Lucene学习总结之七:Lucene搜索过程解析(4)

2.4、搜索查询对象

 

2.4.1.2、创建Weight对象树

BooleanQuery.createWeight(Searcher) 最终返回return new BooleanWeight(searcher),BooleanWeight构造函数的具体实现如下:

 

public BooleanWeight(Searcher searcher) {

  this.similarity = getSimilarity(searcher);

  weights = new ArrayList<Weight>(clauses.size());

  //也是一个递归的过程,沿着新的Query对象树一直到叶子节点

  for (int i = 0 ; i < clauses.size(); i++) {

    weights.add(clauses.get(i).getQuery().createWeight(searcher));

  }

}

对于TermQuery的叶子节点,其TermQuery.createWeight(Searcher) 返回return new TermWeight(searcher)对象,TermWeight构造函数如下:

 

public TermWeight(Searcher searcher) {

  this.similarity = getSimilarity(searcher);

  //此处计算了idf

  idfExp = similarity.idfExplain(term, searcher);

  idf = idfExp.getIdf();

}

//idf的计算完全符合文档中的公式:

 

public IDFExplanation idfExplain(final Term term, final Searcher searcher) {

  final int df = searcher.docFreq(term);

  final int max = searcher.maxDoc();

  final float idf = idf(df, max);

  return new IDFExplanation() {

      public float getIdf() {

        return idf;

      }};

}

public float idf(int docFreq, int numDocs) {

  return (float)(Math.log(numDocs/(double)(docFreq+1)) + 1.0);

}

而ConstantScoreQuery.createWeight(Searcher) 除了创建ConstantScoreQuery.ConstantWeight(searcher)对象外,没有计算idf。

由此创建的Weight对象树如下:

 

weight    BooleanQuery$BooleanWeight  (id=169)   
   |   similarity    DefaultSimilarity  (id=177)   
   |   this$0    BooleanQuery  (id=89)   
   |   weights    ArrayList<E>  (id=188)   
   |      elementData    Object[3]  (id=190)   
   |------[0]    BooleanQuery$BooleanWeight  (id=171)   
   |          |   similarity    DefaultSimilarity  (id=177)   
   |          |   this$0    BooleanQuery  (id=105)   
   |          |   weights    ArrayList<E>  (id=193)   
   |          |      elementData    Object[2]  (id=199)   
   |          |------[0]    ConstantScoreQuery$ConstantWeight  (id=183)   
   |          |               queryNorm    0.0   
   |          |               queryWeight    0.0   
   |          |               similarity    DefaultSimilarity  (id=177)   

   |          |               //ConstantScore(contents:apple*)  
   |          |               this$0    ConstantScoreQuery  (id=123)   
   |          |------[1]    TermQuery$TermWeight  (id=175)   
   |                         idf    2.0986123   
   |                         idfExp    Similarity$1  (id=241)   
   |                         queryNorm    0.0   
   |                         queryWeight    0.0   
   |                         similarity    DefaultSimilarity  (id=177)   

   |                         //contents:boy
   |                        this$0    TermQuery  (id=124)   
   |                         value    0.0   
   |                 modCount    2   
   |                 size    2   
   |------[1]    BooleanQuery$BooleanWeight  (id=179)   
   |          |   similarity    DefaultSimilarity  (id=177)   
   |          |   this$0    BooleanQuery  (id=110)   
   |          |   weights    ArrayList<E>  (id=195)   
   |          |      elementData    Object[2]  (id=204)   
   |          |------[0]    ConstantScoreQuery$ConstantWeight  (id=206)   
   |          |               queryNorm    0.0   
   |          |               queryWeight    0.0   
   |          |               similarity    DefaultSimilarity  (id=177)   

   |          |               //ConstantScore(contents:cat*)
   |          |               this$0    ConstantScoreQuery  (id=135)   
   |          |------[1]    TermQuery$TermWeight  (id=207)   
   |                         idf    1.5389965   
   |                         idfExp    Similarity$1  (id=210)   
   |                         queryNorm    0.0   
   |                         queryWeight    0.0   
   |                         similarity    DefaultSimilarity  (id=177)

