objistrasnb.cpp

ncbi源码

/*
 * ===========================================================================
 * PRODUCTION $Log: objistrasnb.cpp,v $
 * PRODUCTION Revision 1000.3  2004/06/01 19:40:59  gouriano
 * PRODUCTION PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.70
 * PRODUCTION
 * ===========================================================================
 */

/*  $Id: objistrasnb.cpp,v 1000.3 2004/06/01 19:40:59 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: objistrasnb.cpp,v $
* Revision 1000.3  2004/06/01 19:40:59  gouriano
* PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.70
*
* Revision 1.70  2004/05/17 21:03:03  gorelenk
* Added include of PCH ncbi_pch.hpp
*
* Revision 1.69  2004/03/23 15:39:23  gouriano
* Added setup options for skipping unknown data members
*
* Revision 1.68  2004/03/16 17:48:21  gouriano
* make it possible to skip unknown data members
*
* Revision 1.67  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 1.66  2003/10/15 18:00:32  vasilche
* Fixed integer overflow in asn binary input stream after 2GB.
* Added constructor of CObjectIStreamAsnBinary() from CByteSourceReader.
*
* Revision 1.65  2003/08/25 15:59:09  gouriano
* added possibility to use namespaces in XML i/o streams
*
* Revision 1.64  2003/08/19 18:32:38  vasilche
* Optimized reading and writing strings.
* Avoid string reallocation when checking char values.
* Try to reuse old string data when string reference counting is not working.
*
* Revision 1.63  2003/08/14 20:03:58  vasilche
* Avoid memory reallocation when reading over preallocated object.
* Simplified CContainerTypeInfo iterators interface.
*
* Revision 1.62  2003/08/13 15:47:45  gouriano
* implemented serialization of AnyContent objects
*
* Revision 1.61  2003/05/22 20:10:02  gouriano
* added UTF8 strings
*
* Revision 1.60  2003/05/16 18:02:18  gouriano
* revised exception error messages
*
* Revision 1.59  2003/03/26 16:14:23  vasilche
* Removed TAB symbols. Some formatting.
*
* Revision 1.58  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 1.57  2002/01/23 21:58:31  grichenk
* Fixed minor bug in log tags reader
*
* Revision 1.56  2002/01/17 20:52:35  grichenk
* Fixed another bug in long binary tags processing
*
* Revision 1.55  2002/01/14 17:58:11  grichenk
* Fixed long tags processing
*
* Revision 1.54  2001/10/17 20:41:24  grichenk
* Added CObjectOStream::CharBlock class
*
* Revision 1.53  2001/06/07 17:12:51  grichenk
* Redesigned checking and substitution of non-printable characters
* in VisibleString
*
* Revision 1.52  2001/05/17 15:07:08  lavr
* Typos corrected
*
* Revision 1.51  2001/01/22 23:12:57  vakatov
* CObjectIStreamAsnBinary::{Read,Skip}ClassSequential() - use cur.member "pos"
*
* Revision 1.50  2001/01/03 15:22:26  vasilche
* Fixed limited buffer size for REAL data in ASN.1 binary format.
* Fixed processing non ASCII symbols in ASN.1 text format.
*
* Revision 1.49  2000/12/26 22:24:11  vasilche
* Fixed errors of compilation on Mac.
*
* Revision 1.48  2000/12/15 21:29:02  vasilche
* Moved some typedefs/enums from corelib/ncbistd.hpp.
* Added flags to CObjectIStream/CObjectOStream: eFlagAllowNonAsciiChars.
* TByte typedef replaced by Uint1.
*
* Revision 1.47  2000/12/15 15:38:44  vasilche
* Added support of Int8 and long double.
* Enum values now have type Int4 instead of long.
*
* Revision 1.46  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 1.45  2000/10/17 18:45:34  vasilche
* Added possibility to turn off object cross reference detection in
* CObjectIStream and CObjectOStream.
*
* Revision 1.44  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 1.43  2000/09/18 20:00:23  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 1.42  2000/09/01 13:16:18  vasilche
* Implemented class/container/choice iterators.
* Implemented CObjectStreamCopier for copying data without loading into memory.
*
* Revision 1.41  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 1.40  2000/07/03 20:47:22  vasilche
* Removed unused variables/functions.
*
* Revision 1.39  2000/07/03 20:39:56  vasilche
* Fixed comments.
*
* Revision 1.38  2000/07/03 18:42:45  vasilche
* Added interface to typeinfo via CObjectInfo and CConstObjectInfo.
* Reduced header dependency.
*
* Revision 1.37  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 1.36  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 1.35  2000/06/01 19:07:04  vasilche
* Added parsing of XML data.
*
* Revision 1.34  2000/05/24 20:08:47  vasilche
