文章目錄
一．准備工作
二．預覽
- 1.啟動
- 2.解析
- 3.下載中
- 4.下載完成
- 5.結果
三．設計流程
- 1.bilibili_video_spider
- 2.視頻json的查找
四．源代碼
- 1.Bilibili_Video_Downloader-GUI
- 2.bilibili_video_spider
五．總結
由於B站沒有PC客戶端，電腦下載視頻很不方便，遂使用Tk編寫一款B站視頻下載工具，輸入一個網址選擇清晰度之后就能夠下載對應的視頻，可以下載單P、合集、合集單P，使用可視化GUI圖形界面，交互性更強，來吧，展示~
很多人學習python，不知道從何學起。 很多人學習python，掌握了基本語法過后，不知道在哪里尋找案例上手。 很多已經做案例的人，卻不知道如何去學習更加高深的知識。 那么針對這三類人，我給大家提供一個好的學習平台，免費領取視頻教程，電子書籍，以及課程的源代碼！ QQ群：701698587 歡迎加入，一起討論 一起學習！
一．准備工作
tkinter、os系統模塊、re正則模塊、subprocess新的進程模塊、還有本次比較重要的ffmpeg.exe用於視頻和音頻的合並，關於ffmpeg請參考：ffmpeg - 百度百科
二．預覽
1.啟動
2.解析
解析出多個清晰度視頻以供下載
3.下載中
4.下載完成
分別下載完視頻和音頻后，對它們進行合並，最后輸出一個完整的視頻文件
5.結果
1080P+，針不戳
三．設計流程
1.bilibili_video_spider
2.視頻json的查找
- 首先查看網頁源代碼
- 在網頁的這個js里，能夠找到關於視頻的相關視頻、音頻、視頻質量、長度、格式…等信息，直接正則截取就好啦
緊接着，下面這個js里，就是視頻的aid、分P信息、up主信息、相關視頻推薦信息，也用正則就能截取
四．源代碼
1.Bilibili_Video_Downloader-GUI
from tkinter import * from tkinter import ttk from tkinter import messagebox import os import threading from bilibili_video_spider import Bibili_Video_Spider as sp2 import re from my_util import My_Util """ GUI+Spider """ class App: def __init__(self): self.base_dir = './bilibili_videos/' self.start_flag='' self.has_more_flag='' self.spider=sp2() self.create_widget() self.set_widget() self.place_widget() self.window.mainloop() def create_widget(self): self.window = Tk() self.window.title('Bilibili_Video_Downloader-v1.0') width = 450 height = 520 screen_width = self.window.winfo_screenwidth() screen_height = self.window.winfo_screenheight() left = (screen_width - width) / 2 top = (screen_height - height) / 2 self.window.geometry("%dx%d+%d+%d" % (width, height, left, top)) self.window.resizable(0, 0) self.l1 = ttk.Label(self.window, text='請輸入視頻鏈接地址：') self.e1_var=StringVar() self.e1 = ttk.Entry(self.window, width=90,textvariable=self.e1_var) self.l5 = ttk.Label(self.window, text='選擇清晰度：') self.combobox=ttk.Combobox(self.window,state='readonly',width=15,justify='center') self.l2 = ttk.Label(self.window, text='當前狀態：') self.t1 = Text(self.window, width=80, height=20) self.l3_var=StringVar() self.l3 = ttk.Label(self.window, text='當前下載進度：',textvariable=self.l3_var) self.progress=ttk.Progressbar(self.window,orient=HORIZONTAL,length=400,mode='determinate',value=0,maximum=100) self.l4_var = StringVar() self.l4_var.set('0.0%[未下載]') self.l4 = ttk.Label(self.window, textvariable=self.l4_var) self.b1 = ttk.Button(self.window, text='解析', command=lambda: self.thread_it(self.pre_analysis)) self.b2 = ttk.Button(self.window, text='下載', command=lambda: self.thread_it(self.donwload_video)) def set_widget(self): self.window.protocol('WM_DELETE_WINDOW', self.quit_window) self.window.bind('<Escape>', self.escape) self.e1.bind('<Return>', self.enter) self.b2.config(state=DISABLED) self.combobox.config(value=['--請先解析--']) self.combobox.current(0) def place_widget(self): self.l1.pack(anchor="w") self.e1.pack(anchor="w", padx=20) self.l5.pack(anchor="w",pady=5) self.combobox.pack(anchor="center") self.l2.pack(anchor="w") self.t1.pack(anchor="w", padx=20) self.l3.pack(anchor="w",pady=5) self.progress.pack(pady=5) self.l4.pack() self.b1.pack(side='left', padx=90) self.b2.pack(side='left', padx=10) def pre_analysis(self): input_video_link = self.e1.get() input_video_link=input_video_link.strip() if input_video_link.startswith(r'https://www.bilibili.com/video/'): if '&' in input_video_link: raw_link=input_video_link.split('&')[0] else: raw_link=input_video_link try: #av 轉 bv av_number = int(re.findall('https://www.bilibili.com/video/av(\d+)?', raw_link)[0]) url=raw_link.replace(av_number,My_Util().av_convert_bv(av_number)) except IndexError: url=raw_link self.spider.set_start_url(url) self.spider.get_page_html() self.video_number = self.spider.get_video_number() base_title = self.spider.get_video_title() if re.match('https://www.bilibili.com/video/.*\?p=\d+',url): current_num=re.findall('https://www.bilibili.com/video/.*\?p=(\d+)',url) self.has_more_flag=True self.current_video_title=self.spider.part_name_list[int(current_num[0])] else: self.has_more_flag=False self.current_video_title=base_title self.entrace_url=url self.analysis_videos(url) if self.start_flag!