Fix domain validation

jmeridth · jmeridth · commit 12d4df9c3fe2 · 2019-12-03T20:40:45.000-06:00
- Updated regex to not allow numeric only TLDs (examples in tests) - Allow for idna encoded domains and test for them (examples in tests) Fixes #47 Fixes #123
diff --git a/tests/test_domain.py b/tests/test_domain.py
@@ -13,7 +13,9 @@
     '3.cn',
     'a.cn',
     'sub1.sub2.sample.co.uk',
-    'somerandomexample.xn--fiqs8s'
+    'somerandomexample.xn--fiqs8s',
+    'kräuter.com',
+    'über.com'
 ])
 def test_returns_true_on_valid_domain(value):
     assert domain(value)
@@ -29,7 +31,11 @@ def test_returns_true_on_valid_domain(value):
     '_example.com',
     'example_.com',
     'example',
-    'a......b.com'
+    'a......b.com',
+    'a.123',
+    '123.123',
+    '123.123.123',
+    '123.123.123.123'
 ])
 def test_returns_failed_validation_on_invalid_domain(value):
     assert isinstance(domain(value), ValidationFailure)
diff --git a/validators/domain.py b/validators/domain.py
@@ -6,7 +6,7 @@
     r'^(?:[a-z0-9]'  # First character of the domain
     r'(?:[a-z0-9-_]{0,61}[a-z0-9])?\.)'  # Sub domain + hostname
     r'+[a-z0-9][a-z0-9-_]{0,61}'  # First 61 characters of the gTLD
-    r'[a-z0-9]$'  # Last character of the gTLD
+    r'[a-z]$'  # Last character of the gTLD
 )
 
 
@@ -40,4 +40,7 @@ def domain(value):
 
     :param value: domain string to validate
     """
-    return pattern.match(value)
+    try:
+        return pattern.match(value.encode('idna').decode())
+    except UnicodeError as e:
+        return False
