lucene4.7源码研究之索引建立过程(2)

不得不说一句，lucene的源码中大多数地方都留有注释，解释的很详细，阅读起来会很方便

定义分析器

Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_47);

以StandardAnalyzer为例，就只加载了stopword和lucene的版本号

IndexWriterConfig初始化，加载了建立索引需要的配置

IndexWriterConfig config = new IndexWriterConfig(Version.LUCENE_47, analyzer);

IndexWriterConfig继承LiveIndexWriterConfig

// used by IndexWriterConfig
  LiveIndexWriterConfig(Analyzer analyzer, Version matchVersion) {
    this.analyzer = analyzer;//加载分词器
    this.matchVersion = matchVersion;//lucene版本
    ramBufferSizeMB = IndexWriterConfig.DEFAULT_RAM_BUFFER_SIZE_MB;//当内存中达到该值，开始flush数据到硬盘，默认16M
    maxBufferedDocs = IndexWriterConfig.DEFAULT_MAX_BUFFERED_DOCS;//该值与maxBufferedDeleteTerms都为disable_auto_flush，默认值都是-1，自动flush被停止，也就是内存中不允许存储这两个值
    maxBufferedDeleteTerms = IndexWriterConfig.DEFAULT_MAX_BUFFERED_DELETE_TERMS;//同上
    readerTermsIndexDivisor = IndexWriterConfig.DEFAULT_READER_TERMS_INDEX_DIVISOR;//默认为1，读取term文件的除数
    mergedSegmentWarmer = null;//lucene给的解释warm is called before any deletes have been carried over the merged segment | warm是在任何删除操作之前调用，一直延迟到merge完毕
    termIndexInterval = IndexWriterConfig.DEFAULT_TERM_INDEX_INTERVAL; // TODO: this should be private to the codec, not settable here //默认跳跃表间距
    delPolicy = new KeepOnlyLastCommitDeletionPolicy();//针对索引提交删除的策略，实现IndexDeletionPolicy，注解上说明该策略只会保存最近的一次commit，如果一个commit操作完毕，就直接删除
    commit = null;
    useCompoundFile = IndexWriterConfig.DEFAULT_USE_COMPOUND_FILE_SYSTEM;//是否使用复合索引，默认为true
    openMode = OpenMode.CREATE_OR_APPEND;//打开模式为创建和追加
    similarity = IndexSearcher.getDefaultSimilarity();//设置默认打分策略，vsm
    mergeScheduler = new ConcurrentMergeScheduler();//段合并调度器，concurrent线程安全
    writeLockTimeout = IndexWriterConfig.WRITE_LOCK_TIMEOUT;//写锁超时
    indexingChain = DocumentsWriterPerThread.defaultIndexingChain;//索引链
    codec = Codec.getDefault();//lucene索引文件header（magic，codecname，version）校验，根据lucene版本区分，default为当前版本
    if (codec == null) {
      throw new NullPointerException();
    }
    infoStream = InfoStream.getDefault();//默认 NoOutput
    mergePolicy = new TieredMergePolicy();//段合并策略
    flushPolicy = new FlushByRamOrCountsPolicy();//flush策略
    readerPooling = IndexWriterConfig.DEFAULT_READER_POOLING;//默认不使用reader pool
    indexerThreadPool = new ThreadAffinityDocumentsWriterThreadPool(IndexWriterConfig.DEFAULT_MAX_THREAD_STATES);//允许在同一个indexWriter进行index操作的最大线程数量，默认是8
    perThreadHardLimitMB = IndexWriterConfig.DEFAULT_RAM_PER_THREAD_HARD_LIMIT_MB;//默认1945，每个线程在内存中分配最大的容量
  }

其中的各项策略等到真正调用的时候再做分析