objistrxml.cpp

ncbi源码

/*
 * ===========================================================================
 * PRODUCTION $Log: objistrxml.cpp,v $
 * PRODUCTION Revision 1000.3  2004/06/01 19:41:03  gouriano
 * PRODUCTION PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.61
 * PRODUCTION
 * ===========================================================================
 */

/*  $Id: objistrxml.cpp,v 1000.3 2004/06/01 19:41:03 gouriano Exp $
* ===========================================================================
*
*                            PUBLIC DOMAIN NOTICE
*               National Center for Biotechnology Information
*
*  This software/database is a "United States Government Work" under the
*  terms of the United States Copyright Act.  It was written as part of
*  the author's official duties as a United States Government employee and
*  thus cannot be copyrighted.  This software/database is freely available
*  to the public for use. The National Library of Medicine and the U.S.
*  Government have not placed any restriction on its use or reproduction.
*
*  Although all reasonable efforts have been taken to ensure the accuracy
*  and reliability of the software and data, the NLM and the U.S.
*  Government do not and cannot warrant the performance or results that
*  may be obtained by using this software or data. The NLM and the U.S.
*  Government disclaim all warranties, express or implied, including
*  warranties of performance, merchantability or fitness for any particular
*  purpose.
*
*  Please cite the author in any work or product based on this material.
*
* ===========================================================================
*
* Author: Eugene Vasilchenko
*
* File Description:
*   !!! PUT YOUR DESCRIPTION HERE !!!
*
* ---------------------------------------------------------------------------
* $Log: objistrxml.cpp,v $
* Revision 1000.3  2004/06/01 19:41:03  gouriano
* PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.61
*
* Revision 1.61  2004/05/19 17:26:19  gouriano
* Corrected serialization of containers when the code was generated by DTD
*
* Revision 1.60  2004/05/17 21:03:03  gorelenk
* Added include of PCH ncbi_pch.hpp
*
* Revision 1.59  2004/04/28 19:24:29  gouriano
* Corrected reading of containers
*
* Revision 1.58  2004/03/23 15:39:23  gouriano
* Added setup options for skipping unknown data members
*
* Revision 1.57  2004/03/05 20:29:38  gouriano
* make it possible to skip unknown data fields
*
* Revision 1.56  2004/01/30 20:29:56  gouriano
* Corrected reading white spaces
*
* Revision 1.55  2004/01/29 20:48:07  gouriano
* Corrected handling of white spaces in XML tags: preserve insignificant
* white spaces inside the tag.
* Corrected reading of AnyContent object attributes and tags with empty content
*
* Revision 1.54  2004/01/22 20:50:17  gouriano
* corrected reading of undefined attributes, of boolean values, of tags with empty content
*
* Revision 1.53  2004/01/12 16:52:58  gouriano
* Corrected reading EMPTY tag value. Added skipping undescribed class member attributes.
*
* Revision 1.52  2004/01/08 18:16:53  gouriano
* correction when skipping end of line
*
* Revision 1.51  2004/01/08 17:40:53  gouriano
* Added encoding Windows-1252.
* Corrected reading of containers.
* Made it possible to read a choice variant, which is am empty container.
*
* Revision 1.50  2003/11/26 19:59:40  vasilche
* GetPosition() and GetDataFormat() methods now are implemented
