es;q=0.7,de;q=0.6'}

def valid_filename(s):
    return ''.join(c for c in s if c in VALID_FILENAME_CHARS or c.isalpha())

def output_path(path):
    return os.path.join(OUTPUT_PATH, path)

def ensure_folder_exists(path):
    if not os.path.exists(path):
        os.mkdir(path)

def run_bash(command):
    exit_code, output = subprocess.getstatusoutput(command)
    if exit_code != 0:
        print(output)
        raise ValueError(f'failed with exit code {exit_code}')
    return output

def run_bash_check_exist(command):
    exit_code, output = subprocess.getstatusoutput(command)
    return exit_code == 0

def create_cipher(key: bytes, iv: bytes):
    cipher = Cipher(algorithms.AES(key), modes.CBC(iv))
    return cipher

class App(object):

    def __init__(self):
        parser = argparse.ArgumentParser(description='boomstream.com downloader')
        parser.add_argument('--entity', type=str, required=True)
        parser.add_argument('--pin', type=str, required=False)
        parser.add_argument('--use-cache', action='store_true', required=False)
        parser.add_argument('--resolution', type=str, required=False)
        parser.add_argument('--ffmpeg-path', type=str, required=False)
        parser.add_argument('--ffprobe-path', type=str, required=False)
        self.args = parser.parse_args()
        self.ffmpeg = "ffmpeg"
        self.ffprobe = "ffprobe"

    def get_token(self):
        return b64decode(self.config['mediaData']['token']).decode('utf-8')

    def get_m3u8_url(self):
        return b64decode(self.config['mediaData']['links']['hls']).decode('utf-8')

    def get_boomstream_config(self, page):
        """
        Evals value assigned to window.boomstreamConfig variable as JSON. This is ugly,
        but still better than using regular expressions to extract all needed variables
        from HTML page.
        """
        html = fromstring(page)
        result = None

        for script in html.xpath('//script[@type="text/javascript"]'):
            m = re.search("window.boomstreamConfig = ({.*});$", script.text_content(), flags=re.M)
            if m is not None:
                result = json.loads(m.group(1))

        if result is None:
            raise Exception("Could not get boomstreamConfig from the main page")

        with open(output_path('boomstream.config.json'), 'wt') as f:
            del result["translations"]
            f.write(json.dumps(result, ensure_ascii=False, indent=4))

        return result

    def get_playlist(self, url):
        if self.args.use_cache and os.path.exists(output_path('boomstream.playlist.m3u8')):
            with open(output_path('boomstream.playlist.m3u8')) as f:
                return f.read()

        r = requests.get(url, headers=headers)

        with open(output_path('boomstream.playlist.m3u8'), 'wt') as f:
            f.write(r.text)

        return r.text

    def res2int(self, resolution):
        if 'x' in resolution:
            return int(resolution.split('x')[0]) * int(resolution.split('x')[1])
        else:
            return int(resolution)

    def extract_chunklist_urls(self, playlist):
        result = []
        resolution = None

        for line in playlist.split('\n'):
            if line.startswith('#EXT-X-STREAM-INF'):
                m = re.search(r'RESOLUTION=(\d+x\d+)', line)
                if m is not None:
                    resolution = m.group(1)
                else:
                    m = re.search(r'BANDWIDTH=(\d+)', line)
                    if m is not None:
                        resolution = m.group(1)
                    else:
                        raise Exception("Could not get resolution from EXT-X-STREAM-INF")
            elif resolution is not None:
                result.append([resolution, line, self.res2int(resolution)])
                resolution = None

        return result

    def get_chunklist(self, playlist):
        all_chunklists = self.extract_chunklist_urls(playlist)
        print(f"This video is available in the following resolutions: {', '.join(i[0] for i in all_chunklists)}")

        if self.args.resolution is not None:
            url = None
            for item in all_chunklists:
                if item[0] == self.args.resolution:
                    url = item[1]
                    break
            if url is None:
                raise Exception("Playlist for resolution specifeid is --resolution " \
                                "argument is not found")
        else:
            # If the resolution is not specified in args, pick the best one
            url = sorted(all_chunklists, key=lambda x: x[2])[-1][1]

