indexed_export.cpp

来自「ncbi源码」· C++ 代码 · 共 432 行

/*
 * ===========================================================================
 * PRODUCTION $Log: indexed_export.cpp,v $
 * PRODUCTION Revision 1000.5  2004/06/01 20:58:04  gouriano
 * PRODUCTION PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.10
 * PRODUCTION
 * ===========================================================================
 */

/*  $Id: indexed_export.cpp,v 1000.5 2004/06/01 20:58:04 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.
 *
 * ===========================================================================
 *
 * Authors:  Mike DiCuccio
 *
 * File Description:
 *    CDataPlugin_IndexedExport - save a sequence and all of its referenced components
 *                     into a directory
 */

#include <ncbi_pch.hpp>
#include "indexed_export.hpp"

#include <corelib/ncbiapp.hpp>
#include <corelib/ncbireg.hpp>

#include <gui/core/idocument.hpp>
#include <gui/core/plugin_utils.hpp>
#include <gui/core/version.hpp>
#include <gui/plugin/PluginCommandSet.hpp>
#include <gui/plugin/PluginInfo.hpp>
#include <gui/plugin/PluginValueConstraint.hpp>
#include <gui/utils/message_box.hpp>
#include <gui/dialogs/progress/progress_dlg_ex.hpp>

#include <objects/seqset/Seq_entry.hpp>
#include <objmgr/util/sequence.hpp>
#include <objmgr/bioseq_ci.hpp>
#include <objmgr/seq_map_ci.hpp>
#include <serial/objostrasn.hpp>


BEGIN_NCBI_SCOPE
USING_SCOPE(objects);



static const string sc_text_str  ("Text ASN.1");
static const string sc_binary_str("Binary ASN.1");
static const string sc_xml_str   ("XML");
static const string sc_fasta_str ("FastA");
static const string sc_flat_str  ("GenBank Flat-File");


void CDataPlugin_IndexedExport::GetInfo(CPluginInfo& info)
{
    info.Reset();

    // version info macro
    info.SetInfo(CPluginVersion::eMajor, CPluginVersion::eMinor, 0,
                 string(__DATE__) + " " + string(__TIME__),
                 "CDataPlugin_IndexedExport", "Indexed Directory",
                 "Save this record and all its components "
                 "into a directory that can be indexed", "");

    // command info
    CPluginCommandSet& cmds     = info.SetCommands();
    CPluginCommand& save_args   = cmds.AddDataCommand(eDataCommand_save);
    save_args.AddArgument("seqid", "Sequence", CSeq_id::GetTypeInfo());
    save_args.AddDefaultArgument("fmt", "Format", CPluginArg::eString,
                                 sc_binary_str);
    save_args.SetConstraint("fmt",
                            (*CPluginValueConstraint::CreateSet(),
                             sc_text_str,
                             sc_binary_str,
                             sc_xml_str,
                             sc_fasta_str
                             ));

    save_args.AddArgument("file", "File name", CPluginArg::eFile);
    save_args.AddArgument("alias", "Alias", CPluginArg::eString);
}


class CCancelButtonException : EXCEPTION_VIRTUAL_BASE public CException
{
public:
    // Enumerated list of document management errors
    enum EErrCode {
        eCancelPressed
    };

    // Translate the specific error code into a string representations of
    // that error code.
    virtual const char* GetErrCodeString(void) const
    {
        switch (GetErrCode()) {
        case eCancelPressed: return "eCancelPressed";
        default:             return CException::GetErrCodeString();
        }
    }

    NCBI_EXCEPTION_DEFAULT(CCancelButtonException, CException);
};


//
// Save()
// Save the information in a given document into a file
//
void CDataPlugin_IndexedExport::Save(CPluginMessage& msg)
{
    const CPluginCommand& args = msg.GetRequest().GetCommand();
    CPluginReply& reply = msg.SetReply();
    const IDocument& doc = *args["seqid"].GetDocument();
    const CSeq_id* id = dynamic_cast<const CSeq_id*> (&args["seqid"].AsObject());
    if ( !id ) {
        reply.SetStatus(eMessageStatus_failed);
        return;
    }

    CProgressDlgEx dlg;
    dlg.Show();
    try {
        string fmt_str   = args["fmt"].AsString();
        string file_base = args["file"].AsString();
        string alias     = args["alias"].AsString();

        string path;
        string base;
        string sub_path;
        {{
            CDirEntry::SplitPath(file_base, &path, &base);

            sub_path = path;
            sub_path += base + ".data";
            sub_path += CDirEntry::GetPathSeparator();
            CDir dir(sub_path);
            if ( !dir.Exists() ) {
                dir.Create();
            }
        }}


