[3.12] GH-126766: `url2pathname()`: handle 'localhost' authority (GH-127129) (#127131)

GH-126766: `url2pathname()`: handle 'localhost' authority (GH-127129)

Discard any 'localhost' authority from the beginning of a `file:` URI. As a
result, file URIs like `//localhost/etc/hosts` are correctly decoded as
`/etc/hosts`.
(cherry picked from commit ebf564a1d3e2e81b9846535114e481d6096443d2)

Co-authored-by: Barney Gale <barney.gale@gmail.com>

diff --git a/Lib/nturl2path.py b/Lib/nturl2path.py
index 255eb2f547c2ce1b66c5d13d2d1e8bb239cc1c8d..6e0d26c129cb9995aab5c6e43c4ca4a369c02d34 100644 (file)
--- a/Lib/nturl2path.py
+++ b/Lib/nturl2path.py
@@ -15,14 +15,17 @@ def url2pathname(url):
     # become
     #   C:\foo\bar\spam.foo
     import string, urllib.parse
+    if url[:3] == '///':
+        # URL has an empty authority section, so the path begins on the third
+        # character.
+        url = url[2:]
+    elif url[:12] == '//localhost/':
+        # Skip past 'localhost' authority.
+        url = url[11:]
     # Windows itself uses ":" even in URLs.
     url = url.replace(':', '|')
     if not '|' in url:
         # No drive specifier, just convert slashes
-        if url[:3] == '///':
-            # URL has an empty authority section, so the path begins on the
-            # third character.
-            url = url[2:]
         # make sure not to convert quoted slashes :-)
         return urllib.parse.unquote(url.replace('/', '\\'))
     comp = url.split('|')

diff --git a/Lib/test/test_urllib.py b/Lib/test/test_urllib.py
index b06855e7b3a930f6fc13d34a4e14456d3bb61604..4520f8c85e6b03e9e4d54615f91ad054fabe8f61 100644 (file)
--- a/Lib/test/test_urllib.py
+++ b/Lib/test/test_urllib.py
@@ -1604,6 +1604,8 @@ class Pathname_Tests(unittest.TestCase):
         # Localhost paths
         self.assertEqual(fn('//localhost/C:/path/to/file'), 'C:\\path\\to\\file')
         self.assertEqual(fn('//localhost/C|/path/to/file'), 'C:\\path\\to\\file')
+        self.assertEqual(fn('//localhost/path/to/file'), '\\path\\to\\file')
+        self.assertEqual(fn('//localhost//server/path/to/file'), '\\\\server\\path\\to\\file')
         # Percent-encoded forward slashes are preserved for backwards compatibility
         self.assertEqual(fn('C:/foo%2fbar'), 'C:\\foo/bar')
         self.assertEqual(fn('//server/share/foo%2fbar'), '\\\\server\\share\\foo/bar')
@@ -1622,7 +1624,7 @@ class Pathname_Tests(unittest.TestCase):
         self.assertEqual(fn('//foo/bar'), '//foo/bar')
         self.assertEqual(fn('///foo/bar'), '/foo/bar')
         self.assertEqual(fn('////foo/bar'), '//foo/bar')
-        self.assertEqual(fn('//localhost/foo/bar'), '//localhost/foo/bar')
+        self.assertEqual(fn('//localhost/foo/bar'), '/foo/bar')
 
     @unittest.skipUnless(os_helper.FS_NONASCII, 'need os_helper.FS_NONASCII')
     def test_url2pathname_nonascii(self):

diff --git a/Lib/urllib/request.py b/Lib/urllib/request.py
index c89e217b9de1b7b0f8262cec54a41c1361ae5f8f..c521d96d9538833801412a11dfced71684e5547e 100644 (file)
--- a/Lib/urllib/request.py
+++ b/Lib/urllib/request.py
@@ -1685,6 +1685,9 @@ else:
             # URL has an empty authority section, so the path begins on the
             # third character.
             pathname = pathname[2:]
+        elif pathname[:12] == '//localhost/':
+            # Skip past 'localhost' authority.
+            pathname = pathname[11:]
         encoding = sys.getfilesystemencoding()
         errors = sys.getfilesystemencodeerrors()
         return unquote(pathname, encoding=encoding, errors=errors)

diff --git a/Misc/NEWS.d/next/Library/2024-11-22-02-31-55.gh-issue-126766.jfkhBH.rst b/Misc/NEWS.d/next/Library/2024-11-22-02-31-55.gh-issue-126766.jfkhBH.rst
new file mode 100644 (file)
index 0000000..998c99b
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2024-11-22-02-31-55.gh-issue-126766.jfkhBH.rst
@@ -0,0 +1,2 @@
+Fix issue where :func:`urllib.request.url2pathname` failed to discard any
+'localhost' authority present in the URL.