public void TestSentenceBoundary()
{
// LUCENENET specific - using a mock of the JDK BreakIterator class, which is just
// an ICU BreakIterator with custom rules applied. East Asian
// languages are skipped because the DictionaryBasedBreakIterator is not overridden by the rules.
switch (CultureInfo.CurrentCulture.TwoLetterISOLanguageName)
{
case "th": // Thai
case "lo": // Lao
case "my": // Burmese
case "km": // Khmer
case "ja": // Japanese
case "ko": // Korean
case "zh": // Chinese
Assume.That(false, "This test does not apply to East Asian languages.");
break;
}
StringBuilder text = new StringBuilder(TEXT);
// we test this with default locale, its randomized by LuceneTestCase
// LUCENENET specific - using a mock of the JDK BreakIterator class, which is just
// an ICU BreakIterator with custom rules applied.
BreakIterator bi = JdkBreakIterator.GetSentenceInstance(CultureInfo.CurrentCulture);
IBoundaryScanner scanner = new BreakIteratorBoundaryScanner(bi);
int start = TEXT.IndexOf("any application", StringComparison.Ordinal);
int expected = TEXT.IndexOf("It is a", StringComparison.Ordinal);
TestFindStartOffset(text, start, expected, scanner);
expected = TEXT.IndexOf("Apache Lucene is an open source", StringComparison.Ordinal);
TestFindEndOffset(text, start, expected, scanner);
}
public void TestWordBoundary()
{
StringBuilder text = new StringBuilder(TEXT);
BreakIterator bi = JdkBreakIterator.GetWordInstance(CultureInfo.InvariantCulture);
IBoundaryScanner scanner = new BreakIteratorBoundaryScanner(bi);
int start = TEXT.IndexOf("formance");
int expected = TEXT.IndexOf("high-performance");
TestFindStartOffset(text, start, expected, scanner);
expected = TEXT.IndexOf(", full");
TestFindEndOffset(text, start, expected, scanner);
}
public void TestSentenceBoundary()
{
StringBuilder text = new StringBuilder(TEXT);
// we test this with default locale, its randomized by LuceneTestCase
BreakIterator bi = JdkBreakIterator.GetSentenceInstance(CultureInfo.CurrentCulture);
IBoundaryScanner scanner = new BreakIteratorBoundaryScanner(bi);
int start = TEXT.IndexOf("any application");
int expected = TEXT.IndexOf("It is a");
TestFindStartOffset(text, start, expected, scanner);
expected = TEXT.IndexOf("Apache Lucene is an open source");
TestFindEndOffset(text, start, expected, scanner);
}
public void TestWordBoundary()
{
StringBuilder text = new StringBuilder(TEXT);
// LUCENENET specific - using a mock of the JDK BreakIterator class, which is just
// an ICU BreakIterator with custom rules applied.
BreakIterator bi = JdkBreakIterator.GetWordInstance(CultureInfo.InvariantCulture);
IBoundaryScanner scanner = new BreakIteratorBoundaryScanner(bi);
int start = TEXT.IndexOf("formance", StringComparison.Ordinal);
int expected = TEXT.IndexOf("high-performance", StringComparison.Ordinal);
TestFindStartOffset(text, start, expected, scanner);
expected = TEXT.IndexOf(", full", StringComparison.Ordinal);
TestFindEndOffset(text, start, expected, scanner);
}
private BreakIterator GetWordInstance(System.Globalization.CultureInfo locale)
{
return(JdkBreakIterator.GetWordInstance(locale));
}
/// <summary>
/// Returns the <see cref="BreakIterator"/> to use for
/// dividing text into passages. This instantiates an
/// <see cref="BreakIterator.GetSentenceInstance(CultureInfo)"/> by default;
/// subclasses can override to customize.
/// </summary>
protected virtual BreakIterator GetBreakIterator(string field)
{
return(JdkBreakIterator.GetSentenceInstance(CultureInfo.InvariantCulture));
}