   |                         //contents:dog
   |                         this$0    TermQuery  (id=136)   
   |                         value    0.0   
   |                 modCount    2   
   |                 size    2   
   |------[2]    BooleanQuery$BooleanWeight  (id=182)   
              |  similarity    DefaultSimilarity  (id=177)   
              |  this$0    BooleanQuery  (id=113)   
              |  weights    ArrayList<E>  (id=197)   
              |     elementData    Object[2]  (id=216)   
              |------[0]    BooleanQuery$BooleanWeight  (id=181)   
              |          |    similarity    BooleanQuery$1  (id=220)   
              |          |    this$0    BooleanQuery  (id=145)   
              |          |    weights    ArrayList<E>  (id=221)   
              |          |      elementData    Object[2]  (id=224)   
              |          |------[0]    TermQuery$TermWeight  (id=226)   
              |          |                idf    2.0986123   
              |          |                idfExp    Similarity$1  (id=229)   
              |          |                queryNorm    0.0   
              |          |                queryWeight    0.0   
              |          |                similarity    DefaultSimilarity  (id=177)   

              |          |                //contents:eat
              |          |                this$0    TermQuery  (id=150)   
              |          |                value    0.0   
              |          |------[1]    TermQuery$TermWeight  (id=227)   
              |                          idf    1.1823215   
              |                          idfExp    Similarity$1  (id=231)   
              |                          queryNorm    0.0   
              |                          queryWeight    0.0   
              |                          similarity    DefaultSimilarity  (id=177)   

              |                          //contents:cat^0.33333325
              |                          this$0    TermQuery  (id=151)   
              |                          value    0.0   
              |                  modCount    2   
              |                  size    2   
              |------[1]    TermQuery$TermWeight  (id=218)   
                            idf    2.0986123   
                            idfExp    Similarity$1  (id=233)   
                            queryNorm    0.0   
                            queryWeight    0.0   
                            similarity    DefaultSimilarity  (id=177)   

                            //contents:foods
                            this$0    TermQuery  (id=154)   
                            value    0.0   
                    modCount    2   
                    size    2   
        modCount    3   
        size    3   

 

Lucene学习总结之七:Lucene搜索过程解析(4)

 

 

2.4.1.3、计算Term Weight分数

(1) 首先计算sumOfSquaredWeights

按照公式:

 

代码如下:

float sum = weight.sumOfSquaredWeights();

 

 

//可以看出,也是一个递归的过程

public float sumOfSquaredWeights() throws IOException {

  float sum = 0.0f;

  for (int i = 0 ; i < weights.size(); i++) {

    float s = weights.get(i).sumOfSquaredWeights();

    if (!clauses.get(i).isProhibited())

      sum += s;

  }

  sum *= getBoost() * getBoost();  //乘以query boost

  return sum ;

}

对于叶子节点TermWeight来讲,其TermQuery$TermWeight.sumOfSquaredWeights()实现如下:

 

public float sumOfSquaredWeights() {

  //计算一部分打分,idf*t.getBoost(),将来还会用到。

  queryWeight = idf * getBoost();

  //计算(idf*t.getBoost())^2

  return queryWeight * queryWeight;

}

对于叶子节点ConstantWeight来讲,其ConstantScoreQuery$ConstantWeight.sumOfSquaredWeights() 如下:

 

public float sumOfSquaredWeights() {

  //除了用户指定的boost以外,其他都不计算在打分内

  queryWeight = getBoost();

  return queryWeight * queryWeight;

}

(2) 计算queryNorm

其公式如下:

 

其代码如下:

 

public float queryNorm(float sumOfSquaredWeights) {

  return (float)(1.0 / Math.sqrt(sumOfSquaredWeights));

}

(3) 将queryNorm算入打分

代码为:

weight.normalize(norm);

 

//又是一个递归的过程

public void normalize(float norm) {

  norm *= getBoost();

  for (Weight w : weights) {

    w.normalize(norm);

  }

}

其叶子节点TermWeight来讲,其TermQuery$TermWeight.normalize(float) 代码如下:

 

public void normalize(float queryNorm) {

  this.queryNorm = queryNorm;

  //原来queryWeight为idf*t.getBoost(),现在为queryNorm*idf*t.getBoost()。

  queryWeight *= queryNorm;

  //打分到此计算了queryNorm*idf*t.getBoost()*idf = queryNorm*idf^2*t.getBoost()部分。

  value = queryWeight * idf;

}

我们知道,Lucene的打分公式整体如下,到此计算了图中,红色的部分:

你可能感兴趣的:(apple,Lucene)