README.md

@@ -39,6 +39,14 @@ Examples:
 
 ## Changelog
 
+- 1.1.0
+
+  - Try to preserve filename extensions if possible
+
+- 1.0.1
+
+  - First release (as 1.0.1 due to a version number mix-up in 1.0.0)
+
 - 1.0.0-dev3
 
   - Black list low code point characters (<32)

sanitize_filename/sanitize_filename.py

 """A permissive filename sanitizer."""
 import unicodedata
+from os import path
 
 
 def sanitize(filename):
@@ -29,6 +30,12 @@ def sanitize(filename):
     if len(filename) == 0:
         filename = "__"
     if len(filename) > 255:
-        filename = filename[:255]
-        filename = filename.rstrip(". ")  # Re-check last character
+        (base, ext) = path.splitext(filename)
+        if len(ext) > 254:
+            ext = ext[254:]
+        maxl = 255 - len(ext)
+        filename = filename[:maxl]
+        filename = filename + ext
+        # Re-check last character (if there was no extension)
+        filename = filename.rstrip(". ")
     return filename

setup.py

@@ -5,7 +5,7 @@ with open("README.md", "r") as fh:
 
 setuptools.setup(
     name="sanitize_filename",
-    version="1.0.1",
+    version="1.1.0",
     author="Leo Wallentin | J++ Stockholm",
     author_email="mejl@leowallentin.se",
     description="A permissive filename sanitizer.",

test/test_sanitize.py

@@ -12,6 +12,7 @@ def test_invalid_suffix():
     """Dots are not allowed at the end."""
     assert(sanitize("def.") == "def")
     assert(sanitize("def.ghi") == "def.ghi")
+    assert(sanitize("X" * 1000 + ".").endswith("X"))
 
 
 def test_reserved_words():
@@ -23,8 +24,20 @@ def test_reserved_words():
 def test_long_names():
     """Make sure long names are truncated."""
     assert(len(sanitize("X" * 300)) == 255)
+    assert(len(sanitize(".".join(["X" * 100, "X" * 100, "X" * 100]))) == 255)
+    assert(len(sanitize(".".join(["X" * 300, "X" * 300, "X" * 300]))) == 255)
 
 
 def test_unicode_normalization():
     """Names should be NFKD normalized."""
     assert(sanitize("ў") == chr(1091)+chr(774))
+
+
+def test_extensions():
+    """Filename extensions should be preserved when possible."""
+    really_long_name = "X" * 1000 + ".pdf"
+    assert(sanitize(really_long_name).endswith(".pdf"))
+    assert(sanitize("X" * 1000).endswith("X"))
+    assert(sanitize("X" * 100 + "." + "X" * 100 + ".pdf").endswith(".pdf"))
+    assert(sanitize("X" * 100 + "." + "X" * 400).endswith("X"))
+    assert(sanitize("X" * 100 + "." + "X" * 400 + ".pdf").endswith(".pdf"))