import jalview.analysis.AlignSeq;
import jalview.api.DBRefEntryI;
+import jalview.datamodel.features.SequenceFeatures;
+import jalview.datamodel.features.SequenceFeaturesI;
+import jalview.util.Comparison;
+import jalview.util.DBRefUtils;
+import jalview.util.MapList;
import jalview.util.StringUtils;
import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.BitSet;
+import java.util.Collections;
import java.util.Enumeration;
import java.util.List;
+import java.util.ListIterator;
import java.util.Vector;
+import com.stevesoft.pat.Regex;
+
import fr.orsay.lri.varna.models.rna.RNA;
/**
*/
public class Sequence extends ASequence implements SequenceI
{
+ private static final Regex limitrx = new Regex(
+ "[/][0-9]{1,}[-][0-9]{1,}$");
+
+ private static final Regex endrx = new Regex("[0-9]{1,}$");
+
SequenceI datasetSequence;
String name;
String vamsasId;
- DBRefEntryI sourceDBRef;
-
DBRefEntry[] dbrefs;
RNA rna;
*/
int index = -1;
- /** array of sequence features - may not be null for a valid sequence object */
- public SequenceFeature[] sequenceFeatures;
+ private SequenceFeatures sequenceFeatureStore;
+
+ /*
+ * A cursor holding the approximate current view position to the sequence,
+ * as determined by findIndex or findPosition or findPositions.
+ * Using a cursor as a hint allows these methods to be more performant for
+ * large sequences.
+ */
+ private SequenceCursor cursor;
+
+ /*
+ * A number that should be incremented whenever the sequence is edited.
+ * If the value matches the cursor token, then we can trust the cursor,
+ * if not then it should be recomputed.
+ */
+ private int changeCount;
+
+ private GeneLoci geneLoci;
/**
* Creates a new Sequence object.
*/
public Sequence(String name, String sequence, int start, int end)
{
+ this();
initSeqAndName(name, sequence.toCharArray(), start, end);
}
public Sequence(String name, char[] sequence, int start, int end)
{
+ this();
initSeqAndName(name, sequence, start, end);
}
checkValidRange();
}
- com.stevesoft.pat.Regex limitrx = new com.stevesoft.pat.Regex(
- "[/][0-9]{1,}[-][0-9]{1,}$");
-
- com.stevesoft.pat.Regex endrx = new com.stevesoft.pat.Regex("[0-9]{1,}$");
-
void parseId()
{
if (name == null)
{
- System.err
- .println("POSSIBLE IMPLEMENTATION ERROR: null sequence name passed to constructor.");
+ System.err.println(
+ "POSSIBLE IMPLEMENTATION ERROR: null sequence name passed to constructor.");
name = "";
}
// Does sequence have the /start-end signature?
}
/**
+ * default constructor
+ */
+ private Sequence()
+ {
+ sequenceFeatureStore = new SequenceFeatures();
+ }
+
+ /**
* Creates a new Sequence object.
*
* @param name
}
/**
- * Creates a new Sequence object with new features, DBRefEntries,
- * AlignmentAnnotations, and PDBIds but inherits any existing dataset sequence
- * reference.
+ * Creates a new Sequence object with new AlignmentAnnotations but inherits
+ * any existing dataset sequence reference. If non exists, everything is
+ * copied.
*
* @param seq
- * DOCUMENT ME!
+ * if seq is a dataset sequence, behaves like a plain old copy
+ * constructor
*/
public Sequence(SequenceI seq)
{
*/
public Sequence(SequenceI seq, AlignmentAnnotation[] alAnnotation)
{
+ this();
initSeqFrom(seq, alAnnotation);
-
}
+ /**
+ * does the heavy lifting when cloning a dataset sequence, or coping data from
+ * dataset to a new derived sequence.
+ *
+ * @param seq
+ * - source of attributes.
+ * @param alAnnotation
+ * - alignment annotation present on seq that should be copied onto
+ * this sequence
+ */
protected void initSeqFrom(SequenceI seq,
AlignmentAnnotation[] alAnnotation)
{
- initSeqAndName(seq.getName(), seq.getSequence(), seq.getStart(),
- seq.getEnd());
+ char[] oseq = seq.getSequence(); // returns a copy of the array
+ initSeqAndName(seq.getName(), oseq, seq.getStart(), seq.getEnd());
+
description = seq.getDescription();
- if (seq.getSequenceFeatures() != null)
+ if (seq != datasetSequence)
{
- SequenceFeature[] sf = seq.getSequenceFeatures();
- for (int i = 0; i < sf.length; i++)
- {
- addSequenceFeature(new SequenceFeature(sf[i]));
- }
+ setDatasetSequence(seq.getDatasetSequence());
}
- setDatasetSequence(seq.getDatasetSequence());
- if (datasetSequence == null && seq.getDBRef() != null)
+
+ /*
+ * only copy DBRefs and seqfeatures if we really are a dataset sequence
+ */
+ if (datasetSequence == null)
{
- // only copy DBRefs if we really are a dataset sequence
- DBRefEntry[] dbr = seq.getDBRef();
- for (int i = 0; i < dbr.length; i++)
+ if (seq.getDBRefs() != null)
+ {
+ DBRefEntry[] dbr = seq.getDBRefs();
+ for (int i = 0; i < dbr.length; i++)
+ {
+ addDBRef(new DBRefEntry(dbr[i]));
+ }
+ }
+
+ /*
+ * make copies of any sequence features
+ */
+ for (SequenceFeature sf : seq.getSequenceFeatures())
{
- addDBRef(new DBRefEntry(dbr[i]));
+ addSequenceFeature(new SequenceFeature(sf));
}
}
+
if (seq.getAnnotation() != null)
{
AlignmentAnnotation[] sqann = seq.getAnnotation();
}
if (seq.getAllPDBEntries() != null)
{
- Vector ids = seq.getAllPDBEntries();
- Enumeration e = ids.elements();
- while (e.hasMoreElements())
+ Vector<PDBEntry> ids = seq.getAllPDBEntries();
+ for (PDBEntry pdb : ids)
{
- this.addPDBId(new PDBEntry((PDBEntry) e.nextElement()));
+ this.addPDBId(new PDBEntry(pdb));
}
}
}
- /**
- * DOCUMENT ME!
- *
- * @param v
- * DOCUMENT ME!
- */
@Override
- public void setSequenceFeatures(SequenceFeature[] features)
+ public void setSequenceFeatures(List<SequenceFeature> features)
{
- sequenceFeatures = features;
+ if (datasetSequence != null)
+ {
+ datasetSequence.setSequenceFeatures(features);
+ return;
+ }
+ sequenceFeatureStore = new SequenceFeatures(features);
}
@Override
- public synchronized void addSequenceFeature(SequenceFeature sf)
+ public synchronized boolean addSequenceFeature(SequenceFeature sf)
{
- if (sequenceFeatures == null)
+ if (sf.getType() == null)
{
- sequenceFeatures = new SequenceFeature[0];
+ System.err.println("SequenceFeature type may not be null: "
+ + sf.toString());
+ return false;
}
- for (int i = 0; i < sequenceFeatures.length; i++)
+ if (datasetSequence != null)
{
- if (sequenceFeatures[i].equals(sf))
- {
- return;
- }
+ return datasetSequence.addSequenceFeature(sf);
}
- SequenceFeature[] temp = new SequenceFeature[sequenceFeatures.length + 1];
- System.arraycopy(sequenceFeatures, 0, temp, 0, sequenceFeatures.length);
- temp[sequenceFeatures.length] = sf;
-
- sequenceFeatures = temp;
+ return sequenceFeatureStore.add(sf);
}
@Override
public void deleteFeature(SequenceFeature sf)
{
- if (sequenceFeatures == null)
- {
- return;
- }
-
- int index = 0;
- for (index = 0; index < sequenceFeatures.length; index++)
- {
- if (sequenceFeatures[index].equals(sf))
- {
- break;
- }
- }
-
- if (index == sequenceFeatures.length)
- {
- return;
- }
-
- int sfLength = sequenceFeatures.length;
- if (sfLength < 2)
+ if (datasetSequence != null)
{
- sequenceFeatures = null;
+ datasetSequence.deleteFeature(sf);
}
else
{
- SequenceFeature[] temp = new SequenceFeature[sfLength - 1];
- System.arraycopy(sequenceFeatures, 0, temp, 0, index);
-
- if (index < sfLength)
- {
- System.arraycopy(sequenceFeatures, index + 1, temp, index,
- sequenceFeatures.length - index - 1);
- }
-
- sequenceFeatures = temp;
+ sequenceFeatureStore.delete(sf);
}
}
/**
- * Returns the sequence features (if any), looking first on the sequence, then
- * on its dataset sequence, and so on until a non-null value is found (or
- * none). This supports retrieval of sequence features stored on the sequence
- * (as in the applet) or on the dataset sequence (as in the Desktop version).
+ * {@inheritDoc}
*
* @return
*/
@Override
- public SequenceFeature[] getSequenceFeatures()
+ public List<SequenceFeature> getSequenceFeatures()
{
- SequenceFeature[] features = sequenceFeatures;
-
- SequenceI seq = this;
- int count = 0; // failsafe against loop in sequence.datasetsequence...
