Stockholm file for test
[jalview.git] / test / jalview / io / StockholmFileTest.java
index d0561b0..2427167 100644 (file)
@@ -30,24 +30,38 @@ import jalview.datamodel.AlignmentI;
 import jalview.datamodel.Annotation;
 import jalview.datamodel.SequenceFeature;
 import jalview.datamodel.SequenceI;
+import jalview.gui.JvOptionPane;
 
 import java.io.File;
+import java.util.Arrays;
 import java.util.BitSet;
 import java.util.HashMap;
+import java.util.List;
 import java.util.Map;
 
+import org.testng.Assert;
+import org.testng.annotations.BeforeClass;
 import org.testng.annotations.Test;
 
 public class StockholmFileTest
 {
 
+  @BeforeClass(alwaysRun = true)
+  public void setUpJvOptionPane()
+  {
+    JvOptionPane.setInteractiveMode(false);
+    JvOptionPane.setMockResponse(JvOptionPane.CANCEL_OPTION);
+  }
+
   static String PfamFile = "examples/PF00111_seed.stk",
-          RfamFile = "examples/RF00031_folded.stk";
+          RfamFile = "examples/RF00031_folded.stk",
+          RnaSSTestFile = "examples/rna_ss_test.stk";
 
   @Test(groups = { "Functional" })
   public void pfamFileIO() throws Exception
   {
-    testFileIOwithFormat(new File(PfamFile), FileFormat.Stockholm, -1, 0);
+    testFileIOwithFormat(new File(PfamFile), FileFormat.Stockholm, -1, 0,
+            false, false, false);
   }
 
   @Test(groups = { "Functional" })
@@ -72,7 +86,8 @@ public class StockholmFileTest
   @Test(groups = { "Functional" })
   public void rfamFileIO() throws Exception
   {
-    testFileIOwithFormat(new File(RfamFile), FileFormat.Stockholm, 2, 1);
+    testFileIOwithFormat(new File(RfamFile), FileFormat.Stockholm, 2, 1,
+            false, false, false);
   }
 
   /**
@@ -84,10 +99,14 @@ public class StockholmFileTest
    * @param ioformat
    *          - label for IO class used to write and read back in the data from
    *          f
+   * @param ignoreFeatures
+   * @param ignoreRowVisibility
+   * @param allowNullAnnotations
    */
 
   public static void testFileIOwithFormat(File f, FileFormatI ioformat,
-          int naliannot, int nminseqann)
+          int naliannot, int nminseqann, boolean ignoreFeatures,
+          boolean ignoreRowVisibility, boolean allowNullAnnotations)
   {
     System.out.println("Reading file: " + f);
     String ff = f.getPath();
@@ -121,7 +140,8 @@ public class StockholmFileTest
               "Identify routine could not recognise output generated by '"
                       + ioformat + "' writer",
               ioformat.equals(identifyoutput));
-      testAlignmentEquivalence(al, al_input, false);
+      testAlignmentEquivalence(al, al_input, ignoreFeatures,
+              ignoreRowVisibility, allowNullAnnotations);
       int numaliannot = 0, numsqswithali = 0;
       for (AlignmentAnnotation ala : al_input.getAlignmentAnnotation())
       {
@@ -165,6 +185,34 @@ public class StockholmFileTest
   public static void testAlignmentEquivalence(AlignmentI al,
           AlignmentI al_input, boolean ignoreFeatures)
   {
+    testAlignmentEquivalence(al, al_input, ignoreFeatures, false, false);
+  }
+
+  /**
+   * assert alignment equivalence - uses special comparators for RNA structure
+   * annotation rows.
+   * 
+   * @param al
+   *          'original'
+   * @param al_input
+   *          'secondary' or generated alignment from some datapreserving
+   *          transformation
+   * @param ignoreFeatures
+   *          when true, differences in sequence feature annotation are ignored
+   * 
+   * @param ignoreRowVisibility
+   *          when true, do not fail if there are differences in the visibility
+   *          of annotation rows
+   * @param allowNullAnnotation
+   *          when true, positions in alignment annotation that are null will be
+   *          considered equal to positions containing annotation where
+   *          Annotation.isWhitespace() returns true.
+   * 
+   */
+  public static void testAlignmentEquivalence(AlignmentI al,
+          AlignmentI al_input, boolean ignoreFeatures,
+          boolean ignoreRowVisibility, boolean allowNullAnnotation)
+  {
     assertNotNull("Original alignment was null", al);
     assertNotNull("Generated alignment was null", al_input);
 
