public void testAdd()
{
TagExtractor extractor = new TagExtractor(null, 0);
extractor.Add("");
extractor.Add(null); // ignore
}
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);
}
public void testClear()
{
TagExtractor extractor = new TagExtractor("abstract", 10);
extractor.SetTag("abstract");
extractor.Add("This is a sample text.");
Assert.AreEqual(extractor.buf_.ToString(), "This is a sample text.");
Assert.AreEqual(extractor.tag_, "abstract");
extractor.Clear();
Assert.AreEqual(extractor.buf_.ToString(), "");
Assert.AreEqual(extractor.tag_, null);
}
public static LangProfile load(string lang, string file)
{
LangProfile profile = new LangProfile(lang);
TagExtractor tagextractor = new TagExtractor("abstract", 100);
Stream inputStream = null;
try
{
inputStream = File.OpenRead(file);
string extension = Path.GetExtension(file) ?? "";
if (extension.ToUpper() == ".GZ")
{
inputStream = new GZipStream(inputStream, CompressionMode.Decompress);
}
using (XmlReader xmlReader = XmlReader.Create(inputStream))
{
while (xmlReader.Read())
{
switch (xmlReader.NodeType)
{
case XmlNodeType.Element:
tagextractor.SetTag(xmlReader.Name);
break;
case XmlNodeType.Text:
tagextractor.Add(xmlReader.Value);
break;
case XmlNodeType.EndElement:
tagextractor.CloseTag(profile);
break;
}
}
}
}
finally
{
if (inputStream != null)
{
inputStream.Close();
}
}
Console.WriteLine(lang + ": " + tagextractor.Count);
return(profile);
}
/// <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>LangDetectException </exception>
public static LangProfile LoadFromWikipediaAbstract(string lang, string file)
{
LangProfile profile = new LangProfile(lang);
StreamReader br = null;
try
{
Stream strm = File.OpenRead(file);
if (file.EndsWith(".gz"))
{
strm = new GZipStream(strm, CompressionMode.Decompress);
}
br = new StreamReader(strm);
TagExtractor tagextractor = new TagExtractor("abstract", 100);
XmlReader reader = XmlReader.Create(br);
try
{
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;
}
}
}
catch (XmlException e)
{
throw new LangDetectException(ErrorCode.TrainDataFormatError, "Training database file '" + file + "' is an invalid XML.");
}
finally
{
try
{
if (reader != null)
{
reader.Close();
}
}
catch (XmlException e) { }
}
Console.WriteLine(lang + ":" + tagextractor.Count());
}
catch (IOException e)
{
throw new LangDetectException(ErrorCode.CantOpenTrainData, "Can't open training database file '" + file + "'");
}
finally
{
try
{
if (br != null)
{
br.Close();
}
}
catch (IOException e) { }
}
return(profile);
}
