lds_object.cpp

来自「ncbi源码」· C++ 代码 · 共 666 行 · 第 1/2 页

                // Recursively save the parent
                top_level_id = SaveObject(file_id, sniffer, top_obj_info);
            } else {
                top_level_id = top_obj_info->ext_id;
            }

        }}

    }

    const string& type_name = obj_info->info.GetTypeInfo()->GetName();

    map<string, int>::const_iterator it = m_ObjTypeMap.find(type_name);
    if (it == m_ObjTypeMap.end()) {
        LOG_POST(Info << "Unrecognized type: " << type_name);
        return 0;                
    }
    int type_id = it->second;


    string id_str;
    string molecule_title;

    ++m_MaxObjRecId;

    bool is_object = IsObject(*obj_info, &id_str, &molecule_title);
    if (is_object) {
        string all_seq_id; // Space separated list of seq_ids
        const CBioseq* bioseq = CType<CBioseq>().Get(obj_info->info);
        if (bioseq) {
            const CBioseq::TId&  id_list = bioseq->GetId();
            ITERATE(CBioseq::TId, it, id_list) {
                const CSeq_id* seq_id = *it;
                if (seq_id) {
                    all_seq_id.append(seq_id->AsFastaString());
                    all_seq_id.append(" ");
                }
            }
        }

        m_db.object_db.primary_seqid = NStr::ToUpper(id_str);

        obj_info->ext_id = m_MaxObjRecId; // Keep external id for the next scan
        EBDB_ErrCode err;
/*
        m_db.object_attr_db.object_attr_id = m_MaxObjRecId;
        m_db.object_attr_db.object_title = molecule_title;
        m_db.object_attr_db.seq_ids = NStr::ToUpper(all_seq_id);
        EBDB_ErrCode err = m_db.object_attr_db.Insert();
        BDB_CHECK(err, "LDS::ObjectAttr");
*/
        m_db.object_db.object_id = m_MaxObjRecId;
        m_db.object_db.file_id = file_id;
        m_db.object_db.seqlist_id = 0;  // TODO:
        m_db.object_db.object_type = type_id;
        m_db.object_db.file_offset = obj_info->offset;
//        m_db.object_db.object_attr_id = m_MaxObjRecId; 
        m_db.object_db.TSE_object_id = top_level_id;
        m_db.object_db.parent_object_id = parent_id;
        m_db.object_db.object_title = molecule_title;
        m_db.object_db.seq_ids = NStr::ToUpper(all_seq_id);


//        LOG_POST(Info << "Saving object: " << type_name << " " << id_str);

        err = m_db.object_db.Insert();
        BDB_CHECK(err, "LDS::Object");

    } else {
                
        // Set of seq ids referenced in the annotation
        //
        set<string> ref_seq_ids;

        // Check for alignment in annotation
        //
        const CSeq_annot* annot = 
            CType<CSeq_annot>().Get(obj_info->info);
        if (annot && annot->CanGetData()) {
            const CSeq_annot_Base::C_Data& adata = annot->GetData();
            if (adata.Which() == CSeq_annot_Base::C_Data::e_Align) {
                const CSeq_annot_Base::C_Data::TAlign& al_list =
                                            adata.GetAlign();
                ITERATE (CSeq_annot_Base::C_Data::TAlign, it, al_list){
                    if (!(*it)->CanGetSegs())
                        continue;

                    const CSeq_align::TSegs& segs = (*it)->GetSegs();
                    switch (segs.Which())
                    {
                    case CSeq_align::C_Segs::e_Std:
                        {
                        const CSeq_align_Base::C_Segs::TStd& std_list =
                                                    segs.GetStd();
                        ITERATE (CSeq_align_Base::C_Segs::TStd, it2, std_list) {
                            const CRef<CStd_seg>& seg = *it2;
                            const CStd_seg::TIds& ids = seg->GetIds();

                            ITERATE(CStd_seg::TIds, it3, ids) {
                                ref_seq_ids.insert((*it3)->AsFastaString());

                            } // ITERATE

                        } // ITERATE
                        }
                        break;
                    case CSeq_align::C_Segs::e_Denseg:
                        {
                        const CSeq_align_Base::C_Segs::TDenseg& denseg =
                                                        segs.GetDenseg();
                        const CDense_seg::TIds& ids = denseg.GetIds();

                        ITERATE (CDense_seg::TIds, it3, ids) {
                            ref_seq_ids.insert((*it3)->AsFastaString());
                        } // ITERATE
                        
                        }
                        break;
                    case CSeq_align::C_Segs::e_Packed:
                    case CSeq_align::C_Segs::e_Disc:
                        break;
                    }

                } // ITERATE
            }
        }

        // Save all seq ids referred by the alignment
        //
        ITERATE (set<string>, it, ref_seq_ids) {
            m_db.seq_id_list.object_id = m_MaxObjRecId;
            m_db.seq_id_list.seq_id = it->c_str();

            EBDB_ErrCode err = m_db.seq_id_list.Insert();
            BDB_CHECK(err, "LDS::seq_id_list");
        }
        
        obj_info->ext_id = m_MaxObjRecId; // Keep external id for the next scan
                
        m_db.annot_db.annot_id = m_MaxObjRecId;
        m_db.annot_db.file_id = file_id;
        m_db.annot_db.annot_type = type_id;
        m_db.annot_db.file_offset = obj_info->offset;

        LOG_POST(Info << "Saving annotation: " 
                      << type_name 
                      << " " 
                      << id_str
                      << " " 
                      << (const char*)(!top_level_id ? "Top Level. " : " ")
                      << "offs=" 
                      << obj_info->offset
                      );

