默认Facet是统计落入某一组域值的总数的,然后按照总数从大到小排序,判定规则是域值是否相同,其实还可以根据域值是否在某个范围内来判定是否落入某一个分组。这里说的范围就是通过Range定义的,比如:
/**1小时之前的毫秒数*/
final LongRange PAST_HOUR = new LongRange("Past hour", this.nowSec - 3600L,
true, this.nowSec, true);
/**6小时之前的毫秒数*/
final LongRange PAST_SIX_HOURS = new LongRange("Past six hours",
this.nowSec - 21600L, true, this.nowSec, true);
/**24小时之前的毫秒数*/
final LongRange PAST_DAY = new LongRange("Past day", this.nowSec - 86400L,
true, this.nowSec, true);
这里定义了3个数字范围,比如1个小时之前的毫秒数至当前时间毫秒数之间的范围,剩余两个类似,然后就可以根据定义的范围来创建Facet:
//定义3个Facet: 统计
//[过去一小时之前-->当前时间] [过去6小时之前-->当前时间] [过去24小时之前-->当前时间]
Facets facets = new LongRangeFacetCounts("timestamp", fc,
new LongRange[] { this.PAST_HOUR, this.PAST_SIX_HOURS,this.PAST_DAY });
这样就能根据自定义范围去统计了,只要当前域的域值落在这个范围内,则这个facet分组内计数加1.完整示例代码如下:
package com.yida.framework.lucene5.facet;
import java.io.IOException;
import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.LongField;
import org.apache.lucene.document.NumericDocValuesField;
import org.apache.lucene.facet.DrillDownQuery;
import org.apache.lucene.facet.FacetResult;
import org.apache.lucene.facet.Facets;
import org.apache.lucene.facet.FacetsCollector;
import org.apache.lucene.facet.FacetsConfig;
import org.apache.lucene.facet.range.LongRange;
import org.apache.lucene.facet.range.LongRangeFacetCounts;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.search.NumericRangeQuery;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
public class RangeFacetsExample {
private final Directory indexDir = new RAMDirectory();
private IndexSearcher searcher;
/**当前时间的毫秒数*/
private final long nowSec = System.currentTimeMillis();
/**1小时之前的毫秒数*/
final LongRange PAST_HOUR = new LongRange("Past hour", this.nowSec - 3600L,
true, this.nowSec, true);
/**6小时之前的毫秒数*/
final LongRange PAST_SIX_HOURS = new LongRange("Past six hours",
this.nowSec - 21600L, true, this.nowSec, true);
/**24小时之前的毫秒数*/
final LongRange PAST_DAY = new LongRange("Past day", this.nowSec - 86400L,
true, this.nowSec, true);
/**
* 创建测试索引
* @throws IOException
*/
public void index() throws IOException {
IndexWriter indexWriter = new IndexWriter(this.indexDir,
new IndexWriterConfig(new WhitespaceAnalyzer())
.setOpenMode(IndexWriterConfig.OpenMode.CREATE));
/**
* 每次按[1000*i]这个斜率递减创建一个索引
*/
for (int i = 0; i < 100; i++) {
Document doc = new Document();
long then = this.nowSec - i * 1000;
doc.add(new NumericDocValuesField("timestamp", then));
doc.add(new LongField("timestamp", then, Field.Store.YES));
indexWriter.addDocument(doc);
}
this.searcher = new IndexSearcher(DirectoryReader.open(indexWriter,
true));
indexWriter.close();
}
/**
* 获取FacetConfig配置对象
* @return
*/
private FacetsConfig getConfig() {
return new FacetsConfig();
}
public FacetResult search() throws IOException {
/**创建Facet结果收集器*/
FacetsCollector fc = new FacetsCollector();
TopDocs topDocs = FacetsCollector.search(this.searcher, new MatchAllDocsQuery(), 20, fc);
ScoreDoc[] scoreDocs = topDocs.scoreDocs;
for(ScoreDoc scoreDoc : scoreDocs) {
int docId = scoreDoc.doc;
Document doc = searcher.doc(docId);
System.out.println(scoreDoc.doc + "\t" + doc.get("timestamp"));
}
//定义3个Facet: 统计
//[过去一小时之前-->当前时间] [过去6小时之前-->当前时间] [过去24小时之前-->当前时间]
Facets facets = new LongRangeFacetCounts("timestamp", fc,
new LongRange[] { this.PAST_HOUR, this.PAST_SIX_HOURS,
this.PAST_DAY });
return facets.getTopChildren(10, "timestamp", new String[0]);
}
/**
* 使用DrillDownQuery进行Facet统计
* @param range
* @return
* @throws IOException
*/
public TopDocs drillDown(LongRange range) throws IOException {
