Commit 9557316d by win 10

新增了txt文件的来源和大部分平台来源解析

parent 6e7f47cf
...@@ -25,7 +25,7 @@ public class MediaSelfSource { ...@@ -25,7 +25,7 @@ public class MediaSelfSource {
public static void main(String[] args) { public static void main(String[] args) {
ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER); ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER);
List<String> urlList = new ArrayList<>(); List<String> urlList = new ArrayList<>();
urlList.add("https://www.360kuai.com/pc/9277f65f68bba0265?cota=3&kuai_so=1&sign=360_e39369d1&refer_scene=so_3"); urlList.add("http://sh.qihoo.com/pc/9dcfa48989d33df34?cota=1&sign=360_e39369d1&refer_scene=so_3");
List<MediaSelfSourceBean> u = MediaSelfSource.getMediaSelfSource(urlList); List<MediaSelfSourceBean> u = MediaSelfSource.getMediaSelfSource(urlList);
for(MediaSelfSourceBean b : u) { for(MediaSelfSourceBean b : u) {
System.out.println(b.toString()); System.out.println(b.toString());
......
...@@ -81,7 +81,7 @@ public class SourceForward { ...@@ -81,7 +81,7 @@ public class SourceForward {
public static void main(String[] args) { public static void main(String[] args) {
ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER); ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER);
List<String> urlList = new ArrayList<>(); List<String> urlList = new ArrayList<>();
urlList.add("http://www.northnews.cn/2019/0419/3080909.shtml"); urlList.add("https://www.jiemian.com/article/2782869.html");
List<SourceForwardBean> da = SourceForward.getSourceForward(urlList); List<SourceForwardBean> da = SourceForward.getSourceForward(urlList);
for(SourceForwardBean sfb : da) { for(SourceForwardBean sfb : da) {
System.out.println(sfb.toString()); System.out.println(sfb.toString());
......
...@@ -9,6 +9,8 @@ import java.util.Map.Entry; ...@@ -9,6 +9,8 @@ import java.util.Map.Entry;
import org.apache.logging.log4j.LogManager; import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger; import org.apache.logging.log4j.Logger;
import com.zhiwei.common.config.GroupType;
import com.zhiwei.crawler.proxy.ProxyFactory;
import com.zhiwei.source_forward.bean.UrlLiveBean; import com.zhiwei.source_forward.bean.UrlLiveBean;
import com.zhiwei.source_forward.bean.UrlLiveBean.Attribution; import com.zhiwei.source_forward.bean.UrlLiveBean.Attribution;
import com.zhiwei.source_forward.crawler.UrlLiveCrawler; import com.zhiwei.source_forward.crawler.UrlLiveCrawler;
...@@ -69,17 +71,17 @@ public class URLLive { ...@@ -69,17 +71,17 @@ public class URLLive {
return UrlLiveCrawlerThread.getUrlLiveCrawle(urlList); return UrlLiveCrawlerThread.getUrlLiveCrawle(urlList);
} }
// public static void main(String[] args) { public static void main(String[] args) {
// ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER); ProxyFactory.init("zookeeper://192.168.0.36:2181","local",GroupType.PROVIDER);
// List<String> urlList = new ArrayList<>(); List<String> urlList = new ArrayList<>();
// urlList.add("https://www.toutiao.com/a6680674354260345355"); urlList.add("http://www.ebrun.com/ebrungo/zb/316384.shtml");
//
//
// List<UrlLiveBean> u = URLLive.verificationURLLive(urlList); List<UrlLiveBean> u = URLLive.verificationURLLive(urlList);
// for(UrlLiveBean b : u) { for(UrlLiveBean b : u) {
// System.out.println(b.toString()); System.out.println(b.toString());
// } }
// } }
static class UrlLiveCrawlerThread extends Thread{ static class UrlLiveCrawlerThread extends Thread{
......
...@@ -893,6 +893,7 @@ ZOL中关村在线 ...@@ -893,6 +893,7 @@ ZOL中关村在线
华东理工大学 华东理工大学
华东在线 华东在线
华尔街见闻 华尔街见闻
华尔街见闻网
华股财经 华股财经
华龙网 华龙网
华龙网法律频道 华龙网法律频道
...@@ -2488,6 +2489,7 @@ ZOL中关村在线 ...@@ -2488,6 +2489,7 @@ ZOL中关村在线
智慧长沙 智慧长沙
智慧长沙资讯 智慧长沙资讯
智能派 智能派
智通财经
智通财经网 智通财经网
置家网 置家网
中安在线 中安在线
...@@ -3053,3 +3055,26 @@ ZOL中关村在线 ...@@ -3053,3 +3055,26 @@ ZOL中关村在线
最高人民法院网 最高人民法院网
最高人民检察院 最高人民检察院
今日湖北 今日湖北
中国经营报
三言财经
TechWeb.com.cn
中企网
央视新闻移动网
新浪财经-自媒体综合
T媒体
《法人》
国是直通车
科技小肆
雷帝触网
铅笔道
三秦都市报
新浪财经综合
央视财经
第一财经
第一赢销网
国际金融报
A5创业网
运营商世界网讯
中外管理杂志
上游新闻综合
新蓝网·浙江网络广播电视台
\ No newline at end of file
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment