X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fdatamodel%2FSequence.java;h=a442cf063b55044acf047681310a84ff377a5392;hb=136c0793b90b72b928c4d77dc109dd5c644e00d3;hp=5c1fba5da96a747f3b1bfe763bdcbfe0d13b3541;hpb=be32c14cd8e48fe0a207cd7030cb9cd46f894678;p=jalview.git diff --git a/src/jalview/datamodel/Sequence.java b/src/jalview/datamodel/Sequence.java index 5c1fba5..a442cf0 100755 --- a/src/jalview/datamodel/Sequence.java +++ b/src/jalview/datamodel/Sequence.java @@ -21,13 +21,25 @@ 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.List; +import java.util.ListIterator; import java.util.Vector; +import com.stevesoft.pat.Regex; + import fr.orsay.lri.varna.models.rna.RNA; /** @@ -37,8 +49,13 @@ import fr.orsay.lri.varna.models.rna.RNA; * @author $author$ * @version $Revision$ */ -public class Sequence implements SequenceI +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; @@ -51,7 +68,7 @@ public class Sequence implements SequenceI int end; - Vector pdbIds; + Vector pdbIds; String vamsasId; @@ -72,8 +89,22 @@ public class Sequence implements SequenceI */ 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; /** * Creates a new Sequence object. @@ -90,29 +121,36 @@ public class Sequence implements SequenceI */ public Sequence(String name, String sequence, int start, int end) { - this.name = name; - this.sequence = sequence.toCharArray(); - this.start = start; - this.end = end; - parseId(); - checkValidRange(); + this(); + initSeqAndName(name, sequence.toCharArray(), start, end); } public Sequence(String name, char[] sequence, int start, int end) { - this.name = name; - this.sequence = sequence; - this.start = start; - this.end = end; + this(); + initSeqAndName(name, sequence, start, end); + } + + /** + * Stage 1 constructor - assign name, sequence, and set start and end fields. + * start and end are updated values from name2 if it ends with /start-end + * + * @param name2 + * @param sequence2 + * @param start2 + * @param end2 + */ + protected void initSeqAndName(String name2, char[] sequence2, int start2, + int end2) + { + this.name = name2; + this.sequence = sequence2; + this.start = start2; + this.end = end2; parseId(); 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) @@ -121,7 +159,7 @@ public class Sequence implements SequenceI .println("POSSIBLE IMPLEMENTATION ERROR: null sequence name passed to constructor."); name = ""; } - // Does sequence have the /start-end signiature? + // Does sequence have the /start-end signature? if (limitrx.search(name)) { name = limitrx.left(); @@ -159,6 +197,14 @@ public class Sequence implements SequenceI } /** + * default constructor + */ + private Sequence() + { + sequenceFeatureStore = new SequenceFeatures(); + } + + /** * Creates a new Sequence object. * * @param name @@ -172,12 +218,13 @@ public class Sequence implements SequenceI } /** - * 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) { @@ -196,26 +243,56 @@ public class Sequence implements SequenceI */ public Sequence(SequenceI seq, AlignmentAnnotation[] alAnnotation) { - this(seq.getName(), seq.getSequence(), seq.getStart(), seq.getEnd()); + 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) + { + char[] oseq = seq.getSequence(); + initSeqAndName(seq.getName(), Arrays.copyOf(oseq, oseq.length), + 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) { - addDBRef(new DBRefEntry(dbr[i])); + 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()) + { + addSequenceFeature(new SequenceFeature(sf)); } } + if (seq.getAnnotation() != null) { AlignmentAnnotation[] sqann = seq.getAnnotation(); @@ -241,124 +318,99 @@ public class Sequence implements SequenceI } } } - if (seq.getPDBId() != null) + if (seq.getAllPDBEntries() != null) { - Vector ids = seq.getPDBId(); - Enumeration e = ids.elements(); - while (e.hasMoreElements()) + Vector 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! - */ - public void setSequenceFeatures(SequenceFeature[] features) + @Override + public void setSequenceFeatures(List features) { - sequenceFeatures = features; + if (datasetSequence != null) + { + datasetSequence.setSequenceFeatures(features); + return; + } + sequenceFeatureStore = new SequenceFeatures(features); } - public synchronized void addSequenceFeature(SequenceFeature sf) + @Override + 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 */ - public SequenceFeature[] getSequenceFeatures() + @Override + 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; } - public void addPDBId(PDBEntry entry) + @Override + public boolean addPDBId(PDBEntry entry) { if (pdbIds == null) { - pdbIds = new Vector(); + pdbIds = new