=True: self.b2.config(state=NORMAL) # self.b1.config(state=DISABLED) else: messagebox.showwarning('警告', '請輸入正確的分享鏈接！') self.e1_var.set('') def analysis_videos(self,url): """ :param url: :return: """ My_Util().do_makedirs(self.base_dir) self.video_item_ = self.spider.get_video_and_audio(self.spider.get_video_detail_json()) video_quality_list=[] for video_detail in self.video_item_['video_detail']: for data in video_detail.items(): video_quality_list.append(data[0]) self.combobox.config(value=video_quality_list) self.combobox.current(0) self.t1.delete(0.0,END) self.insert_to_t1(f'[視頻標題]：{self.current_video_title}') self.insert_to_t1(f'[視頻時長]：{self.video_item_["video_length"]}') self.insert_to_t1(f'[視頻清晰度]：{" ".join(video_quality_list)}') self.insert_to_t1(f'請選擇清晰度后點擊下載按鈕---------------',time_str=False) def donwload_video(self): self.start_flag=True self.b2.config(state=DISABLED) if self.has_more_flag: ret = messagebox.askyesno('提示', '此視頻包含多P，是否下載全集？') if ret: download_more=True else: download_more=False else: download_more=False for i in range(self.video_number): if download_more: begin_url = self.entrace_url.split('?')[0] + f'?p={i+1}' self.spider.video_title = self.spider.part_name_list[i] current_title=self.spider.part_name_list[i] else: begin_url=self.entrace_url self.spider.video_title = self.current_video_title current_title =self.current_video_title self.insert_to_t1(f'開始下載{current_title}---------------') self.l3_var.set('視頻下載進度：') self.spider.set_start_url(begin_url) video_item_ = self.spider.get_video_and_audio(self.spider.get_video_detail_json()) video_url_list=[] for video_detail in video_item_['video_detail']: for data in video_detail.items(): video_url_list.append(data[1]) download_url = video_url_list[self.combobox.current()] current_video_name=self.spider.part_name_list[i] for progrees, speed in self.spider.down_video(download_url,): self.progress['value'] = progrees self.l4_var.set(f'進度：%.1f%% 速度：%s' % (progrees, speed)) self.progress.update() self.insert_to_t1(f'[{current_video_name}視頻下載完成...') self.l4_var.set('100%[下載完成]') self.insert_to_t1('-' * 30) audio_url = video_item_['audio_url'] self.insert_to_t1(f'開始下載{current_title}音頻---------------') self.l3_var.set('音頻下載進度：') for progrees, speed in self.spider.downlonad_autio(audio_url,): self.progress['value'] = progrees self.l4_var.set(f'進度：%.1f%% 速度：%s' % (progrees, speed)) self.progress.update() self.insert_to_t1(f'[{current_video_name}音頻下載完成...') self.l4_var.set('100%[下載完成]') self.insert_to_t1('-' * 30) self.insert_to_t1(f'開始合並視頻---------------') if (self.spider.mix_video()): self.insert_to_t1(f'清理臨時視頻文件完成---------------') self.insert_to_t1(f'清理臨時音頻文件完成---------------') self.insert_to_t1(f'合並視頻完成---------------') else: self.insert_to_t1(f'發生了異常錯誤！---------------') if not download_more: break self.b1.config(state=NORMAL) self.b2.config(state=NORMAL) def insert_to_t1(self,line,time_str=True): if time_str==True: time_string=My_Util().get_time_string() self.t1.insert(END,f'[{time_string}]'+line+'\n') else: self.t1.insert(END,line+'\n') self.t1.yview_moveto(1) def open_dir(self): abs_path = os.path.abspath(self.base_dir) # 使用絕對路徑打開文件夾 os.startfile(abs_path) def quit_window(self): ret = messagebox.askyesno('提示', '是否要退出？') if ret == True: self.window.destroy() def escape(self,event): self.quit_window() def connect_author(self): messagebox.showinfo('聯系作者', '作者QQ：懷淰メ') def enter(self,event): self.thread_it(self.pre_analysis) def thread_it(self,func, *args): t = threading.Thread(target=func, args=args) self.window.update() t.setDaemon(True) # 設置守護，主線程結束，子線程結束 t.start() if __name__ == '__main__': App() """ test https://www.bilibili.com/video/BV1ML411J7es """
2.bilibili_video_spider