* in parent classes CObjectIStream and CObjectOStream to avoid
* pure virtual method call in destructors.
*
* Revision 1.49  2003/09/25 19:44:08  gouriano
* Corrected serialization of AnyContent object's attributes
*
* Revision 1.48  2003/09/16 14:48:36  gouriano
* Enhanced AnyContent objects to support XML namespaces and attribute info items.
*
* Revision 1.47  2003/08/25 15:59:09  gouriano
* added possibility to use namespaces in XML i/o streams
*
* Revision 1.46  2003/08/14 20:03:58  vasilche
* Avoid memory reallocation when reading over preallocated object.
* Simplified CContainerTypeInfo iterators interface.
*
* Revision 1.45  2003/08/13 15:47:45  gouriano
* implemented serialization of AnyContent objects
*
* Revision 1.44  2003/07/02 13:01:29  gouriano
* added ability to read/write XML files with reference to schema
*
* Revision 1.43  2003/06/30 15:39:48  gouriano
* added encoding (utf-8 or iso-8859-1)
*
* Revision 1.42  2003/06/24 20:57:36  gouriano
* corrected code generation and serialization of non-empty unnamed containers (XML)
*
* Revision 1.41  2003/05/22 20:10:02  gouriano
* added UTF8 strings
*
* Revision 1.40  2003/05/16 18:02:18  gouriano
* revised exception error messages
*
* Revision 1.39  2003/05/05 20:09:10  gouriano
* fixed "skipping" an object
*
* Revision 1.38  2003/03/26 16:14:23  vasilche
* Removed TAB symbols. Some formatting.
*
* Revision 1.37  2003/02/07 16:09:22  gouriano
* correction of GetContainerElementTypeFamily for the case of copying objects
*
* Revision 1.36  2003/02/05 17:08:01  gouriano
* added possibility to read/write objects generated from an ASN.1 spec as "standard" XML - without scope prefixes
*
* Revision 1.35  2003/01/21 19:32:27  gouriano
* corrected reading containers of primitive types
*
* Revision 1.34  2002/12/26 19:32:33  gouriano
* changed XML I/O streams to properly handle object copying
*
* Revision 1.33  2002/12/17 19:04:32  gouriano
* corrected reading/writing of character references
*
* Revision 1.32  2002/12/13 21:50:42  gouriano
* corrected reading of choices
*
* Revision 1.31  2002/12/12 21:08:07  gouriano
* implemented handling of complex XML containers
*
* Revision 1.30  2002/11/26 22:09:02  gouriano
* added HasMoreElements method,
* added unnamed lists of sequences (choices) as container elements
*
* Revision 1.29  2002/11/20 21:22:51  gouriano
* corrected processing of unnamed sequences as choice variants
*
* Revision 1.28  2002/11/19 19:48:52  gouriano
* added support of XML attributes of choice variants
*
* Revision 1.27  2002/11/15 20:33:12  gouriano
* support of XML attributes of empty elements
*
* Revision 1.26  2002/11/14 20:58:54  gouriano
* added support of XML attribute lists
*
* Revision 1.25  2002/10/25 14:49:27  vasilche
* NCBI C Toolkit compatibility code extracted to libxcser library.
* Serial streams flags names were renamed to fXxx.
*
* Names of flags
*
* Revision 1.24  2002/10/15 13:47:59  gouriano
* modified to handle "standard" (generated from DTD) XML i/o
*
* Revision 1.23  2002/09/26 18:11:01  gouriano
* added more checks for MemberId
*
* Revision 1.22  2002/09/25 19:37:36  gouriano
* added the possibility of having no tag prefix in XML I/O streams
*
* Revision 1.21  2001/11/13 20:53:36  grichenk
* Fixed comments reading
*
* Revision 1.20  2001/10/17 20:41:24  grichenk
* Added CObjectOStream::CharBlock class
*
* Revision 1.19  2001/08/08 18:35:09  grichenk
