Commit ee793daf by zhiwei

更新提交

parent d4cb16b0
...@@ -2,7 +2,6 @@ package com.zhiwei.searchhotcrawler.dao; ...@@ -2,7 +2,6 @@ package com.zhiwei.searchhotcrawler.dao;
import java.util.ArrayList; import java.util.ArrayList;
import java.util.Calendar;
import java.util.Date; import java.util.Date;
import java.util.List; import java.util.List;
...@@ -30,7 +29,6 @@ public class HotSearchListDAO extends MongoDBTemplate{ ...@@ -30,7 +29,6 @@ public class HotSearchListDAO extends MongoDBTemplate{
String time = TimeParse.dateFormartString(new Date(), "yyyy-MM-dd"); String time = TimeParse.dateFormartString(new Date(), "yyyy-MM-dd");
String year = time.substring(0,4); String year = time.substring(0,4);
String month = time.substring(5,7); String month = time.substring(5,7);
String collName = Config.collName + year + "_" + month; String collName = Config.collName + year + "_" + month;
super.setCollName(collName); super.setCollName(collName);
} }
......
...@@ -36,7 +36,7 @@ public class HotSearchListTest{ ...@@ -36,7 +36,7 @@ public class HotSearchListTest{
Mongo mongoNew = new MongoClient(address, Arrays.asList(credentialNew)); Mongo mongoNew = new MongoClient(address, Arrays.asList(credentialNew));
DB dbNew = mongoNew.getDB("hot_search_list"); DB dbNew = mongoNew.getDB("hot_search_list");
Map<String,String> timLine = TimeParse.getTimeMap("2019-07-16 00:00:00", "2019-07-16 23:59:59", "HH", 1); Map<String,String> timLine = TimeParse.getTimeMap("2018-02-01 00:00:00", "2019-04-30 23:59:59", "MM", 1);
timLine.forEach((start, end) ->{ timLine.forEach((start, end) ->{
...@@ -49,29 +49,52 @@ public class HotSearchListTest{ ...@@ -49,29 +49,52 @@ public class HotSearchListTest{
System.out.println("collName=========="+collName); System.out.println("collName=========="+collName);
DBCollection collNew = dbNew.getCollection(collName); DBCollection collNew = dbNew.getCollection(collName);
DBObject query = new BasicDBObject(new BasicDBObject("time", DBObject countIndexDoc = new BasicDBObject();
new BasicDBObject("$gte",startDate).append("$lte", endDate))); countIndexDoc.put("count", -1);
DBCursor cur = coll.find(query); DBObject timeIndexDoc = new BasicDBObject();
System.out.println(query +"======="+ cur.count()); timeIndexDoc.put("time", -1);
List<DBObject> dataList = new ArrayList<>(); DBObject rankIndexDoc = new BasicDBObject();
int i = 0; rankIndexDoc.put("rank", -1);
while(cur.hasNext()) { DBObject nameIndexDoc = new BasicDBObject();
DBObject doc = cur.next(); nameIndexDoc.put("name", -1);
DBObject zhihu = new BasicDBObject(); DBObject typeIndexDoc = new BasicDBObject();
zhihu.put("_id", doc.get("_id")); typeIndexDoc.put("type", -1);
zhihu.put("name", doc.get("name"));
zhihu.put("url", doc.get("url")); try {
zhihu.put("count", doc.get("count")); collNew.createIndex(countIndexDoc, new BasicDBObject("name", "count_desc"));
zhihu.put("hot", doc.get("hot")); collNew.createIndex(timeIndexDoc, new BasicDBObject("name", "time_desc"));
zhihu.put("day", doc.get("day")); collNew.createIndex(rankIndexDoc, new BasicDBObject("name", "rank_desc"));
zhihu.put("time", doc.get("time")); collNew.createIndex(nameIndexDoc, new BasicDBObject("name", "name_desc"));
zhihu.put("changeCount", doc.get("changeCount")); collNew.createIndex(typeIndexDoc, new BasicDBObject("name", "type_desc"));
zhihu.put("rank", doc.get("rank")); } catch (Exception e) {
zhihu.put("type", HotSearchType.微博热搜.name()); e.printStackTrace();
collNew.save(zhihu);
dataList.add(zhihu);
} }
// DBObject query = new BasicDBObject(new BasicDBObject("time",
// new BasicDBObject("$gte",startDate).append("$lte", endDate)));
// DBCursor cur = coll.find(query);
// System.out.println(query +"======="+ cur.count());
// List<DBObject> dataList = new ArrayList<>();
// int i = 0;
// while(cur.hasNext()) {
// DBObject doc = cur.next();
// DBObject zhihu = new BasicDBObject();
// zhihu.put("_id", doc.get("_id"));
// zhihu.put("name", doc.get("name"));
// zhihu.put("url", doc.get("url"));
// zhihu.put("count", doc.get("count"));
// zhihu.put("hot", doc.get("hot"));
// zhihu.put("day", doc.get("day"));
// zhihu.put("time", doc.get("time"));
// zhihu.put("changeCount", doc.get("changeCount"));
// zhihu.put("rank", doc.get("rank"));
// zhihu.put("type", HotSearchType.微博热搜.name());
//
// collNew.save(zhihu);
// dataList.add(zhihu);
// }
// System.out.println(collName +"数据量大小" +dataList.size()); // System.out.println(collName +"数据量大小" +dataList.size());
// cur.close(); // cur.close();
// if(!dataList.isEmpty()) { // if(!dataList.isEmpty()) {
......
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