DrillDownQuery q = new DrillDownQuery(getConfig());
q.add("timestamp", NumericRangeQuery.newLongRange("timestamp",
Long.valueOf(range.min), Long.valueOf(range.max),
range.minInclusive, range.maxInclusive));
return this.searcher.search(q, 10);
}
public void close() throws IOException {
this.searcher.getIndexReader().close();
this.indexDir.close();
}
public static void main(String[] args) throws Exception {
RangeFacetsExample example = new RangeFacetsExample();
example.index();
System.out.println("Facet counting example:");
System.out.println("-----------------------");
System.out.println(example.search());
System.out.println("\n");
//只统计6个小时之前的Facet
System.out
.println("Facet drill-down example (timestamp/Past six hours):");
System.out.println("---------------------------------------------");
TopDocs hits = example.drillDown(example.PAST_SIX_HOURS);
System.out.println(hits.totalHits + " totalHits");
example.close();
}
}
我们在通过FacetsCollector.search进行搜索时,其实是可以传入排序器的,这时我们在创建索引时就需要使用SortedSetDocValuesFacetField,这样FacetsCollector.search返回的TopDocs就是经过排序的了,这一点需要注意。具体示例代码如下:
package com.yida.framework.lucene5.facet;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.facet.DrillDownQuery;
import org.apache.lucene.facet.FacetResult;
import org.apache.lucene.facet.Facets;
import org.apache.lucene.facet.FacetsCollector;
import org.apache.lucene.facet.FacetsConfig;
import org.apache.lucene.facet.sortedset.DefaultSortedSetDocValuesReaderState;
import org.apache.lucene.facet.sortedset.SortedSetDocValuesFacetCounts;
import org.apache.lucene.facet.sortedset.SortedSetDocValuesFacetField;
import org.apache.lucene.facet.sortedset.SortedSetDocValuesReaderState;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.search.Sort;
import org.apache.lucene.search.SortField;
import org.apache.lucene.search.SortField.Type;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
public class SimpleSortedSetFacetsExample {
private final Directory indexDir = new RAMDirectory();
private final FacetsConfig config = new FacetsConfig();
private void index() throws IOException {
IndexWriter indexWriter = new IndexWriter(this.indexDir,
new IndexWriterConfig(new WhitespaceAnalyzer())
.setOpenMode(IndexWriterConfig.OpenMode.CREATE));
Document doc = new Document();
doc.add(new SortedSetDocValuesFacetField("Author", "Bob"));
doc.add(new SortedSetDocValuesFacetField("Publish Year", "2010"));
indexWriter.addDocument(this.config.build(doc));
doc = new Document();
doc.add(new SortedSetDocValuesFacetField("Author", "Lisa"));
doc.add(new SortedSetDocValuesFacetField("Publish Year", "2010"));
indexWriter.addDocument(this.config.build(doc));
doc = new Document();
doc.add(new SortedSetDocValuesFacetField("Author", "Lisa"));
doc.add(new SortedSetDocValuesFacetField("Publish Year", "2012"));
indexWriter.addDocument(this.config.build(doc));
doc = new Document();
doc.add(new SortedSetDocValuesFacetField("Author", "Susan"));
doc.add(new SortedSetDocValuesFacetField("Publish Year", "2012"));
indexWriter.addDocument(this.config.build(doc));
doc = new Document();
doc.add(new SortedSetDocValuesFacetField("Author", "Frank"));
doc.add(new SortedSetDocValuesFacetField("Publish Year", "1999"));
indexWriter.addDocument(this.config.build(doc));
indexWriter.close();
}
private List<FacetResult> search() throws IOException {
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(
indexReader);
FacetsCollector fc = new FacetsCollector();
Sort sort = new Sort(new SortField("Author", Type.INT, false));
//FacetsCollector.search(searcher, new MatchAllDocsQuery(), 10, fc);
