pattern.cpp

ncbi源码

/*
 * ===========================================================================
 * PRODUCTION $Log: pattern.cpp,v $
 * PRODUCTION Revision 1000.6  2004/06/01 20:55:24  gouriano
 * PRODUCTION PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.33
 * PRODUCTION
 * ===========================================================================
 */

/*  $Id: pattern.cpp,v 1000.6 2004/06/01 20:55:24 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:  Josh Cherry
 *
 * File Description:  gbench plugin for finding regular expressions in sequences
 *
 */


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

#include <algo/sequence/find_pattern.hpp>
#include <gui/core/plugin_utils.hpp>
#include <gui/core/version.hpp>
#include <gui/dialogs/col/multi_col_dlg.hpp>
#include <gui/plugin/PluginCommandSet.hpp>
#include <gui/plugin/PluginInfo.hpp>
#include <gui/plugin/PluginRequest.hpp>
#include <gui/plugin/PluginValueConstraint.hpp>
#include <gui/objutils/utils.hpp>
#include <objmgr/seq_vector.hpp>
#include <objmgr/util/sequence.hpp>


BEGIN_NCBI_SCOPE
USING_SCOPE(objects);



CAlgoPlugin_Pattern::~CAlgoPlugin_Pattern()
{
}


// standard plugin announce bopilerplate
void CAlgoPlugin_Pattern::GetInfo(CPluginInfo& info)
{
    info.Reset();
    
    // version info macro
    info.SetInfo(CPluginVersion::eMajor, CPluginVersion::eMinor, 0,
                 string(__DATE__) + " " + string(__TIME__),
                 "CAlgoPlugin_Pattern",
                 "Search/Regular Expression Sequence Search",
                 "Find all non-overlapping occurrences of a pattern",
                 "");

    // command info
    CPluginCommandSet& cmds = info.SetCommands();
    CPluginCommand&    args = cmds.AddAlgoCommand(eAlgoCommand_run);
    args.AddArgument("locs", "Locations to evaluate",
                     CSeq_loc::GetTypeInfo(),
                     CPluginArg::TData::e_Array);
    args.AddArgument("pattern", "Pattern to find",
                     CPluginArg::eString);
    args.AddArgument("label", "Label for matches",
                     CPluginArg::eString);
}


void CAlgoPlugin_Pattern::RunCommand(CPluginMessage& msg)
{
    const CPluginCommand& args = msg.GetRequest().GetCommand();
    CPluginReply& reply = msg.SetReply();
    _TRACE("CAlgoPlugin_Pattern::Run()");

    // pattern (r.e.) to find
    string pattern(args["pattern"].AsString());
    // label to appear in feature table
    string label(args["label"].AsString());

    if ( !m_Dialog.get() ) {
        m_Dialog.reset(new CMultiColDlg());

        m_Dialog->SetWindowSize(600, 350);
        m_Dialog->SetTitle("Pattern Search Results");

        m_Dialog->SetColumn(0, "Sequence");
        m_Dialog->SetColumn(1, "Location");
        m_Dialog->SetColumn(2, "Position", FL_ALIGN_CENTER);
        m_Dialog->SetColumn(3, "Matched Sequence", FL_ALIGN_LEFT, 2.0f);
    }

    m_Dialog->SetLabel(string("A search for the pattern\n") + pattern +
                       string("\nproduced:"));

    // clear any previous contents
    m_Dialog->SetRows(0);

    vector<TSeqPos> starts;
    vector<TSeqPos> ends;

    //
    // first, evaluate whole sequences
    //
    int row = 0;
    plugin_args::TLocList locs;
    GetArgValue(args["locs"], locs);
    ITERATE (plugin_args::TLocList, iter, locs) {
        const CSeq_loc&  loc = *iter->second;
        const IDocument& doc = *iter->first;

