lucene4.x的分组实现
至此,已经可以简单的实现分组去重统计的功能了,如果业务比较复杂,例如像报表查询,以及一些特定的统计求和功能,这个就可能需要自己写了
Please read full article from lucene4.x的分组实现
public
static
List<HashMap<String, String>> testGroup(String indexPath,String groupField,String sumField){
List<HashMap<String, String>> map=
new
ArrayList<HashMap<String,String>>();
Directory d1=
null
;
IndexReader read1=
null
;
try
{
d1=FSDirectory.open(
new
File(indexPath));
//磁盘索引
read1=DirectoryReader.open(d1);
//打开流
IndexSearcher sear=
new
IndexSearcher(
new
MultiReader(read1));
//MultiReader此类可以多份索引的读入
//但是得保证各个索引的字段结构一致
GroupingSearch gSearch=
new
GroupingSearch(groupField);
//分组查询按照place分组
Query q=
new
WildcardQuery(
new
Term(groupField,
"*"
));
//查询所有数据
TopGroups t=gSearch.search(sear, q,
0
, Integer.MAX_VALUE);
//设置返回数据
GroupDocs[] g=t.groups;
//获取分组总数
System.out.println(
"总数据数"
+t.totalHitCount);
System.out.println(
"去重复后的数量:"
+g.length);
for
(
int
i=
0
;i<g.length;i++){
ScoreDoc []sd=g[i].scoreDocs;
String str =sear.doc(sd[
0
].doc).get(groupField);
int
total=sumcount(str,groupField,sumField,sear);
//System.out.println("place:"+str+"===>"+"个数:"+g[i].totalHits+);
System.out.println(
"place:"
+str+
"===>"
+
"个数:"
+g[i].totalHits);
HashMap<String, String> m=
new
HashMap<String, String>();
m.put(
"word"
, str);
m.put(
"wx_count"
, total+
""
);
m.put(
"wx_total"
,
"10000"
);
map.add(m);
}
read1.close();
//关闭资源
d1.close();
}
catch
(Exception e){
e.printStackTrace();
}
return
map;
}
Please read full article from lucene4.x的分组实现
No comments:
Post a Comment