TopDocs topDocs = FacetsCollector.search(searcher,new MatchAllDocsQuery(),null,10,sort,true,false,fc);
Facets facets = new SortedSetDocValuesFacetCounts(state, fc);
List<FacetResult> results = new ArrayList<FacetResult>();
results.add(facets.getTopChildren(10, "Author", new String[0]));
results.add(facets.getTopChildren(10, "Publish Year", new String[0]));
indexReader.close();
return results;
}
private FacetResult drillDown() throws IOException {
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
SortedSetDocValuesReaderState state = new DefaultSortedSetDocValuesReaderState(
indexReader);
DrillDownQuery q = new DrillDownQuery(this.config);
q.add("Publish Year", new String[] { "2010" });
FacetsCollector fc = new FacetsCollector();
FacetsCollector.search(searcher, q, 10, fc);
Facets facets = new SortedSetDocValuesFacetCounts(state, fc);
FacetResult result = facets.getTopChildren(10, "Author", new String[0]);
indexReader.close();
return result;
}
public List<FacetResult> runSearch() throws IOException {
index();
return search();
}
public FacetResult runDrillDown() throws IOException {
index();
return drillDown();
}
public static void main(String[] args) throws Exception {
System.out.println("Facet counting example:");
System.out.println("-----------------------");
SimpleSortedSetFacetsExample example = new SimpleSortedSetFacetsExample();
List<FacetResult> results = example.runSearch();
System.out.println("Author: " + results.get(0));
System.out.println("Publish Year: " + results.get(0));
System.out.println("\n");
System.out.println("Facet drill-down example (Publish Year/2010):");
System.out.println("---------------------------------------------");
System.out.println("Author: " + example.runDrillDown());
}
}
前面我们在定义Facet时,即可以不指定域名,比如:
Facets facets = new FastTaxonomyFacetCounts(taxoReader, this.config, fc);
这时默认会根据我们定义的FacetField去决定定义几个Facet进行统计,你可以指定只对指定域定义Facet,这样就可以值返回你关心的Facet统计数据,比如:
Facets author = new FastTaxonomyFacetCounts("author", taxoReader,
this.config, fc);
具体完整示例代码如下:
package com.yida.framework.lucene5.facet;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.facet.FacetField;
import org.apache.lucene.facet.FacetResult;
import org.apache.lucene.facet.Facets;
import org.apache.lucene.facet.FacetsCollector;
import org.apache.lucene.facet.FacetsConfig;
import org.apache.lucene.facet.taxonomy.FastTaxonomyFacetCounts;
import org.apache.lucene.facet.taxonomy.TaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyWriter;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
/**
* 根据指定域加载特定Facet Count
* @author Lanxiaowei
*
*/
public class MultiCategoryListsFacetsExample {
private final Directory indexDir = new RAMDirectory();
private final Directory taxoDir = new RAMDirectory();
private final FacetsConfig config = new FacetsConfig();
public MultiCategoryListsFacetsExample() {
//定义 域别名
this.config.setIndexFieldName("Author", "author");
this.config.setIndexFieldName("Publish Date", "pubdate");
// 设置Publish Date为多值域
this.config.setHierarchical("Publish Date", true);
}
/**
* 创建测试索引
* @throws IOException
*/
private void index() throws IOException {
IndexWriter indexWriter = new IndexWriter(this.indexDir,
new IndexWriterConfig(new WhitespaceAnalyzer())
.setOpenMode(IndexWriterConfig.OpenMode.CREATE));
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(
this.taxoDir);
Document doc = new Document();
doc.add(new FacetField("Author", new String[] { "Bob" }));
doc.add(new FacetField("Publish Date", new String[] { "2010", "10",
