Xport: XHTML Parsing and Objective Reporting Toolkit

/************************************************************************
Xport: XHTML Parsing & Objective Reporting Toolkit
Copyright (C) 2007  Mitchel Haas

This file is part of Xport.

Xport is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

Xport is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with Xport.  If not, see <http://www.gnu.org/licenses/>.

For complete documentation on this library and alternative
licensing options, visit http://www.xportpro.com
Email questions, comments or suggestions to mitchel.haas@xportpro.com
************************************************************************/
#pragma once
#include "stylesheet.h"
#include "common.h"
#include "tag.h"
#include "xhtml_element.h"
#include "xhtml_comment.h"
#include "xhtml_processing_instruction.h"
#include "xhtml_parser.h"
#include "xhtml_formatter.h"
#include <string>
#include <fstream>
#include <sstream>
#include <cassert>

namespace Xport
{
  using namespace css;
  using namespace attribute;
  template<typename T, typename U, typename V, typename W> class xhtml_element;
  template<typename T, typename U, typename V, typename W> class xhtml_doc;
  template<typename T, typename U, typename V, typename W> class xhtml_formatter;
  template<typename T, typename U> class xhtml_stylesheet;

  // convenience typedefs
  // xthml strict
  typedef xhtml_markup<xhtml_strict, char>                                                                      markup;
  typedef xhtml_doc<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                      document;
  typedef xhtml_element<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                  element;
  typedef xhtml_pcdata<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                   pcdata;
  typedef xhtml_comment<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                  comment;
  typedef xhtml_processing_instruction<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                   procinstr;
  typedef xhtml_parser<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                   parser;
  typedef xhtml_formatter<xhtml_strict, char, xhtml_markup<xhtml_strict, char>, xhtml_stylesheet_item<char> >                                formatter;

  typedef xhtml_markup<xhtml_strict, wchar_t>                                                                   wmarkup;
  typedef xhtml_doc<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                                wdocument;
  typedef xhtml_element<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                            welement;
  typedef xhtml_pcdata<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                             wpcdata;
  typedef xhtml_comment<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                            wcomment;
  typedef xhtml_processing_instruction<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >             wprocinstr;
  typedef xhtml_parser<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                             wparser;
  typedef xhtml_formatter<xhtml_strict, wchar_t, xhtml_markup<xhtml_strict, wchar_t>, xhtml_stylesheet_item<wchar_t> >                          wformatter;

  // xhtml transitional
  typedef xhtml_markup<xhtml_transitional, char>                                                                tmarkup;
  typedef xhtml_doc<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                          tdocument;
  typedef xhtml_element<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                      telement;
  typedef xhtml_pcdata<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                       tpcdata;
  typedef xhtml_comment<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                      tcomment;
  typedef xhtml_processing_instruction<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >       tprocinstr;
  typedef xhtml_parser<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                       tparser;
  typedef xhtml_formatter<xhtml_transitional, char, xhtml_markup<xhtml_transitional, char>, xhtml_stylesheet_item<char> >                    tformatter;

  typedef xhtml_markup<xhtml_transitional, wchar_t>                                                             wtmarkup;
  typedef xhtml_doc<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >                    wtdocument;
  typedef xhtml_element<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >                wtelement;
  typedef xhtml_pcdata<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >                 wtpcdata;
  typedef xhtml_comment<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >                wtcomment;
  typedef xhtml_processing_instruction<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> > wtprocinstr;
  typedef xhtml_parser<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >                 wtparser;
  typedef xhtml_formatter<xhtml_transitional, wchar_t, xhtml_markup<xhtml_transitional, wchar_t>, xhtml_stylesheet_item<wchar_t> >              wtformatter;

