Lucene 实战教程第六章 Lucene 的精确、包含、集合查询 Query 的简单使用

所有的搜索基本上都存在精确匹配，包含等操作。Lucene 中同样存在这样的操作，今天我们以 IntPoint 为例，来说说 Lucene 中的精确查询。

IntPoint、LongPoint、FloatPoint、DoublePoint 这个 4 个的操作类似，我就只以 IntPoint 来进行举例。

@Test
public void testIntPointQuery() throws IOException {
    Directory directory = new RAMDirectory();
    IndexWriter indexWriter = new IndexWriter(directory, new IndexWriterConfig(new StandardAnalyzer()));

    // 创建 Document
    Document document = new Document();

    // 创建 Field 字段/域
    Field intPoint = new IntPoint("age", 11);
    document.add(intPoint);
    intPoint = new StoredField("age", 11);
    document.add(intPoint);
    indexWriter.addDocument(document);

    Field intPoint1 = new IntPoint("age", 22);
    document = new Document();
    document.add(intPoint1);
    intPoint1 = new StoredField("age", 22);
    document.add(intPoint1);
    
    indexWriter.addDocument(document);
    indexWriter.close();
    IndexSearcher indexSearcher = new IndexSearcher(DirectoryReader.open(directory));
    //精确查询
    Query query = IntPoint.newExactQuery("age", 11);
    ScoreDoc[] scoreDocs = indexSearcher.search(query, 10).scoreDocs;
    for (ScoreDoc scoreDoc : scoreDocs) {
        System.out.println("精确查询：" + indexSearcher.doc(scoreDoc.doc));
    }
    //范围查询，不包含边界
    query = IntPoint.newRangeQuery("age", Math.addExact(11, 1), Math.addExact(22, -1));
    scoreDocs = indexSearcher.search(query, 10).scoreDocs;
    for (ScoreDoc scoreDoc : scoreDocs) {
        System.out.println("不包含边界：" + indexSearcher.doc(scoreDoc.doc));
    }
    //范围查询，包含边界
    query = IntPoint.newRangeQuery("age", 11, 22);
    scoreDocs = indexSearcher.search(query, 10).scoreDocs;
    for (ScoreDoc scoreDoc : scoreDocs) {
        System.out.println("包含边界：" + indexSearcher.doc(scoreDoc.doc));
    }
    //范围查询，左包含，右不包含
    query = IntPoint.newRangeQuery("age", 11, Math.addExact(22, -1));
    scoreDocs = indexSearcher.search(query, 10).scoreDocs;
    for (ScoreDoc scoreDoc : scoreDocs) {
        System.out.println("左包含右不包含：" + indexSearcher.doc(scoreDoc.doc));
    }
    //集合查询
    query = IntPoint.newSetQuery("age", 11, 22, 33);
    scoreDocs = indexSearcher.search(query, 10).scoreDocs;
    for (ScoreDoc scoreDoc : scoreDocs) {
        System.out.println("集合查询：" + indexSearcher.doc(scoreDoc.doc));
    }
}

IntPoint.newExactQuery 精确查询，使用的是 PointRangeQuery。
IntPoint.newRangeQuery 范围查询，使用的是 PointRangeQuery。
IntPoint.newSetQuery 集合查询，使用的是 PointInSetQuery。

它们都继承自 Query，通过 IntPoint 去创建这些抽象类的匿名实现类。

LongPoint、FloatPoint、DoublePoint 封装的和 IntPoint 都很相似，我就不在列举了。大家主要记住 newExactQuery，newRangeQuery，newSetQuery 三个方法的用法即可。

： » Lucene 实战教程第六章 Lucene 的精确、包含、集合查询 Query 的简单使用