Normalize email local part to NFC

Author: ugexe

CHANGELOG.md

@@ -29,6 +29,8 @@ CHANGELOG
   used. For example, `gmali.com` will become `gmail.com`.
 * When `hashAddress` is used, all trailing periods are now removed from an
   email address domain. Previously only a single period was removed.
+* When `hashAddress` is used, the local part of an email address is now
+  normalized to NFC.
 
 3.3.0 (2023-12-05)
 ------------------

src/main/java/com/maxmind/minfraud/request/Email.java

@@ -7,6 +7,7 @@
 import java.nio.charset.StandardCharsets;
 import java.security.MessageDigest;
 import java.security.NoSuchAlgorithmException;
+import java.text.Normalizer;
 import java.util.Arrays;
 import java.util.Collections;
 import java.util.HashMap;
@@ -422,6 +423,7 @@ private String cleanAddress(String address) {
         String localPart = address.substring(0, domainIndex);
         String domain = address.substring(domainIndex + 1);
 
+        localPart = Normalizer.normalize(localPart, Normalizer.Form.NFC);
         domain = cleanDomain(domain);
 
         int stopChar;

src/test/java/com/maxmind/minfraud/request/EmailTest.java

@@ -228,6 +228,11 @@ public void testNormalizing() throws NoSuchAlgorithmException {
         assertEquals("2b21bc76dab3c8b1622837c1d698936c", e.getAddress(), "MD5");
         e = new Builder().address("example@b\u00FCcher.com").hashAddress().build();
         assertEquals("2b21bc76dab3c8b1622837c1d698936c", e.getAddress(), "MD5");
+
+        e = new Builder().address("bu\u0308cher@example.com").hashAddress().build();
+        assertEquals("53550c712b146287a2d0dd30e5ed6f4b", e.getAddress(), "MD5");
+        e = new Builder().address("b\u00FCcher@example.com").hashAddress().build();
+        assertEquals("53550c712b146287a2d0dd30e5ed6f4b", e.getAddress(), "MD5");
     }
 
     private String toMD5(String s) throws NoSuchAlgorithmException {