* Implemented XML dump.
*
* Revision 1.33  2000/05/09 16:38:39  vasilche
* CObject::GetTypeInfo now moved to CObjectGetTypeInfo::GetTypeInfo to reduce possible errors.
* Added write context to CObjectOStream.
* Inlined most of methods of helping class Member, Block, ByteBlock etc.
*
* Revision 1.32  2000/04/28 16:58:13  vasilche
* Added classes CByteSource and CByteSourceReader for generic reading.
* Added delayed reading of choice variants.
*
* Revision 1.31  2000/04/13 14:50:27  vasilche
* Added CObjectIStream::Open() and CObjectOStream::Open() for easier use.
*
* Revision 1.30  2000/03/29 15:55:28  vasilche
* Added two versions of object info - CObjectInfo and CConstObjectInfo.
* Added generic iterators by class -
* 	CTypeIterator<class>, CTypeConstIterator<class>,
* 	CStdTypeIterator<type>, CStdTypeConstIterator<type>,
* 	CObjectsIterator and CObjectsConstIterator.
*
* Revision 1.29  2000/03/14 14:42:31  vasilche
* Fixed error reporting.
*
* Revision 1.28  2000/03/10 17:59:20  vasilche
* Fixed error reporting.
* Added EOF bug workaround on MIPSpro compiler (not finished).
*
* Revision 1.27  2000/03/07 14:41:34  vasilche
* Removed default argument.
*
* Revision 1.26  2000/03/07 14:06:23  vasilche
* Added stream buffering to ASN.1 binary input.
* Optimized class loading/storing.
* Fixed bugs in processing OPTIONAL fields.
* Added generation of reference counted objects.
*
* Revision 1.25  2000/02/17 20:02:44  vasilche
* Added some standard serialization exceptions.
* Optimized text/binary ASN.1 reading.
* Fixed wrong encoding of StringStore in ASN.1 binary format.
* Optimized logic of object collection.
*
* Revision 1.24  2000/01/10 19:46:40  vasilche
* Fixed encoding/decoding of REAL type.
* Fixed encoding/decoding of StringStore.
* Fixed encoding/decoding of NULL type.
* Fixed error reporting.
* Reduced object map (only classes).
*
* Revision 1.23  2000/01/05 19:43:54  vasilche
* Fixed error messages when reading from ASN.1 binary file.
* Fixed storing of integers with enumerated values in ASN.1 binary file.
* Added TAG support to key/value of map.
* Added support of NULL variant in CHOICE.
*
* Revision 1.22  1999/12/17 19:05:03  vasilche
* Simplified generation of GetTypeInfo methods.
*
* Revision 1.21  1999/10/19 15:41:03  vasilche
* Fixed reference to IOS_BASE
*
* Revision 1.20  1999/10/18 20:21:41  vasilche
* Enum values now have long type.
* Fixed template generation for enums.
*
* Revision 1.19  1999/10/08 21:00:43  vasilche
* Implemented automatic generation of unnamed ASN.1 types.
*
* Revision 1.18  1999/10/04 16:22:17  vasilche
* Fixed bug with old ASN.1 structures.
*
* Revision 1.17  1999/09/24 18:19:18  vasilche
* Removed dependency on NCBI toolkit.
*
* Revision 1.16  1999/09/23 21:16:08  vasilche
* Removed dependence on asn.h
*
* Revision 1.15  1999/09/23 18:56:59  vasilche
* Fixed bugs with overloaded methods in objistr*.hpp & objostr*.hpp
*
* Revision 1.14  1999/09/22 20:11:55  vasilche
* Modified for compilation on IRIX native c++ compiler.
*
* Revision 1.13  1999/09/14 18:54:18  vasilche
* Fixed bugs detected by gcc & egcs.
* Removed unneeded includes.
*
* Revision 1.12  1999/08/16 16:08:31  vasilche
* Added ENUMERATED type.
*
* Revision 1.11  1999/08/13 20:22:57  vasilche
* Fixed lot of bugs in datatool
*
* Revision 1.10  1999/08/13 15:53:51  vasilche
* C++ analog of asntool: datatool
*
* Revision 1.9  1999/07/26 18:31:37  vasilche
* Implemented skipping of unused values.
* Added more useful error report.
*
* Revision 1.8  1999/07/22 20:36:38  vasilche
* Fixed 'using namespace' declaration for MSVC.
*
* Revision 1.7  1999/07/22 19:40:56  vasilche
* Fixed bug with complex object graphs (pointers to members of other objects).
*
* Revision 1.6  1999/07/22 17:33:52  vasilche
* Unified reading/writing of objects in all three formats.
*
* Revision 1.5  1999/07/21 20:02:54  vasilche
* Added embedding of ASN.1 binary output from ToolKit to our binary format.
* Fixed bugs with storing pointers into binary ASN.1
*
* Revision 1.4  1999/07/21 14:20:05  vasilche
* Added serialization of bool.
*
* Revision 1.3  1999/07/09 16:32:54  vasilche
* Added OCTET STRING write/read.
*
* Revision 1.2  1999/07/07 21:15:02  vasilche
* Cleaned processing of string types (string, char*, const char*).
*
* Revision 1.1  1999/07/02 21:31:56  vasilche
* Implemented reading from ASN.1 binary format.
*
* ===========================================================================
*/