import json import requests import re import os import subprocess from my_util import My_Util import time """ 版本2分別下載音頻和視頻，通過ffmpeg合並 三種情況 1.單P 2.多P下載單集 3.多P下載全集 """ class Bibili_Video_Spider(object): def __init__(self,): self.s=requests.session() self.headers={ 'Content-Range': 'bytes 0-xxxxxx', "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36" } self.util=My_Util() def set_start_url(self,start_url): self.start_url=start_url self.get_page_html() def get_video_title(self): """ 起始視頻標題，作為下載視頻的目錄名 :return: """ regx='name="keywords" content="(.*?),' title=re.findall(regx,self.srart_html) title=title[0] return title def get_page_html(self): """ 獲取網頁源代碼 :return: """ headers={ 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36', 'Content-Range': 'bytes 0-xxxxxx', 'Referer': self.start_url } r=self.s.get(self.start_url,headers=headers) if r.status_code==200: r.encoding='utf-8' self.srart_html=r.text def get_video_number(self): """ 是否含有多P，若含有分P，則將所有分P名字存入list :return: """ html_part = re.findall('window.__INITIAL_STATE__=(.*?)</script> <link rel="stylesheet"', self.srart_html) part_json_str = html_part[0].split(';(function(){var')[0] part_json = json.loads(part_json_str.strip()) pages = part_json['videoData']['pages'] self.part_name_list = [part_name['part'] for part_name in pages] if len(pages)!=1: part_number=len(pages) else: part_number=1 return part_number def get_video_detail_json(self): """ 獲取視頻詳情json,里面包括視頻m4a地址，以及audio音頻，版本2主要依賴此Json :return: """ regx='window.__playinfo__=(.*?)</script><script>window.__INITIAL_STATE' video_json_=re.findall(regx,self.srart_html) if video_json_: video_json=json.loads(video_json_[0]) return video_json def get_video_and_audio(self,page_json,): """ 獲取視頻的視頻和音頻，准備合並 :param page_json: :return: """ video_item={} video_data=[] data=page_json['data'] video_definition_list = data.get('accept_description') video_detail_=data.get('dash').get('video') video_link_list=[] for video_detail__ in video_detail_: video_url=video_detail__.get('baseUrl') video_link_list.append(video_url) for v in zip(video_definition_list,video_link_list): item = {} item[v[0]]=v[1] video_data.append(item) video_item['video_length']=self.util.Convert_Millis(page_json["data"]['timelength']) video_item['audio_url'] = page_json["data"]["dash"]["audio"][0]["baseUrl"] video_item['video_detail'] = video_data return video_item def down_video(self,video_url): """ 下載視頻 :param video_url: :param number: 分P的索引從1開始 :return: """ start_time=time.time() headers = { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36', 'Content-Range': 'bytes 0-xxxxxx', 'Referer': self.start_url } #下載視頻 r = self.s.get(video_url, stream=True, headers=headers) file_size=int(r.headers['Content-Length']) count=0 with open(self.video_title+'-temp.mp4','wb')as f: for chunk in r.iter_content(chunk_size=1024): f.write(chunk) count+=len(chunk) progress=float(count/file_size*100) speed = My_Util().format_size((count) / (time.time() - start_time)) + '/S' yield progress,speed def downlonad_autio(self,audio_url): start_time=time.time() headers = { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36', 'Content-Range': 'bytes 0-xxxxxx', 'Referer': self.start_url } r = self.s.get(audio_url, stream=True, headers=headers) file_size=int(r.headers['Content-Length']) count=0 with open(self.video_title+'-temp.aac','wb')as f: for chunk in r.iter_content(chunk_size=1024): f.write(chunk) count+=len(chunk) progress=float(count/file_size*100) speed = My_Util().format_size((count) / (time.time() - start_time)) + '/S' yield progress,speed def mix_video(self,): try: # print(f'開始合並{self.video_title}...') path = "ffmpeg.exe -i " + self.video_title + "-temp.mp4 -i " + self.video_title + "-temp.aac -vcodec copy -acodec copy " + self.video_title + ".mp4" subprocess.call(path, shell=True, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE) os.remove(self.video_title + "-temp.mp4") # print('[清理臨時視頻文件完成]...') os.remove(self.video_title + "-temp.aac") # print('[清理臨時音頻文件完成]...') return True except : return False
五．總結
- 本次使用tkinter加ffmpeg實現了B站視頻的下載，支持所選視頻的所有畫質的下載，tkinter完成GUI的搭建，實現交互，spider實現視頻的解析與下載，ffmpeg實現視頻與音頻的合並，初步實現了B站視頻的下載，當然這只是1.0版本，仍存在一些不足。
1.代碼邏輯混亂，復用率不高。（因為是分幾天寫成的，可能一些想法找不到了）。
2.主要功能較少，GUI的優勢沒有明顯凸顯出來（當前的功能，打包成命令行也能輕易實現）。
3.視頻下載到了一個目錄，應當將帶有分P的全集視頻，新建目錄后下載（這里確實有點吹毛求疵，畢竟是1.0）。