00001 /*========================================================================== 00002 * Copyright (c) 2001 Carnegie Mellon University. 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.lemurproject.org/license.html 00008 * 00009 *========================================================================== 00010 */ 00011 00012 #include "Parser.hpp" 00013 #include "TextHandler.hpp" 00014 #include "LinkedPropertyList.hpp" 00015 00016 #ifndef _WEBPARSER_HPP 00017 #define _WEBPARSER_HPP 00018 00031 00032 // Source code in webparser.lex 00033 00034 class WebParser : public Parser { 00035 00036 public: 00037 static const string identifier; 00038 00039 WebParser(); 00040 00042 void parseFile(const string &filename); 00043 00044 void parseBuffer(char * buf, int len); 00045 00046 long fileTell() const; 00047 00048 private: 00050 void doParse(); 00051 00053 int state; 00054 00056 Property begelem; 00057 Property endelem; 00058 00060 LinkedPropertyList proplist; 00061 }; 00062 00063 #endif