Handles cases where non-numbers (unparseable inputs) are passed to Wo…

…rdsToNumber.
clulab · Aug 9, 2019 · 108116b · 108116b
1 parent d09f2ed
commit 108116b
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 6 deletions.
diff --git a/src/main/scala/org/clulab/timenorm/scate/WordsToNumber.scala b/src/main/scala/org/clulab/timenorm/scate/WordsToNumber.scala
@@ -7,6 +7,8 @@ import org.apache.commons.io.IOUtils
 import org.clulab.timenorm.scfg.SynchronousParser.Tree
 import org.clulab.timenorm.scfg.{SynchronousGrammar, SynchronousParser}
 
+import scala.util.{Failure, Success}
+
 
 object WordsToNumber {
   def apply(languageCode: String): WordsToNumber = languageCode match {
@@ -16,14 +18,14 @@ object WordsToNumber {
 
 class WordsToNumber(grammarStream: InputStream) extends Function[Array[String], Option[Long]] {
 
-  private val grammar = SynchronousGrammar.fromString(IOUtils.toString(grammarStream, Charset.forName("ascii")))
-  private val sourceWords = grammar.sourceSymbols()
-  private val parser = new SynchronousParser(grammar)
+  private val parser = new SynchronousParser(SynchronousGrammar.fromString(
+    IOUtils.toString(grammarStream, Charset.forName("ascii"))))
 
   override def apply(words: Array[String]): Option[Long] = {
-    if (!words.forall(sourceWords)) None else parser.parseAll(words) match {
-      case Array(tree) => Some(this.toDigits(tree).foldLeft(0L) { case (sum, digit) => 10L * sum + digit })
-      case trees => throw new UnsupportedOperationException(
+    parser.tryParseAll(words.toIndexedSeq) match {
+      case Failure(_) => None
+      case Success(IndexedSeq(tree)) => Some(this.toDigits(tree).foldLeft(0L) { case (sum, digit) => 10L * sum + digit })
+      case Success(trees) => throw new UnsupportedOperationException(
         s"Ambiguous grammar for ${words.toList}. Parses:\n${trees.mkString("\n")}")
     }
   }

diff --git a/src/main/scala/org/clulab/timenorm/scfg/SynchronousParser.scala b/src/main/scala/org/clulab/timenorm/scfg/SynchronousParser.scala
@@ -3,6 +3,7 @@ package org.clulab.timenorm.scfg
 import scala.collection.immutable.IndexedSeq
 import scala.collection.mutable
 import scala.collection.mutable.ListBuffer
+import scala.util.{Try, control}
 
 /**
  * A parser for synchronous grammars.
@@ -54,6 +55,16 @@ class SynchronousParser(grammar: SynchronousGrammar) {
     trees
   }
 
+  /**
+    * Attempt to parse the source tokens into a tree of non-terminals and target tokens.
+    *
+    * @param sourceTokens The source tokens to be parsed.
+    * @return Success(trees) if the source tokens could be parsed, Failure otherwise.
+    */
+  def tryParseAll(sourceTokens: IndexedSeq[String]): Try[IndexedSeq[Tree.NonTerminal]] = {
+    control.Exception.catching(classOf[UnsupportedOperationException]).withTry(parseAll(sourceTokens))
+  }
+
   private def parseChart(sourceTokens: IndexedSeq[String]): Array[Array[ChartEntry]] = {
     val nTokens = sourceTokens.size
     val chart = Array.tabulate(nTokens + 1, nTokens) {

diff --git a/src/test/scala/org/clulab/timenorm/scate/WordsToNumberTest.scala b/src/test/scala/org/clulab/timenorm/scate/WordsToNumberTest.scala
@@ -36,5 +36,7 @@ class WordsToNumberTest extends FunSuite with TypesSuite {
 
   test("English invalid numbers") {
     assert(enTextToNumber(Array("several")) === None)
+    assert(enTextToNumber(Array("and")) === None)
+    assert(enTextToNumber(Array.empty) === None)
   }
 }