diff --git a/Lib/nturl2path.py b/Lib/nturl2path.py index 757fd01be..57c7858df 100644 --- a/Lib/nturl2path.py +++ b/Lib/nturl2path.py @@ -3,7 +3,15 @@ This module only exists to provide OS-specific code for urllib.requests, thus do not use directly. """ -# Testing is done through test_urllib. +# Testing is done through test_nturl2path. + +import warnings + + +warnings._deprecated( + __name__, + message=f"{warnings._DEPRECATED_MSG}; use 'urllib.request' instead", + remove=(3, 19)) def url2pathname(url): """OS-specific conversion from a relative URL of the 'file' scheme @@ -14,7 +22,7 @@ def url2pathname(url): # ///C:/foo/bar/spam.foo # become # C:\foo\bar\spam.foo - import string, urllib.parse + import urllib.parse if url[:3] == '///': # URL has an empty authority section, so the path begins on the third # character. @@ -25,19 +33,14 @@ def url2pathname(url): if url[:3] == '///': # Skip past extra slash before UNC drive in URL path. url = url[1:] - # Windows itself uses ":" even in URLs. - url = url.replace(':', '|') - if not '|' in url: - # No drive specifier, just convert slashes - # make sure not to convert quoted slashes :-) - return urllib.parse.unquote(url.replace('/', '\\')) - comp = url.split('|') - if len(comp) != 2 or comp[0][-1] not in string.ascii_letters: - error = 'Bad URL: ' + url - raise OSError(error) - drive = comp[0][-1].upper() - tail = urllib.parse.unquote(comp[1].replace('/', '\\')) - return drive + ':' + tail + else: + if url[:1] == '/' and url[2:3] in (':', '|'): + # Skip past extra slash before DOS drive in URL path. + url = url[1:] + if url[1:2] == '|': + # Older URLs use a pipe after a drive letter + url = url[:1] + ':' + url[2:] + return urllib.parse.unquote(url.replace('/', '\\')) def pathname2url(p): """OS-specific conversion from a file system path to a relative URL @@ -46,6 +49,7 @@ def pathname2url(p): # C:\foo\bar\spam.foo # becomes # ///C:/foo/bar/spam.foo + import ntpath import urllib.parse # First, clean up some special forms. We are going to sacrifice # the additional information anyway @@ -54,16 +58,17 @@ def pathname2url(p): 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 DOS drive specified, just quote the pathname - return urllib.parse.quote(p) - comp = p.split(':', maxsplit=2) - if len(comp) != 2 or len(comp[0]) > 1: - error = 'Bad path: ' + p - raise OSError(error) + drive, root, tail = ntpath.splitroot(p) + if drive: + if drive[1:] == ':': + # DOS drive specified. Add three slashes to the start, producing + # an authority section with a zero-length authority, and a path + # section starting with a single slash. + drive = f'///{drive}' + drive = urllib.parse.quote(drive, safe='/:') + elif root: + # Add explicitly empty authority to path beginning with one slash. + root = f'//{root}' - drive = urllib.parse.quote(comp[0].upper()) - tail = urllib.parse.quote(comp[1]) - return '///' + drive + ':' + tail + tail = urllib.parse.quote(tail) + return drive + root + tail diff --git a/Lib/test/test_nturl2path.py b/Lib/test/test_nturl2path.py new file mode 100644 index 000000000..a6a3422a0 --- /dev/null +++ b/Lib/test/test_nturl2path.py @@ -0,0 +1,107 @@ +import unittest + +from test.support import warnings_helper + + +nturl2path = warnings_helper.import_deprecated("nturl2path") + + +class NTURL2PathTest(unittest.TestCase): + """Test pathname2url() and url2pathname()""" + + def test_basic(self): + # Make sure simple tests pass + expected_path = r"parts\of\a\path" + expected_url = "parts/of/a/path" + result = nturl2path.pathname2url(expected_path) + self.assertEqual(expected_url, result, + "pathname2url() failed; %s != %s" % + (result, expected_url)) + result = nturl2path.url2pathname(expected_url) + self.assertEqual(expected_path, result, + "url2pathame() failed; %s != %s" % + (result, expected_path)) + + def test_pathname2url(self): + # Test special prefixes are correctly handled in pathname2url() + fn = nturl2path.pathname2url + self.assertEqual(fn('\\\\?