+/*
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
+ *
+ * This file is part of Jalview.
+ *
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
package jalview.io;
-import static org.junit.Assert.*;
-import jalview.datamodel.Alignment;
+import static org.testng.AssertJUnit.assertEquals;
+import static org.testng.AssertJUnit.assertNotNull;
+import static org.testng.AssertJUnit.assertTrue;
+import static org.testng.AssertJUnit.fail;
+
import jalview.datamodel.AlignmentAnnotation;
import jalview.datamodel.AlignmentI;
+import jalview.datamodel.Annotation;
import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
import java.io.File;
-import java.io.IOException;
-import java.io.InputStream;
+import java.util.BitSet;
+import java.util.HashMap;
+import java.util.Map;
-import org.junit.Test;
+import org.testng.annotations.Test;
public class StockholmFileTest
{
- public static void main(String argv[])
+ static String PfamFile = "examples/PF00111_seed.stk",
+ RfamFile = "examples/RF00031_folded.stk";
+
+ @Test(groups = { "Functional" })
+ public void pfamFileIO() throws Exception
{
- try {
- new StockholmFileTest().pfamFileIO();
-
- } catch (Exception x)
- {
- x.printStackTrace();
- }
+ testFileIOwithFormat(new File(PfamFile), "STH", -1, 0);
}
- static String PfamFile = "examples/PF00111_seed.stk";
- @Test
- public void pfamFileIO() throws Exception
+ @Test(groups = { "Functional" })
+ public void pfamFileDataExtraction() throws Exception
{
- test(new File(PfamFile));
- AppletFormatAdapter af = new AppletFormatAdapter();
- AlignmentI al = af.readFile(PfamFile, af.FILE, new IdentifyFile().Identify(PfamFile, af.FILE));
- int numpdb=0;
- for (SequenceI sq:al.getSequences())
+ AppletFormatAdapter af = new AppletFormatAdapter();
+ AlignmentI al = af.readFile(PfamFile, af.FILE,
+ new IdentifyFile().identify(PfamFile, af.FILE));
+ int numpdb = 0;
+ for (SequenceI sq : al.getSequences())
+ {
+ if (sq.getAllPDBEntries() != null)
{
- if (sq.getPDBId()!=null)
- {
- numpdb+=sq.getPDBId().size();
- }
+ numpdb += sq.getAllPDBEntries().size();
}
- assertTrue("PF00111 seed alignment has at least 1 PDB file, but the reader found none.",numpdb>0);
+ }
+ assertTrue(
+ "PF00111 seed alignment has at least 1 PDB file, but the reader found none.",
+ numpdb > 0);
}
- private void test(File f)
+ @Test(groups = { "Functional" })
+ public void rfamFileIO() throws Exception
+ {
+ testFileIOwithFormat(new File(RfamFile), "STH", 2, 1);
+ }
+
+ /**
+ * test alignment data in given file can be imported, exported and reimported
+ * with no dataloss
+ *
+ * @param f
+ * - source datafile (IdentifyFile.identify() should work with it)
+ * @param ioformat
+ * - label for IO class used to write and read back in the data from
+ * f
+ */
+
+ public static void testFileIOwithFormat(File f, String ioformat,
+ int naliannot, int nminseqann)
{
System.out.println("Reading file: " + f);
String ff = f.getPath();
try
{
- Alignment al = new AppletFormatAdapter().readFile(ff, AppletFormatAdapter.FILE,
- new IdentifyFile().Identify(ff, AppletFormatAdapter.FILE));
-
+ AppletFormatAdapter rf = new AppletFormatAdapter();
+
+ AlignmentI al = rf.readFile(ff, AppletFormatAdapter.FILE,
+ new IdentifyFile().identify(ff, AppletFormatAdapter.FILE));
+
+ assertNotNull("Couldn't read supplied alignment data.", al);
+
+ // make sure dataset is initialised ? not sure about this
for (int i = 0; i < al.getSequencesArray().length; ++i)
{
- al.getSequenceAt(i).setDatasetSequence(al.getSequenceAt(i));
+ al.getSequenceAt(i).createDatasetSequence();
+ }
+ String outputfile = rf.formatSequences(ioformat, al, true);
+ System.out.println("Output file in '" + ioformat + "':\n"
+ + outputfile + "\n<<EOF\n");
+ // test for consistency in io