        EBDB_ErrCode err = m_db.annot_db.Insert();
        BDB_CHECK(err, "LDS::Annotation");

        m_db.annot2obj_db.object_id = parent_id;
        m_db.annot2obj_db.annot_id = m_MaxObjRecId;

        err = m_db.annot2obj_db.Insert();
        BDB_CHECK(err, "LDS::Annot2Obj");

    }

    return obj_info->ext_id;
}


bool CLDS_Object::IsObject(const CLDS_CoreObjectsReader::SObjectDetails& parse_info,
                           string* object_str_id,
                           string* object_title)
{
    *object_title = "";
    *object_str_id = "";

    if (parse_info.is_top_level) {

        CSeq_entry* seq_entry = CType<CSeq_entry>().Get(parse_info.info);

        if (seq_entry) {
            m_TSE_Manager = new CObjectManager;
            m_Scope = new CScope(*m_TSE_Manager);

            m_Scope->AddTopLevelSeqEntry(*seq_entry);
            return true;
        } else {
            CBioseq* bioseq = CType<CBioseq>().Get(parse_info.info);
            if (bioseq) {
                m_TSE_Manager = new CObjectManager;
                m_Scope = new CScope(*m_TSE_Manager);

                m_Scope->AddBioseq(*bioseq);
            }
        }
    }

    const CBioseq* bioseq = CType<CBioseq>().Get(parse_info.info);
    if (bioseq) {
        const CSeq_id* seq_id = bioseq->GetFirstId();
        if (seq_id) {
            *object_str_id = seq_id->AsFastaString();
        } else {
            *object_str_id = "";
        }

        if (m_Scope) { // we are under OM here
            CBioseq_Handle bio_handle = m_Scope->GetBioseqHandle(*bioseq);
            if (bio_handle) {
                *object_title = sequence::GetTitle(bio_handle);
                //LOG_POST(Info << "object title: " << *object_title);
            } else {
                // the last resort
                bioseq->GetLabel(object_title, CBioseq::eBoth);
            }
            
        } else {  // non-OM controlled object
            bioseq->GetLabel(object_title, CBioseq::eBoth);
        }
    } else {
        const CSeq_annot* annot = 
            CType<CSeq_annot>().Get(parse_info.info);
        if (annot) {
            *object_str_id = "";
            return false;
        } else {
            const CSeq_align* align =  
                CType<CSeq_align>().Get(parse_info.info);
            if (align) {
                *object_str_id = "";
                return false;
            }
        }
    }
    return true;
}


int CLDS_Object::FindMaxObjRecId()
{
    if (m_MaxObjRecId) {
        return m_MaxObjRecId;
    }

    LDS_GETMAXID(m_MaxObjRecId, m_db.object_db, object_id);

    int ann_rec_id = 0;
    LDS_GETMAXID(ann_rec_id, m_db.annot_db, annot_id);

    if (ann_rec_id > m_MaxObjRecId) {
        m_MaxObjRecId = ann_rec_id;
    }

    return m_MaxObjRecId;
}


END_SCOPE(objects)
END_NCBI_SCOPE

/*
 * ===========================================================================
 * $Log: lds_object.cpp,v $
 * Revision 1000.2  2004/06/01 19:46:06  gouriano
 * PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.20
 *
 * Revision 1.20  2004/05/21 21:42:55  gorelenk
 * Added PCH ncbi_pch.hpp
 *
 * Revision 1.19  2004/03/11 18:45:04  kuznets
 * Get correct title for standalone Bioseqs using object manager
 *
 * Revision 1.18  2004/03/09 17:16:59  kuznets
 * Merge object attributes with objects
 *
 * Revision 1.17  2003/10/09 16:48:01  kuznets
 * More LDS logging when parsing files + minor bug fix.
 *
 * Revision 1.16  2003/10/07 20:46:57  kuznets
 * Added diagnostics when some file has been recognized as a file without
 * objects.
 *
 * Revision 1.15  2003/08/21 12:10:04  dicuccio
 * Minor code reformatting
 *
 * Revision 1.14  2003/07/29 19:51:49  kuznets
 * Reflecting changes in the header file
 *
 * Revision 1.13  2003/07/14 19:45:24  kuznets
 * More detailed LOG_POST message
 *
 * Revision 1.12  2003/07/09 19:30:51  kuznets
 * Implemented collection of sequence ids from alignments.
 *
 * Revision 1.11  2003/07/03 19:23:37  kuznets
 * Added recognition of denseg alignments.
 *
 * Revision 1.10  2003/07/02 12:07:42  dicuccio
 * Fix for implicit conversion/assignment in gcc
 *
 * Revision 1.9  2003/07/01 19:27:06  kuznets
 * Added code fragment reading sequence ids from an alignment.
 *
 * Revision 1.8  2003/06/26 16:22:15  kuznets
 * Uppercased all sequence ids before writing into the database.
 *
 * Revision 1.7  2003/06/16 16:24:43  kuznets
 * Fixed #include paths (lds <-> lds_admin separation)
 *
 * Revision 1.6  2003/06/16 15:40:21  kuznets
 * Fixed a bug with collecting all seq_ids from a bioseq
 *
 * Revision 1.5  2003/06/13 16:00:30  kuznets
 * Improved work with sequence ids. Now it keeps all sequence ids bioseq has
 *
 * Revision 1.4  2003/06/10 19:00:32  kuznets
 * Code clean-up
 *
 * Revision 1.3  2003/06/06 20:03:54  kuznets
 * Reflecting new location of fasta reader
 *
 * Revision 1.2  2003/06/04 16:38:45  kuznets
 * Implemented OM-based bioseq title extraction (should work better than
 * CBioseq::GetTitle())
 *
 * Revision 1.1  2003/06/03 14:13:25  kuznets
 * Initial revision
 *
 *
 * ===========================================================================
 */