\\C:\\dir'), '///C:/dir') + self.assertEqual(fn('\\\\?\\unc\\server\\share\\dir'), '//server/share/dir') + self.assertEqual(fn("C:"), '///C:') + self.assertEqual(fn("C:\\"), '///C:/') + self.assertEqual(fn('c:\\a\\b.c'), '///c:/a/b.c') + self.assertEqual(fn('C:\\a\\b.c'), '///C:/a/b.c') + self.assertEqual(fn('C:\\a\\b.c\\'), '///C:/a/b.c/') + self.assertEqual(fn('C:\\a\\\\b.c'), '///C:/a//b.c') + self.assertEqual(fn('C:\\a\\b%#c'), '///C:/a/b%25%23c') + self.assertEqual(fn('C:\\a\\b\xe9'), '///C:/a/b%C3%A9') + self.assertEqual(fn('C:\\foo\\bar\\spam.foo'), "///C:/foo/bar/spam.foo") + # NTFS alternate data streams + self.assertEqual(fn('C:\\foo:bar'), '///C:/foo%3Abar') + self.assertEqual(fn('foo:bar'), 'foo%3Abar') + # No drive letter + self.assertEqual(fn("\\folder\\test\\"), '///folder/test/') + self.assertEqual(fn("\\\\folder\\test\\"), '//folder/test/') + self.assertEqual(fn("\\\\\\folder\\test\\"), '///folder/test/') + self.assertEqual(fn('\\\\some\\share\\'), '//some/share/') + self.assertEqual(fn('\\\\some\\share\\a\\b.c'), '//some/share/a/b.c') + self.assertEqual(fn('\\\\some\\share\\a\\b%#c\xe9'), '//some/share/a/b%25%23c%C3%A9') + # Alternate path separator + self.assertEqual(fn('C:/a/b.c'), '///C:/a/b.c') + self.assertEqual(fn('//some/share/a/b.c'), '//some/share/a/b.c') + self.assertEqual(fn('//?/C:/dir'), '///C:/dir') + self.assertEqual(fn('//?/unc/server/share/dir'), '//server/share/dir') + # Round-tripping + urls = ['///C:', + '///folder/test/', + '///C:/foo/bar/spam.foo'] + for url in urls: + self.assertEqual(fn(nturl2path.url2pathname(url)), url) + + def test_url2pathname(self): + fn = nturl2path.url2pathname + self.assertEqual(fn('/'), '\\') + self.assertEqual(fn('/C:/'), 'C:\\') + self.assertEqual(fn("///C|"), 'C:') + self.assertEqual(fn("///C:"), 'C:') + self.assertEqual(fn('///C:/'), 'C:\\') + self.assertEqual(fn('/C|//'), 'C:\\\\') + self.assertEqual(fn('///C|/path'), 'C:\\path') + # No DOS drive + self.assertEqual(fn("///C/test/"), '\\C\\test\\') + self.assertEqual(fn("////C/test/"), '\\\\C\\test\\') + # DOS drive paths + self.assertEqual(fn('c:/path/to/file'), 'c:\\path\\to\\file') + self.assertEqual(fn('C:/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('C:/path/to/file/'), 'C:\\path\\to\\file\\') + self.assertEqual(fn('C:/path/to//file'), 'C:\\path\\to\\\\file') + self.assertEqual(fn('C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('/C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('///C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn("///C|/foo/bar/spam.foo"), 'C:\\foo\\bar\\spam.foo') + # Colons in URI + self.assertEqual(fn('///\u00e8|/'), '\u00e8:\\') + self.assertEqual(fn('//host/share/spam.txt:eggs'), '\\\\host\\share\\spam.txt:eggs') + self.assertEqual(fn('///c:/spam.txt:eggs'), 'c:\\spam.txt:eggs') + # UNC paths + self.assertEqual(fn('//server/path/to/file'), '\\\\server\\path\\to\\file') + self.assertEqual(fn('////server/path/to/file'), '\\\\server\\path\\to\\file') + self.assertEqual(fn('/////server/path/to/file'), '\\\\server\\path\\to\\file') + # 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') + # Round-tripping + paths = ['C:', + r'\C\test\\', + r'C:\foo\bar\spam.foo'] + for path in paths: + self.assertEqual(fn(nturl2path.pathname2url(path)), path) + + +if __name__ == '__main__': + unittest.main()