From 1361f8e20a0829234025bfeb06d3b8de6c660e0a Mon Sep 17 00:00:00 2001
From: Jim Procter <jprocter@issues.jalview.org>
Date: Mon, 29 Aug 2016 12:02:03 +0100
Subject: [PATCH] JAL-2154 new assertDatasetIsNormalised to trivially test for
 the addition of identical sequence data in the dataset
 (e.g. due to faulty create CDS sequence processes)

---
 test/jalview/datamodel/AlignmentTest.java |  116 +++++++++++++++++++++++++++++
 1 file changed, 116 insertions(+)

diff --git a/test/jalview/datamodel/AlignmentTest.java b/test/jalview/datamodel/AlignmentTest.java
index fe76df5..58e024f 100644
--- a/test/jalview/datamodel/AlignmentTest.java
+++ b/test/jalview/datamodel/AlignmentTest.java
@@ -422,7 +422,123 @@ public class AlignmentTest
     al.getDataset().deleteSequence(sqanotherout);
     assertVerifyAlignment(al, true,
             "verify should have passed after duplicate entry for sequence was removed");
+  }
+
+  /**
+   * checks that the sequence data for an alignment's dataset is non-redundant.
+   * Fails if there are sequences with same id, sequence, start, and.
+   */
+  public static void assertDatasetIsNormalised(AlignmentI al)
+  {
+    if (al.getDataset()!=null)
+    {
+      assertDatasetIsNormalised(al.getDataset());
+      return;
+    }
+    /*
+     * look for pairs of sequences with same ID, start, end, and sequence
+     */
+    List<SequenceI> seqSet = al.getSequences();
+    for (int p=0;p<seqSet.size(); p++)
+    {
+      SequenceI pSeq = seqSet.get(p);
+      for (int q=p+1; q<seqSet.size(); q++)
+      {
+        SequenceI qSeq = seqSet.get(q);
+        if (pSeq.getStart()!=qSeq.getStart())
+        {
+          continue;
+        }
+        if (pSeq.getEnd()!=qSeq.getEnd())
+        {
+          continue;
+        }
+        if (!pSeq.getName().equals(qSeq.getName()))
+        {
+          continue;
+        }
+        if (!Arrays.equals(pSeq.getSequence(), qSeq.getSequence()))
+        {
+          continue;
+        }
+        Assert.fail("Found similar sequences at position "+p+" and "+q+"\n"+pSeq.toString());
+      }
+    }
+  }
+  
+  @Test(groups = { "Functional", "Asserts" })
+  public void testAssertDatasetIsNormalised()
+  {
+    Sequence sq1 = new Sequence("s1/1-4", "asdf");
+    Sequence sq1shift = new Sequence("s1/2-5", "asdf");
+    Sequence sq1seqd = new Sequence("s1/1-4", "asdt");
+    Sequence sq2 = new Sequence("s2/1-4", "asdf");
+    Sequence sq1dup = new Sequence("s1/1-4", "asdf");
 
+    Alignment al = new Alignment(new SequenceI[] { sq1 });
+    al.setDataset(null);
+
+    try
+    {
+      assertDatasetIsNormalised(al);
+    } catch (AssertionError ae)
+    {
+      Assert.fail("Single sequence should be valid normalised dataset.");
+    }
+    al.addSequence(sq2);
+    try
+    {
+      assertDatasetIsNormalised(al);
+    } catch (AssertionError ae)
+    {
+      Assert.fail("Two different sequences should be valid normalised dataset.");
+    }
+    /*
+     * now change sq2's name in the alignment. should still be valid
+     */
+    al.findName(sq2.getName()).setName("sq1");
+    try
+    {
+      assertDatasetIsNormalised(al);
+    } catch (AssertionError ae)
+    {
+      Assert.fail("Two different sequences in dataset, but same name in alignment, should be valid normalised dataset.");
+    }
+
+    al.addSequence(sq1seqd);
+    try
+    {
+      assertDatasetIsNormalised(al);
+    } catch (AssertionError ae)
+    {
+      Assert.fail("sq1 and sq1 with different sequence should be distinct.");
+    }
+
+    al.addSequence(sq1shift);
+    try
+    {
+      assertDatasetIsNormalised(al);
+    } catch (AssertionError ae)
+    {
+      Assert.fail("sq1 and sq1 with different start/end should be distinct.");
+    }
+    /*
+     * finally, the failure case
+     */
+    al.addSequence(sq1dup);
+    boolean ssertRaised = false;
+    try
+    {
+      assertDatasetIsNormalised(al);
+
+    } catch (AssertionError ae)
+    {
+      ssertRaised = true;
+    }
+    if (!ssertRaised)
+    {
+      Assert.fail("Expected identical sequence to raise exception.");
+    }
   }
   /*
    * Read in Stockholm format test data including secondary structure
-- 
1.7.10.2