#include <ncbi_pch.hpp>
#include <corelib/ncbistd.hpp>
#include <corelib/ncbi_limits.hpp>

#include <serial/objistrasnb.hpp>
#include <serial/member.hpp>
#include <serial/memberid.hpp>
#include <serial/enumvalues.hpp>
#include <serial/memberlist.hpp>
#include <serial/objhook.hpp>
#include <serial/classinfo.hpp>
#include <serial/choice.hpp>
#include <serial/continfo.hpp>
#include <serial/objistrimpl.hpp>
#include <serial/pack_string.hpp>

using namespace NCBI_NS_NCBI::CObjectStreamAsnBinaryDefs;

BEGIN_NCBI_SCOPE


CObjectIStream* CObjectIStream::CreateObjectIStreamAsnBinary(void)
{
    return new CObjectIStreamAsnBinary();
}


CObjectIStreamAsnBinary::CObjectIStreamAsnBinary(EFixNonPrint how)
    : CObjectIStream(eSerial_AsnBinary), m_FixMethod(how)
{
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagStart;
    m_CurrentTagLimit = numeric_limits<size_t>::max();
#endif
    m_CurrentTagLength = 0;
}

CObjectIStreamAsnBinary::CObjectIStreamAsnBinary(CNcbiIstream& in,
                                                 EFixNonPrint how)
    : CObjectIStream(eSerial_AsnBinary), m_FixMethod(how)
{
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagStart;
    m_CurrentTagLimit = numeric_limits<size_t>::max();
#endif
    m_CurrentTagLength = 0;
    Open(in);
}

