From: Bob Halley <halley@dnspython.org>
Date: Mon, 26 Sep 2016 19:38:20 +0000 (-0700)
Subject: Add IDNA 2008 Practical mode, since IDNA 2008 is absurdly strict.
X-Git-Tag: v1.15.0~6
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=bb0c9f21f4a6f56f2fe8d7c1fc991080ef89d223;p=thirdparty%2Fdnspython.git

Add IDNA 2008 Practical mode, since IDNA 2008 is absurdly strict.
---

diff --git a/dns/name.py b/dns/name.py
index e1259b95..758a394c 100644
--- a/dns/name.py
+++ b/dns/name.py
@@ -140,25 +140,41 @@ class IDNA2008Codec(IDNACodec):
 
     """IDNA 2008 encoder/decoder."""
 
-    def __init__(self, uts_46=True, transitional=False):
+    def __init__(self, uts_46=False, transitional=False,
+                 allow_pure_ascii=False):
         """Initialize the IDNA 2008 encoder/decoder.
         @param uts_46: If True, apply Unicode IDNA compatibility processing
         as described in Unicode Technical Standard #46
         (U{http://unicode.org/reports/tr46/}).  This parameter is only
         meaningful if IDNA 2008 is in use.  If False, do not apply
-        the mapping.  The default is True.
+        the mapping.  The default is False
         @type uts_46: bool
         @param transitional: If True, use the "transitional" mode described
         in Unicode Technical Standard #46.  This parameter is only
         meaningful if IDNA 2008 is in use.  The default is False.
         @type transitional: bool
+        @param allow_pure_ascii: If True, then a label which
+        consists of only ASCII characters is allowed.  This is less strict
+        than regular IDNA 2008, but is also necessary for mixed names,
+        e.g. a name with starting with "_sip._tcp." and ending in an IDN
+        suffixm which would otherwise be disallowed.  The default is False
+        @type allow_pure_ascii: bool
         """
         self.uts_46 = uts_46
         self.transitional = transitional
+        self.allow_pure_ascii = allow_pure_ascii
+
+    def is_all_ascii(self, label):
+        for c in label:
+            if ord(c) > 0x7f:
+                return False
+        return True
 
     def encode(self, label):
         if label == '':
             return b''
+        if self.allow_pure_ascii and self.is_all_ascii(label):
+            return label.encode('ascii')
         if not have_idna_2008:
             raise NoIDNA2008
         try:
@@ -184,9 +200,11 @@ class IDNA2008Codec(IDNACodec):
 _escaped = bytearray(b'"().;\\@$')
 
 IDNA_2003 = IDNA2003Codec()
-IDNA_2008 = IDNA2008Codec()
-IDNA_2008_Strict = IDNA2008Codec(False)
-IDNA_2008_Transitional = IDNA2008Codec(True, True)
+IDNA_2008_Practical = IDNA2008Codec(True, False, True)
+IDNA_2008_UTS_46 = IDNA2008Codec(True, False, False)
+IDNA_2008_Strict = IDNA2008Codec(False, False, False)
+IDNA_2008_Transitional = IDNA2008Codec(True, True, False)
+IDNA_2008 = IDNA_2008_Practical
 
 def _escapify(label, unicode_mode=False):
     """Escape the characters in label which need it.
diff --git a/tests/test_name.py b/tests/test_name.py
index 11504c71..b9ddd816 100644
--- a/tests/test_name.py
+++ b/tests/test_name.py
@@ -675,6 +675,30 @@ class NameTestCase(unittest.TestCase):
             e2 = dns.name.from_unicode(t, idna_codec=c2)
             self.assertEqual(str(e2), 'xn--knigsgsschen-lcb0w.')
 
+    def testFromUnicodeIDNA2008Mixed(self):
+        # the IDN rules for names are very restrictive, disallowing
+        # practical names like u'_sip._tcp.Königsgäßchen'.  Dnspython
+        # has a "practical" mode which permits labels which are purely
+        # ASCII to go straight through, and thus not invalid useful
+        # things in the real world.
+        if dns.name.have_idna_2008:
+            t = u'_sip._tcp.Königsgäßchen'
+            def bad1():
+                codec = dns.name.IDNA_2008_Strict
+                return dns.name.from_unicode(t, idna_codec=codec)
+            def bad2():
+                codec = dns.name.IDNA_2008_UTS_46
+                return dns.name.from_unicode(t, idna_codec=codec)
+            def bad3():
+                codec = dns.name.IDNA_2008_Transitional
+                return dns.name.from_unicode(t, idna_codec=codec)
+            self.failUnlessRaises(dns.name.IDNAException, bad1)
+            self.failUnlessRaises(dns.name.IDNAException, bad2)
+            self.failUnlessRaises(dns.name.IDNAException, bad3)
+            e = dns.name.from_unicode(t,
+                                      idna_codec=dns.name.IDNA_2008_Practical)
+            self.assertEqual(str(e), '_sip._tcp.xn--knigsgchen-b4a3dun.')
+
     def testToUnicode1(self):
         n = dns.name.from_text(u'foo.bar')
         s = n.to_unicode()