- while (features == null && seq.getDatasetSequence() != null
- && count++ < 10)
+ if (datasetSequence != null)
{
- seq = seq.getDatasetSequence();
- features = ((Sequence) seq).sequenceFeatures;
+ return datasetSequence.getSequenceFeatures();
}
- return features;
+ return sequenceFeatureStore.getAllFeatures();
}
@Override
- public void addPDBId(PDBEntry entry)
+ public SequenceFeaturesI getFeatures()
+ {
+ return datasetSequence != null ? datasetSequence.getFeatures()
+ : sequenceFeatureStore;
+ }
+
+ @Override
+ public boolean addPDBId(PDBEntry entry)
{
if (pdbIds == null)
{
pdbIds = new Vector<PDBEntry>();
+ pdbIds.add(entry);
+ return true;
}
- if (pdbIds.contains(entry))
- {
- updatePDBEntry(pdbIds.get(pdbIds.indexOf(entry)), entry);
- }
- else
- {
- pdbIds.addElement(entry);
- }
- }
- private static void updatePDBEntry(PDBEntry oldEntry, PDBEntry newEntry)
- {
- if (newEntry.getFile() != null)
+ for (PDBEntry pdbe : pdbIds)
{
- oldEntry.setFile(newEntry.getFile());
+ if (pdbe.updateFrom(entry))
+ {
+ return false;
+ }
}
+ pdbIds.addElement(entry);
+ return true;
}
/**
@Override
public Vector<PDBEntry> getAllPDBEntries()
{
- return pdbIds;
+ return pdbIds == null ? new Vector<PDBEntry>() : pdbIds;
}
/**
{
this.sequence = seq.toCharArray();
checkValidRange();
+ sequenceChanged();
}
@Override
@Override
public char[] getSequence()
{
- return sequence;
+ // return sequence;
+ return sequence == null ? null : Arrays.copyOf(sequence,
+ sequence.length);
}
/*
}
/**
- * DOCUMENT ME!
+ * Returns the character of the aligned sequence at the given position (base
+ * zero), or space if the position is not within the sequence's bounds
*
- * @param i
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
+ * @return
*/
@Override
public char getCharAt(int i)
{
- if (i < sequence.length)
+ if (i >= 0 && i < sequence.length)
{
return sequence[i];
}
}
/**
- * DOCUMENT ME!
+ * Sets the sequence description, and also parses out any special formats of
+ * interest
*
* @param desc
- * DOCUMENT ME!
*/
@Override
public void setDescription(String desc)
{
this.description = desc;
+ parseDescription();
}
/**
- * DOCUMENT ME!
+ * Parses and saves fields of an Ensembl-style description e.g.
+ * chromosome:GRCh38:17:45051610:45109016:1
+ */
+ protected void parseDescription()
+ {
+ if (description == null)
+ {
+ return;
+ }
+ String[] tokens = description.split(":");
+ if (tokens.length == 6 && "chromosome".equals(tokens[0])) {
+ String ref = tokens[1];
+ String chrom = tokens[2];
+ try {
+ int chStart = Integer.parseInt(tokens[3]);
+ int chEnd = Integer.parseInt(tokens[4]);
+ boolean forwardStrand = "1".equals(tokens[5]);
+ String species = ""; // dunno yet!
+ int[] from = new int[] { start, end };
+ int[] to = new int[] { forwardStrand ? chStart : chEnd,
+ forwardStrand ? chEnd : chStart };
+ MapList map = new MapList(from, to, 1, 1);
+ GeneLoci gl = new GeneLoci(species, ref, chrom, map);
+ setGeneLoci(gl);
+ } catch (NumberFormatException e)
+ {
+ System.err.println("Bad integers in description " + description);
+ }
+ }
+ }
+
+ public void setGeneLoci(GeneLoci gl)
+ {
+ geneLoci = gl;
+ }
+
+ /**
+ * Returns the gene loci mapping for the sequence (may be null)
*
- * @return DOCUMENT ME!
+ * @return
+ */
+ @Override
+ public GeneLoci getGeneLoci()
+ {
+ return geneLoci;
+ }
+
+ /**
+ * Answers the description
+ *
+ * @return
*/
@Override
public String getDescription()
return this.description;
}
- /*
- * (non-Javadoc)
- *
- * @see jalview.datamodel.SequenceI#findIndex(int)
+ /**
+ * {@inheritDoc}
*/
@Override
public int findIndex(int pos)
{
- // returns the alignment position for a residue
+ /*
+ * use a valid, hopefully nearby, cursor if available
+ */
+ if (isValidCursor(cursor))
+ {
+ return findIndex(pos, cursor);
+ }
+
int j = start;
int i = 0;
- // Rely on end being at least as long as the length of the sequence.
+ int startColumn = 0;
+
+ /*
+ * traverse sequence from the start counting gaps; make a note of
+ * the column of the first residue to save in the cursor
+ */
while ((i < sequence.length) && (j <= end) && (j <= pos))
{
- if (!jalview.util.Comparison.isGap(sequence[i]))
+ if (!Comparison.isGap(sequence[i]))
{
+ if (j == start)
+ {
+ startColumn = i;
+ }
j++;
}
-
i++;
}
- if ((j == end) && (j < pos))
+ if (j == end && j < pos)
{
return end + 1;
}
- else
+
+ updateCursor(pos, i, startColumn);
+ return i;
+ }
+
+ /**
+ * Updates the cursor to the latest found residue and column position
+ *
+ * @param residuePos
+ * (start..)
+ * @param column
+ * (1..)
+ * @param startColumn
+ * column position of the first sequence residue
+ */
+ protected void updateCursor(int residuePos, int column, int startColumn)
+ {
+ /*
+ * preserve end residue column provided cursor was valid
+ */
+ int endColumn = isValidCursor(cursor) ? cursor.lastColumnPosition : 0;
+ if (residuePos == this.end)
{
- return i;
+ endColumn = column;
}
+
+ cursor = new SequenceCursor(this, residuePos, column, startColumn,
+ endColumn, this.changeCount);
}
- @Override
- public int findPosition(int i)
+ /**
+ * Answers the aligned column position (1..) for the given residue position
+ * (start..) given a 'hint' of a residue/column location in the neighbourhood.
+ * The hint may be left of, at, or to the right of the required position.
+ *
+ * @param pos
+ * @param curs
+ * @return
+ */
+ protected int findIndex(int pos, SequenceCursor curs)
{
+ if (!isValidCursor(curs))
+ {
+ /*
+ * wrong or invalidated cursor, compute de novo
+ */
+ return findIndex(pos);
+ }
+
+ if (curs.residuePosition == pos)
+ {
+ return curs.columnPosition;
+ }
+
+ /*
+ * move left or right to find pos from hint.position
+ */
+ int col = curs.columnPosition - 1; // convert from base 1 to 0-based array
+ // index
+ int newPos = curs.residuePosition;
+ int delta = newPos > pos ? -1 : 1;
+
+ while (newPos != pos)
+ {
+ col += delta; // shift one column left or right
+ if (col < 0 || col == sequence.length)
+ {
+ break;
+ }
+ if (!Comparison.isGap(sequence[col]))
+ {
+ newPos += delta;
+ }
+ }
+
+ col++; // convert back to base 1
+ updateCursor(pos, col, curs.firstColumnPosition);
+
+ return col;
+ }
+
+ /**
+ * {@inheritDoc}
+ */
+ @Override
+ public int findPosition(final int column)
+ {
+ /*
+ * use a valid, hopefully nearby, cursor if available
+ */
+ if (isValidCursor(cursor))
+ {
+ return findPosition(column + 1, cursor);
+ }
+
+ // TODO recode this more naturally i.e. count residues only
+ // as they are found, not 'in anticipation'
+
+ /*
+ * traverse the sequence counting gaps; note the column position
+ * of the first residue, to save in the cursor
+ */
+ int firstResidueColumn = 0;
+ int lastPosFound = 0;
+ int lastPosFoundColumn = 0;
+ int seqlen = sequence.length;
+
+ if (seqlen > 0 && !Comparison.isGap(sequence[0]))
+ {
+ lastPosFound = start;
+ lastPosFoundColumn = 0;
+ }
+
int j = 0;
int pos = start;
- int seqlen = sequence.length;
- while ((j < i) && (j < seqlen))
+
+ while (j < column && j < seqlen)
{
- if (!jalview.util.Comparison.isGap(sequence[j]))
+ if (!Comparison.isGap(sequence[j]))
{
+ lastPosFound = pos;
+ lastPosFoundColumn = j;
+ if (pos == this.start)
+ {
+ firstResidueColumn = j;
+ }
pos++;
}
-
j++;
}
+ if (j < seqlen && !Comparison.isGap(sequence[j]))
+ {
+ lastPosFound = pos;
+ lastPosFoundColumn = j;
+ if (pos == this.start)
+ {
+ firstResidueColumn = j;
+ }
+ }
+
+ /*
+ * update the cursor to the last residue position found (if any)
+ * (converting column position to base 1)
+ */
+ if (lastPosFound != 0)
+ {
+ updateCursor(lastPosFound, lastPosFoundColumn + 1,
+ firstResidueColumn + 1);
+ }
return pos;
}
/**
+ * Answers true if the given cursor is not null, is for this sequence object,
+ * and has a token value that matches this object's changeCount, else false.
+ * This allows us to ignore a cursor as 'stale' if the sequence has been
+ * modified since the cursor was created.
+ *
+ * @param curs
+ * @return
+ */
+ protected boolean isValidCursor(SequenceCursor curs)
+ {
+ if (curs == null || curs.sequence != this || curs.token != changeCount)
+ {
+ return false;
+ }
+ /*
+ * sanity check against range
+ */
+ if (curs.columnPosition < 0 || curs.columnPosition > sequence.length)
+ {
+ return false;
+ }
+ if (curs.residuePosition < start || curs.residuePosition > end)
+ {
+ return false;
+ }
+ return true;
+ }
+
+ /**
+ * Answers the sequence position (start..) for the given aligned column
+ * position (1..), given a hint of a cursor in the neighbourhood. The cursor
+ * may lie left of, at, or to the right of the column position.
+ *
+ * @param col
+ * @param curs
+ * @return
+ */
+ protected int findPosition(final int col, SequenceCursor curs)
+ {
+ if (!isValidCursor(curs))
+ {
+ /*
+ * wrong or invalidated cursor, compute de novo
+ */
+ return findPosition(col - 1);// ugh back to base 0
+ }
+
+ if (curs.columnPosition == col)
+ {
+ cursor = curs; // in case this method becomes public
+ return curs.residuePosition; // easy case :-)
+ }
+
+ if (curs.lastColumnPosition > 0 && curs.lastColumnPosition < col)
+ {
+ /*
+ * sequence lies entirely to the left of col
+ * - return last residue + 1
+ */
+ return end + 1;
+ }
+
+ if (curs.firstColumnPosition > 0 && curs.firstColumnPosition > col)
+ {
+ /*
+ * sequence lies entirely to the right of col
+ * - return first residue
+ */
+ return start;
+ }
+
+ // todo could choose closest to col out of column,
+ // firstColumnPosition, lastColumnPosition as a start point
+
+ /*
+ * move left or right to find pos from cursor position
+ */
+ int firstResidueColumn = curs.firstColumnPosition;
+ int column = curs.columnPosition - 1; // to base 0
+ int newPos = curs.residuePosition;
+ int delta = curs.columnPosition > col ? -1 : 1;
+ boolean gapped = false;
+ int lastFoundPosition = curs.residuePosition;
+ int lastFoundPositionColumn = curs.columnPosition;
+
+ while (column != col - 1)
+ {
+ column += delta; // shift one column left or right
+ if (column < 0 || column == sequence.length)
+ {
+ break;
+ }
+ gapped = Comparison.isGap(sequence[column]);
+ if (!gapped)
+ {
+ newPos += delta;
+ lastFoundPosition = newPos;
+ lastFoundPositionColumn = column + 1;
+ if (lastFoundPosition == this.start)
+ {
+ firstResidueColumn = column + 1;
+ }
+ }
+ }
+
+ if (cursor == null || lastFoundPosition != cursor.residuePosition)
+ {
+ updateCursor(lastFoundPosition, lastFoundPositionColumn,
+ firstResidueColumn);
+ }
+
+ /*
+ * hack to give position to the right if on a gap
+ * or beyond the length of the sequence (see JAL-2562)
+ */
+ if (delta > 0 && (gapped || column >= sequence.length))
+ {
+ newPos++;
+ }
+
+ return newPos;
+ }
+
+ /**
+ * {@inheritDoc}
+ */
+ @Override
+ public Range findPositions(int fromColumn, int toColumn)
+ {
+ if (toColumn < fromColumn || fromColumn < 1)
+ {
+ return null;
+ }
+
+ /*
+ * find the first non-gapped position, if any
+ */
+ int firstPosition = 0;
+ int col = fromColumn - 1;
+ int length = sequence.length;
+ while (col < length && col < toColumn)
+ {
+ if (!Comparison.isGap(sequence[col]))
+ {
+ firstPosition = findPosition(col++);
+ break;
+ }
+ col++;
+ }
+
+ if (firstPosition == 0)
+ {
+ return null;
+ }
+
+ /*
+ * find the last non-gapped position
+ */
+ int lastPosition = firstPosition;
+ while (col < length && col < toColumn)
+ {
+ if (!Comparison.isGap(sequence[col++]))
+ {
+ lastPosition++;
+ }
+ }
+
+ return new Range(firstPosition, lastPosition);
+ }
+
+ /**
* Returns an int array where indices correspond to each residue in the
* sequence and the element value gives its position in the alignment
*
}
@Override
+ public BitSet getInsertionsAsBits()
+ {
+ BitSet map = new BitSet();
+ int lastj = -1, j = 0;
+ int pos = start;
+ int seqlen = sequence.length;
+ while ((j < seqlen))
+ {
+ if (jalview.util.Comparison.isGap(sequence[j]))
+ {
+ if (lastj == -1)
+ {
+ lastj = j;
+ }
+ }
+ else
+ {
+ if (lastj != -1)
+ {
+ map.set(lastj, j);
+ lastj = -1;
+ }
+ }
+ j++;
+ }
+ if (lastj != -1)
+ {
+ map.set(lastj, j);
+ lastj = -1;
+ }
+ return map;
+ }
+
+ @Override
public void deleteChars(int i, int j)
{
int newstart = start, newend = end;
start = newstart;
end = newend;
sequence = tmp;
+ sequenceChanged();
}
@Override
}
sequence = tmp;
+ sequenceChanged();
}
@Override
}
@Override
- public void setDBRef(DBRefEntry[] dbref)
+ public void setDBRefs(DBRefEntry[] dbref)
{
+ if (dbrefs == null && datasetSequence != null
+ && this != datasetSequence)
+ {
+ datasetSequence.setDBRefs(dbref);
+ return;
+ }
dbrefs = dbref;
+ if (dbrefs != null)
+ {
+ DBRefUtils.ensurePrimaries(this);
+ }
}
@Override
- public DBRefEntry[] getDBRef()
+ public DBRefEntry[] getDBRefs()
{
if (dbrefs == null && datasetSequence != null
&& this != datasetSequence)
{
- return datasetSequence.getDBRef();
+ return datasetSequence.getDBRefs();
}
return dbrefs;
}
@Override
public void addDBRef(DBRefEntry entry)
{
+ if (datasetSequence != null)
+ {
+ datasetSequence.addDBRef(entry);
+ return;
+ }
+
if (dbrefs == null)
{
dbrefs = new DBRefEntry[0];
}
- int i, iSize = dbrefs.length;
-
- for (i = 0; i < iSize; i++)
+ for (DBRefEntryI dbr : dbrefs)
{
- if (dbrefs[i].equalRef(entry))
+ if (dbr.updateFrom(entry))
{
- if (entry.getMap() != null)
- {
- if (dbrefs[i].getMap() == null)
- {
- // overwrite with 'superior' entry that contains a mapping.
- dbrefs[i] = entry;
- }
- }
+ /*
+ * found a dbref that either matched, or could be
+ * updated from, the new entry - no need to add it
+ */
return;
}
}
- DBRefEntry[] temp = new DBRefEntry[iSize + 1];
- System.arraycopy(dbrefs, 0, temp, 0, iSize);
+ /*
+ * extend the array to make room for one more
+ */
+ // TODO use an ArrayList instead
+ int j = dbrefs.length;
+ DBRefEntry[] temp = new DBRefEntry[j + 1];
+ System.arraycopy(dbrefs, 0, temp, 0, j);
temp[temp.length - 1] = entry;
dbrefs = temp;
+
+ DBRefUtils.ensurePrimaries(this);
}
@Override
public void setDatasetSequence(SequenceI seq)
{
+ if (seq == this)
+ {
+ throw new IllegalArgumentException(
+ "Implementation Error: self reference passed to SequenceI.setDatasetSequence");
+ }
+ if (seq != null && seq.getDatasetSequence() != null)
+ {
+ throw new IllegalArgumentException(
+ "Implementation error: cascading dataset sequences are not allowed.");
+ }
datasetSequence = seq;
}
@Override
public AlignmentAnnotation[] getAnnotation()
{
- return annotation == null ? null : annotation
- .toArray(new AlignmentAnnotation[annotation.size()]);
+ return annotation == null ? null
+ : annotation
+ .toArray(new AlignmentAnnotation[annotation.size()]);
}
@Override
@Override
public SequenceI deriveSequence()
{
- SequenceI seq = new Sequence(this);
- if (datasetSequence != null)
- {
- // duplicate current sequence with same dataset
- seq.setDatasetSequence(datasetSequence);
- }
- else
+ Sequence seq = null;
+ if (datasetSequence == null)
{
if (isValidDatasetSequence())
{
// Use this as dataset sequence
+ seq = new Sequence(getName(), "", 1, -1);
seq.setDatasetSequence(this);
+ seq.initSeqFrom(this, getAnnotation());
+ return seq;
}
else
{
// Create a new, valid dataset sequence
- SequenceI ds = seq;
- ds.setSequence(AlignSeq.extractGaps(
- jalview.util.Comparison.GapChars, new String(sequence)));
- setDatasetSequence(ds);
- ds.setSequenceFeatures(getSequenceFeatures());
- seq = this; // and return this sequence as the derived sequence.
+ createDatasetSequence();
}
}
- return seq;
+ return new Sequence(this);
}
+ private boolean _isNa;
+
+ private int _seqhash = 0;
+
+ /**
+ * Answers false if the sequence is more than 85% nucleotide (ACGTU), else
+ * true
+ */
+ @Override
+ public boolean isProtein()
+ {
+ if (datasetSequence != null)
+ {
+ return datasetSequence.isProtein();
+ }
+ if (_seqhash != sequence.hashCode())
+ {
+ _seqhash = sequence.hashCode();
+ _isNa = Comparison.isNucleotide(this);
+ }
+ return !_isNa;
+ };
+
/*
* (non-Javadoc)
*
{
if (datasetSequence == null)
{
- datasetSequence = new Sequence(getName(), AlignSeq.extractGaps(
- jalview.util.Comparison.GapChars, getSequenceAsString()),
+ Sequence dsseq = new Sequence(getName(),
+ AlignSeq.extractGaps(jalview.util.Comparison.GapChars,
+ getSequenceAsString()),
getStart(), getEnd());
- datasetSequence.setSequenceFeatures(getSequenceFeatures());
- datasetSequence.setDescription(getDescription());
- setSequenceFeatures(null);
- // move database references onto dataset sequence
- datasetSequence.setDBRef(getDBRef());
- setDBRef(null);
- datasetSequence.setPDBId(getAllPDBEntries());
- setPDBId(null);
+
+ datasetSequence = dsseq;
+
+ dsseq.setDescription(description);
+ // move features and database references onto dataset sequence
+ dsseq.sequenceFeatureStore = sequenceFeatureStore;
+ sequenceFeatureStore = null;
+ dsseq.dbrefs = dbrefs;
+ dbrefs = null;
+ // TODO: search and replace any references to this sequence with
+ // references to the dataset sequence in Mappings on dbref
+ dsseq.pdbIds = pdbIds;
+ pdbIds = null;
datasetSequence.updatePDBIds();
if (annotation != null)