* ReadTagData() -- added memory pre-allocation for long strings
*
* Revision 1.18  2001/05/17 15:07:08  lavr
* Typos corrected
*
* Revision 1.17  2000/12/26 22:24:12  vasilche
* Fixed errors of compilation on Mac.
*
* Revision 1.16  2000/12/15 15:38:44  vasilche
* Added support of Int8 and long double.
* Enum values now have type Int4 instead of long.
*
* Revision 1.15  2000/11/07 17:25:40  vasilche
* Fixed encoding of XML:
*     removed unnecessary apostrophes in OCTET STRING
*     removed unnecessary content in NULL
* Added module names to CTypeInfo and CEnumeratedTypeValues
*
* Revision 1.14  2000/10/20 15:51:41  vasilche
* Fixed data error processing.
* Added interface for constructing container objects directly into output stream.
* object.hpp, object.inl and object.cpp were split to
* objectinfo.*, objecttype.*, objectiter.* and objectio.*.
*
* Revision 1.13  2000/10/13 20:59:21  vasilche
* Avoid using of ssize_t absent on some compilers.
*
* Revision 1.12  2000/10/13 20:22:55  vasilche
* Fixed warnings on 64 bit compilers.
* Fixed missing typename in templates.
*
* Revision 1.11  2000/10/03 17:22:44  vasilche
* Reduced header dependency.
* Reduced size of debug libraries on WorkShop by 3 times.
* Fixed tag allocation for parent classes.
* Fixed CObject allocation/deallocation in streams.
* Moved instantiation of several templates in separate source file.
*
* Revision 1.10  2000/09/29 16:18:23  vasilche
* Fixed binary format encoding/decoding on 64 bit compulers.
* Implemented CWeakMap<> for automatic cleaning map entries.
* Added cleaning local hooks via CWeakMap<>.
* Renamed ReadTypeName -> ReadFileHeader, ENoTypeName -> ENoFileHeader.
* Added some user interface methods to CObjectIStream, CObjectOStream and
* CObjectStreamCopier.
*
* Revision 1.9  2000/09/18 20:00:24  vasilche
* Separated CVariantInfo and CMemberInfo.
* Implemented copy hooks.
* All hooks now are stored in CTypeInfo/CMemberInfo/CVariantInfo.
* Most type specific functions now are implemented via function pointers instead of virtual functions.
*
* Revision 1.8  2000/09/01 13:16:18  vasilche
* Implemented class/container/choice iterators.
* Implemented CObjectStreamCopier for copying data without loading into memory.
*
* Revision 1.7  2000/08/15 19:44:49  vasilche
* Added Read/Write hooks:
* CReadObjectHook/CWriteObjectHook for objects of specified type.
* CReadClassMemberHook/CWriteClassMemberHook for specified members.
* CReadChoiceVariantHook/CWriteChoiceVariant for specified choice variants.
* CReadContainerElementHook/CWriteContainerElementsHook for containers.
*
* Revision 1.6  2000/07/03 20:47:23  vasilche
* Removed unused variables/functions.
*
* Revision 1.5  2000/07/03 18:42:45  vasilche
* Added interface to typeinfo via CObjectInfo and CConstObjectInfo.
* Reduced header dependency.
*
* Revision 1.4  2000/06/16 19:24:22  vasilche
* Updated MSVC project.
* Fixed error on MSVC with static const class member.
*
* Revision 1.3  2000/06/16 16:31:20  vasilche
* Changed implementation of choices and classes info to allow use of the same classes in generated and user written classes.
*
* Revision 1.2  2000/06/07 19:45:59  vasilche
* Some code cleaning.
* Macros renaming in more clear way.
* BEGIN_NAMED_*_INFO, ADD_*_MEMBER, ADD_NAMED_*_MEMBER.
*
* Revision 1.1  2000/06/01 19:07:04  vasilche
* Added parsing of XML data.
*
* ===========================================================================
*/