        // find the best ID for this bioseq
        try {
            CBioseq_Handle handle = doc.GetScope().GetBioseqHandle(loc);
            CSeqVector vec =
                handle.GetSequenceView(loc,
                                       CBioseq_Handle::eViewConstructed,
                                       CBioseq_Handle::eCoding_Iupac);

            string seq;
            vec.GetSeqData( (TSeqPos) 0, vec.size(), seq );

            CFindPattern::Find(seq, pattern, starts, ends);

            string& id_str  = m_Dialog->SetCell(row, 0);
            string& loc_str = m_Dialog->SetCell(row, 1);

            const CSeq_id& best_id =
                sequence::GetId(handle, sequence::eGetId_Best);
            id_str.erase();
            best_id.GetLabel(&id_str);
            loc_str = CPluginUtils::GetLabel(loc, &doc.GetScope());

            // preallocate rows in dialog for speed
            m_Dialog->SetRows(row + starts.size());

            for( unsigned int k = 0;  k < starts.size();  k++) {
                string& pos_str = m_Dialog->SetCell(row, 2);
                // 1-based indexing for dialog
                pos_str = NStr::IntToString(starts[k] + 1) + " - "
                    + NStr::IntToString(ends[k] + 1);

                m_Dialog->SetCell(row, 3) =
                    seq.substr(starts[k], ends[k] - starts[k] + 1);
                ++row;
            }

            //
            // add a feature table to doc
            //

            CRef<CSeq_annot> annot(new CSeq_annot());

            for( unsigned int k = 0;  k < starts.size();  k++) {
                // create feature
                CRef<CSeq_feat> feat(new CSeq_feat());

                // set correct location
                {{
                     CSeq_loc& floc = feat->SetLocation();
                     floc.SetInt().SetId().Assign(sequence::GetId(loc));
                     floc.SetInt().SetFrom(starts[k]);
                     floc.SetInt().SetTo(ends[k]);
                     // assume plus strand for now
                     floc.SetInt().SetStrand(eNa_strand_plus);

                     CRef<CSeq_loc> new_loc =
                         CSeqUtils::RemapChildToParent(loc, floc);
                     feat->SetLocation(*new_loc);
                 }}

                // set feature data
                feat->SetData().SetRegion() = label;
    
                // save in annot
                annot->SetData().SetFtable().push_back(feat);
            }
            
            // add description to annot
            annot->AddName("Pattern Matches");
            string comment("Matches to pattern \'");
            comment += pattern + "\'";
            annot->AddComment(comment);

            reply.AddObject(doc, *annot);
            // attach annot to doc
            //const_cast<IDocument&>(doc).AttachAnnot(*annot);
        }
        catch (CException& e) {
            string str = CPluginUtils::GetLabel(loc, &doc.GetScope());
            LOG_POST(Error << "Error processing location " << str
                     << ": " << e.what());
        }
#ifndef _DEBUG
        catch (...) {
            string str = CPluginUtils::GetLabel(loc, &doc.GetScope());
            LOG_POST(Error << "Error processing location " << str);
        }
#endif
    }

    // update all views
    //CDocManager::UpdateAllViews();

    //
    // prepare our dialog box
    //
    m_Dialog->Show();
    reply.AddAction(CPluginReplyAction::e_Add_to_document);
    reply.SetStatus(eMessageStatus_success);
}


