Bilibili 视频下载 Python 实现

前言

本文目标功能基于B站API实现

使用到的API

URL: https://api.bilibili.com/x/player/pagelist
method: GET
remarks: 获取 视频cid 用以下载

url 参数:

参数名	必选	备注
aid	否	视频的av号(avid)
bvid	否	视频的bv号
jsonp	否	默认值"jsonp"无需变动

注: aid 和 bvid必选一个

headers 参数:
目前没有必要的请求头参数

返回值:
以json格式返回, 包含视频cid part (标题) 等等…(包括所有子视频)

例:

URL: https://api.bilibili.com/x/player/playurl
method: GET
remarks: 获取视频资源

url 参数:

参数名	必选	备注
qn	否	指定返回下载链接的视频清晰度
otype	否	默认值 “json”
cid	是	avid bvid同一页面只需变更 cid 来下载子视频
avid	否	av号
bvid	否	bv号

bvid avid必选一个, avid还需要转换

headers 参数:

参数名	备注
Cookie	只需要 SESSDATA 字段
User-Agent	使用postman测试结果表示可以不用设置（建议添加）

例:

内容限制对这些接口的描述比较片面, 如有不明白的地方可以自行研究,也可以加我(俺)一起讨论讨论。

基本思路

1. 通过API 获取 视频cid
   • 可能含有多个视频
2. 使用 cid avid (或 bvid)请求接口返回视频下载链接
3. 下载
   • 这个 playurl 接口返回的视频链接下载速度巨慢…可能需要多线程下载(至少俺就是这么淦的)

具体实现

环境: Python3.9
requests==2.25.1

import requests
import threading
import os, time


from requests.api import head


class CodeChange:
    def __init__(self) -> None:
        self.table='fZodR9XQDSUm21yCkr6zBqiveYah8bt4xsWpHnJE7jL5VG3guMTKNPAwcF'
        self.tr={}
        for i in range(58):
            self.tr[self.table[i]]=i
        self.s=[11,10,3,8,4,6]
        self.xor=177451812
        self.add=8728348608

    def decode(self, x):
        r=0
        for i in range(6):
            r+=self.tr[x[self.s[i]]]*58**i
        return (r-self.add)^self.xor

    def encode(self, x):
        x=(x^self.xor)+self.add
        r=list('BV1  4 1 7  ')
        for i in range(6):
            r[self.s[i]]=self.table[x//58**i%58]
        return ''.join(r)


class BiliBiliVideoDownloader:
    USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/88.0.4324.190 Safari/537.36"
    SESSDATA = "你的SESSDATA"
    # 用于 bv av 互转
    change = CodeChange()
    
    accept_description = []
    accept_quality = []


    def __init__(self, bvid) -> None:
        # bv号
        self.bvid = bvid

        # av号
        self.avid = self.change.decode(bvid)

        # 获取视频信息列表
        self.pagelist = self.__API_pagelist(bvid)


    # 发送 api
    def __SEND_api(self, url, api_name):
        headers = {
            "user-agent": self.USER_AGENT,
            "cookie": f"SESSDATA={self.SESSDATA}"
        }
        try:
            return requests.get(url, headers=headers)
        except:
            print(f"API Error with \"{api_name}\"")


    def __API_pagelist(self, bvid):
        url = f"https://api.bilibili.com/x/player/pagelist?bvid={bvid}&jsonp=jsonp"
        
        response = self.__SEND_api(url, "__API_pagelist")
        data = response.json()["data"]
        # 列表生成器 （方便查看）
        return [
            {
                "cid": item["cid"],
                "page": item["page"],
                "title": item["part"]
            }
            for item in data
        ]
    
    def __API_playurl(self, qn, cid, bvid, otype="json"):
        url = f"https://api.bilibili.com/x/player/playurl?qn={qn}&otype={otype}&cid={cid}&bvid={bvid}"
        # print(url)
        response = self.__SEND_api(url, "__API_playurl")
        return response.json()["data"]

    def Show_Accept_Quality(self):
        if not (self.accept_description and self.accept_quality):
            # use api get the video accept quality
            data = self.__API_playurl("16", self.pagelist[0]["cid"], self.bvid)
            
            self.accept_description = data["accept_description"]
            self.accept_quality = data["accept_quality"]
        
        # output 
        for i in range(len(self.accept_description)):
            print(f"{i}  {self.accept_description[i]}")

    def Show_All_Page_Title(self):
        for i in range(len(self.pagelist)):
            print(f"{i}. {self.pagelist[i]['title']}")


    def __Get_Download_Link(self, index : int, quality : int) -> dict:
        """
        :data["durl"] 为一个列表
        :只有一个元素时视频既是一整段视频
        :大于一个元素每个元素都是一个视频切片 下载完成后还需要进行合并
        """
        cid = self.pagelist[index]["cid"]
        return self.__API_playurl(quality, cid, self.bvid)

    def Start_Download(self, part, qualityIndex):
        data = self.__Get_Download_Link(part, self.accept_quality[qualityIndex])
        title = self.pagelist[part]["title"]
        filename = f"{part}.{title}.{data['format'][:3]}"    # 提取前三个字符作为后缀 [mp3, flv]
        if len(data["durl"]) <= 1:
            filesize = data["durl"][0]["size"]
            url = data["durl"][0]["url"]
            workerThread = DownloadWorkerThread(url, filename, filesize=filesize)
            # 阻塞主线程
            workerThread.setDaemon(True)
            workerThread.start()
            workerThread.join()
            

