public virtual void TestReadBackwards() { File file = CreateTestFile(); TaggedFileRecord record = CreateRecord(file, "tagColumn=0,wordColumn=1,"); IList <IList <TaggedWord> > sentences = new List <IList <TaggedWord> >(); foreach (IList <TaggedWord> sentence in record.Reader()) { sentences.Add(sentence); } NUnit.Framework.Assert.AreEqual(3, sentences.Count); NUnit.Framework.Assert.AreEqual(3, sentences[0].Count); NUnit.Framework.Assert.AreEqual("A", sentences[0][0].Tag()); NUnit.Framework.Assert.AreEqual("B", sentences[0][1].Tag()); NUnit.Framework.Assert.AreEqual("C", sentences[0][2].Tag()); NUnit.Framework.Assert.AreEqual("D", sentences[1][0].Tag()); NUnit.Framework.Assert.AreEqual("E", sentences[1][1].Tag()); NUnit.Framework.Assert.AreEqual("F", sentences[2][0].Tag()); NUnit.Framework.Assert.AreEqual("1", sentences[0][0].Word()); NUnit.Framework.Assert.AreEqual("2", sentences[0][1].Word()); NUnit.Framework.Assert.AreEqual("3", sentences[0][2].Word()); NUnit.Framework.Assert.AreEqual("4", sentences[1][0].Word()); NUnit.Framework.Assert.AreEqual("5", sentences[1][1].Word()); NUnit.Framework.Assert.AreEqual("6", sentences[2][0].Word()); }
internal virtual TaggedFileRecord CreateRecord(File file, string extraArgs) { string description = extraArgs + "format=TSV," + file; Properties props = new Properties(); return(TaggedFileRecord.CreateRecord(props, description)); }
public TextTaggedFileReader(TaggedFileRecord record) { filename = record.file; try { reader = new BufferedReader(new InputStreamReader(new FileInputStream(filename), record.encoding)); } catch (IOException e) { throw new Exception(e); } tagSeparator = record.tagSeparator; PrimeNext(); }
public TSVTaggedFileReader(TaggedFileRecord record) { filename = record.file; try { reader = new BufferedReader(new InputStreamReader(new FileInputStream(filename), record.encoding)); } catch (IOException e) { throw new Exception(e); } wordColumn = ((record.wordColumn == null) ? DefaultWordColumn : record.wordColumn); tagColumn = ((record.tagColumn == null) ? DefaultTagColumn : record.tagColumn); PrimeNext(); }
public virtual void TestError() { File file = CreateBrokenFile(); TaggedFileRecord record = CreateRecord(file, "tagColumn=0,wordColumn=1,"); try { foreach (IList <TaggedWord> sentence in record.Reader()) { throw new AssertionError("Should have thrown an error " + " reading a file with no tags"); } } catch (ArgumentException) { } }
public TreeTaggedFileReader(TaggedFileRecord record) { // int numSentences = 0; filename = record.file; trf = record.trf == null ? new LabeledScoredTreeReaderFactory() : record.trf; transformer = record.treeTransformer; normalizer = record.treeNormalizer; treeFilter = record.treeFilter; treebank = new DiskTreebank(trf, record.encoding); if (record.treeRange != null) { treebank.LoadPath(filename, record.treeRange); } else { treebank.LoadPath(filename); } treeIterator = treebank.GetEnumerator(); FindNext(); }