DatabaseDocument.java

/* DatabaseDocument.java
 *
 * created: 2005
 *
 * This file is part of Artemis
 * 
 * Copyright (C) 2005  Genome Research Limited
 * 
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version 2
 * of the License, or (at your option) any later version.
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 * 
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
 *
 */

package uk.ac.sanger.artemis.util;

import uk.ac.sanger.artemis.io.ChadoCanonicalGene;
import uk.ac.sanger.artemis.io.GFFStreamFeature;
import uk.ac.sanger.artemis.io.ReadFormatException;

import uk.ac.sanger.artemis.chado.ArtemisUtils;
import uk.ac.sanger.artemis.chado.IBatisDAO;
import uk.ac.sanger.artemis.chado.JdbcDAO;
import uk.ac.sanger.artemis.chado.GmodDAO;
import uk.ac.sanger.artemis.chado.ChadoTransaction;
import uk.ac.sanger.artemis.components.database.DatabaseEntrySource;
import uk.ac.sanger.artemis.components.Splash;

import org.gmod.schema.sequence.Feature;
import org.gmod.schema.sequence.FeatureProp;
import org.gmod.schema.sequence.FeatureLoc;
import org.gmod.schema.sequence.FeatureRelationship;
import org.gmod.schema.sequence.FeatureSynonym;
import org.gmod.schema.sequence.FeatureCvTerm;
import org.gmod.schema.sequence.FeatureCvTermProp;
import org.gmod.schema.sequence.FeatureCvTermDbXRef;
import org.gmod.schema.sequence.FeatureCvTermPub;
import org.gmod.schema.cv.CvTerm;
import org.gmod.schema.general.DbXRef;
import org.gmod.schema.organism.Organism;
import org.gmod.schema.pub.PubDbXRef;
import org.gmod.schema.pub.Pub;

import java.sql.*;
import java.text.SimpleDateFormat;
import java.io.*;
import java.net.ConnectException;
import java.util.Hashtable;
import java.util.HashMap;
import java.util.Vector;
import java.util.Enumeration;
import java.util.List;
import java.util.Iterator;
import java.util.Collection;

import javax.swing.JOptionPane;
import javax.swing.JPasswordField;

/**
 * Objects of this class are Documents created from a relational database.
 * 
 */

public class DatabaseDocument extends Document
{
  private String name = null;

  /** source feature_id */
  private String srcFeatureId = "1";

  /** database schema */
  private String schema = "public";

  private static Hashtable cvterms;

  private InputStreamProgressListener progress_listener;

  private HashMap db;

  /** JDBC DAO */
  private JdbcDAO jdbcDAO = null;

  /** iBatis DAO */
  private IBatisDAO connIB = null;

  private ByteBuffer[] gff_buffer;

  private ByteBuffer gff_buff;

  /** entries to split into */
  private String[] types = { "exon", "gene", "CDS", "transcript" };

  /** true if splitting the GFF into entries */
  private boolean splitGFFEntry;

  private boolean iBatis = false;

  private JPasswordField pfield;

  private List schema_list;
  
  private boolean gene_builder;
  
  // controlled vocabulary
  /** controlled_curation controlled vocabulary */
  public static String CONTROLLED_CURATION_TAG_CVNAME = 
                                 "CC_";
  /** product controlled vocabulary */
  public static String PRODUCTS_TAG_CVNAME = "genedb_products";
  
  /**
   * 
   * Create a new Document from a database.
   * 
   * @param location
   *          This should be a URL string giving:
   *          jdbc:postgresql://host:port/datbase_name?user=username
   * 
   */
  public DatabaseDocument(String location, JPasswordField pfield)
  {
    super(location);
    this.pfield = pfield;

    if(location.indexOf('=') > -1)
      this.schema = location.substring( location.indexOf('=')+ 1);
    
    if(System.getProperty("ibatis") != null)
    {
      iBatis = true;
      System.setProperty("chado", location);
    }
  }

  
  /**
   * 
   * Create a new Document from a database.
   * 
   * @param location
   *          This should be a URL string giving:
   *          jdbc:postgresql://host:port/database_name?user=username
   * @param feature_id
   *          ID of a feature to be extracted.
   * 
   */
  public DatabaseDocument(String location, JPasswordField pfield,
                          String srcFeatureId, String schema)
  {
    super(location);
    this.pfield = pfield;

    this.srcFeatureId = srcFeatureId;
    this.schema = schema;

    if(System.getProperty("ibatis") != null)
    {
      iBatis = true;
      System.setProperty("chado", location);
    }
  }

  /**
   * 
   * Create a new Document from a database.
   * 
   * @param location
   *          This should be a URL string giving:
   *          jdbc:postgresql://host:port/datbase_name?user=username
   * @param srcFeatureId
   *          ID of a feature to be extracted.
   * @param splitGFFEntry
   *          split into separate entries based on feature types.
   * @param progress_listener
   *          input stream progress listener
   * 
   */
  public DatabaseDocument(String location, JPasswordField pfield,
                          String srcFeatureId, String schema, boolean splitGFFEntry,
                          InputStreamProgressListener progress_listener)
  {
    super(location);
    this.pfield = pfield;
    this.srcFeatureId = srcFeatureId;
    this.schema = schema;
    this.splitGFFEntry = splitGFFEntry;
    this.progress_listener = progress_listener;
    if(System.getProperty("ibatis") != null)
    {
      iBatis = true;
      System.setProperty("chado", location);
    }
    
    reset(location, schema);
  }

  /**
   * Used by the gene builder to read a database entry
   * for a single gene.
   * @param location
   * @param pfield
   * @param srcFeatureId
   * @param schema
   * @param gene_builder
   */
  public DatabaseDocument(String location, JPasswordField pfield,
          String srcFeatureId, String schema, boolean gene_builder)
  {
    super(location);
    this.pfield = pfield;
    this.srcFeatureId = srcFeatureId;
    this.schema = schema;
    this.gene_builder = gene_builder;

    if(System.getProperty("ibatis") != null)
    {
      iBatis = true;
      System.setProperty("chado", location);
    }
  }
  
  public DatabaseDocument(String location, JPasswordField pfield,
                          String srcFeatureId, String schema,
                          ByteBuffer gff_buff, String name)
  {
    super(location);
    this.pfield = pfield;
    this.srcFeatureId = srcFeatureId;
    this.schema = schema;
    this.gff_buff = gff_buff;
    this.name = name;
    if(System.getProperty("ibatis") != null)
    {
      iBatis = true;
      System.setProperty("chado", location);
    }
  }
  
  /**
   * Reset the schema.
   * @param location
   * @param schema
   */
  private void reset(String location, String schema)
  {
    this.schema = schema;

    if(!location.endsWith("="+schema))
    {
      int index = location.lastIndexOf('=');
      setLocation(location.substring(0,index+1) + schema);
      connIB  = null;
      jdbcDAO = null;
      System.setProperty("chado", (String)getLocation());
    }
  }

  /**
   * Append a String to the Document location.
   * @param name  the name to append.
   */
  public Document append(String name) throws IOException
  {
    return new DatabaseDocument( ((String)getLocation()) + name, pfield);
  }

  /**
   * Return the name of this Document (the last element of the Document
   * location).
   */
  public String getName()
  {
    if(name == null)
    {
      int ind     = ((String) getLocation()).indexOf("?");
      String name = ((String) getLocation()).substring(0, ind);
      ind = name.lastIndexOf("/");
      return name.substring(ind + 1);
    }
    return name;
  }


  /**
  *  Set the name of this document.
  */
  public void setName(String name)
  {
    this.name = name;
  }


  public DatabaseDocument createDatabaseDocument()
  {
    return new DatabaseDocument( (String)getLocation(), pfield,
                                  srcFeatureId, schema );
  }
  
  /**
   * Return true if and only if the Document refered to by this object exists
   * and is readable. Always returns true.
   */
  public boolean readable()
  {
    return true;
  }

  /**
   * Return true if and only if the Document refered to by this object exists
   * and can be written to. Always returns false.
   */
  public boolean writable()
  {
    return true;
  }

  /**
   * Create a new InputStream object from this Document. The contents of the
   * Document can be read from the InputStream.
   * 
   * @exception IOException
   *              Thrown if the Document can't be read from (for example if it
   *              doesn't exist).
   */
  public InputStream getInputStream() throws IOException
  {
    ByteArrayInputStream instream;

    if(gff_buff != null)
    {
      instream = new ByteArrayInputStream(gff_buff.getBytes());
      return instream;
    }

    try
    {
      GmodDAO dao = getDAO();
      ByteBuffer entry = new ByteBuffer();
      
      try
      {
        if(dao instanceof IBatisDAO)
          ((IBatisDAO) dao).startTransaction();

        // if creating a gene builder
        if(gene_builder)
        {
          List schemaList = new Vector();
          schemaList.add(schema);
          return new ByteArrayInputStream(getGeneFeature(srcFeatureId,
              schemaList, dao).getBytes());
        }

        gff_buffer = getGff(dao);

        
        if(splitGFFEntry)
        {
          if(gff_buffer[0].size() > 0)
            entry.append(gff_buffer[0]);

          getChadoSequence(dao, entry);
        }
        else
        {
          for(int i = 0; i < gff_buffer.length; i++)
          {
            if(gff_buffer[i].size() > 0)
              entry.append(gff_buffer[i]);
          }

          getChadoSequence(dao, entry);
        }

        if(dao instanceof IBatisDAO)
          ((IBatisDAO) dao).commitTransaction();
      }
      finally
      {
        if(dao instanceof IBatisDAO)
          ((IBatisDAO) dao).endTransaction();
      }

      instream = new ByteArrayInputStream(entry.getBytes());
      return instream;
    }
    catch(RuntimeException re)
    {
      JOptionPane.showMessageDialog(null, "Problems Reading...\n" +
          re.getMessage(),
          "Problems Reading From the Database ",
          JOptionPane.ERROR_MESSAGE);
      
      re.printStackTrace();
    }
    catch(java.sql.SQLException sqlExp)
    {
      JOptionPane.showMessageDialog(null, "Problems Reading...\n" +
          sqlExp.getMessage(),
          "Problems Reading From the Database ",
          JOptionPane.ERROR_MESSAGE);
      
      sqlExp.printStackTrace();
    }

    return null;
  }

  /**
   * 
   * Called (by DatabaseEntrySource) to retrieve all the documents for each
   * entry created.
   * 
   */
  public DatabaseDocument[] getGffDocuments(String location, String id,
                                            String schema)
  {
    int nentries = 0;
    for(int i = 1; i < gff_buffer.length; i++)
    {
      if(gff_buffer[i].size() > 0)
        nentries++;
    }

    DatabaseDocument[] new_docs = new DatabaseDocument[nentries];
    nentries = 0;
    for(int i = 1; i < gff_buffer.length; i++)
    {
      if(gff_buffer[i].size() == 0)
        continue;

      String name;
      if(i >= types.length)
        name = "other";
      else
        name = types[i];

      new_docs[nentries] = new DatabaseDocument(location, pfield, id, schema,
                                                gff_buffer[i], name);
      nentries++;
    }

    return new_docs;
  }

  /**
   * Create an array of GFF lines.
   * @param dao                 the data access object 
   * @param parentFeatureID     the parent identifier for the features to 
   *                            extract
   * @return   the <code>ByteBuffer</code> array of GFF lines
   */
  private ByteBuffer[] getGff(GmodDAO dao)
  {
    final int srcfeature_id = Integer.parseInt(srcFeatureId);
    
    Splash.logger4j.debug("Build GFF");
    
    // build srcfeature object
    FeatureLoc featureloc = new FeatureLoc();
    Feature srcFeature = new Feature();
    srcFeature.setFeatureId(srcfeature_id);
    featureloc.setFeatureBySrcFeatureId(srcFeature);
    
    //featureloc.setSrcfeature_id(srcfeature_id);
    Feature parent = new Feature();
    parent.setFeatureLoc(featureloc);
    
    List featList = dao.getFeaturesByLocatedOnFeature(parent);

    ByteBuffer[] buffers = new ByteBuffer[types.length + 1];
    for(int i = 0; i < buffers.length; i++)
      buffers[i] = new ByteBuffer();

    final Feature parentFeature = dao.getFeatureById(srcfeature_id);
    
    ByteBuffer this_buff;

    int feature_size = featList.size();
    Hashtable id_store = new Hashtable(feature_size);

    // build feature name store
    for(int i = 0; i < feature_size; i++)
    {
      Feature feat = (Feature)featList.get(i);
      String name       = feat.getUniqueName();
      String featureId = Integer.toString(feat.getFeatureId());

      id_store.put(featureId, name);
    }
    
    // get all dbrefs & synonyms
    Hashtable dbxrefs = IBatisDAO.mergeDbXRef(
                 dao.getFeatureDbXRefsByFeatureUniquename(null));
    Hashtable synonym = getAllFeatureSynonyms(dao, null);

    Hashtable featureCvTerms = getFeatureCvTermsByFeature(dao);
    Hashtable featureCvTermDbXRefs = getFeatureCvTermDbXRef(dao);
    Hashtable featureCvTermPubs = getFeatureCvTermPub(dao);
    
    List pubDbXRefs= dao.getPubDbXRef();

    // create gff byte stream
    for(int i = 0; i < feature_size; i++)
    { 
      // select buffer based on feature type
      Feature feat = (Feature)featList.get(i);
      int type_id = feat.getCvTerm().getCvTermId();
      String typeName = getCvtermName(type_id, dao);
      this_buff = buffers[types.length];
      for(int j = 0; j < types.length; j++)
      {
        if(types[j].equals(typeName))
          this_buff = buffers[j];
      }
      
      
      chadoToGFF(feat, parentFeature.getUniqueName(),
                 dbxrefs, synonym, featureCvTerms,
                 pubDbXRefs, featureCvTermDbXRefs, featureCvTermPubs,
                 id_store, dao, 
                 feat.getFeatureLoc(), this_buff);
       
      if( i%10 == 0 || i == feature_size-1)
        progress_listener.progressMade("Read from database: " + 
                                       feat.getUniqueName());
    }

    return buffers;
  }

  /**
   * Get a <code>Hashtable</code> of feature_id keys and their corresponding 
   * feature_synonym
   * 
   */
  private Hashtable getAllFeatureSynonyms(final GmodDAO dao, 
          final String uniquename) 
  {
    List list = dao.getFeatureSynonymsByFeatureUniquename(uniquename);  
    
    Hashtable synonym = new Hashtable();
    Integer featureId;
    List value;
    FeatureSynonym alias;
    
    for(int i=0; i<list.size(); i++)
    {
      alias = (FeatureSynonym)list.get(i);
      featureId = new Integer(alias.getFeature().getFeatureId());
      if(synonym.containsKey(featureId))
        value = (Vector)synonym.get(featureId);
      else
        value = new Vector();
      
      value.add(alias);
      synonym.put(featureId, value);
    }
    
    return synonym;
  }
  
  private Hashtable getFeatureCvTermsByFeature(final GmodDAO dao)
  {
    List list = dao.getFeatureCvTermsByFeature(null);
    
    Hashtable featureCvTerms = new Hashtable();
    Integer featureId;
    List value;
    FeatureCvTerm feature_cvterm;
    
    for(int i=0; i<list.size(); i++)
    {
      feature_cvterm = (FeatureCvTerm)list.get(i);
      featureId = new Integer(feature_cvterm.getFeature().getFeatureId());
      if(featureCvTerms.containsKey(featureId))
        value = (Vector)featureCvTerms.get(featureId);
      else
        value = new Vector();
      
      value.add(feature_cvterm);
      featureCvTerms.put(featureId, value);
    }
    
    return featureCvTerms;
  }
  
  
  private Hashtable getFeatureCvTermDbXRef(final GmodDAO dao)
  {
    List list = dao.getFeatureCvTermDbXRefByFeature(null);
    if(list == null || list.size() == 0)
      return null;
    
    Integer featureCvTermDbXRefId;
    List value;
    
    Hashtable featureCvTermDbXRefs = new Hashtable(list.size());
    for(int i=0; i<list.size(); i++)
    {
      FeatureCvTermDbXRef featureCvTermDbXRef =
        (FeatureCvTermDbXRef)list.get(i);
      
      featureCvTermDbXRefId = new Integer(
          featureCvTermDbXRef.getFeatureCvTerm().getFeatureCvTermId());
      
      if(featureCvTermDbXRefs.containsKey(featureCvTermDbXRefId))
        value = (Vector)featureCvTermDbXRefs.get(featureCvTermDbXRefId);
      else
        value = new Vector();
      
      value.add(featureCvTermDbXRef);
      featureCvTermDbXRefs.put(featureCvTermDbXRefId, value);
    }
     
    return featureCvTermDbXRefs;
  }
  
  private Hashtable getFeatureCvTermPub(final GmodDAO dao)
  {
    List list = dao.getFeatureCvTermPubByFeature(null);
    if(list == null || list.size() == 0)
      return null;
    
    Integer featureCvTermId;
    List value;
    
    Hashtable featureCvTermPubs = new Hashtable(list.size());
    for(int i=0; i<list.size(); i++)
    {
      FeatureCvTermPub featureCvTermPub =
        (FeatureCvTermPub)list.get(i);
      
      featureCvTermId = new Integer(
          featureCvTermPub.getFeatureCvTerm().getFeatureCvTermId());
      
      if(featureCvTermPubs.containsKey(featureCvTermId))
        value = (Vector)featureCvTermPubs.get(featureCvTermId);
      else
        value = new Vector();
      
      value.add(featureCvTermPub);
      featureCvTermPubs.put(featureCvTermId, value);
    }
     
    return featureCvTermPubs;
  }
  
  /**
   * Use by the gene editor to retrieve the gene and related
   * features
   * @param search_gene     gene uniquename
   * @param schema_search   schema list to search
   * @param dao             data access method
   * @return  GFF byte buffer
   * @throws SQLException
   * @throws ReadFormatException
   * @throws ConnectException 
   */
  private ByteBuffer getGeneFeature(final String search_gene, 
                                    final List schema_search,
                                    GmodDAO dao) 
          throws SQLException, ReadFormatException, ConnectException
  {
    Hashtable id_store = new Hashtable();

    //Feature feature = new Feature();
    //feature.setUniquename(search_gene);

    reset((String)getLocation(), (String)schema_search.get(0));
    dao = getDAO();
    Feature feature = 
      (Feature)(dao.getFeaturesByUniqueName(search_gene).get(0));
    
    ChadoCanonicalGene chado_gene = new ChadoCanonicalGene();
    id_store.put(Integer.toString(feature.getFeatureId()), feature.getUniqueName());

    List featurelocs = new Vector(feature.getFeatureLocsForFeatureId());
    FeatureLoc featureloc = (FeatureLoc) featurelocs.get(0);
    int src_id = featureloc.getFeatureBySrcFeatureId().getFeatureId();

    Feature parent = new Feature();
    parent.setFeatureId(src_id);

    parent = dao.getFeatureById(src_id);  //.getLazyFeature(parent);
    
    chado_gene.setSeqlen(parent.getSeqLen());
    chado_gene.setSrcfeature_id(src_id);

    ByteBuffer buff = new ByteBuffer();
    
    chadoToGFF(feature, null, null, null, null, null, null, null, null, dao,
               featureloc, buff);

    // get children of gene
    List relations = new Vector(feature.getFeatureRelationshipsForObjectId());
    
    for(int i = 0; i < relations.size(); i++)
    {
      //Feature transcript = new Feature();
      
      int id = ((FeatureRelationship) relations.get(i)).getFeatureBySubjectId().getFeatureId();
      //transcript.setId(id);
      Feature transcript = 
        (Feature)dao.getFeatureById(id); //.getLazyFeature(transcript);

      id_store.put(Integer.toString(transcript.getFeatureId()), 
          transcript.getUniqueName());

      FeatureLoc loc = getFeatureLoc(new Vector(
          transcript.getFeatureLocsForFeatureId()), src_id);
      chadoToGFF(transcript, feature.getUniqueName(), null, null, null,
          null, null, null, id_store, dao, loc, buff);

      // get children of transcript - exons and pp
      List transcipt_relations = new Vector(
          transcript.getFeatureRelationshipsForObjectId());

      for(int j = 0; j < transcipt_relations.size(); j++)
      {
        id = ((FeatureRelationship) transcipt_relations.get(j)).getFeatureBySubjectId().getFeatureId();
        //Feature child = new Feature();
        //child.setId(((FeatureRelationship) transcipt_relations.get(j))
        //    .getSubject_id());
        Feature child = 
          (Feature)dao.getFeatureById(id); //dao.getLazyFeature(child);

        id_store.put(Integer.toString(child.getFeatureId()), child.getUniqueName());

        loc = getFeatureLoc(
            new Vector(child.getFeatureLocsForFeatureId()),src_id);
        chadoToGFF(child, transcript.getUniqueName(), null,null, null,
                   null, null, null, id_store, dao, loc, buff);
      }
    }