        print(f"URL: {url}")

        if url is None:
            raise Exception("Could not find chunklist in playlist data")

        if self.args.use_cache and os.path.exists(output_path('boomstream.chunklist.m3u8')):
            with open(output_path('boomstream.chunklist.m3u8')) as f:
                return f.read()

        r = requests.get(url, headers=headers)

        with open(output_path('boomstream.chunklist.m3u8'), 'wt') as f:
            f.write(r.text)

        return r.text

    def get_xmedia_ready(self, chunklist):
        """
        X-MEDIA-READY contains a value that is used to calculate IV for AES-128 and a URL
        to obtain AES-128 encryption key.
        """
        for line in chunklist.split('\n'):
            if line.split(':')[0] == '#EXT-X-MEDIA-READY':
                return line.split(':')[1]

        raise Exception("Could not find X-MEDIA-READY")

    def decrypt(self, source_text, key):
        result = ''
        while len(key) < len(source_text):
            key += key

        for n in range(0, len(source_text), 2):
            c = int(source_text[n:n+2], 16) ^ ord(key[(int(n / 2))])
            result = result + chr(c)

        return result

    def encrypt(self, source_text, key):
        result = ''

        while len(key) < len(source_text):
            key += key

        for i in range(0, len(source_text)):
            result += f'{ord(source_text[i]) ^ ord(key[i]):02x}'

        return result

    def get_aes_key(self, xmedia_ready):
        """
        Returns IV and 16-byte key which will be used to decrypt video chunks
        """
        decr = self.decrypt(xmedia_ready, XOR_KEY)
        print(f'Decrypted X-MEDIA-READY: {decr}')

        iv = decr[20:36]

        key_url = 'https://play.boomstream.com/api/process/' + \
                  self.encrypt(decr[0:20] + self.token, XOR_KEY)

        r = requests.get(key_url, headers=headers)
        key_text = r.text
  
        print(f'Key = {key_text}')
        print(f"IV = {iv}")

        return key_text.encode(), iv.encode()

    def download_chunks(self, chunklist, cipher: Cipher, title):
        valid_name = valid_filename(title)
        ensure_folder_exists(output_path(valid_name))

        filenames = []

        i = 0
        for line in chunklist.split('\n'):
            if not line.startswith('https://'):
                continue

            outf = output_path(os.path.join(valid_name, f"{i:05d}.ts"))
            filenames.append(outf)
            
            if os.path.exists(outf) and os.path.getsize(outf) > 0:
                i += 1
                print(f"Chunk #{i} exists [{outf}]")
                continue

            print(f"Downloading chunk #{i}")
            file_crypt = requests.get(line)
            
            decryptor = cipher.decryptor()
            result = decryptor.update(file_crypt.content) + decryptor.finalize()

            with open(outf, 'ab') as f:
                f.write(result)

            i += 1
        return filenames

    def merge_chunks(self, filenames, expected_result_duration, title):
        """
        Merges all chunks into one file and encodes it to MP4
        """
        source_file_name = valid_filename(title)
        target_file_name = source_file_name
        if not source_file_name.endswith(".mp4"):
            target_file_name = f'{source_file_name}.mp4'

        merged_file_ts_filepath = f"{output_path(source_file_name)}.ts"
        print("Merging chunks...")
        with open(merged_file_ts_filepath, 'wb') as merged:
            for ts_file in filenames:
                with open(ts_file, 'rb') as mergefile:
                    shutil.copyfileobj(mergefile, merged)

        temp_mp4_filepath = merged_file_ts_filepath
        
        if not source_file_name.endswith(".mp4"):
            print("Encoding to MP4")
            temp_mp4_filepath = output_path(target_file_name)
            if self.exist_ffmpeg:
                run_bash(f'{self.ffmpeg} -nostdin -y -i "{merged_file_ts_filepath}" -c copy "{temp_mp4_filepath}"')

