multireader.cpp

Go to the documentation of this file.

Go to the SVN repository for this file.

/*  $Id: multireader.cpp 101850 2024-02-22 16:13:34Z stakhovv $
 * ===========================================================================
 *
 *                            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:  Frank Ludwig, Sergiy Gotvyanskyy, NCBI
 *
 * File Description:
 *   Reader for selected data file formats
 *
 * ===========================================================================
*/
 
#include <ncbi_pch.hpp>
 
#include "fasta_ex.hpp"
 
#include <objtools/readers/gff3_reader.hpp>
#include <objtools/readers/gtf_reader.hpp>
 
#include <objects/seqloc/Seq_id.hpp>
#include <objects/seqloc/Seq_loc.hpp>
#include <objects/seq/Seq_annot.hpp>
#include <objects/seqset/Seq_entry.hpp>
#include <objects/seqfeat/Variation_ref.hpp>
 
#include <objects/seq/Seq_descr.hpp>
#include <objects/seqfeat/Org_ref.hpp>
#include <objects/seqfeat/OrgName.hpp>
#include <objects/submit/Seq_submit.hpp>
#include <objects/submit/Submit_block.hpp>
#include <objects/pub/Pub.hpp>
#include <objects/pub/Pub_equiv.hpp>
#include <objects/biblio/Cit_sub.hpp>
#include <objects/seq/Pubdesc.hpp>
#include <objects/general/Object_id.hpp>
#include <objects/seq/Bioseq.hpp>
#include <objects/seqset/Bioseq_set.hpp>
#include <objects/general/Date.hpp>
#include <objects/biblio/Cit_sub.hpp>
 
#include <corelib/ncbistre.hpp>
 
#include <serial/iterator.hpp>
#include <serial/objistr.hpp>
#include <serial/objostr.hpp>
#include <serial/objostrasn.hpp>
#include <serial/serial.hpp>
#include <objects/seq/Annot_id.hpp>
#include <objects/general/Dbtag.hpp>
#include <objtools/readers/readfeat.hpp>
#include <objtools/readers/aln_reader.hpp>
 
#include <objtools/readers/format_guess_ex.hpp>
#include <objtools/readers/reader_listener.hpp>
 
#include "multireader.hpp"
#include "table2asn_context.hpp"
#include "descr_apply.hpp"
 
#include <objtools/edit/feattable_edit.hpp>
 
#include <corelib/stream_utils.hpp>
#include <common/ncbi_revision.h>
#include "utils.hpp"
 
#ifndef NCBI_SC_VERSION
#   define FLATFILE_PARSER_ENABLED
#elif (NCBI_SC_VERSION == 0)
#   define FLATFILE_PARSER_ENABLED
#endif
 
#ifdef FLATFILE_PARSER_ENABLED
#   include <objtools/flatfile/flatfile_parser.hpp>
#endif
 
#include <common/test_assert.h> /* This header must go last */
 
 
BEGIN_NCBI_SCOPE
USING_SCOPE(objects);
 
 
void g_ModifySeqIds(CSeq_annot& annot, const CSeq_id& match, CRef<CSeq_id> new_id)
{
    CTypeIterator<CSeq_loc> visitor(annot);
 
    CSeq_id& id = *new_id;
    while (visitor) {
        CSeq_loc& loc = *visitor;
        if (loc.GetId()->Compare(match) == CSeq_id::e_YES) {
            loc.SetId(id);
        }
        ++visitor;
    }
}
 
 
namespace
{
 
    struct SCSeqidCompare {
        inline bool operator()(const CSeq_id* left, const CSeq_id* right) const
        {
            return *left < *right;
        };
    };
 
}
 
 
const set<TTypeInfo> CMultiReader::kSupportedTypes = {
    CBioseq_set::GetTypeInfo(),
    CBioseq::GetTypeInfo(),
    CSeq_entry::GetTypeInfo(),
    CSeq_submit::GetTypeInfo(),
};
 
 
CRef<CSerialObject> CMultiReader::xReadASN1Binary(CObjectIStream& pObjIstrm, const string& content_type) const
{
    if (content_type == "Bioseq-set") {
        auto obj = Ref(new CSeq_entry);
        auto& bioseq_set = obj->SetSet();
        pObjIstrm.Read(ObjectInfo(bioseq_set));
        return obj;
    }
 
    if (content_type == "Seq-submit") {
        auto seqsubmit = Ref(new CSeq_submit);
        pObjIstrm.Read(ObjectInfo(*seqsubmit));
        return seqsubmit;
    }
 
    if (content_type == "Seq-entry") {
        auto obj = Ref(new CSeq_entry);
        pObjIstrm.Read(ObjectInfo(*obj));
        return obj;
    }
 
    if (content_type == "Bioseq") {
        auto obj = Ref(new CSeq_entry);
        pObjIstrm.Read(ObjectInfo(obj->SetSeq()));
        return obj;
    };
 
    return {};
}
 
CRef<CSerialObject> CMultiReader::xReadASN1Text(CObjectIStream& pObjIstrm) const
{
    CRef<CSeq_entry> entry;
    CRef<CSeq_submit> submit;
 
    // guess object type
    string sType;
    try {
        sType = pObjIstrm.ReadFileHeader();
    } catch (const CEofException&) {
        sType.clear();
        // ignore EOF exception
    }
 
    // do the right thing depending on the input type
    if (sType == CBioseq_set::GetTypeInfo()->GetName()) {
        entry.Reset(new CSeq_entry);
        pObjIstrm.Read(ObjectInfo(entry->SetSet()), CObjectIStream::eNoFileHeader);
    } else if (sType == CSeq_submit::GetTypeInfo()->GetName()) {
        submit.Reset(new CSeq_submit);
        pObjIstrm.Read(ObjectInfo(*submit), CObjectIStream::eNoFileHeader);
 
        if (submit->GetData().GetEntrys().size() > 1) {
            entry.Reset(new CSeq_entry);
            entry->SetSet().SetSeq_set() = submit->GetData().GetEntrys();
        }
        else
            entry = *submit->SetData().SetEntrys().begin();
    } else if (sType == CSeq_entry::GetTypeInfo()->GetName()) {
        entry.Reset(new CSeq_entry);
        pObjIstrm.Read(ObjectInfo(*entry), CObjectIStream::eNoFileHeader);
    } else if (sType == CSeq_annot::GetTypeInfo()->GetName()) {
        entry.Reset(new CSeq_entry);
        do {
            CRef<CSeq_annot> annot(new CSeq_annot);
            pObjIstrm.Read(ObjectInfo(*annot), CObjectIStream::eNoFileHeader);
            entry->SetSeq().SetAnnot().push_back(annot);
            try {
                sType = pObjIstrm.ReadFileHeader();
            } catch (const CEofException&) {
                sType.clear();
                // ignore EOF exception
            }
        } while (sType == CSeq_annot::GetTypeInfo()->GetName());
    } else {
        return CRef<CSerialObject>();
    }
 
