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.Iterator;
import java.util.List;
+import java.util.ListIterator;
import java.util.Vector;
import fr.orsay.lri.varna.models.rna.RNA;
String vamsasId;
- DBRefEntryI sourceDBRef;
-
DBRefEntry[] dbrefs;
RNA rna;
*/
Vector<AlignmentAnnotation> annotation;
- /**
- * The index of the sequence in a MSA
+ private SequenceFeaturesI 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.
*/
- int index = -1;
+ private SequenceCursor cursor;
- /** array of sequence features - may not be null for a valid sequence object */
- public SequenceFeature[] sequenceFeatures;
+ /*
+ * 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;
/**
* 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,}$");
-
- com.stevesoft.pat.Regex endrx = new com.stevesoft.pat.Regex("[0-9]{1,}$");
-
+ /**
+ * If 'name' ends in /i-j, where i >= j > 0 are integers, extracts i and j as
+ * start and end respectively and removes the suffix from the name
+ */
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?
- if (limitrx.search(name))
+ int slashPos = name.lastIndexOf('/');
+ if (slashPos > -1 && slashPos < name.length() - 1)
{
- name = limitrx.left();
- endrx.search(limitrx.stringMatched());
- setStart(Integer.parseInt(limitrx.stringMatched().substring(1,
- endrx.matchedFrom() - 1)));
- setEnd(Integer.parseInt(endrx.stringMatched()));
+ String suffix = name.substring(slashPos + 1);
+ String[] range = suffix.split("-");
+ if (range.length == 2)
+ {
+ try
+ {
+ int from = Integer.valueOf(range[0]);
+ int to = Integer.valueOf(range[1]);
+ if (from > 0 && to >= from)
+ {
+ name = name.substring(0, slashPos);
+ setStart(from);
+ setEnd(to);
+ checkValidRange();
+ }
+ } catch (NumberFormatException e)
+ {
+ // leave name unchanged if suffix is invalid
+ }
+ }
}
}
+ /**
+ * Ensures that 'end' is not before the end of the sequence, that is,
+ * (end-start+1) is at least as long as the count of ungapped positions. Note
+ * that end is permitted to be beyond the end of the sequence data.
+ */
void checkValidRange()
{
// Note: JAL-774 :
int endRes = 0;
for (int j = 0; j < sequence.length; j++)
{
- if (!jalview.util.Comparison.isGap(sequence[j]))
+ if (!Comparison.isGap(sequence[j]))
{
endRes++;
}
}
/**
+ * default constructor
+ */
+ private Sequence()
+ {
+ sequenceFeatureStore = new SequenceFeatures();
+ }
+
+ /**
* Creates a new Sequence object.
*
* @param name
*/
public Sequence(SequenceI seq, AlignmentAnnotation[] alAnnotation)
{
+ this();
initSeqFrom(seq, alAnnotation);
-
}
/**
protected void initSeqFrom(SequenceI seq,
AlignmentAnnotation[] alAnnotation)
{
- {
- char[] oseq = seq.getSequence();
- initSeqAndName(seq.getName(), Arrays.copyOf(oseq, oseq.length),
- 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 != datasetSequence)
{
setDatasetSequence(seq.getDatasetSequence());
}
- if (datasetSequence == null && seq.getDBRefs() != null)
+
+ /*
+ * only copy DBRefs and seqfeatures if we really are a dataset sequence
+ */
+ if (datasetSequence == null)
{
- // only copy DBRefs and seqfeatures if we really are a dataset sequence
- DBRefEntry[] dbr = seq.getDBRefs();
- for (int i = 0; i < dbr.length; i++)
- {
- addDBRef(new DBRefEntry(dbr[i]));
- }
- if (seq.getSequenceFeatures() != null)