+ AlignmentI al_input = new AppletFormatAdapter().readFile(outputfile,
+ AppletFormatAdapter.PASTE, ioformat);
+ assertNotNull("Couldn't parse reimported alignment data.", al_input);
+
+ String identifyoutput = new IdentifyFile().identify(outputfile,
+ AppletFormatAdapter.PASTE);
+ assertNotNull("Identify routine failed for outputformat " + ioformat,
+ identifyoutput);
+ assertTrue(
+ "Identify routine could not recognise output generated by '"
+ + ioformat + "' writer",
+ ioformat.equals(identifyoutput));
+ testAlignmentEquivalence(al, al_input, false);
+ int numaliannot = 0, numsqswithali = 0;
+ for (AlignmentAnnotation ala : al_input.getAlignmentAnnotation())
+ {
+ if (ala.sequenceRef == null)
+ {
+ numaliannot++;
+ }
+ else
+ {
+ numsqswithali++;
+ }
+ }
+ if (naliannot > -1)
+ {
+ assertEquals("Number of alignment annotations", naliannot,
+ numaliannot);
}
- AlignFile stFile = new StockholmFile(al);
- stFile.setSeqs(al.getSequencesArray());
- String stockholmoutput = stFile.print();
- Alignment al_input = new AppletFormatAdapter().readFile(stockholmoutput,
- AppletFormatAdapter.PASTE, "STH");
- if (al != null && al_input != null)
+ assertTrue(
+ "Number of sequence associated annotations wasn't at least "
+ + nminseqann, numsqswithali >= nminseqann);
+
+ } catch (Exception e)
+ {
+ e.printStackTrace();
+ assertTrue("Couln't format the alignment for output file.", false);
+ }
+ }
+
+ /**
+ * assert alignment equivalence
+ *
+ * @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
+ */
+ public static void testAlignmentEquivalence(AlignmentI al,
+ AlignmentI al_input, boolean ignoreFeatures)
+ {
+ assertNotNull("Original alignment was null", al);
+ assertNotNull("Generated alignment was null", al_input);
+
+ assertTrue("Alignment dimension mismatch: original: " + al.getHeight()
+ + "x" + al.getWidth() + ", generated: " + al_input.getHeight()
+ + "x" + al_input.getWidth(),
+ al.getHeight() == al_input.getHeight()
+ && al.getWidth() == al_input.getWidth());
+
+ // check Alignment annotation
+ AlignmentAnnotation[] aa_new = al_input.getAlignmentAnnotation();
+ AlignmentAnnotation[] aa_original = al.getAlignmentAnnotation();
+
+ // note - at moment we do not distinguish between alignment without any
+ // annotation rows and alignment with no annotation row vector
+ // 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>();
+
+ if (aa_new != null && aa_original != null)
+ {
+ for (int i = 0; i < aa_original.length; i++)
{
- System.out.println("Alignment contains: " + al.getHeight()
- + " and " + al_input.getHeight() + " sequences; "
- + al.getWidth() + " and " + al_input.getWidth()
- + " columns.");
- AlignmentAnnotation[] aa_new = al_input.getAlignmentAnnotation();
- AlignmentAnnotation[] aa_original = al.getAlignmentAnnotation();
-
- // check Alignment annotation
- if (aa_new != null && aa_original != null)
+ if (aa_new.length > i)
{
- System.out.println("Alignment contains: " + aa_new.length
- + " and " + aa_original.length
- + " alignment annotation(s)");
- for (int i = 0; i < aa_original.length; i++)
+ assertTrue("Different alignment annotation at position " + i,
+ equalss(aa_original[i], aa_new[i]));
+ // 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,
+ 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
+ Integer o_ggrp = new Integer(aa_original[i].graphGroup + 2);
+ Integer n_ggrp = new Integer(aa_new[i].graphGroup + 2);
+ BitSet orig_g = orig_groups.get(o_ggrp);
+ BitSet new_g = new_groups.get(n_ggrp);
+ if (orig_g == null)
+ {
+ orig_groups.put(o_ggrp, orig_g = new BitSet());
+ }
+ if (new_g == null)
{
- assertTrue("Different alignment annotation",equalss(aa_original[i], aa_new[i]));
-
+ new_groups.put(n_ggrp, new_g = new BitSet());
}
+ assertEquals("Graph Group pattern differs at annotation " + i,