    if (m_context.m_gapNmin > 0) {
        CGapsEditor gap_edit(
            (CSeq_gap::EType)m_context.m_gap_type,
            m_context.m_DefaultEvidence,
            m_context.m_GapsizeToEvidence,
            m_context.m_gapNmin,
            m_context.m_gap_Unknown_length);
        gap_edit.ConvertNs2Gaps(*entry);
    }
 
    if (submit.Empty())
        return entry;
    else
        return submit;
}
 
//  ----------------------------------------------------------------------------
CRef<CSeq_entry>
CMultiReader::ReadAlignment(CNcbiIstream& instream, const CArgs& args)
//  ----------------------------------------------------------------------------
{
    CAlnReader reader(instream);
    reader.SetAllGap(args["aln-gapchar"].AsString());
    reader.SetMissing(args["aln-gapchar"].AsString());
    if (args["aln-alphabet"].AsString() == "nuc") {
        reader.SetAlphabet(CAlnReader::eAlpha_Nucleotide);
    } else {
        reader.SetAlphabet(CAlnReader::eAlpha_Protein);
    }
 
    reader.Read(0, m_context.m_logger);
    auto pSeqEntry =
        reader.GetSeqEntry(
            CAlnReader::fGenerateLocalIDs,
            m_context.m_logger);
 
    if (pSeqEntry && args["a"]) {
        static const map<string, CBioseq_set::EClass>
            s_StringToClass = {
                { "s", CBioseq_set::eClass_genbank },
                { "s1", CBioseq_set::eClass_pop_set },
                { "s2", CBioseq_set::eClass_phy_set },
                { "s3", CBioseq_set::eClass_mut_set },
                { "s4", CBioseq_set::eClass_eco_set },
                { "s9", CBioseq_set::eClass_small_genome_set }
            };
 
        auto it = s_StringToClass.find(args["a"].AsString());
        if (it != s_StringToClass.end()) {
            pSeqEntry->SetSet().SetClass(it->second);
        }
    }
 
    return pSeqEntry;
}
 
 
//  ----------------------------------------------------------------------------
CRef<CSeq_entry>
CMultiReader::xReadFasta(CNcbiIstream& instream)
{
    if (m_context.m_gapNmin > 0) {
        m_iFlags |= CFastaReader::fParseGaps
                 |  CFastaReader::fLetterGaps;
    } else {
        m_iFlags |= CFastaReader::fNoSplit;
//                 |  CFastaReader::fLeaveAsText;
    }
 
    if (m_context.m_d_fasta) {
        m_iFlags |= CFastaReader::fParseGaps;
    }
 
    m_iFlags |= CFastaReader::fIgnoreMods
             |  CFastaReader::fValidate
             |  CFastaReader::fHyphensIgnoreAndWarn
             |  CFastaReader::fDisableParseRange;
 
    if (m_context.m_allow_accession)
        m_iFlags |= CFastaReader::fParseRawID;
 
 
    m_iFlags |= CFastaReader::fAssumeNuc
             |  CFastaReader::fForceType;
 
    unique_ptr<CFastaReaderEx> pReader(new CFastaReaderEx(m_context, instream, m_iFlags));
    if (! pReader) {
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "File format not supported", 0);
    }
    if (m_context.m_gapNmin > 0) {
        pReader->SetMinGaps(m_context.m_gapNmin, m_context.m_gap_Unknown_length);
    }
 
    // if (m_context.m_gap_evidences.size() > 0 || m_context.m_gap_type >= 0)
    if (! m_context.m_GapsizeToEvidence.empty() ||
        ! m_context.m_DefaultEvidence.empty() ||
        m_context.m_gap_type >= 0) {
        pReader->SetGapLinkageEvidence(
            (CSeq_gap::EType)m_context.m_gap_type,
            m_context.m_DefaultEvidence,
            m_context.m_GapsizeToEvidence);
    }
 
    int max_seqs = kMax_Int;
    CRef<CSeq_entry> result;
    if (m_context.m_di_fasta)
        result = pReader->ReadDeltaFasta(m_context.m_logger);
    else if (m_context.m_d_fasta)
        result = pReader->ReadDeltaFasta(m_context.m_logger);
    else
        result = pReader->ReadSet(max_seqs, m_context.m_logger);
 
    if (result.NotEmpty()) {
        m_context.MakeGenomeCenterId(*result);
    }
 
    if (result->IsSet() && ! m_context.m_HandleAsSet) {
        g_LogGeneralParsingError(
            eDiag_Warning,
            "File " + m_context.m_current_file + " contains multiple sequences",
            *(m_context.m_logger));
    }
    if (result->IsSet()) {
        result->SetSet().SetClass(m_context.m_ClassValue);
    }
 
    return result;
}
 
//  ----------------------------------------------------------------------------
CFormatGuess::EFormat CMultiReader::xInputGetFormat(CNcbiIstream& istr, CFileContentInfo* content_info) const
{
    CFormatGuessEx FG(istr);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eBinaryASN);
    // FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eFasta); // we wouldn't take "no" for an answer anyway
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eTextASN);
    // FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eXml);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eGff3);
    FG.GetFormatHints().DisableAllNonpreferred();
 
    if (! content_info) {
        return FG.GuessFormat();
    }
 
    FG.SetRecognizedGenbankTypes(kSupportedTypes);
    return FG.GuessFormatAndContent(*content_info);
}
 
//  ----------------------------------------------------------------------------
CFormatGuess::EFormat CMultiReader::xAnnotGetFormat(CNcbiIstream& istr) const
{
    CFormatGuess FG(istr);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eBinaryASN);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eTextASN);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eGffAugustus);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eGff3);
    // FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eGff2);
    //  RW-1591: Need at least GFF3 or GTF (plain or Augustus) to properly relate
    //  the features
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eGtf);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eFiveColFeatureTable);
#ifdef FLATFILE_PARSER_ENABLED
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eFlatFileGenbank);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eFlatFileEna);
    FG.GetFormatHints().AddPreferredFormat(CFormatGuess::eFlatFileUniProt);
#endif
    FG.GetFormatHints().DisableAllNonpreferred();
 
    return FG.GuessFormat();
}
 
//  ----------------------------------------------------------------------------
void CMultiReader::WriteObject(
    const CSerialObject& object,
    ostream& ostr)
{
    ostr << MSerial_Format{ m_context.m_binary_asn1_output ? eSerial_AsnBinary : eSerial_AsnText }
         //<< MSerial_VerifyNo
         << object;
    ostr.flush();
}
 