+ if (seq.getDBRefs() != null)
{
- SequenceFeature[] sf = seq.getSequenceFeatures();
- for (int i = 0; i < sf.length; i++)
+ DBRefEntry[] dbr = seq.getDBRefs();
+ for (int i = 0; i < dbr.length; i++)
{
- addSequenceFeature(new SequenceFeature(sf[i]));
+ addDBRef(new DBRefEntry(dbr[i]));
}
}
+
+ /*
+ * make copies of any sequence features
+ */
+ for (SequenceFeature sf : seq.getSequenceFeatures())
+ {
+ addSequenceFeature(new SequenceFeature(sf));
+ }
}
+
if (seq.getAnnotation() != null)
{
AlignmentAnnotation[] sqann = seq.getAnnotation();
}
}
-
@Override
- public void setSequenceFeatures(SequenceFeature[] features)
+ public void setSequenceFeatures(List<SequenceFeature> features)
{
- if (datasetSequence == null)
- {
- sequenceFeatures = features;
- }
- else
+ if (datasetSequence != null)
{
- if (datasetSequence.getSequenceFeatures() != features
- && datasetSequence.getSequenceFeatures() != null
- && datasetSequence.getSequenceFeatures().length > 0)
- {
- new Exception(
- "Warning: JAL-2046 side effect ? Possible implementation error: overwriting dataset sequence features by setting sequence features on alignment")
- .printStackTrace();
- }
datasetSequence.setSequenceFeatures(features);
+ return;
}
+ sequenceFeatureStore = new SequenceFeatures(features);
}
@Override
- public synchronized void addSequenceFeature(SequenceFeature sf)
+ public synchronized boolean addSequenceFeature(SequenceFeature sf)
{
- if (sequenceFeatures==null && datasetSequence != null)
- {
- datasetSequence.addSequenceFeature(sf);
- return;
- }
- 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)
- {
- if (datasetSequence!=null) {
- datasetSequence.deleteFeature(sf);
- }
- 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 SequenceFeaturesI getFeatures()
+ {
+ return datasetSequence != null ? datasetSequence.getFeatures()
+ : sequenceFeatureStore;
}
@Override
- public void addPDBId(PDBEntry entry)
+ public boolean addPDBId(PDBEntry entry)
{
if (pdbIds == null)
{
- pdbIds = new Vector<PDBEntry>();
- }
- if (pdbIds.contains(entry))
- {
- updatePDBEntry(pdbIds.get(pdbIds.indexOf(entry)), entry);
- }
- else
- {
- pdbIds.addElement(entry);
+ pdbIds = new Vector<>();
+ pdbIds.add(entry);
+ return true;
}
- }
- 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 == null ? new Vector<PDBEntry>() : pdbIds;
+ return pdbIds == null ? new Vector<>() : pdbIds;
}
/**
}
/**
- * DOCUMENT ME!
+ * Sets the sequence name. If the name ends in /start-end, then the start-end
+ * values are parsed out and set, and the suffix is removed from the name.
*
- * @param name
- * DOCUMENT ME!
+ * @param theName
*/
@Override
- public void setName(String name)
+ public void setName(String theName)
{
- this.name = name;
+ this.name = theName;
this.parseId();
}
{
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!
+ * 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;
}
+ @Override
+ public void setGeneLoci(String speciesId, String assemblyId,
+ String chromosomeId, MapList map)
+ {
+ addDBRef(new DBRefEntry(speciesId, assemblyId, DBRefEntry.CHROMOSOME
+ + ":" + chromosomeId, new Mapping(map)));
+ }
+
/**
- * DOCUMENT ME!
+ * Returns the gene loci mapping for the sequence (may be null)
*
- * @return DOCUMENT ME!
