Merge remote-tracking branch 'origin/master'

Author: manning

commonbuildjsp.xml

@@ -9,6 +9,7 @@
     </fileset>
     <pathelement location="${build.path}"/>
     <pathelement location="${project.core}/lib/commons-logging.jar"/>
+    <pathelement location="${project.core}/lib/javax.servlet.jar"/>
   </path>
 
   <target name="jsp" depends="classpath,compile">

src/edu/stanford/nlp/ie/machinereading/structure/Span.java

@@ -53,7 +53,23 @@ public static Span fromValues(int val1, int val2) {
       return new Span(val2, val1);
     }
   }
-  
+
+  public static Span fromValues(Object... values) {
+    if (values.length == 1) {
+      return fromValues(values[0], values[0] instanceof Number ? ((Number) values[0]).intValue() + 1 : Integer.parseInt(values[0].toString()) + 1);
+    }
+    if (values.length != 2) { throw new IllegalArgumentException("fromValues() must take an array with 2 elements"); }
+    int val1;
+    if (values[0] instanceof Number) { val1 = ((Number) values[0]).intValue(); }
+    else if (values[0] instanceof String) { val1 = Integer.parseInt((String) values[0]); }
+    else { throw new IllegalArgumentException("Unknown value for span: " + values[0]); }
+    int val2;
+    if (values[1] instanceof Number) { val2 = ((Number) values[1]).intValue(); }
+    else if (values[0] instanceof String) { val2 = Integer.parseInt((String) values[1]); }
+    else { throw new IllegalArgumentException("Unknown value for span: " + values[1]); }
+    return fromValues(val1, val2);
+  }
+
   public int start() { return start; }
   public int end() { return end; }
 

src/edu/stanford/nlp/trees/EnglishGrammaticalRelations.java

@@ -1037,8 +1037,8 @@ public static class NounCompoundModifierGRAnnotation extends GrammaticalRelation
           // which is the normal case for such a pattern.
           "WHNP|WHNP-TMP|WHNP-ADV|NP|NP-TMP|NP-ADV < (NP=target !<: CD $- /^,$/ $-- /^(?:WH)?NP/ !$ CC|CONJP)",
           "WHNP|WHNP-TMP|WHNP-ADV|NP|NP-TMP|NP-ADV < (PRN=target < (NP < /^(?:NN|CD)/ $-- /^-LRB-$/ $+ /^-RRB-$/))",
-          // Maybe delete '@' in next pattern, since not clearly appositional when NP-ADV or NP-TMP. But then what is it?
-          "@WHNP|NP < (@NP=target !<: CD <, /^-LRB-$/ <` /^-RRB-$/ $-- /^(?:WH)?NP/ !$ CC|CONJP)",
+          // NP-ADV is a npadvmod, NP-TMP is a tmod
+          "@WHNP|NP < (NP=target !<: CD <, /^-LRB-$/ <` /^-RRB-$/ $-- /^(?:WH)?NP/ !$ CC|CONJP)",
           // TODO: next pattern with NNP doesn't work because leftmost NNP is deemed head in a
           // structure like (NP (NNP Norway) (, ,) (NNP Verdens_Gang) (, ,))
           "NP|NP-TMP|NP-ADV < (NNP $+ (/^,$/ $+ NNP=target)) !< CC|CONJP",

src/edu/stanford/nlp/util/Characters.java

@@ -63,4 +63,15 @@ public static boolean isPunctuation(char c) {
         cType == Character.INITIAL_QUOTE_PUNCTUATION ||
         cType == Character.FINAL_QUOTE_PUNCTUATION);
   }
+
+  /**
+   * Returns true if a character is a control character, and
+   * false otherwise.
+   * 
+   * @param c
+   * @return
+   */
+  public static boolean isControl(char c) {
+    return Character.getType(c) == Character.CONTROL;
+  }
 }