Merge pull request #179 from ferstar/master

Add a new CDFV2 Microsoft Word filetype and fix tox deps

Author: h2non

filetype/types/document.py

@@ -124,6 +124,11 @@ def match(self, buf):
                 )
             ):
                 return True
+            if (
+                len(buf) > 663 and buf[512:531] == b"R\x00o\x00o\x00t\x00 \x00E\x00n\x00t\x00r\x00y"
+                and buf[640:663] == b"W\x00o\x00r\x00d\x00D\x00o\x00c\x00u\x00m\x00e\x00n\x00t"
+            ):
+                return True
 
         return False
 

tests/fixtures/sample_2.doc

@@ -94,7 +94,7 @@ def test_guess_zstd(self):
             self.assertEqual(kind.extension, 'zst')
 
     def test_guess_doc(self):
-        for name in 'sample.doc', 'sample_1.doc':
+        for name in 'sample.doc', 'sample_1.doc', 'sample_2.doc':
             kind = filetype.guess(os.path.join(FIXTURES, name))
             self.assertIsNotNone(kind)
             self.assertEqual(kind.mime, 'application/msword')

tests/test_types.py

@@ -32,6 +32,7 @@ commands = flake8 {toxinidir} --extend-exclude tests,docs,build,dist,venv,.venv
 basepython = python3
 deps =
      pdoc3
+     setuptools
 commands = pdoc --html --force --output-dir docs filetype
 
 [testenv:clean]