static void AdditionalExamples()
{
var text = "Chris Manning is a nice person. Chris wrote a simple sentence. He also gives oranges to people.";
using (var client = new CoreNLPClient(
//classPath: "%CLASSPATH%",
//classPath: @"C:\Users\user\Dev\ProductSort\src\Libs\CoreNLP\stanford-corenlp-4.0.0",
//args: new string[] {"ssl"},
//kwargs: JObject.Parse("{ \"username\": \"username\", \"password\": \"password\" }"),
preLoad: false,
startServer: true,
annotators: new string[] { "tokenize", "ssplit", "pos", "lemma", "ner", "parse", "depparse", "coref" },
//properties: JObject.Parse("{\"annotators\": \"tokenize,ssplit,pos,lemma,ner,parse,depparse,coref\", \"tokenize.language\": \"en\"}"),
//properties: "english",
//properties: "corenlp_server-stanzaexample.props",
timeout: 30000,
memory: "8G"))
//var text = "José López es una buena persona. José ha escrito una frase simple. También reparte naranjas a la gente.";
//using (var client = new CoreNLPClient(
// properties: "spanish",
// timeout: 30000,
// memory: "8G"))
{
var ann = (Document)client.Annotate(text);
if (ann == null)
{
return;
}
var sentence = ann.Sentence[0];
Console.WriteLine("Annotate OK");
var pattern = "([ner: PERSON]+) /wrote/ /an?/ []{0,3} /sentence|article/";
var matches = client.TokensRegex(text, pattern);
Console.WriteLine("TokensRegex OK");
pattern = "{word:wrote} >nsubj {}=subject >dobj {}=object";
matches = client.Semgrex(text, pattern);
Console.WriteLine("Semgrex OK");
pattern = "NP";
matches = client.TRegex(text, pattern);
Console.WriteLine("TRegex OK");
}
}
static void StanzaExample()
{
var text = "Chris Manning is a nice person. Chris wrote a simple sentence. He also gives oranges to people.";
using (var client = new CoreNLPClient(
annotators: new string[] { "tokenize", "ssplit", "pos", "lemma", "ner", "parse", "depparse", "coref" },
timeout: 30000,
memory: "8G"))
{
var ann = (Document)client.Annotate(text);
var sentence = ann.Sentence[0];
Console.WriteLine("---");
Console.WriteLine("Constituency parse of first sentence");
var constituencyParse = sentence.ParseTree;
Console.WriteLine(constituencyParse);
Console.WriteLine("---");
Console.WriteLine("First subtree of constituency parse");
Console.WriteLine(constituencyParse.Child[0]);
Console.WriteLine("---");
Console.WriteLine("Value of first subtree of constituency parse");
Console.WriteLine(constituencyParse.Child[0].Value);
Console.WriteLine("---");
Console.WriteLine("Dependency parse of first sentence");
var dependencyParse = sentence.BasicDependencies;
Console.WriteLine(dependencyParse);
Console.WriteLine("---");
Console.WriteLine("First token of first sentence");
var token = sentence.Token[0];
Console.WriteLine(token);
Console.WriteLine("---");
Console.WriteLine("Part of Speech tag of token");
Console.WriteLine(token.Pos);
Console.WriteLine("---");
Console.WriteLine("Named Entity tag of token");
Console.WriteLine(token.Ner);
Console.WriteLine("---");
Console.WriteLine("First entity mention in sentence");
Console.WriteLine(sentence.Mentions[0]);
Console.WriteLine("---");
Console.WriteLine("Coref chains for the example");
Console.WriteLine(ann.CorefChain);
var pattern = "([ner: PERSON]+) /wrote/ /an?/ []{0,3} /sentence|article/";
var matches = client.TokensRegex(text, pattern);
Debug.Assert((int)matches["sentences"].Count == 3);
Debug.Assert((int)matches["sentences"][1]["length"] == 1);
Debug.Assert((string)matches["sentences"][1]["0"]["text"] == "Chris wrote a simple sentence");
Debug.Assert((string)matches["sentences"][1]["0"]["1"]["text"] == "Chris");
pattern = "{word:wrote} >nsubj {}=subject >dobj {}=object";
matches = client.Semgrex(text, pattern);
Debug.Assert((int)matches["sentences"].Count == 3);
Debug.Assert((int)matches["sentences"][1]["length"] == 0);
pattern = "NP";
matches = client.TRegex(text, pattern);
foreach (var match in matches)
{
Console.WriteLine(match);
}
}
}