@@ -183,8 +231,8 @@ public class StockholmFileTest
     // we might want to revise this in future
     int aa_new_size = (aa_new == null ? 0 : aa_new.length);
     int aa_original_size = (aa_original == null ? 0 : aa_original.length);
-    Map<Integer, BitSet> orig_groups = new HashMap<Integer, BitSet>();
-    Map<Integer, BitSet> new_groups = new HashMap<Integer, BitSet>();
+    Map<Integer, BitSet> orig_groups = new HashMap<>();
+    Map<Integer, BitSet> new_groups = new HashMap<>();
 
     if (aa_new != null && aa_original != null)
     {
@@ -192,13 +240,18 @@ public class StockholmFileTest
       {
         if (aa_new.length > i)
         {
-          assertTrue("Different alignment annotation at position " + i,
-                  equalss(aa_original[i], aa_new[i]));
+          assertEqualSecondaryStructure(
+                  "Different alignment annotation at position " + i,
+                  aa_original[i], aa_new[i], allowNullAnnotation);
           // compare graphGroup or graph properties - needed to verify JAL-1299
           assertEquals("Graph type not identical.", aa_original[i].graph,
                   aa_new[i].graph);
-          assertEquals("Visibility not identical.", aa_original[i].visible,
+          if (!ignoreRowVisibility)
+          {
+            assertEquals("Visibility not identical.",
+                    aa_original[i].visible,
                   aa_new[i].visible);
+          }
           assertEquals("Threshold line not identical.",
                   aa_original[i].threshold, aa_new[i].threshold);
           // graphGroup may differ, but pattern should be the same
@@ -235,7 +288,8 @@ public class StockholmFileTest
     seq_original = al.getSequencesArray();
     SequenceI[] seq_new = new SequenceI[al_input.getSequencesArray().length];
     seq_new = al_input.getSequencesArray();
-    SequenceFeature[] sequenceFeatures_original, sequenceFeatures_new;
+    List<SequenceFeature> sequenceFeatures_original;
+    List<SequenceFeature> sequenceFeatures_new;
     AlignmentAnnotation annot_original, annot_new;
     //
     for (int i = 0; i < al.getSequencesArray().length; i++)
@@ -271,23 +325,20 @@ public class StockholmFileTest
                   && seq_new[in].getSequenceFeatures() != null)
           {
             System.out.println("There are feature!!!");
-            sequenceFeatures_original = new SequenceFeature[seq_original[i]
-                    .getSequenceFeatures().length];
             sequenceFeatures_original = seq_original[i]
                     .getSequenceFeatures();
-            sequenceFeatures_new = new SequenceFeature[seq_new[in]
-                    .getSequenceFeatures().length];
             sequenceFeatures_new = seq_new[in].getSequenceFeatures();
 
-            assertEquals("different number of features",
-                    seq_original[i].getSequenceFeatures().length,
-                    seq_new[in].getSequenceFeatures().length);
+            assertEquals("different number of features", seq_original[i]
+                    .getSequenceFeatures().size(), seq_new[in]
+                    .getSequenceFeatures().size());
 
-            for (int feat = 0; feat < seq_original[i].getSequenceFeatures().length; feat++)
+            for (int feat = 0; feat < seq_original[i].getSequenceFeatures()
+                    .size(); feat++)
             {
               assertEquals("Different features",
-                      sequenceFeatures_original[feat],
-                      sequenceFeatures_new[feat]);
+                      sequenceFeatures_original.get(feat),
+                      sequenceFeatures_new.get(feat));
             }
           }
           // compare alignment annotation
@@ -301,8 +352,9 @@ public class StockholmFileTest
               {
                 annot_original = al.getSequenceAt(i).getAnnotation()[j];
                 annot_new = al_input.getSequenceAt(in).getAnnotation()[j];
-                assertTrue("Different annotation elements",
-                        equalss(annot_original, annot_new));
+                assertEqualSecondaryStructure(
+                        "Different annotation elements", annot_original,
+                        annot_new, allowNullAnnotation);
               }
             }
           }
@@ -324,39 +376,78 @@ public class StockholmFileTest
     }
   }
 