        // find out what object to serialize
        // in general, we serialize precisely what the document contains
        // the exception to this is that we dereference any seq-id
        // to its largest component (= top-level seq-entry)
        CConstRef<CSerialObject> obj
            (dynamic_cast<const CSerialObject*> (doc.GetObject()));
        if (obj.GetPointer()  &&
            obj->GetThisTypeInfo() == CSeq_id::GetTypeInfo()) {
            const CSeq_id& id =
                dynamic_cast<const CSeq_id&>(*doc.GetObject());
            CBioseq_Handle handle = doc.GetScope().GetBioseqHandle(id);
            obj.Reset(&handle.GetTopLevelSeqEntry());
        }

        //
        // format handling
        //

        EFormat fmt = eFormat_text_asn;
        if (fmt_str == sc_text_str) {
            // ASN.1 text
            fmt = eFormat_text_asn;
        } else if (fmt_str == sc_binary_str) {
            // ASN.1 binary
            fmt = eFormat_binary_asn;
        } else if (fmt_str == sc_xml_str) {
            // XML
            fmt = eFormat_xml;
        } else if (fmt_str == sc_fasta_str) {
            // FastA
            fmt = eFormat_fasta;
        }

        try {
            //
            // first, save the sequence of interest
            //
            set< CConstRef<CSeq_entry> > entries;
            list< CConstRef<CSeq_entry> > entries_for_products;
            dlg.SetMessage("Saving document...");
            CScope& scope = doc.GetScope();
            CBioseq_Handle handle = scope.GetBioseqHandle(*id);
            x_Save(scope, handle.GetTopLevelSeqEntry(), fmt, file_base);
            {{
                CConstRef<CSeq_entry> ref(&handle.GetTopLevelSeqEntry());
                entries.insert(ref);
                entries_for_products.push_back(ref);
            }}

            //
            // now, we iterate all sequences referred to by this sequence
            //
            int count = 0;
            const CSeqMap& seqmap = handle.GetSeqMap();
            CSeqMap::const_iterator map_iter =
                seqmap.begin_resolved(&scope, -1, CSeqMap::fFindRef);
            for ( ;  map_iter;  ++map_iter) {
                if (map_iter.GetType() != CSeqMap::eSeqRef) {
                    continue;
                }

                CSeq_id_Handle idh = map_iter.GetRefSeqid();
                dlg.SetMessage("Retrieving sequence: " + idh.AsString());
                LOG_POST(Info << "retrieving " << idh.AsString()); 
                CBioseq_Handle h = scope.GetBioseqHandle(idh);
                const CSeq_entry& entry = h.GetTopLevelSeqEntry();

                CConstRef<CSeq_entry> ref(&entry);
                if ( !entries.insert(ref).second ) {
                    continue;
                }
                entries_for_products.push_back(ref);

                dlg.SetMessage("Saving sequence: " + idh.AsString());
                LOG_POST(Info << "saving " << idh.AsString()); 

                string file_name =
                    sub_path + base + "." + NStr::IntToString(++count);
                x_Save(scope, entry, fmt, file_name);
                Fl::check();

                if ( !dlg.IsShown() ) {
                    NCBI_THROW(CCancelButtonException,
                               eCancelPressed, "");
                }
            }

            //
            // now, iterate all products in all seq-entries
            //
            ITERATE (list< CConstRef<CSeq_entry> >, iter, entries_for_products) {
                CTypeConstIterator<CSeq_feat> feat_iter(**iter);
                for ( ; feat_iter;  ++feat_iter) {
                    const CSeq_feat& feat = *feat_iter;

                    if ( !feat.IsSetProduct() ) {
                        continue;
                    }

                    string id_label;
                    sequence::GetId(feat.GetProduct()).GetLabel(&id_label);
                    dlg.SetMessage("Retrieving product sequence: " + id_label);
                    LOG_POST(Info << "retrieving " << id_label); 
                    CBioseq_Handle h = scope.GetBioseqHandle(feat.GetLocation());
                    const CSeq_entry& entry = h.GetTopLevelSeqEntry();

                    CConstRef<CSeq_entry> ref(&entry);
                    if ( !entries.insert(ref).second ) {
                        continue;
                    }
                    entries_for_products.push_back(ref);

                    dlg.SetMessage("Saving product sequence: " + id_label);
                    LOG_POST(Info << "saving " << id_label); 

                    string file_name =
                        sub_path + base + "." + NStr::IntToString(++count);
                    x_Save(scope, entry, fmt, file_name);
                    Fl::check();

                    if ( !dlg.IsShown() ) {
                        NCBI_THROW(CCancelButtonException,
                                   eCancelPressed, "");
                    }
                }
            }
        }
        catch (CCancelButtonException&) {
        }