{
+ // annotation is cloned rather than moved, to preserve what's currently
+ // on the alignment
for (AlignmentAnnotation aa : annotation)
{
AlignmentAnnotation _aa = new AlignmentAnnotation(aa);
return null;
}
- Vector subset = new Vector();
- Enumeration e = annotation.elements();
+ Vector<AlignmentAnnotation> subset = new Vector<AlignmentAnnotation>();
+ Enumeration<AlignmentAnnotation> e = annotation.elements();
while (e.hasMoreElements())
{
- AlignmentAnnotation ann = (AlignmentAnnotation) e.nextElement();
+ AlignmentAnnotation ann = e.nextElement();
if (ann.label != null && ann.label.equals(label))
{
subset.addElement(ann);
e = subset.elements();
while (e.hasMoreElements())
{
- anns[i++] = (AlignmentAnnotation) e.nextElement();
+ anns[i++] = e.nextElement();
}
subset.removeAllElements();
return anns;
{
return false;
}
- Vector newpdb = new Vector();
- for (int i = 0; i < dbrefs.length; i++)
+ boolean added = false;
+ for (DBRefEntry dbr : dbrefs)
{
- if (DBRefSource.PDB.equals(dbrefs[i].getSource()))
+ if (DBRefSource.PDB.equals(dbr.getSource()))
{
- PDBEntry pdbe = new PDBEntry();
- pdbe.setId(dbrefs[i].getAccessionId());
- if (pdbIds == null || pdbIds.size() == 0)
- {
- newpdb.addElement(pdbe);
- }
- else
- {
- Enumeration en = pdbIds.elements();
- boolean matched = false;
- while (!matched && en.hasMoreElements())
- {
- PDBEntry anentry = (PDBEntry) en.nextElement();
- if (anentry.getId().equals(pdbe.getId()))
- {
- matched = true;
- }
- }
- if (!matched)
- {
- newpdb.addElement(pdbe);
- }
- }
- }
- }
- if (newpdb.size() > 0)
- {
- Enumeration en = newpdb.elements();
- while (en.hasMoreElements())
- {
- addPDBId((PDBEntry) en.nextElement());
+ /*
+ * 'Add' any PDB dbrefs as a PDBEntry - add is only performed if the
+ * PDB id is not already present in a 'matching' PDBEntry
+ * Constructor parses out a chain code if appended to the accession id
+ * (a fudge used to 'store' the chain code in the DBRef)
+ */
+ PDBEntry pdbe = new PDBEntry(dbr);
+ added |= addPDBId(pdbe);
}
- return true;
}
- return false;
+ return added;
}
@Override
if (entry.getSequenceFeatures() != null)
{
- SequenceFeature[] sfs = entry.getSequenceFeatures();
- for (int si = 0; si < sfs.length; si++)
+ List<SequenceFeature> sfs = entry.getSequenceFeatures();
+ for (SequenceFeature feature : sfs)
{
- SequenceFeature sf[] = (mp != null) ? mp.locateFeature(sfs[si])
- : new SequenceFeature[] { new SequenceFeature(sfs[si]) };
- if (sf != null && sf.length > 0)
+ SequenceFeature sf[] = (mp != null) ? mp.locateFeature(feature)
+ : new SequenceFeature[] { new SequenceFeature(feature) };
+ if (sf != null)
{
for (int sfi = 0; sfi < sf.length; sfi++)
{
// transfer PDB entries
if (entry.getAllPDBEntries() != null)
{
- Enumeration e = entry.getAllPDBEntries().elements();
+ Enumeration<PDBEntry> e = entry.getAllPDBEntries().elements();
while (e.hasMoreElements())
{
- PDBEntry pdb = (PDBEntry) e.nextElement();
+ PDBEntry pdb = e.nextElement();
addPDBId(pdb);
}
}
// transfer database references
- DBRefEntry[] entryRefs = entry.getDBRef();
+ DBRefEntry[] entryRefs = entry.getDBRefs();
if (entryRefs != null)
{
for (int r = 0; r < entryRefs.length; r++)
@Override
public PDBEntry getPDBEntry(String pdbIdStr)
{
- if (getDatasetSequence() == null
- || getDatasetSequence().getAllPDBEntries() == null)
+ if (getDatasetSequence() != null)
+ {
+ return getDatasetSequence().getPDBEntry(pdbIdStr);
+ }
+ if (pdbIds == null)
{
return null;
}
- List<PDBEntry> entries = getDatasetSequence().getAllPDBEntries();
+ List<PDBEntry> entries = getAllPDBEntries();
for (PDBEntry entry : entries)
{
if (entry.getId().equalsIgnoreCase(pdbIdStr))
