/// <summary>
/// Evaluates the specified document.
/// </summary>
/// <param name="factory">The factory used in this analysis.</param>
/// <param name="document">The document to be analyzed.</param>
protected override void Evaluate(ITextFactory factory, IDocument document)
{
if (document.Sentences == null)
{
throw new AnalyzerException(this, "The document does not have the sentences detected.");
}
foreach (var sentence in document.Sentences)
{
var toks = sentence.GetTokens();
var tags = sentence.GetTags();
if (toks == null)
{
throw new AnalyzerException(this, "The document have a sentence without the tokenization.");
}
if (tags == null)
{
throw new AnalyzerException(this, "The document have a sentence without the part-of-speech tags.");
}
string[] chunks;
lock (Chunker) {
chunks = Chunker.Chunk(toks, tags);
}
for (var i = 0; i < chunks.Length; i++)
{
sentence.Tokens[i].ChunkTag = chunks[i];
}
var spans = ChunkSample.PhrasesAsSpanList(toks, tags, chunks);
var list = new List <IChunk>(spans.Length);
foreach (var span in spans)
{
var chunk = factory.CreateChunk(sentence, span);
if (chunk != null)
{
list.Add(chunk);
}
}
sentence.Chunks = new ReadOnlyCollection <IChunk>(list);
}
}
/// <summary>
/// Generates tagged chunk spans for the given sequence returning the result in a span array.
/// </summary>
/// <param name="tokens">An array of the tokens or words of the sequence.</param>
/// <param name="tags">An array of the pos tags of the sequence.</param>
/// <returns>An array of spans with chunk tags for each chunk in the sequence.</returns>
public Span[] ChunkAsSpans(string[] tokens, string[] tags)
{
var preds = Chunk(tokens, tags);
return(ChunkSample.PhrasesAsSpanList(tokens, tags, preds));
}