    return buff;
  }
  

  /**
   * Convert the chado feature into a GFF line
   * @param feat           Chado feature
   * @param parentFeature  parent of this feature
   * @param dbxrefs        hashtable containing dbxrefs
   * @param synonym        hashtable containing synonynms
   * @param featureCvTerms
   * @param pubDbXRefs
   * @param featureCvTermDbXRefs
   * @param id_store       id store for looking up parent names
   * @param dao            chado data access
   * @param featureloc     feature location for this chado feature
   * @param this_buff      byte buffer of GFF line 
   */
  private static void chadoToGFF(final Feature feat,
                                 final String parentFeature,
                                 final Hashtable dbxrefs,
                                 final Hashtable synonym,
                                 final Hashtable featureCvTerms,
                                 final List pubDbXRefs,
                                 final Hashtable featureCvTermDbXRefs,
                                 final Hashtable featureCvTermPubs,
                                 final Hashtable id_store,
                                 final GmodDAO dao,
                                 final FeatureLoc featureloc,
                                 final ByteBuffer this_buff)
  {
    String gff_source = null;
    
    final int fmin          = featureloc.getFmin().intValue() + 1;
    final int fmax          = featureloc.getFmax().intValue();
    final int type_id       = feat.getCvTerm().getCvTermId();
    final Short strand      = featureloc.getStrand();
    final Integer phase     = featureloc.getPhase();
    final String name       = feat.getUniqueName();
    final String typeName   = getCvtermName(type_id, dao);
    final Integer featureId = new Integer(feat.getFeatureId());
    final String timelastmodified = Long.toString(feat.getTimeLastModified().getTime());

    String parent_id = null;
    String parent_relationship = null;
    int rank = -1;
/*    if(feat.getFeatureRelationship() != null)
    {
      FeatureRelationship feat_relationship = feat.getFeatureRelationship();
      parent_id = Integer.toString(feat_relationship.getFeatureByObjectId().getFeatureId());
      long parent_type_id = feat_relationship.getCvTerm().getCvTermId();
      
      parent_relationship = feat_relationship.getCvTerm().getName();
      
      rank= feat_relationship.getRank();
      if(parent_relationship == null)
        parent_relationship = getCvtermName(parent_type_id, dao);
    }
    else */
      
    if(feat.getFeatureRelationshipsForSubjectId() != null)
    {
      List relations = new Vector(feat.getFeatureRelationshipsForSubjectId());
      for(int i=0; i<relations.size(); i++)
      {
        FeatureRelationship feat_relationship = 
                            (FeatureRelationship)relations.get(i);
        parent_id = Integer.toString(feat_relationship.getFeatureByObjectId().getFeatureId());
        rank      = feat_relationship.getRank();
        
        if( feat_relationship.getCvTerm().getName() == null )
        {
          int parent_type_id = feat_relationship.getCvTerm().getCvTermId();
          parent_relationship = getCvtermName(parent_type_id, dao);
        }
        else
          parent_relationship = feat_relationship.getCvTerm().getName();
      }
    }
          
    if(parent_id != null && id_store != null &&  id_store.containsKey(parent_id))
      parent_id = (String)id_store.get(parent_id);
 
    // make gff format
    
    Vector dbxref = null;
    // append dbxrefs
    if(dbxrefs != null &&
       dbxrefs.containsKey(featureId))
    {
      dbxref = (Vector)dbxrefs.get(featureId);
      for(int j=0; j<dbxref.size(); j++)
      {
        if(((String)dbxref.get(j)).startsWith("GFF_source:"))
        {
          gff_source = ((String)dbxref.get(j)).substring(11);
          dbxref.removeElementAt(j);
        }
      }
    }

    this_buff.append(parentFeature + "\t"); // seqid
    
    if(gff_source != null)
      this_buff.append(gff_source+"\t");    // source
    else
      this_buff.append("chado\t");            
    this_buff.append(typeName + "\t");      // type
    this_buff.append(fmin + "\t");          // start
    this_buff.append(fmax + "\t");          // end
    this_buff.append(".\t");                // score
    if(strand.equals( new Short((short)-1)) )                        // strand
      this_buff.append("-\t");
    else if(strand.equals( new Short((short)1)) )
      this_buff.append("+\t");
    else
      this_buff.append(".\t");

    if(phase == null)
      this_buff.append(".\t");               // phase
    else
      this_buff.append(phase+"\t"); 

    this_buff.append("ID=" + name + ";");

   
    if(parent_id != null && !parent_id.equals("0"))
    {
      if(parent_relationship.equals("derives_from"))
        this_buff.append("Derives_from=" + parent_id + ";");
      else
        this_buff.append("Parent=" + parent_id + ";");
    }

    this_buff.append("timelastmodified=" + timelastmodified + ";");
    
    // this is the chado feature_relationship.rank used
    // to order joined features e.g. exons
    if(rank > -1)
      this_buff.append("feature_relationship_rank="+rank+";"); 