        //
        // configure this path to be indexed at app start-up
        //

        CNcbiApplication* app = CNcbiApplication::Instance();
        _ASSERT(app);

        CNcbiRegistry& reg = app->GetConfig();

        string section_name = "LDS";
        {{
            list<string> sections;
            reg.EnumerateSections(&sections);

            size_t count = 0;
            ITERATE (list<string>, iter, sections) {
                if (iter->find("LDS") == 0) {
                    ++count;
                }
            }

            section_name += NStr::IntToString(count + 1);
        }}

        reg.Set(section_name, "Path",  path, CNcbiRegistry::ePersistent);
        reg.Set(section_name, "Alias", alias, CNcbiRegistry::ePersistent);

        reply.SetStatus(eMessageStatus_success);
    }
    catch (CException& e) {
        string msg("Failed to save file:\n");
        msg += e.GetMsg();
        NcbiMessageBox(msg);
        reply.SetStatus(eMessageStatus_failed);
    }
#ifndef _DEBUG
    catch (...) {
        NcbiMessageBox("Failed to save file:\nUnknown error\n");
        reply.SetStatus(eMessageStatus_failed);
    }
#endif
}


void CDataPlugin_IndexedExport::x_Save(CScope& scope, const CSeq_entry& entry,
                            EFormat fmt, const string& file_name)
{
    CNcbiOfstream ostr(file_name.c_str(), ios::binary);
    switch (fmt) {
    case eFormat_text_asn:
        {{
            auto_ptr<CObjectOStream> os
                (CObjectOStream::Open(eSerial_AsnText, ostr));
            *os << entry;
        }}
        break;

    case eFormat_binary_asn:
        {{
            auto_ptr<CObjectOStream> os
                (CObjectOStream::Open(eSerial_AsnBinary, ostr));
            *os << entry;
        }}
        break;

    case eFormat_xml:
        {{
            auto_ptr<CObjectOStream> os
                (CObjectOStream::Open(eSerial_Xml, ostr));
            *os << entry;
        }}
        break;

    case eFormat_fasta:
        {{
            CFastaOstream fasta_ostr(ostr);
            CBioseq_CI bioseq_iter(scope, entry);
            for ( ;  bioseq_iter;  ++bioseq_iter) {
                fasta_ostr.Write(*bioseq_iter);
            }
        }}
        break;
    }
}


END_NCBI_SCOPE


/*
 * ===========================================================================
 * $Log: indexed_export.cpp,v $
 * Revision 1000.5  2004/06/01 20:58:04  gouriano
 * PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.10
 *
 * Revision 1.10  2004/05/25 17:21:59  dicuccio
 * Modified class names.  Fonts to 12 point
 *
 * Revision 1.9  2004/05/21 22:27:48  gorelenk
 * Added PCH ncbi_pch.hpp
 *
 * Revision 1.8  2004/03/11 17:43:41  dicuccio
 * Use new file chooser dialog
 *
 * Revision 1.7  2004/02/05 18:05:07  dicuccio
 * Added better (GUI) handling of sequence download.  Added saving of products as
 * well as main sequence.
 *
 * Revision 1.6  2004/01/27 18:44:25  dicuccio
 * Added progress dialog.  Added missing headers.  Added parameter handling of
 * fiel to save to.  CHanged directory layout of indexed structure to show master
 * record in top-level directory
 *
 * Revision 1.5  2003/12/05 13:07:54  dicuccio
 * Split management interface into a separate plugin.  Fixed linker error
 * introduced with status bar
 *
 * Revision 1.4  2003/11/24 15:45:37  dicuccio
 * Renamed CVersion to CPluginVersion
 *
 * Revision 1.3  2003/11/06 20:12:15  dicuccio
 * Cleaned up handling of USING_SCOPE - removed from all headers
 *
 * Revision 1.2  2003/11/04 17:49:25  dicuccio
 * Changed calling parameters for plugins - pass CPluginMessage instead of paired
 * CPluginCommand/CPluginReply
 *
 * Revision 1.1  2003/10/07 13:41:13  dicuccio
 * Added indexed export plugin
 *
 * Revision 1.1  2003/09/16 14:06:48  dicuccio
 * Initial revision - split from CFileLoader
 *
 * ===========================================================================
 */