}
@Override
- public void setSourceDBRef(DBRefEntryI dbRef)
+ public List<DBRefEntry> getPrimaryDBRefs()
+ {
+ if (datasetSequence != null)
+ {
+ return datasetSequence.getPrimaryDBRefs();
+ }
+ if (dbrefs == null || dbrefs.length == 0)
+ {
+ return Collections.emptyList();
+ }
+ synchronized (dbrefs)
+ {
+ List<DBRefEntry> primaries = new ArrayList<DBRefEntry>();
+ DBRefEntry[] tmp = new DBRefEntry[1];
+ for (DBRefEntry ref : dbrefs)
+ {
+ if (!ref.isPrimaryCandidate())
+ {
+ continue;
+ }
+ if (ref.hasMap())
+ {
+ MapList mp = ref.getMap().getMap();
+ if (mp.getFromLowest() > start || mp.getFromHighest() < end)
+ {
+ // map only involves a subsequence, so cannot be primary
+ continue;
+ }
+ }
+ // whilst it looks like it is a primary ref, we also sanity check type
+ if (DBRefUtils.getCanonicalName(DBRefSource.PDB)
+ .equals(DBRefUtils.getCanonicalName(ref.getSource())))
+ {
+ // PDB dbrefs imply there should be a PDBEntry associated
+ // TODO: tighten PDB dbrefs
+ // formally imply Jalview has actually downloaded and
+ // parsed the pdb file. That means there should be a cached file
+ // handle on the PDBEntry, and a real mapping between sequence and
+ // extracted sequence from PDB file
+ PDBEntry pdbentry = getPDBEntry(ref.getAccessionId());
+ if (pdbentry != null && pdbentry.getFile() != null)
+ {
+ primaries.add(ref);
+ }
+ continue;
+ }
+ // check standard protein or dna sources
+ tmp[0] = ref;
+ DBRefEntry[] res = DBRefUtils.selectDbRefs(!isProtein(), tmp);
+ if (res != null && res[0] == tmp[0])
+ {
+ primaries.add(ref);
+ continue;
+ }
+ }
+ return primaries;
+ }
+ }
+
+ /**
+ * {@inheritDoc}
+ */
+ @Override
+ public List<SequenceFeature> findFeatures(int fromColumn, int toColumn,
+ String... types)
{
- this.sourceDBRef = dbRef;
+ int startPos = findPosition(fromColumn - 1); // convert base 1 to base 0
+ int endPos = fromColumn == toColumn ? startPos
+ : findPosition(toColumn - 1);
+
+ List<SequenceFeature> result = getFeatures().findFeatures(startPos,
+ endPos, types);
+
+ /*
+ * if end column is gapped, endPos may be to the right,
+ * and we may have included adjacent or enclosing features;
+ * remove any that are not enclosing, non-contact features
+ */
+ if (endPos > this.end || Comparison.isGap(sequence[toColumn - 1]))
+ {
+ ListIterator<SequenceFeature> it = result.listIterator();
+ while (it.hasNext())
+ {
+ SequenceFeature sf = it.next();
+ int sfBegin = sf.getBegin();
+ int sfEnd = sf.getEnd();
+ int featureStartColumn = findIndex(sfBegin);
+ if (featureStartColumn > toColumn)
+ {
+ it.remove();
+ }
+ else if (featureStartColumn < fromColumn)
+ {
+ int featureEndColumn = sfEnd == sfBegin ? featureStartColumn
+ : findIndex(sfEnd);
+ if (featureEndColumn < fromColumn)
+ {
+ it.remove();
+ }
+ else if (featureEndColumn > toColumn && sf.isContactFeature())
+ {
+ /*
+ * remove an enclosing feature if it is a contact feature
+ */
+ it.remove();
+ }
+ }
+ }
+ }
+
+ return result;
}
+ /**
+ * Invalidates any stale cursors (forcing recalculation) by incrementing the
+ * token that has to match the one presented by the cursor
+ */
@Override
- public DBRefEntryI getSourceDBRef()
+ public void sequenceChanged()
{
- return this.sourceDBRef;
+ changeCount++;
}
+ /**
+ * {@inheritDoc}
+ */
+ @Override
+ public int replace(char c1, char c2)
+ {
+ if (c1 == c2)
+ {
+ return 0;
+ }
+ int count = 0;
+ synchronized (sequence)
+ {
+ for (int c = 0; c < sequence.length; c++)
+ {
+ if (sequence[c] == c1)
+ {
+ sequence[c] = c2;
+ count++;
+ }
+ }
+ }
+ if (count > 0)
+ {
+ sequenceChanged();
+ }
+
+ return count;
+ }
}
git://source.jalview.org / jalview.git / blobdiff