public void testNormalScenario()
{
TagExtractor extractor = new TagExtractor("abstract", 10);
Assert.AreEqual(extractor.count(), 0);
LangProfile profile = new LangProfile("en");
// normal
extractor.setTag("abstract");
extractor.add("This is a sample text.");
profile.update(extractor.closeTag());
Assert.AreEqual(extractor.count(), 1);
Assert.AreEqual(profile.n_words[0], 17); // Thisisasampletext
Assert.AreEqual(profile.n_words[1], 22); // _T, Th, hi, ...
Assert.AreEqual(profile.n_words[2], 17); // _Th, Thi, his, ...
// too short
extractor.setTag("abstract");
extractor.add("sample");
profile.update(extractor.closeTag());
Assert.AreEqual(extractor.count(), 1);
// other tags
extractor.setTag("div");
extractor.add("This is a sample text which is enough long.");
profile.update(extractor.closeTag());
Assert.AreEqual(extractor.count(), 1);
}
/// <summary>
/// Load Wikipedia abstract database file and generate its language profile
/// </summary>
/// <param name="lang">target language name</param>
/// <param name="file">target database file path</param>
/// <returns>Language profile instance</returns>
/// <exception cref="LangDetectException" />
public static LangProfile loadFromWikipediaAbstract(string lang, string file)
{
LangProfile profile = new LangProfile(lang);
FileInfo fi = new FileInfo(file);
Stream _is = null;
try
{
_is = fi.OpenRead();
if (fi.Name.EndsWith(".gz"))
{
_is = new GZipStream(_is, CompressionMode.Decompress);
}
using (StreamReader br = new StreamReader(_is, System.Text.Encoding.UTF8))
{
TagExtractor tagextractor = new TagExtractor("abstract", 100);
using (XmlReader reader = XmlReader.Create(br))
{
while (reader.Read())
{
switch (reader.NodeType)
{
case XmlNodeType.Element:
tagextractor.setTag(reader.Name);
break;
case XmlNodeType.Text:
tagextractor.add(reader.Value);
break;
case XmlNodeType.EndElement:
string text = tagextractor.closeTag();
if (text != null)
{
profile.update(text);
}
break;
}
}
}
}
}
finally
{
if (null != _is)
{
_is.Close();
_is.Dispose();
}
}
return(profile);
}
public void testCloseTag()
{
TagExtractor extractor = new TagExtractor(null, 0);
extractor.closeTag(); // ignore
}