        if self.exist_ffprobe:
            result_format = run_bash(f'{self.ffprobe}  -i "{temp_mp4_filepath}" -show_format')
            result_duration = float([line[len("duration="):] for line in result_format.split('\n') if line.startswith("duration=")][0])
            print(f"Result duration: {result_duration:.2f}")
            print(f"Expected duration: {expected_result_duration:.2f}")
            if abs(result_duration - expected_result_duration) > 2:
                raise ValueError(f"unexpected result duration: {expected_result_duration:.2f} != {result_duration:.2f}")

        print("Prepare result...")
        ensure_folder_exists(RESULT_OUTPUT_PATH)
        result_filename = os.path.join(RESULT_OUTPUT_PATH, target_file_name)
        if os.path.exists(result_filename):
            os.remove(result_filename)
        shutil.move(temp_mp4_filepath, result_filename)
        print(f"Result: {result_filename}")

    def get_title(self):
        return self.config['entity']['title']

    def get_access_cookies(self):
        pin = self.args.pin
        if pin is None:
            return {}
        r = requests.post("https://play.boomstream.com/api/subscriptions/recovery",
                          headers={'content-type': 'application/json;charset=UTF-8'},
                          data=f'{{"entity":"{self.args.entity}","code":"{pin}"}}')
        response = json.loads(r.text)
        if "data" not in response or "cookie" not in response["data"]:
            if "errors" not in response or "code" not in response:
                raise ValueError(f"unexpected response on authorization: {r.text}")
            else:
                raise ValueError(f"authorization failed: {response['code']} {response['errors']}")
        cookie = response["data"]["cookie"]
        return {cookie["name"]: cookie["value"]}

    def run(self):
        if self.args.ffmpeg_path:
            self.ffmpeg = f'"{self.args.ffmpeg_path}"'
        if self.args.ffprobe_path:
            self.ffprobe = f'"{self.args.ffprobe_path}"'

        self.exist_ffmpeg = run_bash_check_exist(f'{self.ffmpeg} -version')
        self.exist_ffprobe = run_bash_check_exist(f'{self.ffprobe} -version')

        if not self.exist_ffmpeg or not self.exist_ffprobe:
            print("http://ffmpeg.org/download.html")
            if not self.exist_ffmpeg:
                print(f'{self.ffmpeg} not exist. The result will not be converted to *.mp4.\nPlease Download ffmpeg and put it in the same directory as the script or use arg "--ffmpeg-path PATH".')
            if not self.exist_ffprobe:
                print(f'{self.ffprobe} not exist. The result duration will not be checked.\nPlease Download ffprobe and put it in the same directory as the script or use arg "--ffprobe-path PATH".')

        ensure_folder_exists(OUTPUT_PATH)

        cookies = self.get_access_cookies()

        result_path = output_path('result.html')

        if self.args.use_cache and os.path.exists(result_path):
            page = open(result_path).read()
        else:
            r = requests.get(f'https://play.boomstream.com/{self.args.entity}', headers=headers, cookies=cookies)

            with open(result_path, 'wt') as f:
                f.write(r.text)

            page = r.text

        self.config = self.get_boomstream_config(page)
        if "mediaData" not in self.config or "duration" not in self.config['mediaData']:
            raise ValueError(
                "Video config is not available. Probably, the live streaming has not finished yet, or you use "
                "an incorrect pin code. If you're sure that translation is finished and pin code is correct, please "
                "create an issue in project github tracker and attach your boomstream.config.json file.")
        self.token = self.get_token()
        self.m3u8_url = self.get_m3u8_url()
        self.expected_result_duration = float(self.config['mediaData']['duration'])

        print(f"Token = {self.token}")
        print(f"Playlist: {self.m3u8_url}")

        playlist = self.get_playlist(self.m3u8_url)
        chunklist = self.get_chunklist(playlist)

        xmedia_ready = self.get_xmedia_ready(chunklist)

        print(f'X-MEDIA-READY: {xmedia_ready}')

        key, iv = self.get_aes_key(xmedia_ready)
        cipher = create_cipher(key, iv)
        title = self.get_title()
        print(f"Title = {title}")

        filenames = self.download_chunks(chunklist, cipher, title)
        self.merge_chunks(filenames, self.expected_result_duration, title)

if __name__ == '__main__':
    app = App()
    sys.exit(app.run())