X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FSeqsetUtils.java;h=691b905fe282f389be490f5d7ddffbc0a4dec258;hb=173a7ca81c8d8c00f889fd4e3583cc8733d3d179;hp=bd0f45cf73fff7ee5d3dd0bdc4da06b88a8594a7;hpb=f24dacb1da56fccf05d684e2f4899facec2aecf7;p=jalview.git

diff --git a/src/jalview/analysis/SeqsetUtils.java b/src/jalview/analysis/SeqsetUtils.java
index bd0f45c..691b905 100755
--- a/src/jalview/analysis/SeqsetUtils.java
+++ b/src/jalview/analysis/SeqsetUtils.java
@@ -49,10 +49,16 @@ public class SeqsetUtils
     sqinfo.put("Name", seq.getName());
     sqinfo.put("Start", new Integer(seq.getStart()));
     sqinfo.put("End", new Integer(seq.getEnd()));
-    sqinfo.put("SeqFeatures", seq.getSequenceFeatures());
+    Vector sfeat = new Vector();
+    jalview.datamodel.SequenceFeature[] sfarray=seq.getSequenceFeatures();
+    if (sfarray!=null && sfarray.length>0) {
+      for (int i=0;i<sfarray.length;i++)
+        sfeat.add(sfarray[i]);
+    }
+    sqinfo.put("SeqFeatures", sfeat);
     sqinfo.put("PdbId",
-               (seq.getPDBId() != null) ? seq.getPDBId() : new String(""));
-
+               (seq.getPDBId() != null) ? seq.getPDBId() : new Vector());
+    sqinfo.put("datasetSequence", (seq.getDatasetSequence() !=null) ? seq.getDatasetSequence() : new Sequence("THISISAPLACEHOLDER",""));
     return sqinfo;
   }
 
@@ -61,18 +67,20 @@ public class SeqsetUtils
    * TODO: replace these methods with something more elegant.
    * @param sq SequenceI
    * @param sqinfo Hashtable
-   * @return boolean
+   * @return boolean true if name was not updated from sqinfo Name entry
    */
   public static boolean SeqCharacterUnhash(SequenceI sq, Hashtable sqinfo)
   {
     boolean namePresent = true;
+    if (sqinfo==null)
+      return false;
     String oldname = (String) sqinfo.get("Name");
     Integer start = (Integer) sqinfo.get("Start");
     Integer end = (Integer) sqinfo.get("End");
-    java.util.Vector sfeatures = (java.util.Vector) sqinfo.get(
+    Vector sfeatures = (Vector) sqinfo.get(
         "SeqFeatures");
-    String pdbid = (String) sqinfo.get("PdbId");
-
+    Vector pdbid = (Vector) sqinfo.get("PdbId");
+    Sequence seqds = (Sequence) sqinfo.get("datasetSequence");
     if (oldname == null)
     {
       namePresent = false;
@@ -81,8 +89,7 @@ public class SeqsetUtils
     {
       sq.setName(oldname);
     }
-
-    if (!pdbid.equals(""))
+    if (pdbid!=null && pdbid.size()>0)
     {
       sq.setPDBId(pdbid);
     }
@@ -93,9 +100,14 @@ public class SeqsetUtils
       sq.setEnd(end.intValue());
     }
 
-    if (sfeatures != null)
+    if ((sfeatures != null) && (sfeatures.size()>0))
     {
-      sq.setSequenceFeatures(sfeatures);
+      SequenceFeature[] sfarray = (SequenceFeature[]) sfeatures.toArray();
+      sq.setSequenceFeatures(sfarray);
+    }
+
+    if ((seqds!=null) && !(seqds.getName().equals("THISISAPLACEHOLDER") && seqds.getLength()==0)) {
+      sq.setDatasetSequence(seqds);
     }
 
     return namePresent;
@@ -111,6 +123,14 @@ public class SeqsetUtils
     return new String("Sequence" + i);
   }
 
+  /**
+   * Generates a hash of SeqCharacterHash properties for each sequence
+   * in a sequence set, and optionally renames the sequences to an
+   * unambiguous 'safe' name.
+   * @param sequences SequenceI[]
+   * @param write_names boolean set this to rename each of the sequences to its unique_name(index) name
+   * @return Hashtable to be passed to @see deuniquify to recover original names (and properties) for renamed sequences
+   */
   public static Hashtable uniquify(SequenceI[] sequences, boolean write_names)
   {
     // Generate a safely named sequence set and a hash to recover the sequence names
@@ -131,25 +151,72 @@ public class SeqsetUtils
 
     return map;
   }
-
+  /**
+   * recover unsafe sequence names and original properties for a sequence
+   * set using a map generated by @see uniquify(sequences,true)
+   * @param map Hashtable
+   * @param sequences SequenceI[]
+   * @return boolean
+   */
   public static boolean deuniquify(Hashtable map, SequenceI[] sequences)
   {
-    // recover unsafe sequence names for a sequence set
-    boolean allfound = true;
-
-    for (int i = 0; i < sequences.length; i++)
-    {
-      if (map.containsKey(sequences[i].getName()))
-      {
-        Hashtable sqinfo = (Hashtable) map.get(sequences[i].getName());
-        SeqCharacterUnhash(sequences[i], sqinfo);
-      }
-      else
-      {
-        allfound = false;
+    jalview.analysis.SequenceIdMatcher matcher = new SequenceIdMatcher(sequences);
+    SequenceI msq = null;
+    Enumeration keys = map.keys();
+    Vector unmatched = new Vector();
+    for (int i=0, j=sequences.length; i<j; i++)
+      unmatched.add(sequences[i]);
+    while (keys.hasMoreElements()) {
+      Object key = keys.nextElement();
+      if (key instanceof String) {
+        if ((msq = matcher.findIdMatch((String) key))!=null) {
+          Hashtable sqinfo = (Hashtable) map.get(key);
+          unmatched.remove(msq);
+          SeqCharacterUnhash(msq, sqinfo);
+        }
+        else
+        {
+          System.err.println("Can't find '"+((String) key)+"' in uniquified alignment");
+        }
       }
     }
+    if (unmatched.size()>0) {
+      System.err.println("Did not find matches for :");
+      for (Enumeration i = unmatched.elements(); i.hasMoreElements(); System.out.println(((SequenceI) i.nextElement()).getName()))
+           ;
+      return false;
+    }
 
-    return allfound;
+    return true;
+  }
+  /**
+   * returns a subset of the sequenceI seuqences,
+   * including only those that contain at least one residue.
+   * @param sequences SequenceI[]
+   * @return SequenceI[]
+   */
+  public static SequenceI[] getNonEmptySequenceSet(SequenceI[] sequences) {
+      // Identify first row of alignment with residues for prediction
+      boolean ungapped[] = new boolean[sequences.length];
+      int msflen=0;
+      for (int i=0,j=sequences.length; i<j;i++) {
+        String tempseq = jalview.analysis.AlignSeq.extractGaps(jalview.util.Comparison.GapChars, sequences[i].getSequence());
+        if (tempseq.length()==0)
+          ungapped[i]=false;
+        else {
+          ungapped[i]=true;
+          msflen++;
+        }
+      }
+      if (msflen==0)
+        return null; // no minimal set
+      // compose minimal set
+      SequenceI[] mset = new SequenceI[msflen];
+      for (int i=0,j=sequences.length,k=0; i<j;i++) {
+        if (ungapped[i])
+          mset[k++] = sequences[i];
+      }
+      ungapped = null;
+      return mset;
   }
 }