00001 /*========================================================================== 00002 * Copyright (c) 2002 University of Massachusetts. All Rights Reserved. 00003 * 00004 * Use of the Lemur Toolkit for Language Modeling and Information Retrieval 00005 * is subject to the terms of the software license set forth in the LICENSE 00006 * file included with this software, and also available at 00007 * http://www.cs.cmu.edu/~lemur/license.html 00008 * 00009 *========================================================================== 00010 */ 00011 00012 #ifndef _CHINESECHARPARSER_HPP 00013 #define _CHINESECHARPARSER_HPP 00014 #include "Parser.hpp" 00015 #include "TextHandler.hpp" 00016 00021 00022 class ChineseCharParser : public Parser { 00023 00024 public: 00025 ChineseCharParser(); 00026 00028 void parseFile (char * filename); 00029 00031 void parseBuffer (char * buf, int len); 00032 00035 long fileTell(); 00036 00037 00038 private: 00040 void doParse(); 00041 00043 int state; 00044 00045 }; 00046 00047 #endif