CMultiReader::CMultiReader(CTable2AsnContext& context) :
    m_context(context),
    mAtSequenceData(false)
{
}
 
/*
void CMultiReader::ApplyAdditionalProperties(CSeq_entry& entry)
{
    switch(entry.Which()) {
    case CSeq_entry::e_Seq:
        if (!m_context.m_OrganismName.empty() || m_context.m_taxid != 0) {
            CBioSource::TOrg& org(CAutoAddDesc(entry.SetDescr(), CSeqdesc::e_Source).Set().SetSource().SetOrg());
            // we should reset taxid in case new name is different
            if (org.IsSetTaxname() && org.GetTaxId() >0 && org.GetTaxname() != m_context.m_OrganismName) {
                org.SetTaxId(0);
            }
 
            if (!m_context.m_OrganismName.empty())
                org.SetTaxname(m_context.m_OrganismName);
            if (m_context.m_taxid != 0)
                org.SetTaxId(m_context.m_taxid);
        }
        break;
 
    case CSeq_entry::e_Set:
    {
        if (!entry.GetSet().IsSetClass())
            entry.SetSet().SetClass(CBioseq_set::eClass_genbank);
 
        NON_CONST_ITERATE(CBioseq_set_Base::TSeq_set, it, entry.SetSet().SetSeq_set())
        {
            ApplyAdditionalProperties(**it);
        }
    }
        break;
    default:
        break;
    }
}
*/
 
void CMultiReader::LoadDescriptors(const string& ifname, CRef<CSeq_descr>& out_desc) const
{
    out_desc.Reset(new CSeq_descr);
 
    unique_ptr<CObjectIStream> pObjIstrm = xCreateASNStream(ifname);
 
    // guess object type
    //const string sType = pObjIstrm->ReadFileHeader();
 
    // do the right thing depending on the input type
    while (true) {
        try {
            const string sType = pObjIstrm->ReadFileHeader();
            if (sType == CSeq_descr::GetTypeInfo()->GetName()) {
                CRef<CSeq_descr> descr(new CSeq_descr);
                pObjIstrm->Read(ObjectInfo(*descr),
                    CObjectIStream::eNoFileHeader);
                out_desc->Set().insert(out_desc->Set().end(), descr->Get().begin(), descr->Get().end());
            } else if (sType == CSeqdesc::GetTypeInfo()->GetName()) {
                CRef<CSeqdesc> desc(new CSeqdesc);
                pObjIstrm->Read(ObjectInfo(*desc),
                    CObjectIStream::eNoFileHeader);
                out_desc->Set().push_back(desc);
            } else if (sType == CPubdesc::GetTypeInfo()->GetName()) {
                CRef<CSeqdesc> desc(new CSeqdesc);
                pObjIstrm->Read(ObjectInfo(desc->SetPub()),
                    CObjectIStream::eNoFileHeader);
                out_desc->Set().push_back(desc);
            } else {
                throw runtime_error("Descriptor file must contain "
                    "either Seq_descr or Seqdesc elements");
            }
        } catch (CException& ex) {
            if (! NStr::EqualNocase(ex.GetMsg(), "end of file")) {
                throw runtime_error("Unable to read descriptor from file:" + ex.GetMsg());
            }
            break;
        }
    }
}
 
void CMultiReader::LoadTemplate(const string& ifname)
{
    unique_ptr<CObjectIStream> pObjIstrm = xCreateASNStream(ifname);
 
    // guess object type
    string sType = pObjIstrm->ReadFileHeader();
 
    // do the right thing depending on the input type
    if (sType == CSeq_entry::GetTypeInfo()->GetName()) {
        m_context.m_entry_template.Reset( new CSeq_entry );
        pObjIstrm->Read(ObjectInfo(*m_context.m_entry_template), CObjectIStream::eNoFileHeader);
    } else if (sType == CBioseq::GetTypeInfo()->GetName()) {
        CRef<CBioseq> pBioseq( new CBioseq );
        pObjIstrm->Read(ObjectInfo(*pBioseq), CObjectIStream::eNoFileHeader);
        m_context.m_entry_template.Reset( new CSeq_entry );
        m_context.m_entry_template->SetSeq( *pBioseq );
    } else if (sType == CSeq_submit::GetTypeInfo()->GetName()) {
        m_context.m_submit_template.Reset( new CSeq_submit );
        pObjIstrm->Read(ObjectInfo(*m_context.m_submit_template), CObjectIStream::eNoFileHeader);
        if (! m_context.m_submit_template->GetData().IsEntrys()
            || m_context.m_submit_template->GetData().GetEntrys().size() != 1) {
            throw runtime_error("Seq-submit template must contain "
                "exactly one Seq-entry");
        }
    } else if (sType == CSubmit_block::GetTypeInfo()->GetName()) {
        // a Submit-block
        CRef<CSubmit_block> submit_block(new CSubmit_block);
        pObjIstrm->Read(ObjectInfo(*submit_block), CObjectIStream::eNoFileHeader);
 
        // Build a Seq-submit containing this plus a bogus Seq-entry
        m_context.m_submit_template.Reset( new CSeq_submit );
        m_context.m_submit_template->SetSub(*submit_block);
        CRef<CSeq_entry> ent(new CSeq_entry);
        CRef<CSeq_id> dummy_id(new CSeq_id("lcl|dummy_id"));
        ent->SetSeq().SetId().push_back(dummy_id);
        ent->SetSeq().SetInst().SetRepr(CSeq_inst::eRepr_raw);
        ent->SetSeq().SetInst().SetMol(CSeq_inst::eMol_dna);
        m_context.m_submit_template->SetData().SetEntrys().push_back(ent);
    } else if (sType == CSeqdesc::GetTypeInfo()->GetName()) {
        // it's OK
    } else {
        NCBI_USER_THROW_FMT("Template must be Seq-entry, Seq-submit, Bioseq or "
            "Submit-block.  Object seems to be of type: " << sType);
    }
 
    // for submit types, pull out the seq-entry inside and remember it
    if (m_context.m_submit_template.NotEmpty() && m_context.m_submit_template->IsEntrys()) {
        m_context.m_entry_template = m_context.m_submit_template->SetData().SetEntrys().front();
    }
 
