Merge pull request #139 from maxmind/ugexe/email-nfc

Normalize email addresses to NFC

HISTORY.rst

@@ -33,6 +33,8 @@ History
   become ``gmail.com``.
 * Additional ``gmail.com`` typos are now normalized when ``hash_email`` is
   used. For example, ``gmali.com`` will become ``gmail.com``.
+* When ``hash_email`` is used, the local part of an email address is now
+  normalized to NFC.
 
 2.9.0 (2023-12-05)
 ++++++++++++++++++

minfraud/request.py

@@ -8,6 +8,7 @@
 import re
 import warnings
 import hashlib
+import unicodedata
 from typing import Any, Dict
 from voluptuous import MultipleInvalid
 
@@ -364,6 +365,8 @@ def _clean_email(address):
     domain = _clean_domain(address[at_idx + 1 :])  # noqa
     local_part = address[:at_idx]
 
+    local_part = unicodedata.normalize("NFC", local_part)
+
     # Strip off aliased part of email address.
     if domain in _YAHOO_DOMAINS:
         divider = "-"

tests/test_request.py

@@ -141,6 +141,26 @@ def test_maybe_hash_email(self):
                     }
                 },
             },
+            {
+                "name": "email local part nfc normalization form 1",
+                "input": {"email": {"address": "bu\u0308[email protected]"}},
+                "expected": {
+                    "email": {
+                        "address": "53550c712b146287a2d0dd30e5ed6f4b",
+                        "domain": "example.com",
+                    }
+                },
+            },
+            {
+                "name": "email local part nfc normalization form 2",
+                "input": {"email": {"address": "b\u00FC[email protected]"}},
+                "expected": {
+                    "email": {
+                        "address": "53550c712b146287a2d0dd30e5ed6f4b",
+                        "domain": "example.com",
+                    }
+                },
+            },
         ]
 
         for test in tests:
@@ -231,6 +251,8 @@ def test_clean_email():
         {"input": "[email protected]", "output": "[email protected]"},
         {"input": "[email protected].", "output": "[email protected]"},
         {"input": "[email protected]...", "output": "[email protected]"},
+        {"input": "example@bu\u0308cher.com", "output": "[email protected]"},
+        {"input": "example@b\u00FCcher.com", "output": "[email protected]"},
     ]
 
     for test in tests: