Fix the detection logic of the reservation words for the file name on…

… Windows: #57

Signed-off-by: Tsuyoshi Hombashi <tsuyoshi.hombashi@gmail.com>

pathvalidate/_base.py

@@ -4,9 +4,10 @@
 
 import abc
 import os
+import re
 import sys
 from collections.abc import Sequence
-from typing import ClassVar, Optional
+from typing import ClassVar, Final, Optional
 
 from ._common import normalize_platform, unprintable_ascii_chars
 from ._const import DEFAULT_MIN_LEN, Platform
@@ -181,6 +182,8 @@ def sanitize(self, value: PathType, replacement_text: str = "") -> PathType:  #
 
 
 class BaseValidator(AbstractValidator):
+    __RE_ROOT_NAME: Final = re.compile(r"([^\.]+)")
+
     @property
     def min_len(self) -> int:
         return self._min_len
@@ -234,6 +237,10 @@ def _validate_max_len(self) -> None:
         if self.min_len > self.max_len:
             raise ValueError("min_len must be lower than max_len")
 
-    @staticmethod
-    def __extract_root_name(path: str) -> str:
-        return os.path.splitext(os.path.basename(path))[0]
+    @classmethod
+    def __extract_root_name(cls, path: str) -> str:
+        match = cls.__RE_ROOT_NAME.match(os.path.basename(path))
+        if match is None:
+            return ""
+
+        return match.group(1)

pathvalidate/_filename.py

@@ -122,8 +122,16 @@ def _get_sanitize_regexp(self) -> Pattern[str]:
 
 
 class FileNameValidator(BaseValidator):
-    _WINDOWS_RESERVED_FILE_NAMES = ("CON", "PRN", "AUX", "CLOCK$", "NUL") + tuple(
-        f"{name:s}{num:d}" for name, num in itertools.product(("COM", "LPT"), range(1, 10))
+    _WINDOWS_RESERVED_FILE_NAMES = (
+        ("CON", "PRN", "AUX", "CLOCK$", "NUL")
+        + tuple(f"{name:s}{num:d}" for name, num in itertools.product(("COM", "LPT"), range(0, 10)))
+        + tuple(
+            f"{name:s}{ssd:s}"
+            for name, ssd in itertools.product(
+                ("COM", "LPT"),
+                ("\N{SUPERSCRIPT ONE}", "\N{SUPERSCRIPT TWO}", "\N{SUPERSCRIPT THREE}"),
+            )
+        )
     )
     _MACOS_RESERVED_FILE_NAMES = (":",)
 

test/_common.py

@@ -85,7 +85,14 @@
     "clock$",
     "NUL",
     "nul",
-] + [f"{name:s}{num:d}" for name, num in product(["COM", "com", "LPT", "lpt"], range(1, 10))]
+] + [
+    f"{name:s}{num}"
+    for name, num in product(
+        ["COM", "com", "LPT", "lpt"],
+        list(range(0, 10))
+        + ["\N{SUPERSCRIPT ONE}", "\N{SUPERSCRIPT TWO}", "\N{SUPERSCRIPT THREE}"],
+    )
+]
 NTFS_RESERVED_FILE_NAMES = [
     "$Mft",
     "$MftMirr",

test/test_filename.py

@@ -88,6 +88,7 @@ class Test_FileNameValidator:
                 (
                     "AUX",
                     "CLOCK$",
+                    "COM0",
                     "COM1",
                     "COM2",
                     "COM3",
@@ -97,7 +98,11 @@ class Test_FileNameValidator:
                     "COM7",
                     "COM8",
                     "COM9",
+                    "COM²",
+                    "COM³",
+                    "COM¹",
                     "CON",
+                    "LPT0",
                     "LPT1",
                     "LPT2",
                     "LPT3",
@@ -107,6 +112,9 @@ class Test_FileNameValidator:
                     "LPT7",
                     "LPT8",
                     "LPT9",
+                    "LPT²",
+                    "LPT³",
+                    "LPT¹",
                     "NUL",
                     "PRN",
                 ),
@@ -362,6 +370,12 @@ def test_exception_win_invalid_char(self, value, platform):
                 WIN_RESERVED_FILE_NAMES, ["windows", "universal"]
             )
         ]
+        + [
+            [f"{reserved_keyword}.tar.gz", platform, ValidationError]
+            for reserved_keyword, platform in product(
+                WIN_RESERVED_FILE_NAMES, ["windows", "universal"]
+            )
+        ]
         + [
             [reserved_keyword, platform, None]
             for reserved_keyword, platform in product([".", ".."], ["posix", "linux", "macos"])