+ * @return
+ */
+ @Override
+ public GeneLociI getGeneLoci()
+ {
+ DBRefEntry[] refs = getDBRefs();
+ if (refs != null)
+ {
+ for (final DBRefEntry ref : refs)
+ {
+ if (ref.isChromosome())
+ {
+ return new GeneLociI()
+ {
+ @Override
+ public String getSpeciesId()
+ {
+ return ref.getSource();
+ }
+
+ @Override
+ public String getAssemblyId()
+ {
+ return ref.getVersion();
+ }
+
+ @Override
+ public String getChromosomeId()
+ {
+ // strip off "chromosome:" prefix to chrId
+ return ref.getAccessionId().substring(
+ DBRefEntry.CHROMOSOME.length() + 1);
+ }
+
+ @Override
+ public MapList getMap()
+ {
+ return ref.getMap().getMap();
+ }
+ };
+ }
+ }
+ }
+ return null;
+ }
+
+ /**
+ * 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)
+ {
+ endColumn = column;
+ }
+
+ cursor = new SequenceCursor(this, residuePos, column, startColumn,
+ endColumn, this.changeCount);
+ }
+
+ /**
+ * 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(final 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)
+ {
+ break;
+ }
+ if (col == sequence.length)
+ {
+ col--; // return last column if we failed to reach pos
+ break;
+ }
+ if (!Comparison.isGap(sequence[col]))
+ {
+ newPos += delta;
+ }
+ }
+
+ col++; // convert back to base 1
+
+ /*
+ * only update cursor if we found the target position
+ */
+ if (newPos == pos)
{
- return i;
+ updateCursor(pos, col, curs.firstColumnPosition);
}
+
+ return col;
}
+ /**
+ * {@inheritDoc}
+ */
@Override
- public int findPosition(int i)
+ 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 ContiguousI 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
*
return map;
}
+ /**
+ * Build a bitset corresponding to sequence gaps
+ *
+ * @return a BitSet where set values correspond to gaps in the sequence
+ */
+ @Override
+ public BitSet gapBitset()
+ {
+ BitSet gaps = new BitSet(sequence.length);
+ int j = 0;
+ while (j < sequence.length)
+ {
+ if (jalview.util.Comparison.isGap(sequence[j]))
+ {
+ gaps.set(j);
+ }
+ j++;
+ }
+ return gaps;
+ }
+
@Override
public int[] findPositionMap()
{
@Override
public List<int[]> getInsertions()
{
- ArrayList<int[]> map = new ArrayList<int[]>();
+ ArrayList<int[]> map = new ArrayList<>();
int lastj = -1, j = 0;
int pos = start;
int seqlen = sequence.length;
}
@Override
- public void deleteChars(int i, int j)
+ 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(final int i, final int j)
{
int newstart = start, newend = end;
if (i >= sequence.length || i < 0)
boolean createNewDs = false;
// TODO: take a (second look) at the dataset creation validation method for
// the very large sequence case
- int eindex = -1, sindex = -1;
- boolean ecalc = false, scalc = false;
+ int startIndex = findIndex(start) - 1;
+ int endIndex = findIndex(end) - 1;
+ int startDeleteColumn = -1; // for dataset sequence deletions
+ int deleteCount = 0;
+
for (int s = i; s < j; s++)
{
- if (jalview.schemes.ResidueProperties.aaIndex[sequence[s]] != 23)
+ if (Comparison.isGap(sequence[s]))
+ {
+ continue;
+ }
+ deleteCount++;
+ if (startDeleteColumn == -1)
+ {
+ startDeleteColumn = findPosition(s) - start;
+ }
+ if (createNewDs)
+ {
+ newend--;
+ }
+ else
{
- if (createNewDs)
+ if (startIndex == s)
{
- newend--;
+ /*
+ * deleting characters from start of sequence; new start is the
+ * sequence position of the next column (position to the right
+ * if the column position is gapped)
+ */
+ newstart = findPosition(j);
+ break;
}
else
{
- if (!scalc)
+ if (endIndex < j)
{
- sindex = findIndex(start) - 1;
- scalc = true;
- }
- if (sindex == s)
- {
- // delete characters including start of sequence
- newstart = findPosition(j);
- break; // don't need to search for any more residue characters.
+ /*
+ * deleting characters at end of sequence; new end is the sequence
+ * position of the column before the deletion; subtract 1 if this is
+ * gapped since findPosition returns the next sequence position
+ */
+ newend = findPosition(i - 1);
+ if (Comparison.isGap(sequence[i - 1]))
+ {
+ newend--;
+ }
+ break;
}
else
{
- // delete characters after start.
- if (!ecalc)
- {
- eindex = findIndex(end) - 1;
- ecalc = true;
- }
- if (eindex < j)
- {
- // delete characters at end of sequence
- newend = findPosition(i - 1);
- break; // don't need to search for any more residue characters.