"15" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new FacetField("Author", new String[] { "Lisa" }));
doc.add(new FacetField("Publish Date", new String[] { "2010", "10",
"20" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new FacetField("Author", new String[] { "Lisa" }));
doc.add(new FacetField("Publish Date",
new String[] { "2012", "1", "1" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new FacetField("Author", new String[] { "Susan" }));
doc.add(new FacetField("Publish Date",
new String[] { "2012", "1", "7" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new FacetField("Author", new String[] { "Frank" }));
doc.add(new FacetField("Publish Date",
new String[] { "1999", "5", "5" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
indexWriter.close();
taxoWriter.close();
}
private List<FacetResult> search() throws IOException {
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
TaxonomyReader taxoReader = new DirectoryTaxonomyReader(this.taxoDir);
FacetsCollector fc = new FacetsCollector();
FacetsCollector.search(searcher, new MatchAllDocsQuery(), 10, fc);
List<FacetResult> results = new ArrayList<FacetResult>();
//定义author域的Facet,
//FastTaxonomyFacetCounts第一个构造参数指定域名称,
//则只统计指定域的Facet总数,不指定域名称,则默认会统计所有FacetField域的总数
//这就跟SQL中的select * from...和 select name from ...差不多
Facets author = new FastTaxonomyFacetCounts("author", taxoReader,
this.config, fc);
results.add(author.getTopChildren(10, "Author", new String[0]));
//定义pubdate域的Facet
Facets pubDate = new FastTaxonomyFacetCounts("pubdate", taxoReader,
this.config, fc);
results.add(pubDate.getTopChildren(10, "Publish Date", new String[0]));
indexReader.close();
taxoReader.close();
return results;
}
public List<FacetResult> runSearch() throws IOException {
index();
return search();
}
public static void main(String[] args) throws Exception {
System.out
.println("Facet counting over multiple category lists example:");
System.out.println("-----------------------");
List<FacetResult> results = new MultiCategoryListsFacetsExample().runSearch();
System.out.println("Author: " + results.get(0));
System.out.println("Publish Date: " + results.get(1));
}
}
上面说的Facet数字统计方式都是按照出现次数计算总和,其实lucene还提供了js表达式方式自定义Facet数据统计,比如:
/**当前索引文档评分乘以popularity域值的平方根 作为 Facet的统计值,默认是统计命中索引文档总数*/
Expression expr = JavascriptCompiler.compile("_score * sqrt(popularity)");
即表示每个域的域值相同的落入一组,但此时每一组的数字总计不再是计算总个数,而是按照指定的js表达式去计算然后求和,比如上述表达式即表示当前索引文档评分乘以popularity域值的平方根 作为 Facet的统计值,_score是内置变量代指当前索引文档的评分,sqrt是内置函数表示开平方根,lucene expression表达式是通过JavascriptCompiler进行编译的,而JavascriptCompiler是借助大名鼎鼎的开源语法分析器antlr实现的,与antlr类似的还有JFlex,自己google了解去,关于语法分析这个话题就超出本篇主题了。下面说说lucene expression都支持哪些函数和操作符:
大意就是它是基于Javascript语法的特定数字表达式,它支持:
1,Integer,float数字,十六进制和八进制字符
2.支持常用的数学操作符,加减乘除取模等
3.支持位运算符
4.支持布尔运算符
5.支持比较运算符
6.支持常用的数学函数,比如abs取绝对值,log对数,max取最大值,sqrt开平方根等
7.支持三角函数,比如sin,cos等等
8.支持地理函数,如haversin公式即求地球上任意两个点之间的距离,具体请自己去Google haversin公式
9.辅助函数,貌似跟第6条重复,搞什么飞机,我只是翻译下而已,跟我没关系
10.支持任意的外部变量,这一条有点费解,还好我在源码里找到了相关说明
查阅org.apache.lucene.expressions.js包下面的package.html,看到里面有这样一句提示:
打开Bindings类源码,没看到什么有用的信息,然后我无意在JavascriptCompiler类源码中找到如下一段说明,让我茅塞顿开:
/**
* An expression compiler for javascript expressions.
* <p>
* Example:
* <pre class="prettyprint">
* Expression foo = JavascriptCompiler.compile("((0.3*popularity)/10.0)+(0.7*score)");
* </pre>
* <p>
* See the {@link org.apache.lucene.expressions.js package documentation} for
* the supported syntax and default functions.
* <p>
* You can compile with an alternate set of functions via {@link #compile(String, Map, ClassLoader)}.