    // The template may contain a set rather than a seq.
    // That's OK if it contains only one na entry, which we'll use.
    if (m_context.m_entry_template.NotEmpty() && m_context.m_entry_template->IsSet()) {
        CRef<CSeq_entry> tmp(new CSeq_entry);
        for (auto ent_iter: m_context.m_entry_template->GetSet().GetSeq_set()) {
            const CSeq_descr* descr = nullptr;
            if (ent_iter->IsSetDescr()) {
                descr = &ent_iter->GetDescr();
            }
            if (descr) {
                //tmp->Assign(**ent_iter);
                tmp->SetSeq().SetInst();
                // Copy any descriptors from the set to the sequence
                ITERATE(CBioseq_set::TDescr::Tdata, desc_iter, descr->Get())
                {
                    switch ((*desc_iter)->Which()) {
                    case CSeqdesc::e_Pub:
                    case CSeqdesc::e_Source:
                        break;
                    default:
                        continue;
                    }
                    CRef<CSeqdesc> desc(new CSeqdesc);
                    desc->Assign(**desc_iter);
                    tmp->SetSeq().SetDescr().Set().push_back(desc);
                }
                break;
            }
        }
 
        if (tmp->IsSetDescr() && !tmp->GetDescr().Get().empty())
            m_context.m_entry_template = tmp;
    }
 
    // incorporate any Seqdesc's that follow in the file
    if (!pObjIstrm->EndOfData()) {
        if (sType != CSeqdesc::GetTypeInfo()->GetName())
            sType = pObjIstrm->ReadFileHeader();
 
        while (sType == CSeqdesc::GetTypeInfo()->GetName()) {
            CRef<CSeqdesc> desc(new CSeqdesc);
            pObjIstrm->Read(ObjectInfo(*desc), CObjectIStream::eNoFileHeader);
 
            if (m_context.m_entry_template.Empty())
                m_context.m_entry_template.Reset(new CSeq_entry);
 
            {
                if (desc->IsUser() && desc->GetUser().IsDBLink()) {
                    CUser_object& user_obj = desc->SetUser();
                    if (user_obj.IsDBLink()) {
                        user_obj.SetData();
                    }
                }
            }
 
            m_context.m_entry_template->SetSeq().SetDescr().Set().push_back(desc);
 
            if (pObjIstrm->EndOfData())
                break;
 
            try {
                sType = pObjIstrm->ReadFileHeader();
            } catch (CEofException&) {
                break;
            }
        }
    }
 
#if 0
    if (m_context.m_submit_template->IsEntrys()) {
        // Take Seq-submit.sub.cit and put it in the Bioseq
        CRef<CPub> pub(new CPub);
        pub->SetSub().Assign(context.m_submit_template->GetSub().GetCit());
        CRef<CSeqdesc> pub_desc(new CSeqdesc);
        pub_desc->SetPub().SetPub().Set().push_back(pub);
        m_context.m_entry_template->SetSeq().SetDescr().Set().push_back(pub_desc);
    }
#endif
 
    if (m_context.m_entry_template.NotEmpty() && ! m_context.m_entry_template->IsSeq()) {
        throw runtime_error("The Seq-entry must be a Bioseq not a Bioseq-set.");
    }
 
    if (m_context.m_submit_template.NotEmpty()) {
        if (m_context.m_submit_template->IsSetSub() &&
            m_context.m_submit_template->GetSub().IsSetCit()) {
            CRef<CDate> date(new CDate(CTime(CTime::eCurrent), CDate::ePrecision_day));
            m_context.m_submit_template->SetSub().SetCit().SetDate(*date);
        }
    }
 
#if 0
    if (args["output-type"].AsString() == "Seq-entry") {
        // force Seq-entry by throwing out the Seq-submit
        m_context.m_submit_template.Reset( new CSeq_submit );
    }
#endif
}
 
namespace
{
    class AllowedDuplicates : public set<CSeqdesc_Base::E_Choice>
    {
    public:
        AllowedDuplicates()
        {
            insert(CSeqdesc_Base::e_User);
        }
    };
    AllowedDuplicates m_allowed_duplicates;
 
    template <typename _which>
    struct LocateWhich {
        typename _which::E_Choice compare_to;
        bool operator()(_which l) const
        {
            return l.Which() == compare_to;
        }
        bool operator()(const CRef<_which>& l) const
        {
            return l->Which() == compare_to;
        }
    };
}
 
void CMultiReader::MergeDescriptors(CSeq_descr& dest, const CSeq_descr& source) const
{
    ITERATE(CSeq_descr::Tdata, it, source.Get())
    {
        MergeDescriptors(dest, **it);
    }
}
 
void CMultiReader::MergeDescriptors(CSeq_descr & dest, const CSeqdesc & source) const
{
    bool duplicates = (m_allowed_duplicates.find(source.Which()) != m_allowed_duplicates.end());
 
    CAutoAddDesc desc(dest, source.Which());
    desc.Set(duplicates).Assign(source);
}
 
void CMultiReader::ApplyDescriptors(CSeq_entry& entry, const CSeq_descr& source) const
{
    MergeDescriptors(entry.SetDescr(), source);
    //g_ApplyDescriptors(source.Get(), entry);
}
 
namespace
{
    void CopyDescr(CSeq_entry& dest, const CSeq_entry& src) {
        if (src.IsSetDescr() && ! src.GetDescr().Get().empty()) {
            dest.SetDescr().Set().insert(dest.SetDescr().Set().end(),
                src.GetDescr().Get().begin(),
                src.GetDescr().Get().end());
        }
    }
    void CopyAnnot(CSeq_entry& dest, const CSeq_entry& src) {
        if (src.IsSetAnnot() && ! src.GetAnnot().empty()) {
            dest.SetAnnot().insert(dest.SetAnnot().end(),
                src.GetAnnot().begin(),
                src.GetAnnot().end());
        }
    }
}
 
void CMultiReader::LoadGFF3Fasta(istream& in, TAnnots& annots)
{
    g_LogDiagMessage(
        m_context.m_logger,
        EDiagSev::eDiag_Info,
        string("Recognized input file as format: ") + CFormatGuess::GetFormatName(CFormatGuess::eGff3));
 
    bool post_process = false;
    annots = xReadGFF3(in, post_process); // initializes m_gff3_reader!
    if (! AtSeqenceData()) {
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "Specified GFF3 file does not include any sequence data", 0);
    }
    x_PostProcessAnnots(annots);
}
 
 
void CMultiReader::LoadGFF3Fasta(istream& in, TAnnotMap& annotMap)
{
    TAnnots annots;
    LoadGFF3Fasta(in, annots);
    for (auto pAnnot : annots) {
        AddAnnotToMap(pAnnot, annotMap);
    }
}
 