- }
- else
- {
- createNewDs = true;
- newend--; // decrease end position by one for the deleted residue
- // and search further
- }
+ createNewDs = true;
+ newend--;
}
}
}
}
- // deletion occured in the middle of the sequence
+
if (createNewDs && this.datasetSequence != null)
{
- // construct a new sequence
+ /*
+ * if deletion occured in the middle of the sequence,
+ * construct a new dataset sequence and delete the residues
+ * that were deleted from the aligned sequence
+ */
Sequence ds = new Sequence(datasetSequence);
+ ds.deleteChars(startDeleteColumn, startDeleteColumn + deleteCount);
+ datasetSequence = ds;
// TODO: remove any non-inheritable properties ?
// TODO: create a sequence mapping (since there is a relation here ?)
- ds.deleteChars(i, j);
- datasetSequence = ds;
}
start = newstart;
end = newend;
sequence = tmp;
+ sequenceChanged();
}
@Override
}
sequence = tmp;
+ sequenceChanged();
}
@Override
@Override
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
@Override
public void addDBRef(DBRefEntry entry)
{
- // TODO add to dataset sequence instead if there is one?
+ if (datasetSequence != null)
+ {
+ datasetSequence.addDBRef(entry);
+ return;
+ }
+
if (dbrefs == null)
{
dbrefs = new DBRefEntry[0];
temp[temp.length - 1] = entry;
dbrefs = temp;
+
+ DBRefUtils.ensurePrimaries(this);
}
@Override
public void setDatasetSequence(SequenceI seq)
{
- // TODO check for circular reference before setting?
+ 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
{
if (this.annotation == null)
{
- this.annotation = new Vector<AlignmentAnnotation>();
+ this.annotation = new Vector<>();
}
if (!this.annotation.contains(annotation))
{
@Override
public SequenceI deriveSequence()
{
- Sequence seq=null;
+ Sequence seq = null;
if (datasetSequence == null)
{
if (isValidDatasetSequence())
else
{
// Create a new, valid dataset sequence
- createDatasetSequence();
+ createDatasetSequence();
}
}
return new Sequence(this);
private boolean _isNa;
- private long _seqhash = 0;
+ private int _seqhash = 0;
+ /**
+ * Answers false if the sequence is more than 85% nucleotide (ACGTU), else
+ * true
+ */
@Override
public boolean isProtein()
{
if (_seqhash != sequence.hashCode())
{
_seqhash = sequence.hashCode();
- _isNa=jalview.util.Comparison.isNucleotide(new SequenceI[] { this });
+ _isNa = Comparison.isNucleotide(this);
}
return !_isNa;
};
{
if (datasetSequence == null)
{
- Sequence dsseq = 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 = dsseq;
dsseq.setDescription(description);
// move features and database references onto dataset sequence
- dsseq.sequenceFeatures = sequenceFeatures;
- sequenceFeatures=null;
+ dsseq.sequenceFeatureStore = sequenceFeatureStore;
+ sequenceFeatureStore = null;
dsseq.dbrefs = dbrefs;
- dbrefs=null;
+ dbrefs = null;
// TODO: search and replace any references to this sequence with
// references to the dataset sequence in Mappings on dbref
dsseq.pdbIds = pdbIds;
return null;
}
- Vector subset = new Vector();
- Enumeration e = annotation.elements();
+ Vector<AlignmentAnnotation> subset = new Vector<>();
+ 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++)
- {
- if (DBRefSource.PDB.equals(dbrefs[i].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)
+ boolean added = false;
+ for (DBRefEntry dbr : dbrefs)
{
- Enumeration en = newpdb.elements();
- while (en.hasMoreElements())
+ if (DBRefSource.PDB.equals(dbr.getSource()))
{
- 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);
}
}
}
}
- /**
- * @return The index (zero-based) on this sequence in the MSA. It returns
- * {@code -1} if this information is not available.
- */
- @Override
- public int getIndex()
- {
- return index;
- }
-
- /**
- * Defines the position of this sequence in the MSA. Use the value {@code -1}
- * if this information is undefined.
