X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fdatamodel%2FSequence.java;h=950ee0515ce493393a9051afae74a768749d6eac;hb=17e4ea278bc9a5fb280db1252ce78b7a295215f5;hp=29d114de1ba8b7522c5ae5001955fab4abf6d013;hpb=1b4069135fa518efcb11d9cb729e5031347b8fe0;p=jalview.git diff --git a/src/jalview/datamodel/Sequence.java b/src/jalview/datamodel/Sequence.java index 29d114d..950ee05 100755 --- a/src/jalview/datamodel/Sequence.java +++ b/src/jalview/datamodel/Sequence.java @@ -21,48 +21,77 @@ package jalview.datamodel; 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; /** * - * Implements the SequenceI interface for a char[] based sequence object. - * - * @author $author$ - * @version $Revision$ + * Implements the SequenceI interface for a char[] based sequence object */ public class Sequence extends ASequence implements SequenceI { + + /** + * A subclass that gives us access to modCount, which tracks whether there + * have been any changes. We use this to update + * + * @author hansonr + * + * @param + */ + @SuppressWarnings("serial") + public class DBModList extends ArrayList + { + + protected int getModCount() + { + return modCount; + } + + } + SequenceI datasetSequence; - String name; + private String name; private char[] sequence; - String description; + private String description; - int start; + private int start; - int end; + private int end; - Vector pdbIds; + private Vector pdbIds; - String vamsasId; + private String vamsasId; - DBRefEntry[] dbrefs; + private DBModList dbrefs; // controlled access - RNA rna; + /** + * a flag to let us know that elements have changed in dbrefs + * + * @author Bob Hanson + */ + private int refModCount = 0; + + private RNA rna; /** * This annotation is displayed below the alignment but the positions are tied @@ -70,15 +99,24 @@ public class Sequence extends ASequence implements SequenceI * * TODO: change to List<> */ - Vector annotation; + private Vector 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. @@ -95,11 +133,13 @@ public class Sequence extends ASequence implements SequenceI */ 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); } @@ -123,30 +163,49 @@ public class Sequence extends ASequence implements SequenceI 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 : @@ -155,7 +214,7 @@ public class Sequence extends ASequence implements SequenceI int endRes = 0; for (int j = 0; j < sequence.length; j++) { - if (!jalview.util.Comparison.isGap(sequence[j])) + if (!Comparison.isGap(sequence[j])) { endRes++; } @@ -174,6 +233,14 @@ public class Sequence extends ASequence implements SequenceI } /** + * default constructor + */ + private Sequence() + { + sequenceFeatureStore = new SequenceFeatures(); + } + + /** * Creates a new Sequence object. * * @param name @@ -212,8 +279,8 @@ public class Sequence extends ASequence implements SequenceI */ public Sequence(SequenceI seq, AlignmentAnnotation[] alAnnotation) { + this(); initSeqFrom(seq, alAnnotation); - } /** @@ -229,34 +296,38 @@ public class Sequence extends ASequence implements SequenceI 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) + List dbr = seq.getDBRefs(); + if (dbr != null) { - SequenceFeature[] sf = seq.getSequenceFeatures(); - for (int i = 0; i < sf.length; i++) + for (int i = 0, n = dbr.size(); i < n; i++) { - addSequenceFeature(new SequenceFeature(sf[i])); + addDBRef(new DBRefEntry(dbr.get(i))); } } + + /* + * make copies of any sequence features + */ + for (SequenceFeature sf : seq.getSequenceFeatures()) + { + addSequenceFeature(new SequenceFeature(sf)); + } } + if (seq.getAnnotation() != null) { AlignmentAnnotation[] sqann = seq.getAnnotation(); @@ -292,123 +363,68 @@ public class Sequence extends ASequence implements SequenceI } } - @Override - public void setSequenceFeatures(SequenceFeature[] features) + public void setSequenceFeatures(List 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 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 @@ -416,7 +432,7 @@ public class Sequence extends ASequence implements SequenceI { if (pdbIds == null) { - pdbIds = new Vector(); + pdbIds = new Vector<>(); pdbIds.add(entry); return true; } @@ -452,7 +468,7 @@ public class Sequence extends ASequence implements SequenceI @Override public Vector getAllPDBEntries() { - return pdbIds == null ? new Vector() : pdbIds; + return pdbIds == null ? new Vector<>() : pdbIds; } /** @@ -473,15 +489,15 @@ public class Sequence extends ASequence implements SequenceI } /** - * 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(); } @@ -564,6 +580,7 @@ public class Sequence extends ASequence implements SequenceI { this.sequence = seq.toCharArray(); checkValidRange(); + sequenceChanged(); } @Override @@ -581,7 +598,9 @@ public class Sequence extends ASequence implements SequenceI @Override public char[] getSequence() { - return sequence; + // return sequence; + return sequence == null ? null + : Arrays.copyOf(sequence, sequence.length); } /* @@ -662,10 +681,10 @@ public class Sequence extends ASequence implements SequenceI } /** - * 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) @@ -673,10 +692,40 @@ public class Sequence extends ASequence implements SequenceI this.description = desc; } + @Override + public void setGeneLoci(String speciesId, String assemblyId, + String chromosomeId, MapList map) + { + addDBRef(new GeneLocus(speciesId, assemblyId, 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() + { + List refs = getDBRefs(); + if (refs != null) + { + for (final DBRefEntry ref : refs) + { + if (ref instanceof GeneLociI) + { + return (GeneLociI) ref; + } + } + } + return null; + } + + /** + * Answers the description + * + * @return */ @Override public String getDescription() @@ -684,58 +733,382 @@ public class Sequence extends ASequence implements SequenceI 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 base 0 + 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 * @@ -764,6 +1137,27 @@ public class Sequence extends ASequence implements SequenceI 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() { @@ -787,9 +1181,9 @@ public class Sequence extends ASequence implements SequenceI @Override public List getInsertions() { - ArrayList map = new ArrayList(); + ArrayList map = new ArrayList<>(); int lastj = -1, j = 0; - int pos = start; + // int pos = start; int seqlen = sequence.length; while ((j < seqlen)) { @@ -819,7 +1213,41 @@ public class Sequence extends ASequence implements SequenceI } @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) @@ -831,66 +1259,81 @@ public class Sequence extends ASequence implements SequenceI 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; - for (int s = i; s < j; s++) + + 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 < sequence.length; 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 @@ -921,6 +1364,7 @@ public class Sequence extends ASequence implements SequenceI } sequence = tmp; + sequenceChanged(); } @Override @@ -941,24 +1385,22 @@ public class Sequence extends ASequence implements SequenceI vamsasId = id; } + @Deprecated @Override - public void setDBRefs(DBRefEntry[] dbref) + public void setDBRefs(DBModList newDBrefs) { if (dbrefs == null && datasetSequence != null && this != datasetSequence) { - datasetSequence.setDBRefs(dbref); + datasetSequence.setDBRefs(newDBrefs); return; } - dbrefs = dbref; - if (dbrefs != null) - { - DBRefUtils.ensurePrimaries(this); - } + dbrefs = newDBrefs; + refModCount = 0; } @Override - public DBRefEntry[] getDBRefs() + public DBModList getDBRefs() { if (dbrefs == null && datasetSequence != null && this != datasetSequence) @@ -979,12 +1421,12 @@ public class Sequence extends ASequence implements SequenceI if (dbrefs == null) { - dbrefs = new DBRefEntry[0]; + dbrefs = new DBModList<>(); } - for (DBRefEntryI dbr : dbrefs) + for (int ib = 0, nb = dbrefs.size(); ib < nb; ib++) { - if (dbr.updateFrom(entry)) + if (dbrefs.get(ib).updateFrom(entry)) { /* * found