CODEC-312: Fix possible StringIndexOutOfBoundException thrown by Matc…

…hRatingApproachEncoder.encode() method (#220) * CODEC-312: Fix possible StringIndexOutOfBoundException Signed-off-by: Arthur Chan <arthur.chan@adalogics.com> * CODEC-312: Add unit test Signed-off-by: Arthur Chan <arthur.chan@adalogics.com> * Remove unmaintained comments --------- Signed-off-by: Arthur Chan <arthur.chan@adalogics.com> Co-authored-by: Gary Gregory <garydgregory@users.noreply.github.com>
apache · Nov 25, 2023 · 047d247 · 047d247
1 parent a2e542d
commit 047d247
Show file tree

Hide file tree

Showing 2 changed files with 24 additions and 14 deletions.
diff --git a/src/main/java/org/apache/commons/codec/language/MatchRatingApproachEncoder.java b/src/main/java/org/apache/commons/codec/language/MatchRatingApproachEncoder.java
@@ -126,10 +126,20 @@ public final String encode(String name) {
         // Preprocessing
         name = cleanName(name);
 
+        // Bulletproof if name becomes empty after cleanName(name)
+        if (SPACE.equals(name) || name.isEmpty()) {
+            return EMPTY;
+        }
+
         // BEGIN: Actual encoding part of the algorithm...
         // 1. Delete all vowels unless the vowel begins the word
         name = removeVowels(name);
 
+        // Bulletproof if name becomes empty after removeVowels(name)
+        if (SPACE.equals(name) || name.isEmpty()) {
+            return EMPTY;
+        }
+
         // 2. Remove second consonant from any double consonant
         name = removeDoubleConsonants(name);
 

diff --git a/src/test/java/org/apache/commons/codec/language/MatchRatingApproachEncoderTest.java b/src/test/java/org/apache/commons/codec/language/MatchRatingApproachEncoderTest.java
@@ -35,8 +35,6 @@
  */
 public class MatchRatingApproachEncoderTest extends AbstractStringEncoderTest<MatchRatingApproachEncoder> {
 
-    // ********** BEGIN REGION - TEST SUPPORT METHODS
-
     @Override
     protected MatchRatingApproachEncoder createStringEncoder() {
         return new MatchRatingApproachEncoder();
@@ -248,10 +246,6 @@ public final void testCompare_Surname_OSULLIVAN_OSUILLEABHAIN_SuccessfulMatch()
         assertTrue(this.getStringEncoder().isEncodeEquals("O'Sullivan", "Ó ' Súilleabháin"));
     }
 
-    // ***** END REGION - TEST SUPPORT METHODS
-
-    // ***** BEGIN REGION - TEST GET MRA ENCODING
-
     @Test
     public final void testCompare_Surname_PRZEMYSL_PSHEMESHIL_SuccessfullyMatched() {
         assertTrue(this.getStringEncoder().isEncodeEquals(" P rz e m y s l", " P sh e m e sh i l"));
@@ -297,10 +291,6 @@ public final void testCompare_ZACH_ZAKARIA_SuccessfullyMatched() {
         assertTrue(this.getStringEncoder().isEncodeEquals("Zach", "Zacharia"));
     }
 
-    // ***** END REGION - TEST GET MRA ENCODING
-
-    // ***** BEGIN REGION - TEST GET MRA COMPARISONS
-
     @Test
     public final void testCompareNameNullSpace_ReturnsFalseSuccessfully() {
         assertFalse(getStringEncoder().isEncodeEquals(null, " "));
@@ -433,8 +423,6 @@ public final void testIsEncodeEquals_CornerCase_FirstNameNothing_ReturnsFalse()
         assertFalse(this.getStringEncoder().isEncodeEquals("", "test"));
     }
 
-    // **** BEGIN YIDDISH/SLAVIC SECTION ****
-
     @Test
     public final void testIsEncodeEquals_CornerCase_FirstNameNull_ReturnsFalse() {
         assertFalse(this.getStringEncoder().isEncodeEquals(null, "test"));
@@ -470,8 +458,6 @@ public final void testIsVowel_SingleVowel_ReturnsTrue() {
         assertTrue(this.getStringEncoder().isVowel("I"));
     }
 
-    // **** END YIDDISH/SLAVIC SECTION ****
-
     @Test
     public final void testIsVowel_SmallD_ReturnsFalse() {
         assertFalse(this.getStringEncoder().isVowel("d"));
@@ -519,4 +505,18 @@ public final void testRemoveVowel_ALESSANDRA_Returns_ALSSNDR() {
 
     // ***** END REGION - TEST GET MRA COMPARISONS
 
+    @Test
+    public final void testPunctuationOnly() {
+        assertEquals(this.getStringEncoder().encode(".,-"), "");
+    }
+
+    @Test
+    public final void testVowelOnly() {
+        assertEquals(this.getStringEncoder().encode("aeiouAEIOU"), "A");
+    }
+
+    @Test
+    public final void testVowelAndPunctuationOnly() {
+        assertEquals(this.getStringEncoder().encode("uoiea.,-AEIOU"), "U");
+    }
 }