[Python-checkins] bpo-43607: Fix urllib handling of Windows paths with \\?\ prefix (GH-25539)

[zooba]

https://github.com/python/cpython/commit/3513d55a617012002c3f82dbf3cec7ec1abd7090
commit: 3513d55a617012002c3f82dbf3cec7ec1abd7090
branch: master
author: Steve Dower <steve.dower at python.org>
committer: zooba <steve.dower at microsoft.com>
date: 2021-04-23T18:02:47+01:00
summary:

bpo-43607: Fix urllib handling of Windows paths with \\?\ prefix (GH-25539)

files:
A Misc/NEWS.d/next/Library/2021-04-22-22-39-58.bpo-43607.7IYDkG.rst
M Lib/nturl2path.py
M Lib/test/test_urllib.py

diff --git a/Lib/nturl2path.py b/Lib/nturl2path.py
index 853e6608380e9..61852aff58912 100644
--- a/Lib/nturl2path.py
+++ b/Lib/nturl2path.py
@@ -50,6 +50,14 @@ def pathname2url(p):
     # becomes
     #   ///C:/foo/bar/spam.foo
     import urllib.parse
+    # First, clean up some special forms. We are going to sacrifice
+    # the additional information anyway
+    if p[:4] == '\\\\?\\':
+        p = p[4:]
+        if p[:4].upper() == 'UNC\\':
+            p = '\\' + p[4:]
+        elif p[1:2] != ':':
+            raise OSError('Bad path: ' + p)
     if not ':' in p:
         # No drive specifier, just convert slashes and quote the name
         if p[:2] == '\\\\':
@@ -59,7 +67,7 @@ def pathname2url(p):
             p = '\\\\' + p
         components = p.split('\\')
         return urllib.parse.quote('/'.join(components))
-    comp = p.split(':')
+    comp = p.split(':', maxsplit=2)
     if len(comp) != 2 or len(comp[0]) > 1:
         error = 'Bad path: ' + p
         raise OSError(error)
diff --git a/Lib/test/test_urllib.py b/Lib/test/test_urllib.py
index f41fa2a950686..82f1d9dc2e7bb 100644
--- a/Lib/test/test_urllib.py
+++ b/Lib/test/test_urllib.py
@@ -1526,6 +1526,24 @@ def test_quoting(self):
                          "url2pathname() failed; %s != %s" %
                          (expect, result))
 
+    @unittest.skipUnless(sys.platform == 'win32',
+                         'test specific to the nturl2path functions.')
+    def test_prefixes(self):
+        # Test special prefixes are correctly handled in pathname2url()
+        given = '\\\\?\\C:\\dir'
+        expect = '///C:/dir'
+        result = urllib.request.pathname2url(given)
+        self.assertEqual(expect, result,
+                         "pathname2url() failed; %s != %s" %
+                         (expect, result))
+        given = '\\\\?\\unc\\server\\share\\dir'
+        expect = '/server/share/dir'
+        result = urllib.request.pathname2url(given)
+        self.assertEqual(expect, result,
+                         "pathname2url() failed; %s != %s" %
+                         (expect, result))
+
+
     @unittest.skipUnless(sys.platform == 'win32',
                          'test specific to the urllib.url2path function.')
     def test_ntpath(self):
diff --git a/Misc/NEWS.d/next/Library/2021-04-22-22-39-58.bpo-43607.7IYDkG.rst b/Misc/NEWS.d/next/Library/2021-04-22-22-39-58.bpo-43607.7IYDkG.rst
new file mode 100644
index 0000000000000..fa62846acaa98
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2021-04-22-22-39-58.bpo-43607.7IYDkG.rst
@@ -0,0 +1,2 @@
+:mod:`urllib` can now convert Windows paths with ``\\?\`` prefixes into URL
+paths.