From: Jim Procter <jprocter@issues.jalview.org>
Date: Thu, 16 Nov 2017 16:45:35 +0000 (+0000)
Subject: Merge branch 'bug/JAL-2829deleteCharsWithGaps' into develop
X-Git-Tag: Release_2_10_3b1~30
X-Git-Url: http://source.jalview.org/gitweb/?a=commitdiff_plain;h=2b4cc5fb6217217761b4a8bd2529e51b4775a57d;hp=-c;p=jalview.git

Merge branch 'bug/JAL-2829deleteCharsWithGaps' into develop
---

2b4cc5fb6217217761b4a8bd2529e51b4775a57d
diff --combined src/jalview/datamodel/Sequence.java
index d254a17,2feb9f6..15d1378
--- a/src/jalview/datamodel/Sequence.java
+++ b/src/jalview/datamodel/Sequence.java
@@@ -77,6 -77,11 +77,6 @@@ public class Sequence extends ASequenc
     */
    Vector<AlignmentAnnotation> annotation;
  
 -  /**
 -   * The index of the sequence in a MSA
 -   */
 -  int index = -1;
 -
    private SequenceFeaturesI sequenceFeatureStore;
  
    /*
@@@ -1160,7 -1165,7 +1160,7 @@@
    }
  
    @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)
@@@ -1172,62 -1177,75 +1172,75 @@@
      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)
-           {
-             sindex = findIndex(start) - 1;
-             scalc = true;
-           }
-           if (sindex == s)
+           if (endIndex < j)
            {
-             // 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;
@@@ -1677,6 -1695,30 +1690,6 @@@
      }
    }
  
 -  /**
 -   * @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)
    {
diff --combined src/jalview/datamodel/SequenceI.java
index 9ad0a61,857f206..2f3e925
--- a/src/jalview/datamodel/SequenceI.java
+++ b/src/jalview/datamodel/SequenceI.java
@@@ -192,13 -192,14 +192,14 @@@ public interface SequenceI extends ASeq
    public int findIndex(int pos);
  
    /**
-    * Returns the sequence position for an alignment position.
+    * Returns the sequence position for an alignment (column) position. If at a
+    * gap, returns the position of the next residue to the right. If beyond the
+    * end of the sequence, returns 1 more than the last residue position.
     * 
     * @param i
     *          column index in alignment (from 0..<length)
     * 
-    * @return TODO: JAL-2562 - residue number for residue (left of and) nearest
-    *         ith column
+    * @return
     */
    public int findPosition(int i);
  
@@@ -449,6 -450,17 +450,6 @@@
    public void transferAnnotation(SequenceI entry, Mapping mp);
  
    /**
 -   * @param index
 -   *          The sequence index in the MSA
 -   */
 -  public void setIndex(int index);
 -
 -  /**
 -   * @return The index of the sequence in the alignment
 -   */
 -  public int getIndex();
 -
 -  /**
     * @return The RNA of the sequence in the alignment
     */