From: Raymond Hettinger <python@rcn.com>
Date: Sat, 15 Oct 2005 16:41:53 +0000 (+0000)
Subject: Teach unquote() to handle unicode inputs
X-Git-Tag: v2.5a0~1288
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=4b0f20def3528469f3579445783d2d5db3f50c48;p=python

Teach unquote() to handle unicode inputs
---

diff --git a/Lib/test/test_urllib.py b/Lib/test/test_urllib.py
index 94f0f9efb9..4579c479d1 100644
--- a/Lib/test/test_urllib.py
+++ b/Lib/test/test_urllib.py
@@ -415,6 +415,10 @@ class UnquotingTests(unittest.TestCase):
         self.assertEqual(expect, result,
                          "using unquote_plus(): %s != %s" % (expect, result))
 
+    def test_unquote_with_unicode(self):
+        r = urllib.unquote(u'br%C3%BCckner_sapporo_20050930.doc')
+        self.assertEqual(r, u'br\xc3\xbcckner_sapporo_20050930.doc')
+
 class urlencode_Tests(unittest.TestCase):
     """Tests for urlencode()"""
 
diff --git a/Lib/urllib.py b/Lib/urllib.py
index bc16be0162..f00d02ff9e 100644
--- a/Lib/urllib.py
+++ b/Lib/urllib.py
@@ -1061,6 +1061,8 @@ def unquote(s):
             res[i] = _hextochr[item[:2]] + item[2:]
         except KeyError:
             res[i] = '%' + item
+        except UnicodeDecodeError:
+            res[i] = unichr(int(item[:2], 16)) + item[2:]
     return "".join(res)
 
 def unquote_plus(s):