-  /*
-   * compare annotations
+  /**
+   * compare two annotation rows, with special support for secondary structure
+   * comparison. With RNA, only the value and the secondaryStructure symbols are
+   * compared, displayCharacter and description are ignored. Annotations where
+   * Annotation.isWhitespace() is true are always considered equal.
+   * 
+   * @param message
+   *          - not actually used yet..
+   * @param annot_or
+   *          - the original annotation
+   * @param annot_new
+   *          - the one compared to the original annotation
+   * @param allowNullEquivalence
+   *          when true, positions in alignment annotation that are null will be
+   *          considered equal to non-null positions for which
+   *          Annotation.isWhitespace() is true.
    */
-  private static boolean equalss(AlignmentAnnotation annot_or,
-          AlignmentAnnotation annot_new)
+  private static void assertEqualSecondaryStructure(String message,
+          AlignmentAnnotation annot_or, AlignmentAnnotation annot_new,
+          boolean allowNullEqivalence)
   {
+    // TODO: test to cover this assert behaves correctly for all allowed
+    // variations of secondary structure annotation row equivalence
     if (annot_or.annotations.length != annot_new.annotations.length)
     {
-      System.err.println("Different lengths for annotation row elements: "
+      fail("Different lengths for annotation row elements: "
               + annot_or.annotations.length + "!="
               + annot_new.annotations.length);
-      return false;
     }
+    boolean isRna = annot_or.isRNA();
+    assertTrue("Expected " + (isRna ? " valid RNA " : " no RNA ")
+            + " secondary structure in the row.",
+            isRna == annot_new.isRNA());
     for (int i = 0; i < annot_or.annotations.length; i++)
     {
       Annotation an_or = annot_or.annotations[i], an_new = annot_new.annotations[i];
       if (an_or != null && an_new != null)
       {
-        if (!an_or.displayCharacter.trim().equals(
-                an_new.displayCharacter.trim())
-                || !("" + an_or.secondaryStructure).trim().equals(
-                        ("" + an_new.secondaryStructure).trim())
-                || (an_or.description != an_new.description && !((an_or.description == null && an_new.description
-                        .trim().length() == 0)
-                        || (an_new.description == null && an_or.description
-                                .trim().length() == 0) || an_or.description
-                        .trim().equals(an_new.description.trim()))))
+
+        if (isRna)
         {
-          System.err.println("Annotation Element Mismatch\nElement " + i
-                  + " in original: " + annot_or.annotations[i].toString()
-                  + "\nElement " + i + " in new: "
-                  + annot_new.annotations[i].toString());
-          return false;
+          if (an_or.secondaryStructure != an_new.secondaryStructure
+                  || ((Float.isNaN(an_or.value) != Float
+                          .isNaN(an_new.value)) || an_or.value != an_new.value))
+          {
+            fail("Different RNA secondary structure at column " + i
+                    + " expected: [" + annot_or.annotations[i].toString()
+                    + "] but got: [" + annot_new.annotations[i].toString()
+                    + "]");
+          }
+        }
+        else
+        {
+          // not RNA secondary structure, so expect all elements to match...
+          if ((an_or.isWhitespace() != an_new.isWhitespace())
+                  || !an_or.displayCharacter.trim().equals(
+                  an_new.displayCharacter.trim())
+                  || !("" + an_or.secondaryStructure).trim().equals(
+                          ("" + an_new.secondaryStructure).trim())
+                  || (an_or.description != an_new.description && !((an_or.description == null && an_new.description
+                          .trim().length() == 0)
+                          || (an_new.description == null && an_or.description
+                                  .trim().length() == 0) || an_or.description
+                          .trim().equals(an_new.description.trim())))
+                  || !((Float.isNaN(an_or.value) && Float
+                          .isNaN(an_new.value)) || an_or.value == an_new.value))
+          {
