Treat a string path as a one-file group in newer_pairwise_group

distutils/_modified.py

@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-import functools
+import os
 import os.path
 from collections.abc import Callable, Iterable
 from typing import Literal, TypeVar
@@ -91,4 +91,22 @@ def missing_as_newer(source):
     )
 
 
-newer_pairwise_group = functools.partial(newer_pairwise, newer=newer_group)
+def _as_source_group(source):
+    if isinstance(source, (str, bytes, os.PathLike)):
+        return (source,)
+    return source
+
+
+def newer_pairwise_group(sources, targets, missing: Literal["error", "ignore", "newer"] = "error"):
+    """
+    Filter source groups that are newer than their paired targets.
+
+    Each item in *sources* is a group of files passed to ``newer_group``.
+    A single path is treated as a one-file group, so a string is not
+    iterated as individual characters.
+    """
+
+    def _group_newer(source, target):
+        return newer_group(_as_source_group(source), target, missing=missing)
+
+    return newer_pairwise(sources, targets, newer=_group_newer)

distutils/tests/test_modified.py

@@ -119,6 +119,24 @@ def test_newer_pairwise_group(groups_target):
     assert newer == ([groups_target.newer], [groups_target.target])
 
 
+def test_newer_pairwise_group_single_path(tmp_path):
+    """A string path is a one-file group, not an iterable of characters."""
+    src = tmp_path / "source.c"
+    old_tgt = tmp_path / "old.o"
+    new_tgt = tmp_path / "new.o"
+    src.write_text("", encoding="utf-8")
+    old_tgt.write_text("", encoding="utf-8")
+    new_tgt.write_text("", encoding="utf-8")
+    os.utime(src, (10, 10))
+    os.utime(old_tgt, (1, 1))
+    os.utime(new_tgt, (20, 20))
+
+    stale = newer_pairwise_group([str(src)], [str(new_tgt)])
+    dirty = newer_pairwise_group([str(src)], [str(old_tgt)])
+    assert stale == ([], [])
+    assert dirty == ([str(src)], [str(old_tgt)])
+
+
 def test_newer_group_no_sources_no_target(tmp_path):
     """
     Consider no sources and no target "newer".

newsfragments/411.bugfix.rst

@@ -0,0 +1 @@
+``newer_pairwise_group`` now treats a string path as a one-file group instead of iterating its characters. (pypa/distutils#411)