Vector(); + pdbIds.add(entry); + return true; } - if (!pdbIds.contains(entry)) + + for (PDBEntry pdbe : pdbIds) { - pdbIds.addElement(entry); + if (pdbe.updateFrom(entry)) + { + return false; + } } + pdbIds.addElement(entry); + return true; } /** @@ -367,7 +419,8 @@ public class Sequence implements SequenceI * @param id * DOCUMENT ME! */ - public void setPDBId(Vector id) + @Override + public void setPDBId(Vector id) { pdbIds = id; } @@ -377,9 +430,10 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ - public Vector getPDBId() + @Override + public Vector getAllPDBEntries() { - return pdbIds; + return pdbIds == null ? new Vector() : pdbIds; } /** @@ -387,6 +441,7 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @Override public String getDisplayId(boolean jvsuffix) { StringBuffer result = new StringBuffer(name); @@ -404,6 +459,7 @@ public class Sequence implements SequenceI * @param name * DOCUMENT ME! */ + @Override public void setName(String name) { this.name = name; @@ -415,6 +471,7 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @Override public String getName() { return this.name; @@ -426,6 +483,7 @@ public class Sequence implements SequenceI * @param start * DOCUMENT ME! */ + @Override public void setStart(int start) { this.start = start; @@ -436,6 +494,7 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @Override public int getStart() { return this.start; @@ -447,6 +506,7 @@ public class Sequence implements SequenceI * @param end * DOCUMENT ME! */ + @Override public void setEnd(int end) { this.end = end; @@ -457,6 +517,7 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @Override public int getEnd() { return this.end; @@ -467,6 +528,7 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @Override public int getLength() { return this.sequence.length; @@ -478,22 +540,27 @@ public class Sequence implements SequenceI * @param seq * DOCUMENT ME! */ + @Override public void setSequence(String seq) { this.sequence = seq.toCharArray(); checkValidRange(); + sequenceChanged(); } + @Override public String getSequenceAsString() { return new String(sequence); } + @Override public String getSequenceAsString(int start, int end) { return new String(getSequence(start, end)); } + @Override public char[] getSequence() { return sequence; @@ -504,6 +571,7 @@ public class Sequence implements SequenceI * * @see jalview.datamodel.SequenceI#getSequence(int, int) */ + @Override public char[] getSequence(int start, int end) { if (start < 0) @@ -557,16 +625,15 @@ public class Sequence implements SequenceI } /** - * 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]; } @@ -582,6 +649,7 @@ public class Sequence implements SequenceI * @param desc * DOCUMENT ME! */ + @Override public void setDescription(String desc) { this.description = desc; @@ -592,68 +660,333 @@ public class Sequence implements SequenceI * * @return DOCUMENT ME! */ + @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) + { + int endColumn = cursor == null ? 0 : cursor.lastColumnPosition; + 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(int pos, SequenceCursor curs) + { + if (!isValidCursor(curs)) + { + /* + * wrong or invalidated cursor, compute de novo + */ + return findIndex(pos); + } + + if (curs.residuePosition == pos) { - return i; + 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(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; + } + + /** * Returns an int array where indices correspond to each residue in the * sequence and the element value gives its position in the alignment * * @return int[SequenceI.getEnd()-SequenceI.getStart()+1] or null if no * residues in SequenceI object */ + @Override public int[] gapMap() { String seq = jalview.analysis.AlignSeq.extractGaps( @@ -715,8 +1048,41 @@ public class Sequence implements SequenceI { if (lastj != -1) { - map.add(new int[] - { lastj, j - 1 }); + map.add(new int[] { lastj, j - 1 }); + lastj = -1; + } + } + j++; + } + if (lastj != -1) + { + map.add(new int[] { lastj, j - 1 }); + lastj = -1; + } + return map; + } + + @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; } } @@ -724,8 +1090,7 @@ public class Sequence implements SequenceI } if (lastj != -1) { - map.add(new int[] - { lastj, j - 1 }); + map.set(lastj, j); lastj = -1; } return map; @@ -735,7 +1100,7 @@ public class Sequence implements SequenceI public void deleteChars(int i, int j) { int newstart = start, newend = end; - if (i >= sequence.length) + if (i >= sequence.length || i < 0) { return; } @@ -804,6 +1169,7 @@ public class Sequence implements SequenceI start = newstart; end = newend; sequence = tmp; + sequenceChanged(); } @Override @@ -834,6 +1200,7 @@ public class Sequence implements SequenceI } sequence = tmp; + sequenceChanged(); } @Override @@ -855,18 +1222,28 @@ public class Sequence implements SequenceI } @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; } @@ -874,39 +1251,56 @@ public class Sequence implements SequenceI @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; } @@ -923,7 +1317,6 @@ public class Sequence implements SequenceI .toArray(new AlignmentAnnotation[annotation.size()]); } - @Override public boolean hasAnnotation(AlignmentAnnotation ann) { @@ -935,7 +1328,7 @@ public class Sequence implements SequenceI { if (this.annotation == null) { - this.annotation = new Vector(); + this.annotation = new Vector(); } if (!this.annotation.contains(annotation)) { @@ -944,6 +1337,7 @@ public class Sequence implements SequenceI annotation.setSequenceRef(this); } + @Override public void removeAlignmentAnnotation(AlignmentAnnotation annotation) { if (this.annotation != null) @@ -979,56 +1373,80 @@ public class Sequence implements SequenceI @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 long _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) * * @see jalview.datamodel.SequenceI#createDatasetSequence() */ + @Override public SequenceI createDatasetSequence() { if (datasetSequence == null) { - datasetSequence = new Sequence(getName(), AlignSeq.extractGaps( + 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(getPDBId()); - 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); @@ -1049,6 +1467,7 @@ public class Sequence implements SequenceI * jalview.datamodel.SequenceI#setAlignmentAnnotation(AlignmmentAnnotation[] * annotations) */ + @Override public void setAlignmentAnnotation(AlignmentAnnotation[] annotations) { if (annotation != null) @@ -1075,11 +1494,11 @@ public class Sequence 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); @@ -1094,7 +1513,7 @@ public class Sequence implements SequenceI e = subset.elements(); while (e.hasMoreElements()) { - anns[i++] = (AlignmentAnnotation) e.nextElement(); + anns[i++] = e.nextElement(); } subset.removeAllElements(); return anns; @@ -1112,46 +1531,22 @@ public class Sequence implements SequenceI { 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 @@ -1171,13 +1566,12 @@ public class Sequence 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++) { @@ -1188,17 +1582,17 @@ public class Sequence implements SequenceI } // transfer PDB entries - if (entry.getPDBId() != null) + if (entry.getAllPDBEntries() != null) { - Enumeration e = entry.getPDBId().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.getDBRef(); + DBRefEntry[] entryRefs = entry.getDBRefs(); if (entryRefs != null) { for (int r = 0; r < entryRefs.length; r++) @@ -1222,6 +1616,7 @@ public class Sequence 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; @@ -1235,16 +1630,19 @@ public class Sequence implements SequenceI * 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) { rna = r; } + @Override public RNA getRNA() { return rna; @@ -1269,4 +1667,162 @@ public class Sequence implements SequenceI return result; } + @Override + public String toString() + { + return getDisplayId(false); + } + + @Override + public PDBEntry getPDBEntry(String pdbIdStr) + { + if (getDatasetSequence() != null) + { + return getDatasetSequence().getPDBEntry(pdbIdStr); + } + if (pdbIds == null) + { + return null; + } + List entries = getAllPDBEntries(); + for (PDBEntry entry : entries) + { + if (entry.getId().equalsIgnoreCase(pdbIdStr)) + { + return entry; + } + } + return null; + } + + @Override + public List getPrimaryDBRefs() + { + if (datasetSequence != null) + { + return datasetSequence.getPrimaryDBRefs(); + } + if (dbrefs == null || dbrefs.length == 0) + { + return Collections.emptyList(); + } + synchronized (dbrefs) + { + List primaries = new ArrayList(); + 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 findFeatures(int fromColumn, int toColumn, + 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 result = new ArrayList<>(); + if (datasetSequence != null) + { + result = datasetSequence.getFeatures().findFeatures(startPos, endPos, + types); + } + else + { + result = sequenceFeatureStore.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; + * remove any that are not enclosing, non-contact features + */ + if (endPos > this.end || Comparison.isGap(sequence[fromColumn - 1]) + || Comparison.isGap(sequence[toColumn - 1])) + { + ListIterator 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) + { + noOverlap = true; + } + if (noOverlap) + { + 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++; + } }