+            fail("Annotation Element Mismatch\nElement " + i
+                    + " in original: " + annot_or.annotations[i].toString()
+                    + "\nElement " + i + " in new: "
+                    + annot_new.annotations[i].toString());
+          }
         }
       }
       else if (annot_or.annotations[i] == null
@@ -366,19 +457,202 @@ public class StockholmFileTest
       }
       else
       {
-        System.err.println("Annotation Element Mismatch\nElement "
-                + i
-                + " in original: "
-                + (annot_or.annotations[i] == null ? "is null"
-                        : annot_or.annotations[i].toString())
-                + "\nElement "
-                + i
-                + " in new: "
-                + (annot_new.annotations[i] == null ? "is null"
-                        : annot_new.annotations[i].toString()));
-        return false;
+        if (allowNullEqivalence)
+        {
+          if (an_or != null && an_or.isWhitespace())
+
+          {
+            continue;
+          }
+          if (an_new != null && an_new.isWhitespace())
+          {
+            continue;
+          }
+        }
+        // need also to test for null in one, non-SS annotation in other...
+        fail("Annotation Element Mismatch\nElement " + i + " in original: "
+                + (an_or == null ? "is null" : an_or.toString())
+                + "\nElement " + i + " in new: "
+                + (an_new == null ? "is null" : an_new.toString()));
       }
     }
-    return true;
+  }
+
+  /**
+   * @see assertEqualSecondaryStructure - test if two secondary structure
+   *      annotations are not equal
+   * @param message
+   * @param an_orig
+   * @param an_new
+   * @param allowNullEquivalence
+   */
+  public static void assertNotEqualSecondaryStructure(String message,
+          AlignmentAnnotation an_orig, AlignmentAnnotation an_new,
+          boolean allowNullEquivalence)
+  {
+    boolean thrown = false;
+    try
+    {
+      assertEqualSecondaryStructure("", an_orig, an_new,
+              allowNullEquivalence);
+    } catch (AssertionError af)
+    {
+      thrown = true;
+    }
+    if (!thrown)
+    {
+      fail("Expected difference for [" + an_orig + "] and [" + an_new + "]");
+    }
+  }
+  private AlignmentAnnotation makeAnnot(Annotation ae)
+  {
+    return new AlignmentAnnotation("label", "description", new Annotation[]
+    { ae });
+  }
+
+  @Test(groups={"Functional"})
+  public void testAnnotationEquivalence()
+  {
+    AlignmentAnnotation one = makeAnnot(new Annotation("", "", ' ', 1));
+    AlignmentAnnotation anotherOne = makeAnnot(new Annotation("", "", ' ',
+            1));
+    AlignmentAnnotation sheet = makeAnnot(new Annotation("","",'E',0f));
+    AlignmentAnnotation anotherSheet = makeAnnot(new Annotation("","",'E',0f)); 
+    AlignmentAnnotation sheetWithLabel = makeAnnot(new Annotation("1", "",
+            'E', 0f));
+    AlignmentAnnotation anotherSheetWithLabel = makeAnnot(new Annotation(
+            "1", "", 'E', 0f));
+    AlignmentAnnotation rnaNoDC = makeAnnot(new Annotation("","",'<',0f));
+    AlignmentAnnotation anotherRnaNoDC = makeAnnot(new Annotation("","",'<',0f));
+    AlignmentAnnotation rnaWithDC = makeAnnot(new Annotation("B", "", '<',
+            0f));
+    AlignmentAnnotation anotherRnaWithDC = makeAnnot(new Annotation("B",
+            "", '<', 0f));
+    
+    // check self equivalence
+    for (boolean allowNull : new boolean[] { true, false })
+    {
+      assertEqualSecondaryStructure("Should be equal", one, anotherOne,
+              allowNull);
+      assertEqualSecondaryStructure("Should be equal", sheet, anotherSheet,
+              allowNull);
+      assertEqualSecondaryStructure("Should be equal", sheetWithLabel,
+              anotherSheetWithLabel, allowNull);
+      assertEqualSecondaryStructure("Should be equal", rnaNoDC,
+              anotherRnaNoDC, allowNull);
+      assertEqualSecondaryStructure("Should be equal", rnaWithDC,
+              anotherRnaWithDC, allowNull);
+      // display character doesn't matter for RNA structure (for 2.10.2)
+      assertEqualSecondaryStructure("Should be equal", rnaWithDC, rnaNoDC,
+              allowNull);
+      assertEqualSecondaryStructure("Should be equal", rnaNoDC, rnaWithDC,
+              allowNull);
+    }
+
+    // verify others are different
+    List<AlignmentAnnotation> aaSet = Arrays.asList(one, sheet,