    
        

class DownloadWorkerThread(threading.Thread):
    fileLock = threading.Lock()
    threadCount = 30
    maxBlockSize = 1024*1024*5
    curRangNum = 0
    curRangNumLock = threading.Lock()
    headers = {
        "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/88.0.4324.190 Safari/537.36",
        "referer": "https://www.bilibili.com/"
    }

    EMPTY_STR = ""
    ERROR_STATUS_VALUE = -1

    def __init__(self, url, filename, **kwargs) -> None:
        threading.Thread.__init__(self)
        self.url = url
        self.filename = filename
        self.curSize = 0


        self.filesize = kwargs.get("filesize") or self.Auto_Get_Filesize()
        self.maxBlockSize = kwargs.get("maxBlockSize") or self.maxBlockSize
        self.headers = kwargs.get("headers") or self.headers
        self.threadCount = kwargs.get("threadCount") or self.threadCount

        # check filename 
        if not os.path.exists(filename):
            # Create file 
            with open(filename, "w+")as f:
                f.close()
            
        

    def run(self):
        tLst = []
        print(f"""
*******************************************************************
文件名: {self.filename}
文件大小：{self.filesize//(1024**2)} MB
启用线程: {self.threadCount}
*******************************************************************
        """)

        for i in range(self.threadCount):
            t = threading.Thread(target=self.Range_Worker, args=(self, ))
            t.setDaemon(True)
            # print("Start thread: " + t.getName())
            t.start()
            tLst.append(t)
        print("Thread strat success")
        t = threading.Thread(target=self.Display_Progress, args=())
        t.setDaemon(True)
        t.start()
        tLst.append(t)

        for t in tLst:
            t.join()
        
    def Display_Progress(self):
        num = 0
        size = self.filesize // (1024**2)
        preSize = 0
        while num < 50:
            num = int(50*(self.curSize/self.filesize))
            # \033[32m  将后面文字设为绿色   \033[0m 设为默认
            print(f"\r [\033[32m{'■'*num}{' '*(50-num)}\033[0m]{self.curSize//(1024**2)}/{size} MB    {(self.curSize-preSize)//(1024**2)} MB/S", end="")
            preSize = self.curSize
            time.sleep(1)
            


    def Range_Worker(self, downloader):
        reqHeaders = self.headers.copy()
        
        while True:
            if self.curRangNum != 0 and self.curRangNum*self.maxBlockSize > self.filesize:
                break
            self.curRangNumLock.acquire()
            rag = f"{self.curRangNum * self.maxBlockSize}-{(lambda x : x if x < self.filesize else self.EMPTY_STR)((self.curRangNum+1) * self.maxBlockSize-1)}"
            self.curRangNum += 1
            self.curRangNumLock.release()
            reqHeaders["Range"] = "bytes=" + rag
            response = requests.get(url=self.url, headers=reqHeaders, stream=True)
            if response.ok:
                # print(rag)
                dataLst = []
                for chunk in response.iter_content(chunk_size=4096):
                    self.fileLock.acquire()
                    self.curSize += len(chunk)
                    self.fileLock.release()
                    dataLst.append(chunk)
                downloader.Write_Content(b''.join(dataLst), rag)
                # print(response.headers.get("Content-Range") + "\n", reqHeaders, "\n" + rag)
            else:
                print(rag + f": Error     status:{response.status_code}" )
                response.raise_for_status()

    def Write_Content(self, content, content_range):
        seekIndex = int(content_range.split("-")[0])
        self.fileLock.acquire()
        with open(self.filename, "rb+")as f:
            f.seek(seekIndex)
            f.write(content)
        self.fileLock.release()
        # print(f"{content_range}")

    def Auto_Get_Filesize(self) -> int:
        headers = self.headers.copy()
        headers["range"] = "bytes=0-0"
        response = requests.get(self,url, headers=headers, stream=True)
        if response.ok:
            return (lambda x : int(x.split("-")[-1]) if x != None else self.ERROR_STATUS_VALUE)(response.headers.get("Content-Range"))
        else:
            response.raise_for_status()

    def Auto_Get_Blocksize(self) -> int:
        pass

if __name__ == "__main__":    

    # 下载测试
    video = BiliBiliVideoDownloader("BV1w54y1y7Di")
    video.Show_All_Page_Title()
    # pageIndex = int(input("\nSelect page:"))
    video.Show_Accept_Quality()
    qualityIndex = int(input("\nSelect download quality:"))
    for pageIndex in range(len(video.pagelist)):
        print(f"\n{pageIndex}.{video.pagelist[pageIndex]['title']}")
        video.Start_Download(pageIndex, qualityIndex)