a dbref that either matched, or could be @@ -994,18 +1436,19 @@ public class Sequence extends ASequence implements SequenceI } } - /* - * 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; + // /// BH OUCH! + // /* + // * extend the array to make room for one more + // */ + // // TODO use an ArrayList instead + // int j = dbrefs.length; + // List temp = new DBRefEntry[j + 1]; + // System.arraycopy(dbrefs, 0, temp, 0, j); + // temp[temp.length - 1] = entry; + // + // dbrefs = temp; - dbrefs = temp; - - DBRefUtils.ensurePrimaries(this); + dbrefs.add(entry); } @Override @@ -1013,12 +1456,12 @@ public class Sequence extends ASequence implements SequenceI { if (seq == this) { - throw new Error( + throw new IllegalArgumentException( "Implementation Error: self reference passed to SequenceI.setDatasetSequence"); } if (seq != null && seq.getDatasetSequence() != null) { - throw new Error( + throw new IllegalArgumentException( "Implementation error: cascading dataset sequences are not allowed."); } datasetSequence = seq; @@ -1033,8 +1476,9 @@ public class Sequence extends ASequence implements SequenceI @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 @@ -1048,7 +1492,7 @@ public class Sequence extends ASequence implements SequenceI { if (this.annotation == null) { - this.annotation = new Vector(); + this.annotation = new Vector<>(); } if (!this.annotation.contains(annotation)) { @@ -1093,7 +1537,7 @@ public class Sequence extends ASequence implements SequenceI @Override public SequenceI deriveSequence() { - Sequence seq=null; + Sequence seq = null; if (datasetSequence == null) { if (isValidDatasetSequence()) @@ -1107,7 +1551,7 @@ public class Sequence extends ASequence implements SequenceI else { // Create a new, valid dataset sequence - createDatasetSequence(); + createDatasetSequence(); } } return new Sequence(this); @@ -1115,8 +1559,14 @@ public class Sequence extends ASequence implements SequenceI private boolean _isNa; - private long _seqhash = 0; + private int _seqhash = 0; + + private List primaryRefs; + /** + * Answers false if the sequence is more than 85% nucleotide (ACGTU), else + * true + */ @Override public boolean isProtein() { @@ -1127,10 +1577,10 @@ public class Sequence extends ASequence implements SequenceI if (_seqhash != sequence.hashCode()) { _seqhash = sequence.hashCode(); - _isNa=jalview.util.Comparison.isNucleotide(new SequenceI[] { this }); + _isNa = Comparison.isNucleotide(this); } return !_isNa; - }; + } /* * (non-Javadoc) @@ -1142,18 +1592,19 @@ public class Sequence extends ASequence implements SequenceI { 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; @@ -1210,11 +1661,11 @@ public class Sequence extends ASequence implements SequenceI return null; } - Vector subset = new Vector(); - Enumeration e = annotation.elements(); + Vector subset = new Vector<>(); + Enumeration 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); @@ -1229,7 +1680,7 @@ public class Sequence extends ASequence implements SequenceI e = subset.elements(); while (e.hasMoreElements()) { - anns[i++] = (AlignmentAnnotation) e.nextElement(); + anns[i++] = e.nextElement(); } subset.removeAllElements(); return anns; @@ -1243,13 +1694,14 @@ public class Sequence extends ASequence implements SequenceI // TODO: could merge DBRefs return datasetSequence.updatePDBIds(); } - if (dbrefs == null || dbrefs.length == 0) + if (dbrefs == null || dbrefs.size() == 0) { return false; } boolean added = false; - for (DBRefEntry dbr : dbrefs) + for (int ib = 0, nb = dbrefs.size(); ib < nb; ib++) { + DBRefEntry dbr = dbrefs.get(ib); if (DBRefSource.PDB.equals(dbr.getSource())) { /* @@ -1282,12 +1734,13 @@ public class Sequence extends ASequence implements SequenceI if (entry.getSequenceFeatures() != null) { - SequenceFeature[] sfs = entry.getSequenceFeatures(); - for (int si = 0; si < sfs.length; si++) + List 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++) { @@ -1300,20 +1753,20 @@ public