java实现抖音、快手、微视等短视频解析去除水印下载无水印视频

import cn.hutool.core.map.MapUtil;
import cn.hutool.http.HttpUtil;
import com.alibaba.fastjson.JSONObject;
import org.jsoup.Connection;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.select.Elements;
import java.io.*;
import java.net.HttpURLConnection;
import java.net.URL;import java.util.Date;
import java.util.HashMap;
import java.util.Map;
/**  org.jsoup jsoup 1.12.2  */
public class VideoService { 
/** * 方法描述: 快手解析下载视频 * 用Jsoup 爬快手视频
 * 1、首先获取js信息 
* 2、取出window.pageData的信息 
* @param url 
* @author lxb
 */ 
public static String kuaiShouParseUrl(String url) { 
HashMap headers = new HashMap<>();
 headers.put("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); 
headers.put("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); 
headers.put("Accept-Encoding", "gzip, deflate, sdch, br"); 
headers.put("Accept-Language", "zh-cn,zh;q=0.8"); 
headers.put("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); 
String redirectUrl = HttpUtil.createGet(url).addHeaders(headers).execute().header("Location"); 
System.out.println(redirectUrl); String body = HttpUtil.createGet(redirectUrl).addHeaders(headers).execute().body(); 
Document doc = Jsoup.parse(body); Elements videoElement = doc.select("script[type=text/javascript]"); 
String videoInfo = videoElement.get(3).data().replaceAll("window.pageData= ", ""); 
// 获取当前视频用户信息 JSONObject json = JSONObject.parseObject(videoInfo); 
// 获取当前视频信息 JSONObject video = json.getJSONObject("video"); 
// 无水印视频地址 String videoUrl = video.getString("srcNoMark");
 // 截取前面的数据 videoUrl = videoUrl.substring(0, videoUrl.indexOf("?")); return videoUrl; } 
/** * 抖音解析下载视频 * 用Jsoup 解析
 * 1、先获取该抖音视频ID 
* 2、调用抖音接口获取视频真实地址 https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids= + 抖音视频ID 
* 3、获取抖音无水印地址 把 https://aweme.snssdk.com/aweme/v1/playwm/ 的playwm替换成play 
* 4、重新请求一次抖音无水印地址获取该视频原地址 * @param url 
* @author lxb 
*/ 
public static String douYinParseUrl(String url) { 
String videoUrl = ""; 
try { Connection con= Jsoup.connect(url); 
con.header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); 
con.header("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); 
con.header("Accept-Encoding", "gzip, deflate, sdch, br"); 
con.header("Accept-Language", "zh-cn,zh;q=0.8"); 
con.header("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); Connection.Response resp= con.method(Connection.Method.GET).execute(); String respUrl = resp.url().toString(); /** * 加上/video/这个7位数 * 获取 视频ID = 获取/video/ 和 / 中间的数据 */ String itemId = respUrl.split("/video/")[1].split("/")[0]; /** * （固定）https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids= * videoUrl = 视频链接 + 视频ID */ String trueAddress= "https://www.iesdouyin.com/web/api/v2/aweme/iteminfo/?item_ids=" + itemId; String jsonStr = Jsoup.connect(trueAddress).ignoreContentType(true).execute().body(); JSONObject obj = JSONObject.parseObject(jsonStr); // 获取当前的视频的真实url String videoAddress= obj.getJSONArray("item_list").getJSONObject(0).getJSONObject("video").getJSONObject("play_addr").getJSONArray("url_list").get(0).toString(); /** * https://aweme.snssdk.com/aweme/v1/playwm/?video_id= * 把playwm替换成play */ videoAddress = videoAddress.replaceAll("playwm", "play"); // 重新请求一次抖音无水印地址 Map headers = MapUtil.newHashMap(); headers.put("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); headers.put("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); headers.put("Accept-Encoding", "gzip, deflate, sdch, br"); headers.put("Accept-Language", "zh-cn,zh;q=0.8"); headers.put("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); // 该视频的原地址 videoUrl = HttpUtil.createGet(videoAddress).addHeaders(headers).execute().header("Location"); }catch (Exception e){ e.printStackTrace(); } return videoUrl; } /** * 抖音火山解析下载视频 * @param url * @author lxb */ public static String huoShanParseUrl(String url) { String videoUrl = ""; try { Connection con= Jsoup.connect(url); con.header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); con.header("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); con.header("Accept-Encoding", "gzip, deflate, sdch, br"); con.header("Accept-Language", "zh-cn,zh;q=0.8"); con.header("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); Connection.Response resp= con.method(Connection.Method.GET).execute(); String respUrl = resp.url().toString(); String itemId = respUrl.split("item_id=")[1].split("&")[0]; /** * （固定）https://share.huoshan.com/api/item/info?item_id= * videoUrl = 视频链接 + 视频ID */ String trueAddress= "https://share.huoshan.com/api/item/info?item_id=" + itemId; String jsonStr = Jsoup.connect(trueAddress).ignoreContentType(true).execute().body(); JSONObject obj = JSONObject.parseObject(jsonStr); // 获取当前的视频的真实url String video_url= obj.getJSONObject("data").getJSONObject("item_info").getString("url"); String video_id = video_url.split("video_id=")[1].split("&")[0]; String real_url = "http://hotsoon.snssdk.com/hotsoon/item/video/_playback/?video_id=" + video_id; // 重新请求一次无水印地址 Map headers = MapUtil.newHashMap(); headers.put("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); headers.put("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); headers.put("Accept-Encoding", "gzip, deflate, sdch, br"); headers.put("Accept-Language", "zh-cn,zh;q=0.8"); headers.put("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); // 该视频的原地址 videoUrl = HttpUtil.createGet(real_url).addHeaders(headers).execute().header("Location"); }catch (Exception e){ e.printStackTrace(); } return videoUrl; } /** * 微视解析下载视频 * @param url * @author lxb */ public static String weiShiParseUrl(String url) { String videoUrl = ""; try { Connection con= Jsoup.connect(url); con.header("Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8"); con.header("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7"); con.header("Accept-Encoding", "gzip, deflate, sdch, br"); con.header("Accept-Language", "zh-cn,zh;q=0.8"); con.header("User-Agent", "Mozilla/5.0 (iPhone; CPU iPhone OS 12_1_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Mobile/16D57 Version/12.0 Safari/604.1"); Connection.Response resp= con.method(Connection.Method.GET).execute(); String respUrl = resp.url().toString(); String itemId = respUrl.split("feedid=")[1].split("&")[0]; /** * （固定）https://h5.qzone.qq.com/webapp/json/weishi/WSH5GetPlayPage?feedid * videoUrl = 视频链接 + 视频ID */ String trueAddress= "https://h5.qzone.qq.com/webapp/json/weishi/WSH5GetPlayPage?feedid=" + itemId; String jsonStr = Jsoup.connect(trueAddress).ignoreContentType(true).execute().body(); JSONObject obj = JSONObject.parseObject(jsonStr); // 获取当前的视频的真实url videoUrl= obj.getJSONObject("data").getJSONArray("feeds").getJSONObject(0).getString("video_url"); }catch (Exception e){ e.printStackTrace(); } return videoUrl; } public static void main(String[] args){// System.out.println(weiShiParseUrl(Utils.filterUrl("嗨，我正在参加微视“谁是王者人气英雄”挑战赛，快来帮我投票吧！>>https://h5.weishi.qq.com/weishi/wsplay/challenge2?feedid=78sKriqyF1KyJk2cT&challegeid=100543&spid=1558186608003629&qua=v1_iph_weishi_8.2.7_241_app_a&chid=127081004&pkg=&attach=cp_reserves3_1190370002")));// System.out.println(huoShanParseUrl(filterUrl("金名美食（家常菜)在火山分享了视频，快来围观！传送门戳我>>https://share.huoshan.com/hotsoon/s/FJ8Wl9cV0d8/ 复制此链接，打开【抖音火山版】，直接观看视频~")));// System.out.println(kuaishouParseUrl(filterUrl("出去玩就这？就这？ #黑丝 https://v.kuaishou.com/6lIPLQ 复制此消息，打开【快手】直接观看！")));// System.out.println(douYinParseUrl(filterUrl("二三十岁人了，都是老样子，人很普通，说话又不够甜，长得也一般般，重要是，还没什么钱。#情感文字视频 https://v.douyin.com/JPoaJry/ 复制此链接，打开抖音，直接看观视频！")));// System.out.println(filterUrl("二三十岁人了，都是老样子，人很普通，说话又不够甜，长得也一般般，重要是，还没什么钱。#情感文字视频 https://v.douyin.com/JPoaJry/ 复制此链接，打开抖音，直接看观视频！")); } // 保存视频 public static void save(String s) throws Exception { long start = System.currentTimeMillis(); // 打开和URL之间的连接 HttpURLConnection conn2 = (HttpURLConnection) new URL(s).openConnection(); // 设置通用的请求属性 conn2.setRequestProperty("accept", "*/*"); conn2.setRequestProperty("connection", "Keep-Alive"); conn2.setRequestProperty("Accept-Charset", "utf-8"); conn2.setRequestProperty("Content-Type", "audio/mp3"); conn2.setRequestProperty("user-agent", "Mozilla/5.0 (compatible; MSIE 6.0; Windows NT 5.1;SV1)"); conn2.setRequestMethod("GET"); conn2.connect(); // 定义BufferedReader输入流来读取URL的响应 InputStream in2 = conn2.getInputStream(); String last = ".mp4"; if (s.contains("mp4")) { last = ".mp4"; } else if (s.contains("mp3")) { last = ".mp3"; } else { last = ".mp4"; } //文件保存的位置 FileOutputStream fileOutputStream = new FileOutputStream(new File("D:\\" + new Date().getTime() + last)); byte[] b = new byte[1024 * 5]; int len; while ((len = in2.read(b)) != -1) { fileOutputStream.write(b, 0, len); } long end = System.currentTimeMillis(); System.out.println("用时间：" + (end - start) + "毫秒"); fileOutputStream.flush(); fileOutputStream.close(); }}