 
CRef<CSerialObject> CMultiReader::FetchEntry(
    const CFormatGuess::EFormat& format,
    const string&                objectType,
    unique_ptr<istream>&         pIstr,
    TAnnotMap&                   annotMap)
{
    CRef<CSerialObject> pInputObject;
    switch (format) {
    case CFormatGuess::eBinaryASN:
        m_obj_stream.reset(CObjectIStream::Open(eSerial_AsnBinary, *pIstr.release(), eTakeOwnership));
        pInputObject = xReadASN1Binary(*m_obj_stream, objectType);
        break;
    case CFormatGuess::eTextASN:
        m_obj_stream.reset(CObjectIStream::Open(eSerial_AsnText, *pIstr.release(), eTakeOwnership));
        pInputObject = xReadASN1Text(*m_obj_stream);
        break;
    case CFormatGuess::eGff3:
        LoadGFF3Fasta(*pIstr, annotMap);
    case CFormatGuess::eFasta: // What about buffered input?
    default:
        m_iFlags = CFastaReader::fNoUserObjs;
        pInputObject = xReadFasta(*pIstr);
    }
 
    if (! pInputObject) {
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "File format not recognized", 0);
    }
    // RW-617: apply template descriptors only if input is *not* ASN1:
    // What about binary ASN.1?
    bool merge_template_descriptors = (format != CFormatGuess::eTextASN);
    return xApplyTemplate(pInputObject, merge_template_descriptors);
}
 
 
CFormatGuess::EFormat CMultiReader::OpenFile(const string& filename, CRef<CSerialObject>& input_sequence, TAnnotMap& annotMap)
{
    CFormatGuess::EFormat format;
    CFileContentInfo content_info;
    {
        unique_ptr<istream> istream(new CNcbiIfstream(filename));
        format = xInputGetFormat(*istream, &content_info);
    }
 
    switch (format)
    {
    case CFormatGuess::eBinaryASN:
        m_obj_stream.reset(CObjectIStream::Open(eSerial_AsnBinary, filename));
        input_sequence = xReadASN1Binary(*m_obj_stream, content_info.mInfoGenbank.mObjectType);
        break;
    case CFormatGuess::eTextASN:
        m_obj_stream.reset(CObjectIStream::Open(eSerial_AsnText, filename));
        input_sequence = xReadASN1Text(*m_obj_stream);
        break;
    case CFormatGuess::eGff3:
    {
        unique_ptr<istream> in(new CNcbiIfstream(filename));
        LoadGFF3Fasta(*in, annotMap);
        m_iFlags = 0;
        m_iFlags |= CFastaReader::fNoUserObjs;
        input_sequence = xReadFasta(*in);
    }
        break;
    default: // RW-616 - Assume FASTA
    {
        format = CFormatGuess::eFasta;
        m_iFlags = 0;
        m_iFlags |= CFastaReader::fNoUserObjs;
 
        CBufferedInput istream;
        istream.get().open(filename);
        input_sequence = xReadFasta(istream);
    }
        break;
    }
    if (input_sequence.Empty())
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "File format not recognized", 0);
    //rw-617: apply template descriptors only if input is *not* ASN1:
    bool merge_template_descriptors = (format != CFormatGuess::eTextASN);
    input_sequence = xApplyTemplate(input_sequence, merge_template_descriptors);
 
    return format;
}
 
void CMultiReader::GetSeqEntry(CRef<CSeq_entry>& entry, CRef<CSeq_submit>& submit, CRef<CSerialObject> obj)
{
    if (obj->GetThisTypeInfo() == CSeq_submit::GetTypeInfo()) {
        submit.Reset(static_cast<CSeq_submit*>(obj.GetPointer()));
        entry = submit->SetData().SetEntrys().front();
    } else if (obj->GetThisTypeInfo() == CSeq_entry::GetTypeInfo()) {
        entry.Reset(static_cast<CSeq_entry*>(obj.GetPointer()));
    }
}
 
CRef<CSerialObject> CMultiReader::xApplyTemplate(CRef<CSerialObject> obj, bool merge_template_descriptors) const
{
    CRef<CSeq_entry> entry;
    CRef<CSeq_submit> submit;
 
    GetSeqEntry(entry, submit, obj);
 
    if (entry.NotEmpty()) // &&
    {
        if (submit.Empty())
        if (entry->IsSet() && entry->GetSet().GetSeq_set().size() < 2 &&
            entry->GetSet().GetSeq_set().front()->IsSeq())
        {
            CRef<CSeq_entry> seq = entry->SetSet().SetSeq_set().front();
            CopyDescr(*seq, *entry);
            CopyAnnot(*seq, *entry);
            entry = seq;
        }
        entry->ResetParentEntry();
        entry->Parentize();
 
        if (merge_template_descriptors) {
            m_context.MergeWithTemplate(*entry);
        } else {
            if (m_context.m_t && m_context.m_logger) {
                string msg(
                    "Template file descriptors are ignored if input is ASN.1");
                m_context.m_logger->PutError(
                    *unique_ptr<CLineError>(
                        CLineError::Create(ILineError::eProblem_Unset,
                            eDiag_Warning, "", 0, "", "", "", msg)));
            }
        }
    }
 
    if (submit.Empty())
        return entry;
    else
        return submit;
}
 
CRef<CSerialObject> CMultiReader::ReadNextEntry()
{
    if (m_obj_stream)
        return xReadASN1Text(*m_obj_stream);
    else
        return CRef<CSerialObject>();
}
 
CMultiReader::TAnnots CMultiReader::xReadGFF3(CNcbiIstream& instream, bool post_process)
{
    int flags = 0;
    flags |= CGff3Reader::fGenbankMode;
    flags |= CGff3Reader::fRetainLocusIds;
    flags |= CGff3Reader::fGeneXrefs;
    flags |= CGff3Reader::fAllIdsAsLocal;
 
    CReaderListener readerListener;
    CGff3Reader reader(flags, m_AnnotName, m_AnnotTitle, CReadUtil::AsSeqId, &readerListener);
 
