Index establishment process of lucene4.7 source Research (2)

Have to say, Lucene source code in most places are left with comments, explained in detail, reading will be very convenient

Defining analyzers

New StandardAnalyzer (version.lucene_47);

Take StandardAnalyzer as an example, only the version number of Stopword and Lucene is loaded

Indexwriterconfig initialization, loading the configuration required to build the index

New Indexwriterconfig (version.lucene_47, analyzer);

Indexwriterconfig Inheritance Liveindexwriterconfig

//used by IndexwriterconfigLiveindexwriterconfig (Analyzer Analyzer, Version matchversion) { This. Analyzer = Analyzer;//load a word breaker     This. matchversion = matchversion;//Lucene versionRambuffersizemb = INDEXWRITERCONFIG.DEFAULT_RAM_BUFFER_SIZE_MB;//when this value is reached in memory, the flush data is started to the hard disk, the default 16MMaxbuffereddocs = Indexwriterconfig.default_max_buffered_docs;//This value is Disable_auto_flush with Maxbuffereddeleteterms and the default value is-1, auto flush is stopped, that is, the memory is not allowed to store both valuesMaxbuffereddeleteterms = indexwriterconfig.default_max_buffered_delete_terms;//Ibid .Readertermsindexdivisor = Indexwriterconfig.default_reader_terms_index_divisor;//The default is 1, which reads the divisor of the term fileMergedsegmentwarmer =NULL;//Lucene gives explanations warm is called before any deletes has been carried over the merged segment | Warm was called before any delete operation, and was deferred until the merge was completed. /c4>Termindexinterval = Indexwriterconfig.default_term_index_interval;//Todo:this should is private to the codec, not settable here//default Jump table spacingDelpolicy =NewKeeponlylastcommitdeletionpolicy ();//for the policy of the index commit deletion, implement Indexdeletionpolicy, note that the policy will only save the most recent commit, if a commit operation is complete, directly deletecommit =NULL; Usecompoundfile= Indexwriterconfig.default_use_compound_file_system;//whether to use a composite index, which is true by defaultOpenMode = Openmode.create_or_append;//open mode for creating and appendingSimilarity = Indexsearcher.getdefaultsimilarity ();//set the default scoring policy, VSMMergescheduler =NewConcurrentmergescheduler ();//Segment Merge Scheduler, concurrent thread safetyWritelocktimeout = Indexwriterconfig.write_lock_timeout;//Write lock TimeoutIndexingchain = Documentswriterperthread.defaultindexingchain;//Index Chaincodec = Codec.getdefault ();//Lucene Index file header (magic,codecname,version) checksum, according to Lucene version, default is the current version    if(codec = =NULL) {      Throw NewNullPointerException (); } Infostream= Infostream.getdefault ();//default NooutputMergepolicy =NewTieredmergepolicy ();//Segment Merge PolicyFlushpolicy =NewFlushbyramorcountspolicy ();//Flush Policyreaderpooling = indexwriterconfig.default_reader_pooling;//Reader pool is not used by defaultIndexerthreadpool =NewThreadaffinitydocumentswriterthreadpool (indexwriterconfig.default_max_thread_states);//The maximum number of threads allowed to index operations on the same indexwriter, default is 8PERTHREADHARDLIMITMB = INDEXWRITERCONFIG.DEFAULT_RAM_PER_THREAD_HARD_LIMIT_MB;//default 1945, each thread allocates the maximum capacity in memory}

Each of these strategies waits for the actual call to be analyzed.

Index establishment process of lucene4.7 source Research (2)