Lucnen.net 实现高性能Group By 分组统计

Lucene要实现高性能的搜索结果分类统计功能，看了网上大多数都没有一个很好的解决方案。如果自己写循环统计那性能太差了，如果改动Lucene的dll又很麻烦，需要了解整个Lucene的搜索原理。在此我提供一个类，可以很好的解决此问题。

首先你在建立索引的时候必须添加一个额外的字段供分类统计使用，它必须是Field.Index.UN_TOKENIZED ；

//this field for search doc.Add(new Field("CompanyName", job.CompanyName, Field.Store.YES, Field.Index.TOKENIZED)); //this field for aggregation doc.Add(new Field("CompanyName_UN", job.CompanyName, Field.Store.YES, Field.Index.UN_TOKENIZED));

下面是实现高性能的分组统计类：

using System; using System.Collections; using System.Collections.Generic; using Lucene.Net; using Lucene.Net.Index; using Lucene.Net.QueryParsers; using Lucene.Net.Search; public class SimpleFacets { public static Dictionary<string, int> Facet(Query query, IndexSearcher s, string field) { StringIndex stringIndex = FieldCache_Fields.DEFAULT.GetStringIndex(s.GetIndexReader(), field); int[] c = new int[stringIndex.lookup.Length]; FacetCollector results = new FacetCollector(c, stringIndex); s.Search(query, results); DictionaryEntryQueue queue = new DictionaryEntryQueue(stringIndex.lookup.Length); for (int i = 1; i < stringIndex.lookup.Length; i++) { if (c[i] > 0 && stringIndex.lookup[i] != null && stringIndex.lookup[i] != "0") { queue.Insert(new FacetEntry(stringIndex.lookup[i], -c[i])); } } int dictionaryEntrySize = queue.Size(); Dictionary<string, int> dics = new Dictionary<string, int>(); for (int j = dictionaryEntrySize - 1; j >= 0; j--) { FacetEntry entry = queue.Pop() as FacetEntry; dics.Add(entry.Value, -entry.Count); } return dics; } /// <summary>Helper class for order the resulting array in value order /// </summary> private sealed class DictionaryEntryQueue : Lucene.Net.Util.PriorityQueue { internal DictionaryEntryQueue(int size) : base() { Initialize(size); } public override bool LessThan(object a, object b) { FacetEntry de1 = (FacetEntry)a; FacetEntry de2 = (FacetEntry)b; return (int)de1.Count < (int)de2.Count; } } /// <summary>class for work in the priority queue and avoid some boxing. /// </summary> private class FacetEntry { // Fields private int count; private string value; // Methods public FacetEntry(string v, int c) { this.value = v; this.count = c; } public int Count { get { return count; } set { count = value; } } public string Value { get { return this.value; } set { this.value = value; } } } /// <summary>collector that count the hits for every token /// </summary> private class FacetCollector : HitCollector { // Fields private int[] counter; private StringIndex si; // Methods public FacetCollector(int[] c, StringIndex s) { this.counter = c; this.si = s; } public override void Collect(int doc, float score) { this.counter[this.si.order[doc]]++; } } }

调用也很简单:

Dictionary<string, int> companyNameAggregate = new Dictionary<string, int>(); companyNameAggregate = SimpleFacets.Facet(query, currentIndexSearcher, "CompanyName_UN");