Stanford NLP Java Method Translation Issue

Question

I'm using the Stanford NLP tools in C#, via the IKVM Java interface. Also getting ideas from https://sergey-tihon.github.io/Stanford.NLP.NET/StanfordCoreNLP.html

String text = "This is a test sentence.";
var props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, parse");
var curDir = Environment.CurrentDirectory;
Directory.SetCurrentDirectory(modelsDirectory);
var pipeline = new StanfordCoreNLP(props);
Directory.SetCurrentDirectory(curDir);
var annotation = new Annotation(text);
pipeline.annotate(annotation);

This code works fine to get my Annotation. However, when I try to access the annotation to extract various entities in the annotation, I run into trouble. Using code from this: How can I split a text into sentences using the Stanford parser?

List<CoreMap> sentences = annotation.get(SentencesAnnotation.class);

Not clear how to translate SentencesAnnotation.class into something that C# will accept.

score 2 · Accepted Answer · answered Jun 26 '19 at 21:42

2

In general, Java Foo.class translates to C# typeof(Foo), so C# should accept the following:

IList<CoreMap> sentences = annotation.get(typeof(SentencesAnnotation));

answered Jun 26 '19 at 21:42

Dave Doknjas

6,394
1
15
28

I could only get the code to compile and `sentences` to populate by using the `var` type. – Ric Gaudet Jul 09 '19 at 19:33

Stanford NLP Java Method Translation Issue

1 Answers1