* For example:
* <pre class="prettyprint">
* Map<String,Method> functions = new HashMap<>();
* // add all the default functions
* functions.putAll(JavascriptCompiler.DEFAULT_FUNCTIONS);
* // add cbrt()
* functions.put("cbrt", Math.class.getMethod("cbrt", double.class));
* // call compile with customized function map
* Expression foo = JavascriptCompiler.compile("cbrt(score)+ln(popularity)",
* functions,
* getClass().getClassLoader());
* </pre>
*
* @lucene.experimental
*/
重点在这里:
Map<String,Method> functions = new HashMap<String,Method>();
// add all the default functions
functions.putAll(JavascriptCompiler.DEFAULT_FUNCTIONS);
// add cbrt()
functions.put("cbrt", Math.class.getMethod("cbrt", double.class));
// call compile with customized function map
Expression foo = JavascriptCompiler.compile("cbrt(score)+ln(popularity)",
functions,
getClass().getClassLoader());
通过示例代码,不难理解,意思就是我们可以把我们Java里类的某个方法通过JavascriptCompilter进行编译,就能像直接调用js内置函数一样,毕竟js内置函数是有限的,但内置的js函数不能满足你的要求,你就可以把java里自定义方法进行编译在js里执行。示例的代码就是把Math类里的cbrt方法(求一个double数值的立方根)编译为js内置函数,然后你就可以像使用js内置函数一样使用java里任意类的方法了。下面是一个使用lucene expressions完成Facet统计的完整示例:
package com.yida.framework.lucene5.facet;
import java.io.IOException;
import java.text.ParseException;
import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.NumericDocValuesField;
import org.apache.lucene.document.TextField;
import org.apache.lucene.expressions.Expression;
import org.apache.lucene.expressions.SimpleBindings;
import org.apache.lucene.expressions.js.JavascriptCompiler;
import org.apache.lucene.facet.FacetField;
import org.apache.lucene.facet.FacetResult;
import org.apache.lucene.facet.Facets;
import org.apache.lucene.facet.FacetsCollector;
import org.apache.lucene.facet.FacetsConfig;
import org.apache.lucene.facet.taxonomy.TaxonomyFacetSumValueSource;
import org.apache.lucene.facet.taxonomy.TaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyWriter;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.search.SortField;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
public class ExpressionAggregationFacetsExample {
private final Directory indexDir = new RAMDirectory();
private final Directory taxoDir = new RAMDirectory();
private final FacetsConfig config = new FacetsConfig();
private void index() throws IOException {
IndexWriter indexWriter = new IndexWriter(this.indexDir,
new IndexWriterConfig(new WhitespaceAnalyzer())
.setOpenMode(IndexWriterConfig.OpenMode.CREATE));
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(
this.taxoDir);
Document doc = new Document();
doc.add(new TextField("c", "foo bar", Field.Store.NO));
doc.add(new NumericDocValuesField("popularity", 5L));
doc.add(new FacetField("A", new String[] { "B" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new TextField("c", "foo foo bar", Field.Store.NO));
doc.add(new NumericDocValuesField("popularity", 3L));
doc.add(new FacetField("A", new String[] { "C" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
doc.add(new TextField("c", "foo foo bar", Field.Store.NO));
doc.add(new NumericDocValuesField("popularity", 8L));
doc.add(new FacetField("A", new String[] { "B" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
indexWriter.close();
taxoWriter.close();
}
private FacetResult search() throws IOException, ParseException {
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
TaxonomyReader taxoReader = new DirectoryTaxonomyReader(this.taxoDir);
/**当前索引文档评分乘以popularity域值的平方根 作为 Facet的统计值,默认是统计命中索引文档总数*/
Expression expr = JavascriptCompiler.compile("_score * sqrt(popularity)");
SimpleBindings bindings = new SimpleBindings();
bindings.add(new SortField("_score", SortField.Type.SCORE));
bindings.add(new SortField("popularity", SortField.Type.LONG));
FacetsCollector fc = new FacetsCollector(true);
FacetsCollector.search(searcher, new MatchAllDocsQuery(), 10, fc);
//以Expression表达式的计算值定义一个Facet
Facets facets = new TaxonomyFacetSumValueSource(taxoReader,
this.config, fc, expr.getValueSource(bindings));
FacetResult result = facets.getTopChildren(10, "A", new String[0]);
indexReader.close();
taxoReader.close();
return result;
}
public FacetResult runSearch() throws IOException, ParseException {
index();
return search();
}
public static void main(String[] args) throws Exception {
System.out.println("Facet counting example:");
System.out.println("-----------------------");
FacetResult result = new ExpressionAggregationFacetsExample()
.runSearch();
System.out.println(result);
}
}
前面说到的Lucene Expression是通过一个表达式来计算统计值然后求和,我们也可以使用AssociationFacetField域来关联一个任意值,然后根据这个值来求和统计,类似这样:
//3 --> lucene[不再是统计lucene这个域值的出现总次数,而是统计IntAssociationFacetField的第一个构造参数assoc总和]