CObjectIStreamAsnBinary::CObjectIStreamAsnBinary(CNcbiIstream& in,
                                                 bool deleteIn,
                                                 EFixNonPrint how)
    : CObjectIStream(eSerial_AsnBinary), m_FixMethod(how)
{
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagStart;
    m_CurrentTagLimit = numeric_limits<size_t>::max();
#endif
    m_CurrentTagLength = 0;
    Open(in, deleteIn);
}

CObjectIStreamAsnBinary::CObjectIStreamAsnBinary(CByteSourceReader& reader,
                                                 EFixNonPrint how)
    : CObjectIStream(eSerial_AsnBinary), m_FixMethod(how)
{
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagStart;
    m_CurrentTagLimit = numeric_limits<size_t>::max();
#endif
    m_CurrentTagLength = 0;
    Open(reader);
}

#if CHECK_STREAM_INTEGRITY
Uint1 CObjectIStreamAsnBinary::PeekTagByte(size_t index)
{
    if ( m_CurrentTagState != eTagStart )
        ThrowError(fIllegalCall,
            "illegal PeekTagByte call: only allowed at tag start");
    return m_Input.PeekChar(index);
}

Uint1 CObjectIStreamAsnBinary::StartTag(void)
{
    if ( m_CurrentTagLength != 0 )
        ThrowError(fIllegalCall,
            "illegal StartTag call: current tag length != 0");
    return PeekTagByte();
}
#endif

TTag CObjectIStreamAsnBinary::PeekTag(void)
{
    Uint1 byte = StartTag();
    ETag sysTag = ExtractTag(byte);
    if ( sysTag != eLongTag ) {
        m_CurrentTagLength = 1;
#if CHECK_STREAM_INTEGRITY
        m_CurrentTagState = eTagParsed;
#endif
        return sysTag;
    }
    TTag tag = 0;
    size_t i = 1;
    const size_t KBitsInByte = 8; // ?
    const size_t KTagBits = sizeof(tag) * KBitsInByte - 1;
    do {
        if ( tag >= (1 << (KTagBits - 7)) ) {
            ThrowError(fOverflow, "tag number is too big: "+NStr::UIntToString(tag));
        }
        byte = PeekTagByte(i++);
        tag = (tag << 7) | (byte & 0x7f);
    } while ( (byte & 0x80) != 0 );
    m_CurrentTagLength = i;
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagParsed;
#endif
    return tag;
}

TTag CObjectIStreamAsnBinary::PeekTag(EClass cls, bool constructed)
{
    if ( ExtractClassAndConstructed(PeekTagByte()) !=
         MakeTagByte(cls, constructed) ) {
        ThrowError(fFormatError, "unexpected tag class/constructed: #"
            + NStr::UIntToString(PeekTagByte()) + ", should be #"
            + NStr::UIntToString(MakeTagByte(cls, constructed)));
    }
    return PeekTag();
}

string CObjectIStreamAsnBinary::PeekClassTag(void)
{
    Uint1 byte = StartTag();
    if ( ExtractTag(byte) != eLongTag ) {
        ThrowError(fFormatError, "long tag expected");
    }
    string name;
    size_t i = 1;
    Uint1 c;
    while ( ((c = PeekTagByte(i++)) & 0x80) != 0 ) {
        name += char(c & 0x7f);
        if ( i > 1024 ) {
            ThrowError(fOverflow, "tag number is too big (greater than 1024)");
        }
    }
    m_CurrentTagLength = i;
#if CHECK_STREAM_INTEGRITY
    m_CurrentTagState = eTagParsed;
#endif
    name += char(c & 0x7f);
    return name;
}

Uint1 CObjectIStreamAsnBinary::PeekAnyTag(void)
{
    Uint1 fByte = StartTag();
    if ( ExtractTag(fByte) != eLongTag ) {
        m_CurrentTagLength = 1;
#if CHECK_STREAM_INTEGRITY