class Sequence extends ASequence implements SequenceI // transfer PDB entries if (entry.getAllPDBEntries() != null) { - Enumeration e = entry.getAllPDBEntries().elements(); + Enumeration e = entry.getAllPDBEntries().elements(); while (e.hasMoreElements()) { - PDBEntry pdb = (PDBEntry) e.nextElement(); + PDBEntry pdb = e.nextElement(); addPDBId(pdb); } } // transfer database references - DBRefEntry[] entryRefs = entry.getDBRefs(); + List entryRefs = entry.getDBRefs(); if (entryRefs != null) { - for (int r = 0; r < entryRefs.length; r++) + for (int r = 0, n = entryRefs.size(); r < n; r++) { - DBRefEntry newref = new DBRefEntry(entryRefs[r]); + DBRefEntry newref = new DBRefEntry(entryRefs.get(r)); if (newref.getMap() != null && mp != null) { // remap ref using our local mapping @@ -1328,30 +1781,6 @@ public class Sequence extends ASequence implements SequenceI } } - /** - * @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) { @@ -1368,7 +1797,7 @@ public class Sequence extends ASequence implements SequenceI public List getAlignmentAnnotations(String calcId, String label) { - List result = new ArrayList(); + List result = new ArrayList<>(); if (this.annotation != null) { for (AlignmentAnnotation ann : annotation) @@ -1411,24 +1840,38 @@ public class Sequence extends ASequence implements SequenceI return null; } + private List tmpList; @Override public List getPrimaryDBRefs() { - if (datasetSequence!=null) + if (datasetSequence != null) { return datasetSequence.getPrimaryDBRefs(); } - if (dbrefs==null || dbrefs.length==0) + if (dbrefs == null || dbrefs.size() == 0) { return Collections.emptyList(); } synchronized (dbrefs) { - List primaries = new ArrayList(); - DBRefEntry[] tmp = new DBRefEntry[1]; - for (DBRefEntry ref : dbrefs) + if (refModCount == dbrefs.getModCount() && primaryRefs != null) + { + return primaryRefs; // no changes + } + refModCount = dbrefs.getModCount(); + List primaries = (primaryRefs == null + ? (primaryRefs = new ArrayList<>()) + : primaryRefs); + primaries.clear(); + if (tmpList == null) { + tmpList = new ArrayList<>(); + tmpList.add(null); // for replacement + } + for (int i = 0, n = dbrefs.size(); i < n; i++) + { + DBRefEntry ref = dbrefs.get(i); if (!ref.isPrimaryCandidate()) { continue; @@ -1443,8 +1886,8 @@ public class Sequence extends ASequence implements SequenceI } } // 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 (DBRefSource.PDB_CANONICAL_NAME + .equals(ref.getCanonicalSourceName())) { // PDB dbrefs imply there should be a PDBEntry associated // TODO: tighten PDB dbrefs @@ -1453,23 +1896,203 @@ public class Sequence extends ASequence implements SequenceI // 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) + if (pdbentry == null || pdbentry.getFile() == null) { - primaries.add(ref); + continue; } - continue; } - // check standard protein or dna sources - tmp[0] = ref; - DBRefEntry[] res = DBRefUtils.selectDbRefs(!isProtein(), tmp); - if (res != null && res[0] == tmp[0]) + else { - primaries.add(ref); - continue; + // check standard protein or dna sources + tmpList.set(0, ref); + List res = DBRefUtils.selectDbRefs(!isProtein(), + tmpList); + if (res == null || res.get(0) != tmpList.get(0)) + { + continue; + } } + primaries.add(ref); } + + // version must be not null, as otherwise it will not be a candidate, + // above + DBRefUtils.ensurePrimaries(this, primaries); return primaries; } } + /** + * {@inheritDoc} + */ + @Override + public List 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 result = getFeatures().findFeatures(startPos, + endPos, types); + if (datasetSequence != null) + { + result = datasetSequence.getFeatures().findFeatures(startPos, endPos, + types); + } + else + { + result = sequenceFeatureStore.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 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 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 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; + } }