Merge pull request #26 from nexB/preserve_spaces_in_filenames

Add preserve_spaces argument to portable_filename

Author: pombredanne

CHANGELOG.rst

@@ -4,6 +4,9 @@ Release notes
 vNext
 -----
 
+- Add preserve_spaces argument in commoncode.paths.portable_filename.
+  This argument will prevent the replacement of spaces in filenames.
+
 
 Version 21.6.11
 ---------------
@@ -32,7 +35,7 @@ Version 21.5.12
 
 - Add new function to find a command or shared object file in the PATH (e.g. in
   environment variables). See commoncode.command.find_in_path()
-- Add new simplified the commoncode.command.execute() function. 
+- Add new simplified the commoncode.command.execute() function.
 - Add support for Python 3.10
 - Update tests to cope with Python 3.6 bug https://bugs.python.org/issue26919
 - Adopt latest skeleton with configure scripts updates

src/commoncode/paths.py

@@ -26,14 +26,16 @@
 # Build OS-portable and safer paths
 
 
-def safe_path(path, posix=False):
+def safe_path(path, posix=False, preserve_spaces=False):
     """
     Convert `path` to a safe and portable POSIX path usable on multiple OSes.
     The returned path is an ASCII-only byte string, resolved for relative
     segments and itself relative.
 
     The `path` is treated as a POSIX path if `posix` is True or as a Windows
     path with blackslash separators otherwise.
+
+    If `preserve_spaces` is True, then the spaces in `path` will not be replaced.
     """
     # if the path is UTF, try to use unicode instead
     if not isinstance(path, str):
@@ -50,7 +52,7 @@ def safe_path(path, posix=False):
     _pathmod, path_sep = path_handlers(path, posix)
 
     segments = [s.strip() for s in path.split(path_sep) if s.strip()]
-    segments = [portable_filename(s) for s in segments]
+    segments = [portable_filename(s, preserve_spaces=preserve_spaces) for s in segments]
 
     if not segments:
         return '_'
@@ -133,12 +135,16 @@ def resolve(path, posix=True):
 
 
 legal_punctuation = r"!\#$%&\(\)\+,\-\.;\=@\[\]_\{\}\~"
+legal_spaces = r" "
 legal_chars = r'A-Za-z0-9' + legal_punctuation
+legal_chars_inc_spaces = legal_chars + legal_spaces
 illegal_chars_re = r'[^' + legal_chars + r']'
+illegal_chars_exc_spaces_re = r'[^' + legal_chars_inc_spaces + r']'
 replace_illegal_chars = re.compile(illegal_chars_re).sub
+replace_illegal_chars_exc_spaces = re.compile(illegal_chars_exc_spaces_re).sub
 
 
-def portable_filename(filename):
+def portable_filename(filename, preserve_spaces=False):
     """
     Return a new name for `filename` that is portable across operating systems.
 
@@ -156,13 +162,18 @@ def portable_filename(filename):
 
     Also inspired by Werkzeug:
     https://raw.githubusercontent.com/pallets/werkzeug/8c2d63ce247ba1345e1b9332a68ceff93b2c07ab/werkzeug/utils.py
+
+    If `preserve_spaces` is True, then spaces in `filename` will not be replaced.
     """
     filename = toascii(filename, translit=True)
 
     if not filename:
         return '_'
 
-    filename = replace_illegal_chars('_', filename)
+    if preserve_spaces:
+        filename = replace_illegal_chars_exc_spaces('_', filename)
+    else:
+        filename = replace_illegal_chars('_', filename)
 
     # these are illegal both upper and lowercase and with or without an extension
     # we insert an underscore after the base name.

tests/test_paths.py

@@ -127,6 +127,10 @@ def test_portable_filename(self):
         expected = 'A___file__with_Spaces.mov'
         assert paths.portable_filename("A:\\ file/ with Spaces.mov") == expected
 
+        # Test `preserve_spaces` option. Spaces should not be replaced
+        expected = 'Program Files (x86)'
+        assert paths.portable_filename("Program Files (x86)", preserve_spaces=True) == expected
+
         # Unresolved relative paths will be treated as a single filename. Use
         # resolve instead if you want to resolve paths:
         expected = '___.._.._etc_passwd'