Be stricter about parsing hostname labels

Labels must start with an alphanumeric character, may contain
alphanumeric characters or hyphens, but must not end with a hyphen.
We enforce those rules, but allow underscores wherever hyphens are
accepted, and allow alphanumeric ranges anywhere.

We relax the definition of "alphanumeric" to include Unicode characters
even though such inventory hostnames cannot be used in practice unless
an ansible_ssh_host is set for each of them.

We still don't enforce length restrictions—the fact that we have to
accept ranges makes it more complex, and it doesn't seem especially
worthwhile.

lib/ansible/parsing/utils/addresses.py

@@ -61,6 +61,17 @@ ipv4_component = r'''
     )
 '''.format(range=numeric_range)
 
+# A hostname label, e.g. 'foo' in 'foo.example.com'. Consists of alphanumeric
+# characters plus dashes (and underscores) or valid ranges. The label may not
+# start or end with a hyphen or an underscore. This is interpolated into the
+# hostname pattern below. We don't try to enforce the 63-char length limit.
+
+label = r'''
+    (?:[\w]|{range})                    # Starts with an alphanumeric or a range
+    (?:[\w_-]|{range})*                 # Then zero or more of the same or [_-]
+    (?<![_-])                           # ...as long as it didn't end with [_-]
+'''.format(range=alphanumeric_range)
+
 patterns = {
     # This matches a square-bracketed expression with a port specification. What
     # is inside the square brackets is validated later.
@@ -138,14 +149,11 @@ patterns = {
     # 253 characters total) or make any attempt to process IDNs.
 
     'hostname': re.compile(
-        r'''^                           # We need at least one label,
-            (?:                         # which comprises:
-                [a-z0-9_-]|             # (a valid domain label character
-                {0}                     # or a bracketed alphanumeric range)
-            )+
-            (?:\.(?:[a-z0-9_-]|{0})+)*  # Followed by zero or more .labels
+        r'''^
+            {label}                     # We must have at least one label
+            (?:\.{label})*              # Followed by zero or more .labels
             $
-        '''.format(alphanumeric_range), re.X|re.I
+        '''.format(label=label), re.X|re.I|re.UNICODE
     ),
 }
 

test/units/parsing/test_addresses.py

@@ -1,3 +1,5 @@
+# -*- coding: utf-8 -*-
+
 import unittest
 
 from ansible.parsing.utils.addresses import parse_address
@@ -21,12 +23,19 @@ class TestParseAddress(unittest.TestCase):
         'some-host:80': ['some-host', 80],
         'some.host.com:492': ['some.host.com', 492],
         '[some.host.com]:493': ['some.host.com', 493],
+        'a-b.3foo_bar.com:23': ['a-b.3foo_bar.com', 23],
+        u'fóöbär': [u'fóöbär', None],
+        u'fóöbär:32': [u'fóöbär', 32],
+        u'fóöbär.éxàmplê.com:632': [u'fóöbär.éxàmplê.com', 632],
 
         # Various errors
         '': [None, None],
         'some..host': [None, None],
         'some.': [None, None],
         '[example.com]': [None, None],
+        'some-': [None, None],
+        'some-.foo.com': [None, None],
+        'some.-foo.com': [None, None],
     }
 
     range_tests = {
@@ -34,7 +43,11 @@ class TestParseAddress(unittest.TestCase):
         '192.0.2.[3:10]:23': ['192.0.2.[3:10]', 23],
         'abcd:ef98::7654:[1:9]': ['abcd:ef98::7654:[1:9]', None],
         '[abcd:ef98::7654:[6:32]]:2222': ['abcd:ef98::7654:[6:32]', 2222],
+        u'fóöb[a:c]r.éxàmplê.com:632': [u'fóöb[a:c]r.éxàmplê.com', 632],
+        '[a:b]foo.com': ['[a:b]foo.com', None],
+        'foo[a:b].com': ['foo[a:b].com', None],
         'foo[a:b]:42': ['foo[a:b]', 42],
+        'foo[a-b]-.com': [None, None],
         'foo[a-b]:32': [None, None],
         'foo[x-y]': [None, None],
     }