Code Primer

The Top-level classes that run SolrMarc are in the package org.solrmarc.driver

• Boot.java -- The top-level Main routine that dynamically loads the required jars, and then call the main routine specified as the first command line argument

• BootableMain.java -- The base class for main routines that can be called by the Boot class. Does all of the command-line argument processing, and additional dynamic jar loading.

• ConfigDriver.java -- A backwards-compatibiity driver program. Reads an old-style config.properties file, and builds the correct command-line to call the IndexDriver class.

• IndexDriver.java -- Uses the command-line arguments to create a MarcReader, a collection of AbstractValueIndexer objects, and a SolrProxy object and then passes them to the Indexer class.

• Indexer.java -- The single-threaded reference implementation of the indexing process. Reads a record, builds the SolrInputDocument, and sends it to the SolrProxy.

• ThreadedIndexer.java -- The multi-threaded, producer-consumer implementation of the Indexer class, AKA where the magic happens.

• MarcReaderThread.java -- Used by ThreadedIndexer. Reads records, puts them in the readQ queue.

• IndexerWorker.java -- Used by ThreadedIndexer. Takes a record from the readQ queue, processes the index specifications to produce a SolrInputDocument, puts that in docQ queue.

• ChunkIndexerWorker.java -- Used by ThreadedIndexer. Takes a chunk of SolrInputDocuments from the docQ queue, send them to Solr.

• RecordAndDoc.java -- Used by the Indexing Queues, consists of a Record and a SolrInputDocument.

The heart of the index specification processing in in the package org.solrmarc.index.indexer

• ValueIndexerFactory.java -- The primary driver for reading index specifications, and building AbstractValueIndexer objects. A global singleton instance.

• FullScanner.lex -- The Jflex specification for the lexical tokenizer of the Index Specification Parser.

• FullConditionalParser.cup -- The Java Cup specification for the LALR parser generator

• VerboseSymbolFactory.java -- Used by the parser and scanner in debug mode.

• FullConditionalParser.java -- generated Java code for the LALR parser.

• FullConditionalScanner.java -- generated Java code for the JFlex tokenizer.

• FullSym.java -- generated Java code used by the parser and scanner.

• IndexerSpecException.java -- A runtime exception that can be used to indicate a problem with the Index Specification

• AbstractValueIndexer.java -- The abstract class created by the ValueIndexerFactory for each Index Specification.

• MultiValueIndexer.java -- The concrete implementation of AbstractValueIndexer, consists of a MultiValue Extractor, zero or more MultiValue maps, plus a MultiValue collector.