X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FCrossRef.java;h=4ba7e416e7d115d183d68ce50878e21f1377b892;hb=aba253e57b22ce7d1f4fe376935e42aeb4f6d591;hp=71b0aa0b9adb5cd92d39b5f7908fa3dcc349e4ed;hpb=655b78299307682a4c7a6e5af0ed4618cbc9c924;p=jalview.git

diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java
index 71b0aa0..4ba7e41 100644
--- a/src/jalview/analysis/CrossRef.java
+++ b/src/jalview/analysis/CrossRef.java
@@ -24,6 +24,7 @@ import jalview.datamodel.AlignedCodonFrame;
 import jalview.datamodel.Alignment;
 import jalview.datamodel.AlignmentI;
 import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.DBRefSource;
 import jalview.datamodel.Mapping;
 import jalview.datamodel.Sequence;
 import jalview.datamodel.SequenceFeature;
@@ -106,6 +107,16 @@ public class CrossRef
         findXrefSourcesForSequence(seq, dna, sources);
       }
     }
+    sources.remove(DBRefSource.EMBL); // hack to prevent EMBL xrefs resulting in
+                                      // redundant datasets
+    if (dna)
+    {
+      sources.remove(DBRefSource.ENSEMBL); // hack to prevent Ensembl and
+                                           // EnsemblGenomes xref option shown
+                                           // from cdna panel
+      sources.remove(DBRefSource.ENSEMBLGENOMES);
+    }
+    // redundant datasets
     return sources;
   }
 
@@ -209,8 +220,7 @@ public class CrossRef
 
     rseqs = new ArrayList<SequenceI>();
     AlignedCodonFrame cf = new AlignedCodonFrame();
-    matcher = new SequenceIdMatcher(
-            dataset.getSequences());
+    matcher = new SequenceIdMatcher(dataset.getSequences());
 
     for (SequenceI seq : fromSeqs)
     {
@@ -422,13 +432,16 @@ public class CrossRef
     if (retrieved != null)
     {
       boolean addedXref = false;
+      List<SequenceI> newDsSeqs = new ArrayList<SequenceI>(), doNotAdd = new ArrayList<SequenceI>();
+
       for (SequenceI retrievedSequence : retrieved)
       {
         // dataset gets contaminated ccwith non-ds sequences. why ??!
         // try: Ensembl -> Nuc->Ensembl, Nuc->Uniprot-->Protein->EMBL->
         SequenceI retrievedDss = retrievedSequence.getDatasetSequence() == null ? retrievedSequence
                 : retrievedSequence.getDatasetSequence();
-        addedXref |= importCrossRefSeq(cf, dss, retrievedDss);
+        addedXref |= importCrossRefSeq(cf, newDsSeqs, doNotAdd, dss,
+                retrievedDss);
       }
       if (!addedXref)
       {
@@ -441,7 +454,17 @@ public class CrossRef
           // try: Ensembl -> Nuc->Ensembl, Nuc->Uniprot-->Protein->EMBL->
           SequenceI retrievedDss = retrievedSequence.getDatasetSequence() == null ? retrievedSequence
                   : retrievedSequence.getDatasetSequence();
-          addedXref |= importCrossRefSeq(cf, dss, retrievedDss);
+          addedXref |= importCrossRefSeq(cf, newDsSeqs, doNotAdd, dss,
+                  retrievedDss);
+        }
+      }
+      for (SequenceI newToSeq : newDsSeqs)
+      {
+        if (!doNotAdd.contains(newToSeq)
+                && dataset.findIndex(newToSeq) == -1)
+        {
+          dataset.addSequence(newToSeq);
+          matcher.add(newToSeq);
         }
       }
     }
@@ -494,6 +517,7 @@ public class CrossRef
    * @return true if retrieveSequence was imported
    */
   private boolean importCrossRefSeq(AlignedCodonFrame cf,
+          List<SequenceI> newDsSeqs, List<SequenceI> doNotAdd,
           SequenceI sourceSequence, SequenceI retrievedSequence)
   {
     /**
@@ -502,7 +526,6 @@ public class CrossRef
      */
     boolean imported = false;
     DBRefEntry[] dbr = retrievedSequence.getDBRefs();
-    List<SequenceI> newDsSeqs = new ArrayList<SequenceI>();
     if (dbr != null)
     {
       for (DBRefEntry dbref : dbr)
@@ -574,9 +597,15 @@ public class CrossRef
                    */
                   for (DBRefEntry ref : toRefs)
                   {
+                    if (dbref.getSrcAccString().equals(
+                            ref.getSrcAccString()))
+                    {
+                      continue; // avoid overwriting the ref on source sequence
+                    }
                     matched.addDBRef(ref); // add or update mapping
                   }
                 }
+                doNotAdd.add(map.getTo());
                 map.setTo(matched);
 
                 /*
@@ -600,8 +629,7 @@ public class CrossRef
                      * attribute in equality test; this avoids creating many
                      * otherwise duplicate exon features on genomic sequence
                      */
-                    SequenceFeature newFeature = new SequenceFeature(
-                            feat)
+                    SequenceFeature newFeature = new SequenceFeature(feat)
                     {
                       @Override
                       public boolean equals(Object o)
@@ -634,18 +662,10 @@ public class CrossRef
         dataset.addSequence(retrievedSequence);
         matcher.add(retrievedSequence);
       }
-      for (SequenceI newToSeq : newDsSeqs)
-      {
-
-        if (dataset.findIndex(newToSeq) == -1)
-        {
-          dataset.addSequence(newToSeq);
-          matcher.add(newToSeq);
-        }
-      }
     }
     return imported;
   }
+
   /**
    * Sets the inverse sequence mapping in the corresponding dbref of the mapped
    * to sequence (if any). This is used after fetching a cross-referenced
@@ -845,8 +865,8 @@ public class CrossRef
     MapList mapping = null;
     SequenceI dsmapFrom = mapFrom.getDatasetSequence() == null ? mapFrom
             : mapFrom.getDatasetSequence();
-    SequenceI dsmapTo = mapTo.getDatasetSequence() == null ? mapTo
-            : mapTo.getDatasetSequence();
+    SequenceI dsmapTo = mapTo.getDatasetSequence() == null ? mapTo : mapTo
+            .getDatasetSequence();
     /*
      * look for a reverse mapping, if found make its inverse. 
      * Note - we do this on dataset sequences only.