move dependency conversion utility to GrammaticalStructureConversionU…

…tils, use same method for all converters
stanfordnlp · Jul 9, 2016 · 3a0b91c · 3a0b91c
1 parent 49c6e52
commit 3a0b91c
Show file tree

Hide file tree

Showing 15 changed files with 1,161 additions and 1,549 deletions.
diff --git a/src/edu/stanford/nlp/ie/AbstractSequenceClassifier.java b/src/edu/stanford/nlp/ie/AbstractSequenceClassifier.java
@@ -657,21 +657,20 @@ public List<Triple<String, Integer, Integer>> classifyToCharacterOffsets(String
   }
 
   /**
-   * Have a word segmenter segment a String into a list of words.
+   * ONLY USE IF LOADED A CHINESE WORD SEGMENTER!!!!!
-   * ONLY USE IF YOU LOADED A CHINESE WORD SEGMENTER!!!!!
    *
-   * @param sentence The string to be classified
+   * @param sentence
+   *          The string to be classified
    * @return List of words
    */
-  // This method is currently [2016] only called in a very small number of places:
-  // the parser's jsp webapp, ChineseSegmenterAnnotator, and SegDemo.
-  // Maybe we could eliminate it?
   public List<String> segmentString(String sentence) {
     return segmentString(sentence, defaultReaderAndWriter);
   }
 
-  public List<String> segmentString(String sentence, DocumentReaderAndWriter<IN> readerAndWriter) {
+  public List<String> segmentString(String sentence,
-    ObjectBank<List<IN>> docs = makeObjectBankFromString(sentence, readerAndWriter);
+                                    DocumentReaderAndWriter<IN> readerAndWriter) {
+    ObjectBank<List<IN>> docs = makeObjectBankFromString(sentence,
+                                                         readerAndWriter);
 
     StringWriter stringWriter = new StringWriter();
     PrintWriter stringPrintWriter = new PrintWriter(stringWriter);
@@ -686,7 +685,7 @@ public List<String> segmentString(String sentence, DocumentReaderAndWriter<IN> r
     return Arrays.asList(segmented.split("\\s"));
   }
 
-  /*
+  /**
    * Classify the contents of {@link SeqClassifierFlags scf.testFile}. The file
    * should be in the format expected based on {@link SeqClassifierFlags
    * scf.documentReader}.
@@ -708,10 +707,7 @@ public List<String> segmentString(String sentence, DocumentReaderAndWriter<IN> r
    * @return The same {@link List}, but with the elements annotated with their
    *         answers (stored under the
    *         {@link edu.stanford.nlp.ling.CoreAnnotations.AnswerAnnotation}
-   *         key). The answers will be the class labels defined by the CRF
+   *         key).
-   *         Classifier. They might be things like entity labels (in BIO
-   *         notation or not) or something like "1" vs. "0" on whether to
-   *         begin a new token here or not (in word segmentation).
    */
   public abstract List<IN> classify(List<IN> document);
 

diff --git a/src/edu/stanford/nlp/ie/crf/CRFClassifier.java b/src/edu/stanford/nlp/ie/crf/CRFClassifier.java
@@ -960,7 +960,7 @@ protected static Index<CRFLabel> allLabels(int window, Index<String> classIndex)
    * Makes a CRFDatum by producing features and a label from input data at a
    * specific position, using the provided factory.
    *
-   * @param info The input data. Particular feature factories might look for arbitrary keys in the IN items.
+   * @param info The input data
    * @param loc The position to build a datum at
    * @param featureFactories The FeatureFactories to use to extract features
    * @return The constructed CRFDatum