+ orig_g, new_g);
+ orig_g.set(i);
+ new_g.set(i);
+ }
+ else
+ {
+ System.err.println("No matching annotation row for "
+ + aa_original[i].toString());
}
+ }
+ }
+ assertEquals(
+ "Generated and imported alignment have different annotation sets",
+ aa_original_size, aa_new_size);
- // check sequences, annotation and features
- SequenceI[] seq_original = new SequenceI[al.getSequencesArray().length];
- seq_original = al.getSequencesArray();
- SequenceI[] seq_new = new SequenceI[al_input.getSequencesArray().length];
- seq_new = al_input.getSequencesArray();
- SequenceFeature[] sequenceFeatures_original, sequenceFeatures_new;
- AlignmentAnnotation annot_original, annot_new;
- //
- for (int i = 0; i < al.getSequencesArray().length; i++)
+ // check sequences, annotation and features
+ SequenceI[] seq_original = new SequenceI[al.getSequencesArray().length];
+ seq_original = al.getSequencesArray();
+ SequenceI[] seq_new = new SequenceI[al_input.getSequencesArray().length];
+ seq_new = al_input.getSequencesArray();
+ SequenceFeature[] sequenceFeatures_original, sequenceFeatures_new;
+ AlignmentAnnotation annot_original, annot_new;
+ //
+ for (int i = 0; i < al.getSequencesArray().length; i++)
+ {
+ String name = seq_original[i].getName();
+ int start = seq_original[i].getStart();
+ int end = seq_original[i].getEnd();
+ System.out.println("Check sequence: " + name + "/" + start + "-"
+ + end);
+
+ // search equal sequence
+ for (int in = 0; in < al_input.getSequencesArray().length; in++)
+ {
+ if (name.equals(seq_new[in].getName())
+ && start == seq_new[in].getStart()
+ && end == seq_new[in].getEnd())
{
- String name = seq_original[i].getName();
- int start = seq_original[i].getStart();
- int end = seq_original[i].getEnd();
- System.out.println("Check sequence: " + name + "/" + start + "-"
- + end);
-
- // search equal sequence
- for (int in = 0; in < al_input.getSequencesArray().length; in++)
+ String ss_original = seq_original[i].getSequenceAsString();
+ String ss_new = seq_new[in].getSequenceAsString();
+ assertEquals("The sequences " + name + "/" + start + "-" + end
+ + " are not equal", ss_original, ss_new);
+
+ assertTrue(
+ "Sequence Features were not equivalent"
+ + (ignoreFeatures ? " ignoring." : ""),
+ ignoreFeatures
+ || (seq_original[i].getSequenceFeatures() == null && seq_new[in]
+ .getSequenceFeatures() == null)
+ || (seq_original[i].getSequenceFeatures() != null && seq_new[in]
+ .getSequenceFeatures() != null));
+ // compare sequence features
+ if (seq_original[i].getSequenceFeatures() != null
+ && seq_new[in].getSequenceFeatures() != null)
{
- if (name.equals(seq_new[in].getName())
- && start == seq_new[in].getStart()
- && end == seq_new[in].getEnd())
- {
- String ss_original = seq_original[i].getSequenceAsString();
- String ss_new = seq_new[in].getSequenceAsString();
- assertTrue("The sequences " + name + "/" + start
- + "-" + end + " are not equal", ss_original.equals(ss_new));
-
- assertTrue("Sequence Features were not equivalent", (seq_original[i].getSequenceFeatures()==null && seq_new[in].getSequenceFeatures()==null) || (seq_original[i].getSequenceFeatures()!=null && seq_new[in].getSequenceFeatures()!=null));
- // compare sequence features
- if (seq_original[i].getSequenceFeatures() != null
- && 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();
-
- assertTrue("different number of features", seq_original[i].getSequenceFeatures().length == seq_new[in]
- .getSequenceFeatures().length);
-
- for (int feat = 0; feat < seq_original[i]
- .getSequenceFeatures().length; feat++)
- {
- assertTrue("Different features",sequenceFeatures_original[feat]
- .equals(sequenceFeatures_new[feat]));
- }
- }
+ 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();