  // xhtml frameset
  typedef xhtml_markup<xhtml_frameset, char>                                                                    fmarkup;
  typedef xhtml_doc<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                                  fdocument;
  typedef xhtml_element<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                              felement;
  typedef xhtml_pcdata<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                               fpcdata;
  typedef xhtml_comment<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                              fcomment;
  typedef xhtml_processing_instruction<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >               fprocinstr;
  typedef xhtml_parser<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                               fparser;
  typedef xhtml_formatter<xhtml_frameset, char, xhtml_markup<xhtml_frameset, char>, xhtml_stylesheet_item<char> >                            fformatter;

  typedef xhtml_markup<xhtml_frameset, wchar_t>                                                                 wfmarkup;
  typedef xhtml_doc<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                            wfdocument;
  typedef xhtml_element<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                        wfelement;
  typedef xhtml_pcdata<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                         wfpcdata;
  typedef xhtml_comment<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                        wfcomment;
  typedef xhtml_processing_instruction<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >         wfprocinstr;
  typedef xhtml_parser<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                         wfparser;
  typedef xhtml_formatter<xhtml_frameset, wchar_t, xhtml_markup<xhtml_frameset, wchar_t>, xhtml_stylesheet_item<wchar_t> >                      wfformatter;

  enum document_formation {root_doc = 1};
}

template <typename DT, typename CT, typename MBT, typename SBT>
class Xport::xhtml_doc
{
public:
  inline explicit xhtml_doc(document_formation formation = document_formation(0));
  xhtml_doc(const xhtml_doc& rhs);
  virtual ~xhtml_doc();
  xhtml_doc& operator =(const xhtml_doc& rhs);

  typedef DT doc_type;
  typedef typename MBT::iterator        base_iterator_type;         // vc6 compatibility
  typedef typename MBT::const_iterator  const_base_iterator_type;   // vc6 compatibility

  // interface
  MBT* body() { return get_body(); }
  const MBT* body() const { return get_body(); }
  MBT* frameset() { return get_frameset(); }
  const MBT* frameset() const { return get_frameset(); }
  MBT* head() { return get_head(); }
  const MBT* head() const { return get_head(); }
  MBT* title() { return get_title(); }
  const MBT* title() const { return get_title(); }
  MBT* html() { return pHtml; }
  const MBT* html() const { return pHtml; }

  void clear(document_formation formation = document_formation(0)); 
  std::basic_string<CT> document_type() const { std::basic_string<CT> temp; DT::name(temp); return temp; }
  bool parse(const std::basic_string<CT>& path);
  bool parse(const xhtml_parser<DT, CT, MBT, SBT>& prsr);
  bool write(const xhtml_formatter<DT, CT, MBT, SBT>& fmtr) const;
  bool write(const std::basic_string<char>& path) const;

  typename MBT::iterator insert(const MBT& mkup) { return insert(pRoot->end(), mkup); }
  base_iterator_type insert(const const_base_iterator_type& pos, const MBT& mkup);
  base_iterator_type insert(const xhtml_stylesheet<CT, SBT>& ss);
  xhtml_doc& operator << (const MBT& mkup) { insert(mkup); return *this; }

  // iterator accessors
  typename MBT::iterator begin() { return pRoot->begin(); }
  typename MBT::iterator end() { return pRoot->end(); }
  typename MBT::const_iterator begin() const { return pRoot->begin(); }
  typename MBT::const_iterator end() const { return pRoot->end(); }
  
  // descendant iterator accessors
  typename MBT::const_descendant_iterator descendant_begin() const { return pRoot->descendant_begin(); }
  typename MBT::const_descendant_iterator descendant_end() const { return pRoot->descendant_end(); }
  typename MBT::descendant_iterator descendant_begin()  { return pRoot->descendant_begin(); }
  typename MBT::descendant_iterator descendant_end()  { return pRoot->descendant_end(); }

private:
  // implementation
  std::basic_string<CT> get_doctype_definition() const;
  MBT* get_main_frameset_frame();
  base_iterator_type insert_html(const MBT& htmlElem);
  base_iterator_type insert_root(const MBT& rootElem);
  void create_root_doc_elements();
  void assign_root_doc_elements();
  MBT* get_body() const;
  MBT* get_head() const;
  MBT* get_title() const;
  MBT* get_frameset() const;

// data
private:
  MBT* pRoot;
  MBT* pHtml;
  mutable MBT* pHead;
  mutable MBT* pTitle;
  mutable MBT* pBody;
  mutable MBT* pFrameset;
};


#include "xhtml_doc.inl"