From d11914fc106b8cfbcbf14176b49f9ae71992a8b8 Mon Sep 17 00:00:00 2001
From: Jim Procter <jprocter@issues.jalview.org>
Date: Mon, 29 Aug 2016 22:15:58 +0100
Subject: [PATCH] JAL-2154 filter out dbrefs that already match primary refs
 for sequences already retrieved. This is a patch for ENA
 records where protein products were not reconstructed
 (AC006347) resulting in duplicate sequence fetches

---
 src/jalview/analysis/CrossRef.java |   31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java
index 46256a6..1295b46 100644
--- a/src/jalview/analysis/CrossRef.java
+++ b/src/jalview/analysis/CrossRef.java
@@ -390,6 +390,37 @@ public class CrossRef
     SequenceI[] retrieved = null;
     SequenceI dss = seq.getDatasetSequence() == null ? seq : seq
             .getDatasetSequence();
+    // first filter in case we are retrieving crossrefs that have already been
+    // retrieved. this happens for cases where a database record doesn't yield
+    // protein products for CDS
+    DBRefEntry[] dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
+    for (SequenceI sq : dataset.getSequences())
+    {
+      boolean dupeFound = false;
+      // !fromDna means we are looking only for nucleotide sequences, not
+      // protein
+      if (sq.isProtein() == fromDna)
+      {
+        for (DBRefEntry dbr : sq.getPrimaryDBRefs())
+        {
+          for (DBRefEntry found : DBRefUtils.searchRefs(dbrSourceSet, dbr))
+          {
+            sourceRefs.remove(found);
+            dupeFound = true;
+          }
+        }
+      }
+      if (dupeFound)
+      {
+        dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
+      }
+    }
+    if (sourceRefs.size() == 0)
+    {
+      // no more work to do! We already had all requested sequence records in
+      // the dataset.
+      return;
+    }
     try
     {
       retrieved = sftch.getSequences(sourceRefs, !fromDna);
-- 
1.7.10.2