- // compare alignment annotation
- if (al.getSequenceAt(i).getAnnotation() != null
- && al_input.getSequenceAt(in).getAnnotation() != null)
- {
- for (int j = 0; j < al.getSequenceAt(i).getAnnotation().length; j++)
- {
- if (al.getSequenceAt(i).getAnnotation()[j] != null
- && al_input.getSequenceAt(in).getAnnotation()[j] != null)
- {
- annot_original = al.getSequenceAt(i).getAnnotation()[j];
- annot_new = al_input.getSequenceAt(in).getAnnotation()[j];
- assertTrue("Different annotation", equalss(annot_original, annot_new));
- }
- }
- }
- else if (al.getSequenceAt(i).getAnnotation() == null
- && al_input.getSequenceAt(in).getAnnotation() == null)
- {
- System.out.println("No annotations");
- }
- else if (al.getSequenceAt(i).getAnnotation() != null
- && al_input.getSequenceAt(in).getAnnotation() == null)
+ assertEquals("different number of features",
+ seq_original[i].getSequenceFeatures().length,
+ seq_new[in]
+ .getSequenceFeatures().length);
+
+ for (int feat = 0; feat < seq_original[i].getSequenceFeatures().length; feat++)
+ {
+ assertEquals("Different features",
+ sequenceFeatures_original[feat],
+ sequenceFeatures_new[feat]);
+ }
+ }
+ // compare alignment annotation
+ if (al.getSequenceAt(i).getAnnotation() != null
+ && al_input.getSequenceAt(in).getAnnotation() != null)
+ {
+ for (int j = 0; j < al.getSequenceAt(i).getAnnotation().length; j++)
+ {
+ if (al.getSequenceAt(i).getAnnotation()[j] != null
+ && al_input.getSequenceAt(in).getAnnotation()[j] != null)
{
- assertTrue("Annotations differed between sequences ("+al.getSequenceAt(i).getName()+") and ("+al_input.getSequenceAt(i).getName()+")", false);
+ annot_original = al.getSequenceAt(i).getAnnotation()[j];
+ annot_new = al_input.getSequenceAt(in).getAnnotation()[j];
+ assertTrue("Different annotation elements",
+ equalss(annot_original, annot_new));
}
- break;
}
}
+ else if (al.getSequenceAt(i).getAnnotation() == null
+ && al_input.getSequenceAt(in).getAnnotation() == null)
+ {
+ System.out.println("No annotations");
+ }
+ else if (al.getSequenceAt(i).getAnnotation() != null
+ && al_input.getSequenceAt(in).getAnnotation() == null)
+ {
+ fail("Annotations differed between sequences ("
+ + al.getSequenceAt(i).getName() + ") and ("
+ + al_input.getSequenceAt(i).getName() + ")");
+ }
+ break;
}
}
- else
- {
- assertTrue("Couldn't read alignment", false);
- }
- } catch (Exception e)
- {
- e.printStackTrace();
- assertTrue("Couln't format the alignment for output file.", false);
}
}
/*
* compare annotations
*/
- private boolean equalss(AlignmentAnnotation annot_or,
+ private static boolean equalss(AlignmentAnnotation annot_or,
AlignmentAnnotation annot_new)
{
if (annot_or.annotations.length != annot_new.annotations.length)
{
+ System.err.println("Different lengths for annotation row elements: "
+ + annot_or.annotations.length + "!="
+ + annot_new.annotations.length);
return false;
}
for (int i = 0; i < annot_or.annotations.length; i++)
{
- if (annot_or.annotations[i] != null
- && annot_new.annotations[i] != null)
+ Annotation an_or = annot_or.annotations[i], an_new = annot_new.annotations[i];
+ if (an_or != null && an_new != null)
{
- // Jim's comment - shouldn't the conditional here be using || not && for all these clauses ?
- if (!annot_or.annotations[i].displayCharacter
- .equals(annot_new.annotations[i].displayCharacter)
- && annot_or.annotations[i].secondaryStructure != annot_new.annotations[i].secondaryStructure
- && !annot_or.annotations[i].description
- .equals(annot_new.annotations[i].description))
+ 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()))))
{
+ 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;
}
}
}
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;
}
}
git://source.jalview.org / jalview.git / blobdiff