+            sheetWithLabel, rnaWithDC);
+    for (int p = 0; p < aaSet.size(); p++)
+    {
+      for (int q = 0; q < aaSet.size(); q++)
+      {
+        if (p != q)
+        {
+          assertNotEqualSecondaryStructure("Should be different",
+                    aaSet.get(p), aaSet.get(q), false);
+        }
+        else
+        {
+          assertEqualSecondaryStructure("Should be same", aaSet.get(p),
+                  aaSet.get(q), false);
+          assertEqualSecondaryStructure("Should be same", aaSet.get(p),
+                  aaSet.get(q), true);
+          assertNotEqualSecondaryStructure(
+                  "Should be different to empty anot", aaSet.get(p),
+                  makeAnnot(Annotation.EMPTY_ANNOTATION), false);
+          assertNotEqualSecondaryStructure(
+                  "Should be different to empty annot",
+                  makeAnnot(Annotation.EMPTY_ANNOTATION), aaSet.get(q),
+                  true);
+          assertNotEqualSecondaryStructure("Should be different to null",
+                  aaSet.get(p), makeAnnot(null), false);
+          assertNotEqualSecondaryStructure("Should be different to null",
+                  makeAnnot(null), aaSet.get(q), true);
+        }
+      }
+    }
+
+    // test null
+
+  }
+
+  String aliFile = ">Dm\nAAACCCUUUUACACACGGGAAAGGG";
+  String annFile = "JALVIEW_ANNOTATION\n# Created: Thu May 04 11:16:52 BST 2017\n\n"
+          + "SEQUENCE_REF\tDm\nNO_GRAPH\tsecondary structure\tsecondary structure\t"
+          + "(|(|(|(|, .|, .|, .|, .|)|)|)|)|\t0.0\nROWPROPERTIES\t"
+          + "secondary structure\tscaletofit=true\tshowalllabs=true\tcentrelabs=false";
+
+  String annFileCurlyWuss = "JALVIEW_ANNOTATION\n# Created: Thu May 04 11:16:52 BST 2017\n\n"
+          + "SEQUENCE_REF\tDm\nNO_GRAPH\tsecondary structure\tsecondary structure\t"
+          + "(|(|(|(||{|{||{|{||)|)|)|)||}|}|}|}|\t0.0\nROWPROPERTIES\t"
+          + "secondary structure\tscaletofit=true\tshowalllabs=true\tcentrelabs=false";
+  String annFileFullWuss = "JALVIEW_ANNOTATION\n# Created: Thu May 04 11:16:52 BST 2017\n\n"
+          + "SEQUENCE_REF\tDm\nNO_GRAPH\tsecondary structure\tsecondary structure\t"
+          + "(|(|(|(||{|{||[|[||)|)|)|)||}|}|]|]|\t0.0\nROWPROPERTIES\t"
+          + "secondary structure\tscaletofit=true\tshowalllabs=true\tcentrelabs=false";
+
+  @Test(groups = { "Functional" })
+  public void secondaryStructureForRNASequence() throws Exception
+  {
+    roundTripSSForRNA(aliFile, annFile);
+  }
+
+  @Test(groups = { "Functional" })
+  public void curlyWUSSsecondaryStructureForRNASequence() throws Exception
+  {
+    roundTripSSForRNA(aliFile, annFileCurlyWuss);
+  }
+
+  @Test(groups = { "Functional" })
+  public void fullWUSSsecondaryStructureForRNASequence() throws Exception
+  {
+    roundTripSSForRNA(aliFile, annFileFullWuss);
+  }
+
+  @Test(groups = { "Functional" })
+  public void detectWussBrackets()
+  {
+    for (char ch : new char[] { '{', '}', '[', ']', '(', ')', '<', '>' })
+    {
+      Assert.assertTrue(StockholmFile.DETECT_BRACKETS.matchAt("" + ch, 0),
+              "Didn't recognise " + ch + " as a WUSS bracket");
+    }
+    for (char ch : new char[] { '@', '!', 'V', 'Q', '*', ' ', '-', '.' })
+    {
+      Assert.assertFalse(StockholmFile.DETECT_BRACKETS.matchAt("" + ch, 0),
+              "Shouldn't recognise " + ch + " as a WUSS bracket");
+    }
+  }
+  private static void roundTripSSForRNA(String aliFile, String annFile)
+          throws Exception
+  {
+    AlignmentI al = new AppletFormatAdapter().readFile(aliFile,
+            DataSourceType.PASTE, jalview.io.FileFormat.Fasta);
+    AnnotationFile aaf = new AnnotationFile();
+    aaf.readAnnotationFile(al, annFile, DataSourceType.PASTE);
+    al.getAlignmentAnnotation()[0].visible = true;
+
+    // TODO: create a better 'save as <format>' pattern
+    StockholmFile sf = new StockholmFile(al);
+
+    String stockholmFile = sf.print(al.getSequencesArray(), true);
+
+    AlignmentI newAl = new AppletFormatAdapter().readFile(stockholmFile,
+            DataSourceType.PASTE, jalview.io.FileFormat.Stockholm);
+    // AlignmentUtils.showOrHideSequenceAnnotations(newAl.getViewport()
+    // .getAlignment(), Arrays.asList("Secondary Structure"), newAl
+    // .getViewport().getAlignment().getSequences(), true, true);
+    testAlignmentEquivalence(al, newAl, true, true, true);
+
   }
 }