plist can be either binary or text

identify/extensions.py

@@ -97,7 +97,6 @@
     'phtml': {'text', 'php'},
     'pl': {'text', 'perl'},
     'plantuml': {'text', 'plantuml'},
-    'plist': {'text', 'plist'},
     'png': {'binary', 'image', 'png'},
     'po': {'text', 'pofile'},
     'pp': {'text', 'puppet'},
@@ -161,6 +160,9 @@
     'zip': {'binary', 'zip'},
     'zsh': {'text', 'shell', 'zsh'},
 }
+EXTENSIONS_NEED_BINARY_CHECK = {
+    'plist': {'plist'},
+}
 
 NAMES = {
     '.babelrc': {'text', 'json', 'babelrc'},

identify/identify.py

@@ -22,6 +22,7 @@
 
 ALL_TAGS = {DIRECTORY, SYMLINK, FILE, EXECUTABLE, NON_EXECUTABLE, TEXT, BINARY}
 ALL_TAGS.update(*extensions.EXTENSIONS.values())
+ALL_TAGS.update(*extensions.EXTENSIONS_NEED_BINARY_CHECK.values())
 ALL_TAGS.update(*extensions.NAMES.values())
 ALL_TAGS.update(*interpreters.INTERPRETERS.values())
 ALL_TAGS = frozenset(ALL_TAGS)
@@ -54,6 +55,9 @@ def tags_from_path(path):
             if len(shebang) > 0:
                 tags.update(tags_from_interpreter(shebang[0]))
 
+    # some extensions can be both binary and text
+    # see EXTENSIONS_NEED_BINARY_CHECK
+    if not {TEXT, BINARY} & tags:
         if file_is_text(path):
             tags.add(TEXT)
         else:
@@ -80,6 +84,8 @@ def tags_from_filename(filename):
         ext = ext[1:].lower()
         if ext in extensions.EXTENSIONS:
             ret.update(extensions.EXTENSIONS[ext])
+        elif ext in extensions.EXTENSIONS_NEED_BINARY_CHECK:
+            ret.update(extensions.EXTENSIONS_NEED_BINARY_CHECK[ext])
 
     return ret
 

tests/extensions_test.py

@@ -11,3 +11,16 @@
 def test_extensions_have_binary_or_text(extension):
     tags = extensions.EXTENSIONS[extension]
     assert len({'text', 'binary'} & tags) == 1, tags
+
+
+@pytest.mark.parametrize('extension', extensions.EXTENSIONS_NEED_BINARY_CHECK)
+def test_need_binary_check_do_not_specify_text_binary(extension):
+    tags = extensions.EXTENSIONS_NEED_BINARY_CHECK[extension]
+    assert len({'text', 'binary'} & tags) == 0, tags
+
+
+def test_mutually_exclusive_check_types():
+    assert not (
+        set(extensions.EXTENSIONS) &
+        set(extensions.EXTENSIONS_NEED_BINARY_CHECK)
+    )

tests/identify_test.py

@@ -16,6 +16,13 @@ def test_all_tags_includes_basic_ones():
     assert 'directory' in identify.ALL_TAGS
 
 
+def test_all_tags_contains_each_type():
+    assert 'xml' in identify.ALL_TAGS  # extension
+    assert 'plist' in identify.ALL_TAGS  # extension, needs binary check
+    assert 'dockerfile' in identify.ALL_TAGS  # by file convention
+    assert 'python3' in identify.ALL_TAGS  # by shebang
+
+
 def test_tags_from_path_does_not_exist(tmpdir):
     x = tmpdir.join('foo')
     with pytest.raises(ValueError):
@@ -73,6 +80,35 @@ def test_tags_from_path_binary(tmpdir):
     }
 
 
+def test_tags_from_path_plist_binary(tmpdir):
+    x = tmpdir.join('t.plist')
+    x.write_binary(
+        b'bplist00\xd1\x01\x02_\x10\x0fLast Login NameWDefault\x08\x0b\x1d\x00'
+        b'\x00\x00\x00\x00\x00\x01\x01\x00\x00\x00\x00\x00\x00\x00\x03\x00\x00'
+        b'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00%',
+    )
+    assert identify.tags_from_path(x.strpath) == {
+        'file', 'plist', 'binary', 'non-executable',
+    }
+
+
+def test_tags_from_path_plist_text(tmpdir):
+    x = tmpdir.join('t.plist')
+    x.write(
+        '<?xml version="1.0" encoding="UTF-8"?>\n'
+        '<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">\n'
+        '<plist version="1.0">\n'
+        '<dict>\n'
+        '\t<key>Last Login Name</key>\n'
+        '\t<string>Default</string>\n'
+        '</dict>\n'
+        '</plist>\n',
+    )
+    assert identify.tags_from_path(x.strpath) == {
+        'file', 'plist', 'text', 'non-executable',
+    }
+
+
 @pytest.mark.parametrize(
     ('filename', 'expected'),
     (
@@ -85,6 +121,9 @@ def test_tags_from_path_binary(tmpdir):
         ('mod/test.py', {'text', 'python'}),
         ('mod/Dockerfile', {'text', 'dockerfile'}),
 
+        # does not set binary / text
+        ('f.plist', {'plist'}),
+
         # case of extension should be ignored
         ('f.JPG', {'binary', 'image', 'jpeg'}),
         # but case of name checks should still be honored