- *
- * @param The
- * position for this sequence. This value is zero-based (zero for
- * this first sequence)
- */
- @Override
- public void setIndex(int value)
- {
- index = value;
- }
-
@Override
public void setRNA(RNA r)
{
public List<AlignmentAnnotation> getAlignmentAnnotations(String calcId,
String label)
{
- List<AlignmentAnnotation> result = new ArrayList<AlignmentAnnotation>();
+ List<AlignmentAnnotation> result = new ArrayList<>();
if (this.annotation != null)
{
for (AlignmentAnnotation ann : annotation)
return null;
}
-
@Override
public List<DBRefEntry> getPrimaryDBRefs()
{
- if (datasetSequence!=null)
+ if (datasetSequence != null)
{
return datasetSequence.getPrimaryDBRefs();
}
- if (dbrefs==null || dbrefs.length==0)
+ if (dbrefs == null || dbrefs.length == 0)
{
- return Arrays.asList(new DBRefEntry[0]);
+ return Collections.emptyList();
}
synchronized (dbrefs)
{
- List<DBRefEntry> primaries = new ArrayList<DBRefEntry>();
- DBRefEntry tmp[] = new DBRefEntry[1], res[] = null;
+ List<DBRefEntry> primaries = new ArrayList<>();
+ DBRefEntry[] tmp = new DBRefEntry[1];
for (DBRefEntry ref : dbrefs)
{
- if (!ref.isPrimary())
+ if (!ref.isPrimaryCandidate())
{
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())))
+ 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 downlaoded and
+ // 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.getType() != null
- && pdbentry.getType().equalsIgnoreCase("PDB"))
+ if (pdbentry != null && pdbentry.getFile() != null)
{
primaries.add(ref);
}
}
// check standard protein or dna sources
tmp[0] = ref;
- res = DBRefUtils.selectDbRefs(!isProtein(), tmp);
+ DBRefEntry[] res = DBRefUtils.selectDbRefs(!isProtein(), tmp);
if (res != null && res[0] == tmp[0])
{
primaries.add(ref);
}
}
+ /**
+ * {@inheritDoc}
+ */
+ @Override
+ public List<SequenceFeature> findFeatures(int fromColumn, int toColumn,
+ String... types)
+ {
+ 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
+ */
+ boolean endColumnIsGapped = toColumn > 0 && toColumn <= sequence.length
+ && Comparison.isGap(sequence[toColumn - 1]);
+ if (endPos > this.end || endColumnIsGapped)
+ {
+ 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 void sequenceChanged()
+ {
+ 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;
+ }
+
+ @Override
+ public String getSequenceStringFromIterator(Iterator<int[]> it)
+ {
+ StringBuilder newSequence = new StringBuilder();
+ while (it.hasNext())
+ {
+ int[] block = it.next();
+ if (it.hasNext())
+ {
+ newSequence.append(getSequence(block[0], block[1] + 1));
+ }
+ else
+ {
+ newSequence.append(getSequence(block[0], block[1]));
+ }
+ }
+
+ return newSequence.toString();
+ }
+
+ @Override
+ public int firstResidueOutsideIterator(Iterator<int[]> regions)
+ {
+ int start = 0;
+
+ if (!regions.hasNext())
+ {
+ return findIndex(getStart()) - 1;
+ }
+
+ // Simply walk along the sequence whilst watching for region
+ // boundaries
+ int hideStart = getLength();
+ int hideEnd = -1;
+ boolean foundStart = false;
+
+ // step through the non-gapped positions of the sequence
+ for (int i = getStart(); i <= getEnd() && (!foundStart); i++)
+ {
+ // get alignment position of this residue in the sequence
+ int p = findIndex(i) - 1;
+
+ // update region start/end
+ while (hideEnd < p && regions.hasNext())
+ {
+ int[] region = regions.next();
+ hideStart = region[0];
+ hideEnd = region[1];
+ }
+ if (hideEnd < p)
+ {
+ hideStart = getLength();
+ }
+ // update boundary for sequence
+ if (p < hideStart)
+ {
+ start = p;
+ foundStart = true;
+ }
+ }
+
+ if (foundStart)
+ {
+ return start;
+ }
+ // otherwise, sequence was completely hidden
+ return 0;
+ }
}