More unit tests

1 files changed, 20 insertions, 1 deletions

diff --git a/linguistics/src/test/java/com/yahoo/language/sentencepiece/SentencePieceTest.java b/linguistics/src/test/java/com/yahoo/language/sentencepiece/SentencePieceTest.java
index 70361f55750..7d0c1c5c78e 100644
--- a/linguistics/src/test/java/com/yahoo/language/sentencepiece/SentencePieceTest.java
+++ b/linguistics/src/test/java/com/yahoo/language/sentencepiece/SentencePieceTest.java
@@ -17,7 +17,7 @@ import static org.junit.Assert.assertArrayEquals;
 public class SentencePieceTest {
 
     @Test
-    public void testEnglishTokenization() throws IOException {
+    public void testEnglishTokenization() {
         var tester = new SentencePieceTester(new File("src/test/models/sentencepiece/en.wiki.bpe.vs10000.model").toPath());
         tester.assertSegmented("h", "▁h");
         tester.assertSegmented("he", "▁he");
@@ -42,6 +42,25 @@ public class SentencePieceTest {
     }
 
     @Test
+    public void testNoCollapse() {
+        var tester = new SentencePieceTester(new SentencePieceEncoder.Builder()
+                                                     .addDefaultModel(new File("src/test/models/sentencepiece/en.wiki.bpe.vs10000.model").toPath())
+                                                     .setCollapseUnknowns(false));
+        tester.assertSegmented("KHJ hello", "▁", "K", "H", "J", "▁hel", "lo");
+    }
+
+    @Test
+    public void testHighestScore() {
+        var tester = new SentencePieceTester(new SentencePieceEncoder.Builder()
+                                                     .addDefaultModel(new File("src/test/models/sentencepiece/en.wiki.bpe.vs10000.model").toPath())
+                                                     .setScoring(SentencePieceEncoder.Scoring.highestScore));
+        tester.assertSegmented("h", "▁h");
+        tester.assertSegmented("he", "▁he");
+        tester.assertSegmented("hel", "▁h", "el");
+        tester.assertSegmented("hello", "▁h", "el", "lo");
+    }
+
+    @Test
     public void testJapaneseTokenization() throws IOException {
         SentencePieceEncoder.Builder builder = new SentencePieceEncoder.Builder();
         builder.addModel(Language.JAPANESE, new File("src/test/models/sentencepiece/ja.wiki.bpe.vs5000.model").toPath());