include/openbabel/xml.h

/**********************************************************************
xml.h Declaration of XMLConversion,
declaration and definition of XMLBaseFormat and XMLMoleculeFormat
Copyright (C) 2005-2006 by Chris Morley

This program 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 version 2 of the License.

This program 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.
***********************************************************************/

#ifndef OB_XML_H
#define OB_XML_H

#include <typeinfo>

#include <openbabel/obconversion.h>
#include <openbabel/obmolecformat.h>
#include <openbabel/mol.h>

#include <libxml/xmlreader.h>
#include <libxml/xmlwriter.h>
#include <typeinfo>

namespace OpenBabel
{


  //forward declaration
  class XMLBaseFormat;

  //******************************************************
  /** \class XMLConversion xml.h <openbabel/xml.h>
      \brief A subclass for conversion of XML formats

      An extended OBConversion class which includes a libxml2 reader for use
      with XML formats. Copies an OBConversion and then extends it
      with a XML parser. Instances made on the heap are deleted when
      the original OBConversion object is.

      This class is not intended to be used externally -- instead use
      OBConversion which will find both XML and non-XML OBFormats.

      Instead, this subclass also has support for handling specific
      needs in XML formats. For example, an XML file may include
      multiple namespaces, and the conversion should call appropriate
      XMLBaseFormat formats as needed.
  **/
  class XMLConversion : public OBConversion
    {
    public:
      ///Existing OBConversion instance copied
      XMLConversion(OBConversion* pConv);

      ///Frees reader and writer if necessary
      ~XMLConversion();

      bool SetupReader();///< opens libxml2 reader
      bool SetupWriter();///< opens libxml2 writer

      ///Parses the input xml stream and sends each element to the format's callback routines
      bool ReadXML(XMLBaseFormat* pFormat, OBBase* pOb);

      ///Read and discard XML text up to the next occurrence of the tag e.g."/molecule>"
      ///This is left as the current node. Returns 1 on success, 0 if not found, -1 if failed.
      int SkipXML(const char* ctag);

      typedef std::map<std::string, XMLBaseFormat*> NsMapType;

      ///This static function returns a reference to the map
      ///Avoids "static initialization order fiasco"
      static NsMapType& Namespaces()
        {
          static NsMapType ns;
          return ns;

          //static NsMapType* nsm = NULL;
          //if (!nsm)
          //  nsm = new NsMapType;
          //return *nsm;
        };

      static void RegisterXMLFormat(XMLBaseFormat* pFormat,
                                    bool IsDefault=false, const char* uri=nullptr);

      ///Returns the extended OBConversion class, making it if necessary
      static XMLConversion* GetDerived(OBConversion* pConv, bool ForReading=true);

      ///Because OBConversion::Convert is still using the unextended OBConversion object
      ///we need to obtain the conversion parameters from it when requested
      bool IsLast()
        { return _pConv->IsLast(); }
      int GetOutputIndex()
        { return  _pConv->GetOutputIndex(); }


      xmlTextReaderPtr GetReader() const
        { return _reader;   };

      xmlTextWriterPtr GetWriter() const
        { return _writer;   };

      void OutputToStream()
        {
          xmlOutputBufferFlush(_buf);
        }

      static XMLBaseFormat* GetDefaultXMLClass() //TODO make dependent on object type
        { return _pDefault;};

      void LookForNamespace()
        { _LookingForNamespace = true; };

      ///Static callback functions for xmlReaderForIO()
      static int ReadStream(void * context, char * buffer, int len);
      static int WriteStream(void * context, const char * buffer, int len);
      //static int CloseStream(void* context);

      std::string GetAttribute(const char* attrname);

      ///Sets value to element content. Returns false if there is no content.
      std::string GetContent();

      ///Sets value to element content as an integer. Returns false if there is no content.
      bool    GetContentInt(int& value);

      ///Sets value to element content as an double. Returns false if there is no content.
      bool GetContentDouble(double& value);

    private:
      static XMLBaseFormat* _pDefault;
      OBConversion* _pConv;
      std::streampos  _requestedpos, _lastpos;
      xmlTextReaderPtr _reader;
      xmlTextWriterPtr _writer;
      xmlOutputBufferPtr _buf;
      //    xmlBufferPtr _buf;
      bool _LookingForNamespace;
    public:
      bool _SkipNextRead;
    };

  //*************************************************
  /// \class XMLBaseFormat xml.h <openbabel/xml.h>
  /// \brief Abstract class containing common functionality for XML formats.
  class XMLBaseFormat : public OBFormat
    {
    protected:
      XMLConversion* _pxmlConv;

      //formating for output
      std::string _prefix;
      int baseindent, ind;
      std::string nsdecl;
      int _embedlevel;

    public:
      ~XMLBaseFormat(){}
      virtual const char* NamespaceURI()const=0;
      virtual bool DoElement(const std::string& ElName){return false;};
      virtual bool EndElement(const std::string& ElName){return false;};
      /// The tag at the end of the chemical object e.g. "/molecule>"
      virtual const char* EndTag(){return ">";};

    protected:
      xmlTextReaderPtr reader() const
        {
          return _pxmlConv->GetReader();
        }

      xmlTextWriterPtr writer() const
        {
          return _pxmlConv->GetWriter();
        }

      void OutputToStream()
        {
          _pxmlConv->OutputToStream();
        }

      ///Skip past first n objects in input stream (or current one with n=0)
      /// Returns 1 on success, -1 on error and 0 if not implemented
      virtual int SkipObjects(int n, OBConversion* pConv)
        {
          //don't implement on base class
          if(*EndTag()=='>')
            return 0;

          //Set up XMLConversion class with reader
          _pxmlConv = XMLConversion::GetDerived(pConv,true);
          if(!_pxmlConv)
            return -1;

          //always find the end of at least 1 object
          if(n==0)++n;

          //Skip n objects, returning -1 if not successful
          int i;
          for(i=0; i<n; ++i)
            if(_pxmlConv->SkipXML(EndTag())!=1)
              return -1;

          return 1;
        }

    };

  //*************************************************
  /// \class XMLMoleculeFormat xml.h <openbabel/xml.h>
  /// \brief Abstract class for XML formats which represent molecules
  class XMLMoleculeFormat : public XMLBaseFormat
    {
    protected:
      OBMol* _pmol;

    public:
      ~XMLMoleculeFormat(){}
      virtual bool ReadChemObject(OBConversion* pConv)
        {
          return OBMoleculeFormat::ReadChemObjectImpl(pConv, this);
        };

      virtual bool WriteChemObject(OBConversion* pConv)
        {
          return OBMoleculeFormat::WriteChemObjectImpl(pConv, this);
        };

      virtual bool ReadMolecule(OBBase* pOb, OBConversion* pConv)
        {
          _pmol = dynamic_cast<OBMol*>(pOb);
          if(!_pmol)
            return false;
          _pxmlConv = XMLConversion::GetDerived(pConv,true);
          if(!_pxmlConv)
            return false;
          _embedlevel = -1;
          return _pxmlConv->ReadXML(this,pOb);
        };

      const std::type_info& GetType()
        {
          return typeid(OBMol*);
        };

    };


}//namespace

//! \file
//! \brief Declaration of XMLConversion,
//!  declaration and definition of XMLBaseFormat and XMLMoleculeFormat

#endif