#include <ncbi_pch.hpp>
#include <corelib/ncbistd.hpp>
#include <serial/objistrxml.hpp>
#include <serial/enumvalues.hpp>
#include <serial/objhook.hpp>
#include <serial/classinfo.hpp>
#include <serial/choice.hpp>
#include <serial/ptrinfo.hpp>
#include <serial/continfo.hpp>
#include <serial/memberlist.hpp>
#include <serial/memberid.hpp>

BEGIN_NCBI_SCOPE

CObjectIStream* CObjectIStream::CreateObjectIStreamXml()
{
    return new CObjectIStreamXml();
}

CObjectIStreamXml::CObjectIStreamXml(void)
    : CObjectIStream(eSerial_Xml),
      m_TagState(eTagOutside), m_Attlist(false),
      m_StdXml(false), m_EnforcedStdXml(false),
      m_Encoding( eEncoding_Unknown )
{
}

CObjectIStreamXml::~CObjectIStreamXml(void)
{
}

CObjectIStreamXml::EEncoding CObjectIStreamXml::GetEncoding(void) const
{
    return m_Encoding;
}

string CObjectIStreamXml::GetPosition(void) const
{
    return "line "+NStr::UIntToString(m_Input.GetLine());
}

void CObjectIStreamXml::SetEnforcedStdXml(bool set)
{
    m_EnforcedStdXml = set;
    if (m_EnforcedStdXml) {
        m_StdXml = false;
    }
}

static inline
bool IsBaseChar(char c)
{
    return
        c >= 'A' && c <='Z' ||
        c >= 'a' && c <= 'z' ||
        c >= '\xC0' && c <= '\xD6' ||
        c >= '\xD8' && c <= '\xF6' ||
        c >= '\xF8' && c <= '\xFF';
}

static inline
bool IsDigit(char c)
{
    return c >= '0' && c <= '9';
}

static inline
bool IsIdeographic(char /*c*/)
{
    return false;
}

static inline
bool IsLetter(char c)
{
    return IsBaseChar(c) || IsIdeographic(c);
}

static inline
bool IsFirstNameChar(char c)
{
    return IsLetter(c) || c == '_' || c == ':';
}

static inline
bool IsCombiningChar(char /*c*/)
{
    return false;
}

static inline
bool IsExtender(char c)
{
    return c == '\xB7';
}

static inline
bool IsNameChar(char c)
{
    return IsFirstNameChar(c) ||
        IsDigit(c) || c == '.' || c == '-' ||
        IsCombiningChar(c) || IsExtender(c);
}

static inline
bool IsWhiteSpace(char c)
{
    return c == ' ' || c == '\t' || c == '\n' || c == '\r';
}

static inline
bool IsEndOfTagChar(char c)
{
    return c == '>' || c == '/';
}

char CObjectIStreamXml::SkipWS(void)
{
    _ASSERT(InsideTag());
    for ( ;; ) {
        char c = m_Input.SkipSpaces();
        switch ( c ) {
        case '\t':
            m_Input.SkipChar();
            continue;
        case '\r':
        case '\n':
            m_Input.SkipChar();
            m_Input.SkipEndOfLine(c);
            continue;
        default:
            return c;
        }
    }
}

char CObjectIStreamXml::SkipWSAndComments(void)
{
    _ASSERT(OutsideTag());
    for ( ;; ) {
        char c = m_Input.SkipSpaces();
        switch ( c ) {
        case '\t':
            m_Input.SkipChar();
            continue;
        case '\r':
        case '\n':
            m_Input.SkipChar();
            m_Input.SkipEndOfLine(c);
            continue;
        case '<':
            if ( m_Input.PeekChar(1) == '!' &&
                 m_Input.PeekChar(2) == '-' &&
                 m_Input.PeekChar(3) == '-' &&
                 m_Input.PeekChar(4) != '-' ) {
                // start of comment
                m_Input.SkipChars(4);
                for ( ;; ) {
                    m_Input.FindChar('-');
                    if ( m_Input.PeekChar(1) == '-' ) {
                        // --
                        if ( m_Input.PeekChar(2) == '>' ) {
                            // -->
                            m_Input.SkipChars(3);
                            break;
                        }
                        else {
                            // --[^>]
                            ThrowError(fFormatError,
                                       "double dash '--' is not allowed in XML comments");
                        }
                    }
                    else {
                        // -[^-]
                        m_Input.SkipChars(2);
                    }
                    
                }
                continue; // skip the next WS or comment
            }