Commit d2860d69 by zhiwei

修复自媒体号中今日头条匹配规则

parent b7e2a3dc
......@@ -130,31 +130,54 @@ public class TreateData {
if(url.contains("toutiao.com")){
//今日头条帐号匹配
if(html.contains(" name: '")){
source = "今日头条-" + html.split("name: '")[1].split("',")[0];
source = html.split("name: '")[1].split("',")[0].trim();
}else if(html.contains("screen_name:")){
source = html.split("screen_name:'")[1].split("',")[0].trim();
}
if(source!=null && !source.equals("")){
source = "今日头条-" + source;
}
}else if(url.contains("sohu.com")){
//搜狐自媒体号
if(html.contains("<meta name=\"mediaid\"")){
source = "搜狐-" + html.split("<meta name=\"mediaid\" content=\"")[1].split("\"")[0];
source = "搜狐-" + html.split("<meta name=\"mediaid\" content=\"")[1].split("\"")[0].trim();
if(source!=null && !source.equals("")){
source = "搜狐-" + source;
}
}
}else if(url.contains("a.mini.eastday.com")){
//处理东方头条网-自媒体号匹配
source = "东方头条-" + document.select("[class=\"share_cnt_p clearfix\"]").select("div.fl").select("i").get(1).text();
source = document.select("[class=\"share_cnt_p clearfix\"]").select("div.fl").select("i").get(1).text().trim();
if(source!=null && !source.equals("")){
source = "东方头条-" + source;
}
}else if(url.contains("sh.qihoo.com")){
//今日报点解析
source = "今日爆点-" + document.select("p.info").select("span.source").text();
source = document.select("p.info").select("span.source").text().trim();
if(source!=null && !source.equals("")){
source = "今日爆点-" + source;
}
}else if(url.contains("cj.sina.com.cn")){
//新浪财经头条号
if(html.contains("<meta name=\"mediaid\"")){
source = "财经头条-" + html.split("<meta name=\"mediaid\" content=\"")[1].split("\"")[0];
source = html.split("<meta name=\"mediaid\" content=\"")[1].split("\"")[0].trim();
if(source!=null && !source.equals("")){
source = "财经头条-" + source;
}
}
}else if(url.contains("baijia.baidu.com")){
//百度百家
source = "百家号-" + document.select("section.info").select("span.author").text();
source = document.select("section.info").select("span.author").text().trim();
if(source!=null && !source.equals("")){
source = "百家号-" + source;
}
}else if(url.contains("yidianzixun.com")){
//一点资讯
if(html.contains("related_wemedia")){
source = "一点号-" + html.split("media_name\":\"")[1].split("\",\"")[0];
source = html.split("media_name\":\"")[1].split("\",\"")[0].trim();
if(source!=null && !source.equals("")){
source = "一点号-" + source;
}
}else{
source = html.split("source\":\"")[1].split("\",\"")[0];
}
......
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