package com.Utill;

import com.alibaba.fastjson.JSONObject;
import com.ckms.comp.manager.im.govManage.service.GovManageServiceImpl;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.io.*;
import java.util.*;


/**
 * @Author:
 * @Date: 2019/10/1 13:43
 * @Describe:
 */
public class JsoupDemo {

    private static final Logger logger = LoggerFactory.getLogger(JsoupDemo.class);

    private String keyword = null;
    private String endtime = null;
    private String begintime = null;
    private String zzUrl = null;
    private boolean start = false;
    int isSize = 1;
    Integer pageSize = 0;
    Integer timeType = 6;  // 时间类型   6: 指定时间  5：近半年  0：今日  4：近3月


    /**
     * 爬虫调用主方法
     *
     * @param id        关键词的饿id
     * @param kw        关键词
     * @param startTime 开始时间
     * @param endTime   结束时间
     */
    public void fetch(Integer id, String kw, String startTime, String endTime) {
        logger.info("开始拉取网站数据》》》》》》》》》》》》》》》》》》》》》》》》》》》");
        Integer totalSize = null;
        try {
            keyword = toUtf8String(kw);
            begintime = startTime.replace("/", "%3A");
            endtime = endTime.replaceAll("/", "%3A");
            pageSize = 1;
            zzUrl = "http://search.ccgp.gov.cn/bxsearch?searchtype=1&page_index=" + pageSize + "&bidSort=0&buyerName=&projectId=&pinMu=0&bidType=0&dbselect=bidx&kw=" + keyword + "&start_time=" + begintime + "&end_time=" + endtime + "&timeType=" + timeType + "&displayZone=&zoneId=&pppStatus=0&agentName=";
            start = true;
            //获取总页数
            totalSize = fetchHomePageUrl(zzUrl);
            if (totalSize != null && totalSize > 0) {
                while (start) {

                    if (isSize > totalSize) {
                        start = false;
                    } else {
                        pageSize = isSize;
                        zzUrl = "http://search.ccgp.gov.cn/bxsearch?searchtype=1&page_index=" + isSize + "&bidSort=0&buyerName=&projectId=&pinMu=0&bidType=0&dbselect=bidx&kw=" + keyword + "&start_time=" + begintime + "&end_time=" + endtime + "&timeType=" + timeType + "&displayZone=&zoneId=&pppStatus=0&agentName=";
                        //开始遍历每一页数据
                        fetchingPageUrl(id, isSize, zzUrl);
                        isSize++;
                    }
                }
            } else {
                logger.info("关键词：" + kw + "没有查询到相关数据-" + new Date());
                isSize = 1;
            }
        } catch (Exception er) {
            logger.error("========================获取数据失败，重新获取中.....=================================");
        }
    }


    //URL汉字转码
    public static String toUtf8String(String s) {
        StringBuffer sb = new StringBuffer();
        for (int i = 0; i < s.length(); i++) {
            char c = s.charAt(i);
            if (c >= 0 && c <= 255) {
                sb.append(c);
            } else {
                byte[] b;
                try {
                    b = String.valueOf(c).getBytes("utf-8");
                } catch (Exception ex) {
                    System.out.println(ex);
                    b = new byte[0];
                }
                for (int j = 0; j < b.length; j++) {
                    int k = b[j];
                    if (k < 0)
                        k += 256;
                    sb.append("%" + Integer.toHexString(k).toUpperCase());
                }
            }
        }
        return sb.toString();
    }

    // 首次访问url ,需查找当前查询词所在的总页数
    public Integer fetchHomePageUrl(String ur) {
        Integer size = null;
        try {
            //maxBodySize : 0  不限制大小， 1： 1m    userAgent: 伪装成浏览器
            Document doc = Jsoup.connect(ur).header("Accept-Encoding", "gzip, deflate")
                    .userAgent("Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.31 (KHTML, like Gecko) Chrome/26.0.1410.64 Safari/537.31")
                    .maxBodySize(0).timeout(600000).get();
            Elements eli = doc.select(".vT-srch-result-list-con2 .vT-srch-result-list .vt-srch-result-list-bid>li");
            if (eli.size() > 0) {
                //获取分页html
                Elements allElements = doc.select(".vT_z .pager>script");
                if (allElements.size() > 0) {
                    String pager = allElements.get(0).childNodes().toString();
                    String startStr = pager.substring(7, pager.length());
                    String objectStr = startStr.substring(0, startStr.lastIndexOf(")"));
                    //string转json对象
                    JSONObject jsonObject = JSONObject.parseObject(objectStr);
                    //  共有多少页
                    size = Integer.valueOf((Integer) jsonObject.get("size"));
                }
            }
        } catch (IOException e1) {
            e1.printStackTrace();
        }
        return size;
    }

    // 遍历页数
    public void fetchingPageUrl(Integer id, Integer pageSize, String fetchUrl) {
        Integer is_homePage = 1;  // 0 ： 首页  1： 否

        String noticeOutline = null;
        try {
            Document pDoc = Jsoup.connect(fetchUrl).header("Accept-Encoding", "gzip, deflate")
                    .userAgent("Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.31 (KHTML, like Gecko) Chrome/26.0.1410.64 Safari/537.31")
                    .maxBodySize(0)
                    .timeout(600000).get();
            //根据css样式 获取当前页的所有url
            Elements elements = pDoc.select(".vT_z .vT-srch-result-list-bid>li");
            // Elements hrefs = doc.select("a[href]");
            //遍历当前页的所有url
            for (Element element : elements) {
                //根据Jsoup解析，解析每一页的详细数据，见下一篇博客

            }
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}