#ifndef _HTMLARTICLEPARSER_H #define _HTMLARTICLEPARSER_H class DBObject; #include "HTMLParser.h" #include "StringMap.h" #include #include #include using namespace std; class HTMLArticleParser: public HTMLParser { //regex filters for selecting chunks based on type in the codestring static Filter fArticle, fTitle, fDescription, fKeywords; public: //produces an Article from random HTML HTMLArticleParser(const int _pagegroupid, TIPsDatabase *_db, const Domain *_domain, const int _type, short _minimumByteDensity = 50); const int parsertype() const; const size_t parse(const InternetResource *ir, vector *objects) const; }; #endif