END_NCBI_SCOPE


/*
 * ===========================================================================
 * $Log: pattern.cpp,v $
 * Revision 1000.6  2004/06/01 20:55:24  gouriano
 * PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.33
 *
 * Revision 1.33  2004/05/21 22:27:47  gorelenk
 * Added PCH ncbi_pch.hpp
 *
 * Revision 1.32  2004/05/03 13:05:42  dicuccio
 * gui/utils --> gui/objutils where needed
 *
 * Revision 1.31  2004/04/01 14:14:02  lavr
 * Spell "occurred", "occurrence", and "occurring"
 *
 * Revision 1.30  2004/03/05 17:35:37  dicuccio
 * Use sequence::GetId() instead of CSeq_id::GetStringDescr()
 *
 * Revision 1.29  2004/02/13 15:08:50  mjohnson
 * Removed local URL for plugin help.
 *
 * Revision 1.28  2004/01/27 18:37:53  dicuccio
 * Code clean-up.  Use standard names for plugins.  Removed unnecessary #includes
 *
 * Revision 1.27  2004/01/07 15:50:37  dicuccio
 * Adjusted for API change in CPluginUtils::GetLabel().  Standardized exception
 * reporting in algorithms.
 *
 * Revision 1.26  2003/12/17 17:58:50  jcherry
 * Set dialog label every time
 *
 * Revision 1.25  2003/12/16 22:25:09  jcherry
 * Added plugin for searching against patterns loaded from file
 *
 * Revision 1.24  2003/12/15 20:16:08  jcherry
 * Changed CFindPattern::Find to take a string rather than a CSeqVector
 *
 * Revision 1.23  2003/12/15 19:53:05  jcherry
 * Preallocate rows in dialog for speed
 *
 * Revision 1.22  2003/11/24 15:45:26  dicuccio
 * Renamed CVersion to CPluginVersion
 *
 * Revision 1.21  2003/11/18 17:48:37  dicuccio
 * Added standard processing of return values
 *
 * Revision 1.20  2003/11/14 00:20:29  jcherry
 * Added url for help
 *
 * Revision 1.19  2003/11/06 20:12:12  dicuccio
 * Cleaned up handling of USING_SCOPE - removed from all headers
 *
 * Revision 1.18  2003/11/04 17:49:23  dicuccio
 * Changed calling parameters for plugins - pass CPluginMessage instead of paired
 * CPluginCommand/CPluginReply
 *
 * Revision 1.17  2003/10/27 17:46:48  dicuccio
 * Removed dead #includes
 *
 * Revision 1.16  2003/10/14 16:24:37  dicuccio
 * Correctly remap new feature locations through the parent location to the master
 * sequence
 *
 * Revision 1.15  2003/10/07 13:47:00  dicuccio
 * Renamed CPluginURL* to CPluginValue*
 *
 * Revision 1.14  2003/09/25 17:21:35  jcherry
 * Added name to annot
 *
 * Revision 1.13  2003/09/04 14:05:24  dicuccio
 * Use IDocument instead of CDocument
 *
 * Revision 1.12  2003/09/03 14:46:53  rsmith
 * change namespace name from args to plugin_args to avoid clashes with variable names.
 *
 * Revision 1.11  2003/08/21 12:03:07  dicuccio
 * Make use of new typedef in plugin_utils.hpp for argument values.
 *
 * Revision 1.10  2003/08/11 18:03:53  jcherry
 * Fixed to use 0-based indexing for feature table (but retain
 * 1-based indexing for dialog box)
 *
 * Revision 1.9  2003/08/05 17:03:55  dicuccio
 * Made multi-column output dialog a member variable - allows non-modal operation
 *
 * Revision 1.8  2003/07/28 11:51:48  dicuccio
 * Rewrote CTablePanel<> to be more flexible and better contained.  Added standard
 * multicolumn list dialog.  Deprecated use of COutputm_Dialog->
 *
 * Revision 1.7  2003/07/22 15:32:16  dicuccio
 * Changed to make use of new API in plugin_utils.hpp - GetArgValue()
 *
 * Revision 1.6  2003/07/14 11:13:05  shomrat
 * Plugin messageing system related changes
 *
 * Revision 1.5  2003/07/09 17:20:07  jcherry
 * Added code to add comment to annotation.
 * Removed some gratuitous code.
 *
 * Revision 1.4  2003/07/08 22:09:00  jcherry
 * Added code to attach results as feature table to document handed
 *
 * Revision 1.3  2003/07/03 19:14:12  jcherry
 * Initial version
 *
 * Revision 1.1  2003/07/03 19:06:39  jcherry
 * Initial version
 *
 * ===========================================================================
 */