Merge pull request #166 from SethMMorton/fix-windows-os-sorted

Disable extension splitting on Windows for os_sorted

CHANGELOG.md

@@ -6,6 +6,13 @@ Unreleased
 - The changelog now only explictly exists in the repository
   (issue [#162](https://github.com/SethMMorton/natsort/issues/162))
 
+### Fixed
+
+- Inputs with spaces near adjascent to the extension are now sorted
+  correctly for `os_sorted` on Windows (issues
+  [#165](https://github.com/SethMMorton/natsort/issues/165) and
+  [#166](https://github.com/SethMMorton/natsort/pull/166))
+
 [8.3.1] - 2023-03-01
 ---
 

natsort/natsort.py

@@ -670,13 +670,13 @@ def numeric_regex_chooser(alg: NSType) -> str:
 
 
 def _split_apply(
-    v: Any, key: Optional[Callable[[T], NatsortInType]] = None
+    v: Any, key: Optional[Callable[[T], NatsortInType]] = None, treat_base: bool = True
 ) -> Iterator[str]:
     if key is not None:
         v = key(v)
     if not isinstance(v, (str, PurePath)):
         v = str(v)
-    return utils.path_splitter(v)
+    return utils.path_splitter(v, treat_base=treat_base)
 
 
 # Choose the implementation based on the host OS
@@ -694,7 +694,7 @@ def os_sort_keygen(
     ) -> Callable[[Any], NatsortOutType]:
         return cast(
             Callable[[Any], NatsortOutType],
-            lambda x: tuple(map(_winsort_key, _split_apply(x, key))),
+            lambda x: tuple(map(_winsort_key, _split_apply(x, key, treat_base=False))),
         )
 
 else:

natsort/utils.py

@@ -876,7 +876,7 @@ def do_decoding(s: Any, encoding: str) -> Any:
 
 # noinspection PyIncorrectDocstring
 def path_splitter(
-    s: PathArg, _d_match: MatchFn = re.compile(r"\.\d").match
+    s: PathArg, treat_base: bool = True, _d_match: MatchFn = re.compile(r"\.\d").match
 ) -> Iterator[str]:
     """
     Split a string into its path components.
@@ -886,6 +886,10 @@ def path_splitter(
     Parameters
     ----------
     s : str | pathlib.Path
+    treat_base: bool, optional
+        If True, treat the base of component of the file path as
+        special and split off extensions. If False, do not do this.
+        The default is True.
 
     Returns
     -------
@@ -909,17 +913,18 @@ def path_splitter(
         path_parts = []
         base = str(s)
 
-    # Now, split off the file extensions until
-    #  - we reach a decimal number at the beginning of the suffix
-    #  - more than two suffixes have been seen
-    #  - a suffix is more than five characters (including leading ".")
-    #  - there are no more extensions
     suffixes = []
-    for i, suffix in enumerate(reversed(PurePath(base).suffixes)):
-        if _d_match(suffix) or i > 1 or len(suffix) > 5:
-            break
-        suffixes.append(suffix)
-    suffixes.reverse()
+    if treat_base:
+        # Now, split off the file extensions until
+        #  - we reach a decimal number at the beginning of the suffix
+        #  - more than two suffixes have been seen
+        #  - a suffix is more than five characters (including leading ".")
+        #  - there are no more extensions
+        for i, suffix in enumerate(reversed(PurePath(base).suffixes)):
+            if _d_match(suffix) or i > 1 or len(suffix) > 5:
+                break
+            suffixes.append(suffix)
+        suffixes.reverse()
 
     # Remove the suffixes from the base component
     base = base.replace("".join(suffixes), "")

tests/test_os_sorted.py

@@ -24,14 +24,24 @@ def test_os_sorted_compound() -> None:
         "/p/Folder (1)/file.tar.gz",
         "/p/Folder/file.x1.10.tar.gz",
     ]
-    expected = [
-        "/p/Folder/file.x1.9.tar.gz",
-        "/p/Folder/file.x1.10.tar.gz",
-        "/p/Folder (1)/file.tar.gz",
-        "/p/Folder (1)/file (1).tar.gz",
-        "/p/Folder (2)/file.tar.gz",
-        "/p/Folder (10)/file.tar.gz",
-    ]
+    if platform.system() == "Windows":
+        expected = [
+            "/p/Folder/file.x1.9.tar.gz",
+            "/p/Folder/file.x1.10.tar.gz",
+            "/p/Folder (1)/file (1).tar.gz",
+            "/p/Folder (1)/file.tar.gz",
+            "/p/Folder (2)/file.tar.gz",
+            "/p/Folder (10)/file.tar.gz",
+        ]
+    else:
+        expected = [
+            "/p/Folder/file.x1.9.tar.gz",
+            "/p/Folder/file.x1.10.tar.gz",
+            "/p/Folder (1)/file.tar.gz",
+            "/p/Folder (1)/file (1).tar.gz",
+            "/p/Folder (2)/file.tar.gz",
+            "/p/Folder (10)/file.tar.gz",
+        ]
     result = natsort.os_sorted(given)
     assert result == expected