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;
*/
Vector<AlignmentAnnotation> annotation;
- /**
- * The index of the sequence in a MSA
- */
- int index = -1;
-
- private SequenceFeatures sequenceFeatureStore;
+ private SequenceFeaturesI sequenceFeatureStore;
/*
* A cursor holding the approximate current view position to the sequence,
checkValidRange();
}
+ /**
+ * 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++;
}
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)
}
/**
- * 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();
}
@Override
public char[] getSequence()
{
- return sequence;
+ // return sequence;
+ return sequence == null ? null : Arrays.copyOf(sequence,
+ sequence.length);
}
/*
*/
protected void updateCursor(int residuePos, int column, int startColumn)
{
- int endColumn = cursor == null ? 0 : cursor.lastColumnPosition;
+ /*
+ * preserve end residue column provided cursor was valid
+ */
+ int endColumn = isValidCursor(cursor) ? cursor.lastColumnPosition : 0;
if (residuePos == this.end)
{
endColumn = column;
return curs.residuePosition; // easy case :-)
}
- if (curs.lastColumnPosition > 0 && curs.lastColumnPosition <= col)
+ if (curs.lastColumnPosition > 0 && curs.lastColumnPosition < col)
{
/*
* sequence lies entirely to the left of col
return end + 1;
}
- if (curs.firstColumnPosition > 0 && curs.firstColumnPosition >= col)
+ if (curs.firstColumnPosition > 0 && curs.firstColumnPosition > col)
{
/*
* sequence lies entirely to the right of col
}
/**
+ * {@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 void deleteChars(int i, int j)
+ 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]))
{
- if (createNewDs)
+ continue;
+ }
+ deleteCount++;
+ if (startDeleteColumn == -1)
+ {
+ startDeleteColumn = findPosition(s) - start;
+ }
+ if (createNewDs)
+ {
+ newend--;
+ }
+ else
+ {
+ 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;
@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
private boolean _isNa;
- private long _seqhash = 0;
+ private int _seqhash = 0;
/**
* Answers false if the sequence is more than 85% nucleotide (ACGTU), else
{
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;
List<SequenceFeature> sfs = entry.getSequenceFeatures();
for (SequenceFeature feature : sfs)
{
- SequenceFeature sf[] = (mp != null) ? mp.locateFeature(feature)
+ SequenceFeature sf[] = (mp != null) ? mp.locateFeature(feature)
: new SequenceFeature[] { new SequenceFeature(feature) };
if (sf != null)
{
}
}
- /**
- * @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)
{
}
}
// 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
String... types)
{
int startPos = findPosition(fromColumn - 1); // convert base 1 to base 0
- int endPos = findPosition(toColumn - 1);
- // to trace / debug behaviour:
- // System.out
- // .println(String
- // .format("%s.findFeatures columns [%d-%d] positions [%d-%d] leaves cursor %s",
- // getName(), fromColumn, toColumn, startPos,
- // endPos, cursor));
- List<SequenceFeature> result = new ArrayList<>();
- if (datasetSequence != null)
- {
- result = datasetSequence.getFeatures().findFeatures(startPos, endPos,
- types);
- }
- else
- {
- result = sequenceFeatureStore.findFeatures(startPos, endPos, types);
- }
+ int endPos = fromColumn == toColumn ? startPos
+ : findPosition(toColumn - 1);
+
+ List<SequenceFeature> result = getFeatures().findFeatures(startPos,
+ endPos, types);
/*
- * if the start or end column is gapped, startPos or endPos may be to the
- * left or right, and we may have included adjacent or enclosing features;
+ * 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[fromColumn - 1])
- || Comparison.isGap(sequence[toColumn - 1]))
+ 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 featureStartColumn = findIndex(sf.getBegin());
- int featureEndColumn = findIndex(sf.getEnd());
- boolean noOverlap = featureStartColumn > toColumn
- || featureEndColumn < fromColumn;
-
- /*
- * reject an 'enclosing' feature if it is actually a contact feature
- */
- if (sf.isContactFeature() && featureStartColumn < fromColumn
- && featureEndColumn > toColumn)
+ int sfBegin = sf.getBegin();
+ int sfEnd = sf.getEnd();
+ int featureStartColumn = findIndex(sfBegin);
+ if (featureStartColumn > toColumn)
{
- noOverlap = true;
+ it.remove();
}
- if (noOverlap)
+ else if (featureStartColumn < fromColumn)
{
- it.remove();
+ 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();
+ }
}
}
}
{
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;
+ }
}