Add a command line interface to the UniversalEnhancer

Author: AngledLuffa

src/edu/stanford/nlp/trees/ud/ProcessUniversalEnhancerRequest.java

@@ -0,0 +1,88 @@
+package edu.stanford.nlp.trees.ud;
+
+import java.io.InputStream;
+import java.io.IOException;
+import java.io.OutputStream;
+import java.util.regex.Pattern;
+
+import edu.stanford.nlp.ling.CoreAnnotations;;
+import edu.stanford.nlp.pipeline.Annotation;
+import edu.stanford.nlp.pipeline.CoreNLPProtos;
+import edu.stanford.nlp.pipeline.ProtobufAnnotationSerializer;
+import edu.stanford.nlp.semgraph.SemanticGraph;
+import edu.stanford.nlp.semgraph.SemanticGraphCoreAnnotations;
+import edu.stanford.nlp.trees.EnglishPatterns;
+import edu.stanford.nlp.util.CoreMap;
+import edu.stanford.nlp.util.ProcessProtobufRequest;
+
+
+public class ProcessUniversalEnhancerRequest extends ProcessProtobufRequest {
+
+  /**
+   * Enhance the dependencies on a single sentence, using the given relative pronouns pattern
+   */
+  public static void enhanceDependencies(Pattern relativePronounsPattern, Annotation annotation) {
+    for (CoreMap sentence : annotation.get(CoreAnnotations.SentencesAnnotation.class)) {
+      SemanticGraph basic = sentence.get(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation.class);
+      SemanticGraph enhanced = UniversalEnhancer.enhanceGraph(basic, null, false, null, relativePronounsPattern);
+      sentence.set(SemanticGraphCoreAnnotations.EnhancedDependenciesAnnotation.class, enhanced);
+    }
+  }
+
+  /**
+   * Process all sentences in the document, enhancing the basic dependencies on each sentence
+   */
+  public static CoreNLPProtos.Document processRequest(Pattern relativePronounsPattern, CoreNLPProtos.DependencyEnhancerRequest request) {
+    ProtobufAnnotationSerializer serializer = new ProtobufAnnotationSerializer();
+    Annotation annotation = serializer.fromProto(request.getDocument());
+
+    enhanceDependencies(relativePronounsPattern, annotation);
+
+    return serializer.toProto(annotation);
+  }
+
+  /**
+   * Figure out a relative patterns pronoun to use based on the information given in the request
+   */
+  public static Pattern getRelativePronouns(CoreNLPProtos.DependencyEnhancerRequest request) {
+    if (request.hasRelativePronouns()) {
+      return Pattern.compile(request.getRelativePronouns());
+    }
+    if (request.hasLanguage()) {
+      switch(request.getLanguage()) {
+      case English:
+      case UniversalEnglish:
+        return EnglishPatterns.RELATIVIZING_WORD_PATTERN;
+      case Chinese:
+      case UniversalChinese:
+        // there are no relative pronouns in Chinese!
+        return null;
+      default:
+        throw new IllegalArgumentException("Relative word pattern not defined for " + request.getLanguage());
+      }
+    }
+    throw new IllegalArgumentException("Could not find Language or predefined relative pronouns pattern in the request");
+  }
+
+  /**
+   * Process a single request, adding enhanced dependencies to each sentence in the document
+   */
+  @Override
+  public void processInputStream(InputStream in, OutputStream out) throws IOException {
+    CoreNLPProtos.DependencyEnhancerRequest request = CoreNLPProtos.DependencyEnhancerRequest.parseFrom(in);
+
+    Pattern relativePronounsPattern = getRelativePronouns(request);
+
+    CoreNLPProtos.Document response = processRequest(relativePronounsPattern, request);
+    response.writeTo(out);
+  }
+
+  /**
+   * The inherited main program will either enhance a single document,
+   * or will listen to stdin and enhance every document that comes in
+   * until a terminator is sent or the stream closes
+   */
+  public static void main(String[] args) throws IOException {
+    ProcessProtobufRequest.process(new ProcessUniversalEnhancerRequest(), args);
+  }
+}