    CStreamLineReader lr(instream);
    TAnnots annots;
 
    try {
        reader.ReadSeqAnnots(annots, lr, m_context.m_logger);
        m_gff3_merger = reader.GetLocationMerger();
        mAtSequenceData = reader.AtSequenceData();
 
        if (post_process) {
            x_PostProcessAnnots(annots);
        }
 
        for (const auto& msg : readerListener) {
            m_context.m_logger->PutMessage(msg);
        }
    }
    catch (const CReaderMessage& msg) {
        m_context.m_logger->PutMessage(msg);
    }
 
    return annots;
}
 
 
static void s_RemoveGenBankDbxrefs(list<CRef<CSeq_feat>>& ftable)
{
    for (auto pFeat : ftable) {
        if (pFeat->IsSetDbxref()) {
            auto& dbxrefs = pFeat->SetDbxref();
            auto it = remove_if(dbxrefs.begin(), dbxrefs.end(),
                [](const CRef<CDbtag>& pDbtag) {
                return(pDbtag && pDbtag->IsSetDb() &&
                        NStr::EqualNocase(pDbtag->GetDb(), "GenBank"));
                });
            dbxrefs.erase(it, dbxrefs.end());
            if (dbxrefs.empty()) {
                pFeat->ResetDbxref();
            }
        }
    }
}
 
 
void CMultiReader::x_PostProcessAnnots(TAnnots& annots) const
{
    unsigned int startingLocusTagNumber = 1;
    unsigned int startingFeatureId = 1;
    for (auto it = annots.begin(); it != annots.end(); ++it) {
 
        auto& annot = **it;
        auto& data = annot.SetData();
        if (! data.IsFtable() || data.GetFtable().empty()) {
            continue; // all that follows applies to feature tables only
        }
 
        s_RemoveGenBankDbxrefs(data.SetFtable()); // RW-1861
 
        edit::CFeatTableEdit fte(
            annot, 0, m_context.m_locus_tag_prefix, startingLocusTagNumber, startingFeatureId, m_context.m_logger);
        //fte.InferPartials();
        fte.GenerateMissingParentFeatures(m_context.m_eukaryote, m_gff3_merger.get());
        if (m_context.m_locus_tags_needed) {
            if (m_context.m_locus_tag_prefix.empty() && !fte.AnnotHasAllLocusTags()) {
                NCBI_THROW(CArgException, eNoArg,
                    "GFF annotation requires locus tags, which are missing from one or more genes, so the command line argument -locus-tag-prefix is needed");
            }
            fte.GenerateLocusTags();
        }
        fte.GenerateProteinAndTranscriptIds();
        //fte.InstantiateProducts();
        fte.ProcessCodonRecognized();
        fte.EliminateBadQualifiers();
        fte.SubmitFixProducts();
 
        startingLocusTagNumber = fte.PendingLocusTagNumber();
        startingFeatureId = fte.PendingFeatureId();
    }
}
 
 
unique_ptr<CObjectIStream> CMultiReader::xCreateASNStream(const string& filename) const
{
    unique_ptr<istream> instream(new CNcbiIfstream(filename));
    return xCreateASNStream(CFormatGuess::eUnknown, instream);
}
 
unique_ptr<CObjectIStream> CMultiReader::xCreateASNStream(CFormatGuess::EFormat format, unique_ptr<istream>& instream) const
{
    // guess format
    ESerialDataFormat eSerialDataFormat = eSerial_None;
    {
        if (format == CFormatGuess::eUnknown)
            format = xInputGetFormat(*instream);
 
        switch(format) {
        case CFormatGuess::eBinaryASN:
            eSerialDataFormat = eSerial_AsnBinary;
            break;
        case CFormatGuess::eUnknown:
        case CFormatGuess::eTextASN:
            eSerialDataFormat = eSerial_AsnText;
            break;
        case CFormatGuess::eXml:
            eSerialDataFormat = eSerial_Xml;
            break;
        default:
            NCBI_USER_THROW_FMT(
                "Descriptor file seems to be in an unsupported format: "
                << CFormatGuess::GetFormatName(format) );
            break;
        }
 
        //instream.seekg(0);
    }
 
    unique_ptr<CObjectIStream> pObjIstrm(
        CObjectIStream::Open(eSerialDataFormat, *instream.release(), eTakeOwnership));
 
    return pObjIstrm;
}
 
CMultiReader::~CMultiReader()
{
}
 
class CAnnotationLoader
{
public:
    using TAnnots = CMultiReader::TAnnots;
 
    bool Init(const TAnnots& annots)
    {
        if (annots.empty()) {
            return false;
        }
 
        m_Annots         = annots;
        m_annot_iterator = m_Annots.begin();
        return true;
    }
 
    bool Init(const string& seqid_prefix, unique_ptr<istream>& instream, ILineErrorListener* logger)
    {
        m_seqid_prefix = seqid_prefix;
        m_line_reader = ILineReader::New(*instream, eTakeOwnership);
        instream.release();
        m_logger = logger;
        return true;
    }
 
    CRef<CSeq_annot> GetNextAnnot()
    {
        if (!m_Annots.empty()) {
            if (m_annot_iterator != m_Annots.end()) {
                return *m_annot_iterator++;
            }
        } else if (m_line_reader.NotEmpty()) {
            while (!m_line_reader->AtEOF()) {
                CRef<CSeq_annot> annot = CFeature_table_reader::ReadSequinFeatureTable(
                    *m_line_reader,
                    CFeature_table_reader::fLeaveProteinIds |
                    CFeature_table_reader::fCreateGenesFromCDSs |
                    CFeature_table_reader::fAllIdsAsLocal |
                    CFeature_table_reader::fPreferGenbankId,
                    m_logger, nullptr/*filter*/, m_seqid_prefix);
 
                if (annot.NotEmpty() && annot->IsSetData() && annot->GetData().IsFtable() &&
                    ! annot->GetData().GetFtable().empty()) {
                    return annot;
                }
            }
        }
        return CRef<CSeq_annot>();
    }
 
private:
    TAnnots m_Annots;
    TAnnots::iterator m_annot_iterator;
    string m_seqid_prefix;
    CRef<ILineReader> m_line_reader;
    ILineErrorListener* m_logger;
};
 
bool CMultiReader::xGetAnnotLoader(CAnnotationLoader& loader, const string& filename)
{
    unique_ptr<istream> in(new CNcbiIfstream(filename));
 