    //this_buff.append("feature_id="+feature_id+";");
    
    // attributes
    if(feat.getFeatureProps() != null &&
       feat.getFeatureProps().size() > 0)
    {
      Collection featureprops = feat.getFeatureProps();
      Iterator it = featureprops.iterator();
      
      while(it.hasNext())
      {
        FeatureProp featprop = (FeatureProp)it.next();
        String qualifier_name = getCvtermName(featprop.getCvTerm().getCvTermId(), dao);
        if(qualifier_name == null)
          continue;
        if(featprop.getValue() != null)
          this_buff.append(qualifier_name+ "=" +
                           GFFStreamFeature.encode(featprop.getValue())+";");
      }
    }

    // append dbxrefs
    if(dbxref != null && dbxref.size() > 0)
    {
      this_buff.append("Dbxref=");
      for(int j=0; j<dbxref.size(); j++)
      {
        this_buff.append((String)dbxref.get(j));
        if(j<dbxref.size()-1)
          this_buff.append(",");
      }
      this_buff.append(";");
    }
    
    // append synonyms
    if(synonym != null &&
       synonym.containsKey(featureId))
    {   
      FeatureSynonym alias;
      Vector v_synonyms = (Vector)synonym.get(featureId);
      for(int j=0; j<v_synonyms.size(); j++)
      {
        alias = (FeatureSynonym)v_synonyms.get(j);
        
        this_buff.append( getCvtermName(alias.getSynonym().getCvTerm().getCvTermId(), dao) + "=" );
        //this_buff.append(alias.getSynonym().getCvterm().getName()+"=");
        this_buff.append(alias.getSynonym().getName());
        
        if(j<v_synonyms.size()-1)
          this_buff.append(";");
      }
    }
    
    // GO, controlled_curation, product
    if(featureCvTerms != null && 
       featureCvTerms.containsKey(featureId))
    {
      FeatureCvTerm feature_cvterm;
      Vector v_feature_cvterms = (Vector)featureCvTerms.get(featureId);
      for(int j=0; j<v_feature_cvterms.size(); j++)
      {
        feature_cvterm = (FeatureCvTerm)v_feature_cvterms.get(j);
        
        Integer featureCvTermId = new Integer( feature_cvterm.getFeatureCvTermId() );
        
        List featureCvTermDbXRefList = null;
        
        if(featureCvTermDbXRefs != null)
          featureCvTermDbXRefList = (List)featureCvTermDbXRefs.get(featureCvTermId);
        
        List featureCvTermPubList = null;
        
        if(featureCvTermPubs != null)
          featureCvTermPubList = (List)featureCvTermPubs.get(featureCvTermId);
          
        appendControlledVocabulary(this_buff, dao, feature_cvterm,
                                   featureCvTermDbXRefList,featureCvTermPubList, pubDbXRefs);
      }
      //System.out.println(new String(this_buff.getBytes()));
    }
      
    this_buff.append("\n");
  }
  
  /**
   * Appends controlled vocabulary terms to the buffer
   * @param attr_buff
   * @param dao
   * @param feature_cvterm
   * @param featureCvTermDbXRef
   */
  public static void appendControlledVocabulary(final ByteBuffer attr_buff,
                                          final GmodDAO dao,
                                          final FeatureCvTerm feature_cvterm,
                                          final List featureCvTermDbXRefs,
                                          final List featureCvTermPubs,
                                          final List pubDbXRefs)
  {
    CvTerm cvterm =  getCvTerm( feature_cvterm.getCvTerm().getCvTermId(), dao);
    DbXRef dbXRef = feature_cvterm.getCvTerm().getDbXRef();
    
    if(cvterm.getCv().getName().startsWith(DatabaseDocument.CONTROLLED_CURATION_TAG_CVNAME))
    {
      attr_buff.append("controlled_curation=");
      
      attr_buff.append("term="+feature_cvterm.getCvTerm().getName()+"%3B");
      attr_buff.append("cv="+feature_cvterm.getCvTerm().getCv().getName()+"%3B");   
      
      // N.B. the db_xref may be a FeatureCvTermDbXRef or a Pub for /controlled_curation
      int nfound_dbxref = 0;
      if(feature_cvterm.getPub().getUniqueName() != null &&
         !feature_cvterm.getPub().getUniqueName().equals("NULL"))
      {