doc.add(new IntAssociationFacetField(3, "tags", new String[] { "lucene" }));
doc.add(new FloatAssociationFacetField(0.87F, "genre", new String[] { "computing" }));
其实创建一个AssociationFacetField等价于创建了一个StringField和一个BinaryDocValuesField(创建这个域是为了辅助统计Facet Count) ,有关AssociationFacetField的完整示例代码如下:
package com.yida.framework.lucene5.facet;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.facet.DrillDownQuery;
import org.apache.lucene.facet.FacetResult;
import org.apache.lucene.facet.Facets;
import org.apache.lucene.facet.FacetsCollector;
import org.apache.lucene.facet.FacetsConfig;
import org.apache.lucene.facet.taxonomy.FloatAssociationFacetField;
import org.apache.lucene.facet.taxonomy.IntAssociationFacetField;
import org.apache.lucene.facet.taxonomy.TaxonomyFacetSumFloatAssociations;
import org.apache.lucene.facet.taxonomy.TaxonomyFacetSumIntAssociations;
import org.apache.lucene.facet.taxonomy.TaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader;
import org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyWriter;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
public class AssociationsFacetsExample {
private final Directory indexDir = new RAMDirectory();
private final Directory taxoDir = new RAMDirectory();
private final FacetsConfig config;
public AssociationsFacetsExample()
{
this.config = new FacetsConfig();
this.config.setMultiValued("tags", true);
this.config.setIndexFieldName("tags", "$tags");
this.config.setMultiValued("genre", true);
this.config.setIndexFieldName("genre", "$genre");
}
private void index() throws IOException
{
IndexWriterConfig iwc = new IndexWriterConfig(new WhitespaceAnalyzer()).setOpenMode(IndexWriterConfig.OpenMode.CREATE);
IndexWriter indexWriter = new IndexWriter(this.indexDir, iwc);
DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(this.taxoDir);
Document doc = new Document();
//3 --> lucene[不再是统计lucene这个域值的出现总次数,而是统计IntAssociationFacetField的第一个构造参数assoc总和]
doc.add(new IntAssociationFacetField(3, "tags", new String[] { "lucene" }));
doc.add(new FloatAssociationFacetField(0.87F, "genre", new String[] { "computing" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
doc = new Document();
//1 --> lucene
doc.add(new IntAssociationFacetField(1, "tags", new String[] { "lucene" }));
doc.add(new IntAssociationFacetField(2, "tags", new String[] { "solr" }));
doc.add(new FloatAssociationFacetField(0.75F, "genre", new String[] { "computing" }));
doc.add(new FloatAssociationFacetField(0.34F, "genre", new String[] { "software" }));
indexWriter.addDocument(this.config.build(taxoWriter, doc));
indexWriter.close();
taxoWriter.close();
}
private List<FacetResult> sumAssociations() throws IOException
{
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
TaxonomyReader taxoReader = new DirectoryTaxonomyReader(this.taxoDir);
FacetsCollector fc = new FacetsCollector();
FacetsCollector.search(searcher, new MatchAllDocsQuery(), 10, fc);
//定义了两个Facet
Facets tags = new TaxonomyFacetSumIntAssociations("$tags", taxoReader, this.config, fc);
Facets genre = new TaxonomyFacetSumFloatAssociations("$genre", taxoReader, this.config, fc);
List<FacetResult> results = new ArrayList<FacetResult>();
results.add(tags.getTopChildren(10, "tags", new String[0]));
results.add(genre.getTopChildren(10, "genre", new String[0]));
indexReader.close();
taxoReader.close();
return results;
}
private FacetResult drillDown() throws IOException
{
DirectoryReader indexReader = DirectoryReader.open(this.indexDir);
IndexSearcher searcher = new IndexSearcher(indexReader);
TaxonomyReader taxoReader = new DirectoryTaxonomyReader(this.taxoDir);
DrillDownQuery q = new DrillDownQuery(this.config);
q.add("tags", new String[] { "solr" });
FacetsCollector fc = new FacetsCollector();
FacetsCollector.search(searcher, q, 10, fc);
Facets facets = new TaxonomyFacetSumFloatAssociations("$genre", taxoReader, this.config, fc);
FacetResult result = facets.getTopChildren(10, "$genre", new String[0]);
indexReader.close();
taxoReader.close();
return result;
}
public List<FacetResult> runSumAssociations() throws IOException
{
index();
return sumAssociations();
}
public FacetResult runDrillDown() throws IOException
{
index();
return drillDown();
}
public static void main(String[] args) throws Exception
{
System.out.println("Sum associations example:");
System.out.println("-------------------------");
List<FacetResult> results = new AssociationsFacetsExample().runSumAssociations();
System.out.println("tags: " + results.get(0));
System.out.println("genre: " + results.get(1));
}
}
Facet就说这么多了,一些细节方面大家请看demo代码里面的注释,如果我哪里没说到的而你又感觉很迷惑无法理解的,请提出来,大家一起交流学习。如果我有哪里说的不够准确,也希望大家能够积极指正多多批评。 Demo源码请看底下的附件!!!
如果你还有什么问题请加我Q-Q:7-3-6-0-3-1-3-0-5,
或者加裙一起交流学习!