    CFormatGuess::EFormat uFormat = xAnnotGetFormat(*in);
 
    if (uFormat == CFormatGuess::eUnknown) {
        string ext;
        CDirEntry::SplitPath(filename, nullptr, nullptr, &ext);
        NStr::ToLower(ext);
        if (ext == ".gff" || ext == ".gff3")
            uFormat = CFormatGuess::eGff3;
        else if (ext == ".gtf")
            uFormat = CFormatGuess::eGtf;
        else if (ext == ".tbl")
            uFormat = CFormatGuess::eFiveColFeatureTable;
        else if (ext == ".asn" || ext == ".sqn" || ext == ".sap")
            uFormat = CFormatGuess::eTextASN;
 
        if (uFormat != CFormatGuess::eUnknown) {
            g_LogDiagMessage(
                m_context.m_logger,
                EDiagSev::eDiag_Info,
                string("Presuming annotation format by filename suffix: ") + CFormatGuess::GetFormatName(uFormat));
        }
    } else {
        g_LogDiagMessage(
            m_context.m_logger,
            EDiagSev::eDiag_Info,
            string("Recognized annotation format: ") + CFormatGuess::GetFormatName(uFormat));
    }
 
    TAnnots annots;
    switch (uFormat) {
    case CFormatGuess::eFiveColFeatureTable: {
        string seqid_prefix;
        if (!m_context.m_genome_center_id.empty())
            seqid_prefix = "gnl|" + m_context.m_genome_center_id + "|";
        return loader.Init(seqid_prefix, in, m_context.m_logger);
    } break;
    case CFormatGuess::eTextASN: {
        auto obj_stream = xCreateASNStream(uFormat, in);
        CRef<CSerialObject> obj = xReadASN1Text(*obj_stream);
        CRef<CSeq_submit> unused;
        CRef<CSeq_entry> pEntry;
        GetSeqEntry(pEntry, unused, obj);
        if (pEntry && pEntry->IsSetAnnot()) {
            annots = pEntry->GetAnnot();
        }
    } break;
    case CFormatGuess::eGff3:
        annots = xReadGFF3(*in, true);
        break;
    case CFormatGuess::eGtf:
    case CFormatGuess::eGffAugustus:
        annots = xReadGTF(*in);
        break;
#ifdef FLATFILE_PARSER_ENABLED
    case CFormatGuess::eFlatFileGenbank:
    case CFormatGuess::eFlatFileEna:
    case CFormatGuess::eFlatFileUniProt: {
        in.reset();
        auto pEntry = xReadFlatfile(uFormat, filename);
        if (pEntry && pEntry->IsSetAnnot()) {
            annots = pEntry->GetAnnot();
        }
    } break;
#endif
 
    default:
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "Annotation file format not recognized. Run format validator on your annotation file", 1);
    }
 
    if (! annots.empty()) {
        loader.Init(annots);
        return true;
    }
    return false;
}
 
 
static CRef<CSeq_id> s_GetAnnotId(const CSeq_annot& annot)
{
    CRef<CSeq_id> pAnnotId;
    if (annot.IsSetId()) {
        pAnnotId.Reset(new CSeq_id());
        const CAnnot_id& firstId = *(annot.GetId().front());
        if (firstId.IsLocal()) {
            pAnnotId->SetLocal().Assign(firstId.GetLocal());
        } else if (firstId.IsGeneral()) {
            pAnnotId->SetGeneral().Assign(firstId.GetGeneral());
        } else {
            return pAnnotId;
        }
    } else if (! annot.GetData().GetFtable().empty()) {
        // get a reference to CSeq_id instance, we'd need to update it recently
        // 5 column feature reader has a single shared instance for all features
        // update one at once would change all the features
        pAnnotId.Reset(const_cast<CSeq_id*>(annot.GetData().GetFtable().front()->GetLocation().GetId()));
    }
 
    return pAnnotId;
}
 
 
void CMultiReader::AddAnnotToMap(CRef<CSeq_annot> pAnnot, TAnnotMap& annotMap)
{
    auto pAnnotId = s_GetAnnotId(*pAnnot);
    if (! pAnnotId) {
        return;
    }
 
    auto idString = pAnnotId->GetSeqIdString();
    NStr::ToLower(idString);
    auto it = annotMap.find(idString);
    if (it == annotMap.end()) {
        annotMap.emplace(idString, list<CRef<CSeq_annot>>{ pAnnot });
    } else {
        it->second.push_back(pAnnot);
    }
}
 
 
void CMultiReader::LoadAnnotMap(const string& filename, TAnnotMap& annotMap)
{
    CAnnotationLoader annot_loader;
    if (! xGetAnnotLoader(annot_loader, filename)) {
        return;
    }
 
    CRef<CSeq_annot> pAnnot;
    while ((pAnnot = annot_loader.GetNextAnnot()).NotEmpty()) {
        AddAnnotToMap(pAnnot, annotMap);
    }
}
 
 
static CRef<CSeq_annot> s_GetBioseqAnnot(CBioseq& bioseq)
{
    CRef<CSeq_annot> pBioseqAnnot;
 
    if (bioseq.IsSetAnnot()) {
        auto& bioseqAnnots = bioseq.SetAnnot();
        auto it = find_if(bioseqAnnots.begin(),
                          bioseqAnnots.end(),
                          [](CRef<CSeq_annot> pAnnot) {
                              return (pAnnot && pAnnot->IsFtable());
                          });
        if (it != bioseqAnnots.end()) {
            pBioseqAnnot = *it;
        }
    }
    return pBioseqAnnot;
}
 
static void s_AddAnnotsToBioseq(
    list<CRef<CSeq_annot>>& annots,
    CBioseq& bioseq,
    CRef<CSeq_annot>& pBioseqAnnot)
{
    if (pBioseqAnnot) {
        for (auto pAnnot : annots) {
            objects::edit::CFeatTableEdit featEdit(*pBioseqAnnot);
            featEdit.MergeFeatures(pAnnot->SetData().SetFtable());
        }
        return;
    }
 
    pBioseqAnnot = s_GetBioseqAnnot(bioseq);
 
    if (! pBioseqAnnot) {
        pBioseqAnnot = annots.front();
        bioseq.SetAnnot().push_back(pBioseqAnnot);
        auto it = next(annots.begin());
        while (it != annots.end()) {
            objects::edit::CFeatTableEdit featEdit(*pBioseqAnnot);
            featEdit.MergeFeatures((*it)->SetData().SetFtable());
            ++it;
        }
    } else {
        for (auto pAnnot : annots) {
            objects::edit::CFeatTableEdit featEdit(*pBioseqAnnot);
            featEdit.MergeFeatures(pAnnot->SetData().SetFtable());
        }
    }
}
 
static bool s_HasPrefixMatch(
    const string& idString,
    CMultiReader::TAnnotMap& annotMap,
    map<string, CMultiReader::TAnnotMap::iterator>& matchMap)
{
    matchMap.clear();
    auto it = annotMap.lower_bound(idString);
    while (it != annotMap.end() && NStr::StartsWith(it->first, idString)) {
        matchMap.emplace(it->first, it);
        ++it;
    }
    return ! matchMap.empty();
}
 
bool CMultiReader::x_HasMatch(
    bool matchVersions,
    const string& idString,
    CMultiReader::TAnnotMap& annotMap,
    set<string>& matchedAnnots,
    list<CRef<CSeq_annot>>& annots) const
{
    if (matchVersions) {
        return x_HasExactMatch(idString, annotMap, matchedAnnots, annots);
    }
 
    bool hasMatch = false;
    map<string, TAnnotMap::iterator> matchMap;
    shared_lock<shared_mutex> sLock{m_Mutex};
    if (! s_HasPrefixMatch(idString, annotMap, matchMap)) {
        return false;
    }
    sLock.unlock();
    {
        unique_lock<shared_mutex> uLock{ m_Mutex };
        for (auto match : matchMap) {
            const auto& annotId = match.first;
            auto it = match.second;
            if (matchedAnnots.insert(annotId).second) {
                hasMatch = true;
                annots.splice(annots.end(), it->second);
                annotMap.erase(it);
            }
        }
    }
 
    return hasMatch;
}
 
bool CMultiReader::x_HasExactMatch(
    const string& idString,
    CMultiReader::TAnnotMap& annotMap,
    set<string>& matchedAnnots,
    list<CRef<CSeq_annot>>& annots) const
{
    shared_lock<shared_mutex> sLock{ m_Mutex };
    auto it = annotMap.find(idString);
    if (it == annotMap.end()) {
        return false;
    }
    string annotId = it->first;
    sLock.unlock();
 
    {
        unique_lock<shared_mutex> uLock{m_Mutex};
        if (matchedAnnots.insert(annotId).second) {
            annots = move(it->second);
            annotMap.erase(it);
            return true;
        }
    }
 
    return false;
}
 
 
void CMultiReader::AddAnnots(TAnnotMap& annotMap,
        set<string>& matchedAnnots,
        CBioseq& bioseq) const
{
    CRef<CSeq_annot> pBioseqAnnot;
    for (auto pSeqId : bioseq.GetId()) {
        list<CRef<CSeq_annot>> annots;
        bool hasMatch = false;
        bool matchVersions = (pSeqId->GetTextseq_Id() == nullptr);
        auto idString = pSeqId->GetSeqIdString();
        NStr::ToLower(idString);
        hasMatch = x_HasMatch(matchVersions, idString, annotMap, matchedAnnots, annots);
 
        if (! hasMatch &&
            pSeqId->IsGeneral() &&
            pSeqId->GetGeneral().IsSetDb() &&
            (pSeqId->GetGeneral().GetDb() == m_context.m_genome_center_id) &&
            pSeqId->GetGeneral().IsSetTag() && pSeqId->GetGeneral().GetTag().IsStr()) {
            matchVersions = true;
            idString = pSeqId->GetGeneral().GetTag().GetStr();
            NStr::ToLower(idString);
            hasMatch = x_HasMatch(matchVersions, idString, annotMap, matchedAnnots, annots);
        }
 
        if (! hasMatch) {
            continue;
        }
 
        for (auto pAnnot : annots) {
            auto pAnnotId = s_GetAnnotId(*pAnnot);
            g_ModifySeqIds(*pAnnot, *pAnnotId, pSeqId);
        }
 
        s_AddAnnotsToBioseq(annots, bioseq, pBioseqAnnot);
    }
}
 
 
CMultiReader::TAnnots CMultiReader::xReadGTF(CNcbiIstream& instream)
{
    int flags = 0;
    flags |= CGtfReader::fGenbankMode;
    flags |= CGtfReader::fAllIdsAsLocal;
    flags |= CGtfReader::fGenerateChildXrefs;
 
    CGtfReader reader(flags, m_AnnotName, m_AnnotTitle);
    CStreamLineReader lr(instream);
    TAnnots annots;
    try {
        reader.ReadSeqAnnots(annots, lr, m_context.m_logger);
        x_PostProcessAnnots(annots);
    } catch (CReaderMessage& msg) {
        m_context.m_logger->PutMessage(msg);
    }
 
    return annots;
}
 
#ifdef FLATFILE_PARSER_ENABLED
CRef<CSeq_entry> CMultiReader::xReadFlatfile(CFormatGuess::EFormat format, const string& filename)
{
    unique_ptr<Parser> pp(new Parser);
    switch (format)
    {
    case CFormatGuess::eFlatFileGenbank:
        pp->format = Parser::EFormat::GenBank;
        pp->source = Parser::ESource::GenBank;
        pp->seqtype = CSeq_id::e_Genbank;
        break;
    case CFormatGuess::eFlatFileEna:
        pp->format = Parser::EFormat::EMBL;
        pp->source = Parser::ESource::EMBL;
        pp->acprefix = ParFlat_EMBL_AC;
        pp->seqtype = CSeq_id::e_Embl;
        break;
    case CFormatGuess::eFlatFileUniProt:
        pp->format = Parser::EFormat::SPROT;
        pp->source = Parser::ESource::SPROT;
        pp->seqtype = CSeq_id::e_Swissprot;
        break;
    default:
        NCBI_THROW2(CObjReaderParseException, eFormat,
            "This flat file format is not supported: " + filename, 0);
        break;
    }
/*
#ifdef WIN32
    pp->ifp = fopen(filename.c_str(), "rb");
#else
    pp->ifp = fopen(filename.c_str(), "r");
#endif
*/
    pp->output_format = Parser::EOutput::BioseqSet;
 
    CFlatFileParser ffparser(m_context.m_logger);
    auto obj = ffparser.Parse(*pp, filename);
    if (obj.NotEmpty()) {
        if (obj->GetThisTypeInfo() == CBioseq_set::GetTypeInfo()) {
            auto bioseq_set = Ref(CTypeConverter<CBioseq_set>::SafeCast(obj.GetPointerOrNull()));
            auto entry = Ref(new CSeq_entry);
            entry->SetSeq();
            auto& annot = entry->SetAnnot();
            for (auto& bioseq : bioseq_set->SetSeq_set()) {
                if (bioseq->IsSetAnnot())
                    annot.splice(annot.end(), bioseq->SetAnnot());
            }
            if (entry->IsSetAnnot())
                return entry;
        }
    }
    return {};
}
#endif
 
END_NCBI_SCOPE