fsn2bytes and bytes2fsn now default to utf-8/wtf-8

I've never needed something else, let's make it the default.

NEWS

@@ -1,3 +1,10 @@
+1.3.4 - 2018-01-23
+------------------
+
+* :func:`fsn2bytes` and :func:`bytes2fsn` now default to "wtf-8" for the
+  Windows path encoding instead of having no default.
+
+
 1.3.3 - 2018-01-03
 ------------------
 

docs/tutorial.rst

@@ -12,7 +12,7 @@ There are various ways to create fsnative instances:
     >>> senf.fsnative(u"foo")
     'foo'
 
-    # create form some serialized format
+    # create from some serialized format
     >>> senf.bytes2fsn(b"foo", "utf-8")
     'foo'
 

senf/_fsnative.py

@@ -473,22 +473,21 @@ def text2fsn(text):
     return fsnative(text)
 
 
-def fsn2bytes(path, encoding):
+def fsn2bytes(path, encoding="utf-8"):
     """
     Args:
         path (fsnative): The path to convert
-        encoding (`str` or `None`): `None` if you don't care about Windows
+        encoding (`str`): encoding used for Windows
     Returns:
         `bytes`
     Raises:
         TypeError: If no `fsnative` path is passed
-        ValueError: If encoding fails or no encoding is given
+        ValueError: If encoding fails or the encoding is invalid
 
     Converts a `fsnative` path to `bytes`.
 
     The passed *encoding* is only used on platforms where paths are not
-    associated with an encoding (Windows for example). If you don't care about
-    Windows you can pass `None`.
+    associated with an encoding (Windows for example).
 
     For Windows paths, lone surrogates will be encoded like normal code points
     and surrogate pairs will be merged before encoding. In case of ``utf-8``
@@ -510,22 +509,24 @@ def fsn2bytes(path, encoding):
         return path
 
 
-def bytes2fsn(data, encoding):
+def bytes2fsn(data, encoding="utf-8"):
     """
     Args:
         data (bytes): The data to convert
-        encoding (`str` or `None`): `None` if you don't care about Windows
+        encoding (`str`): encoding used for Windows
     Returns:
         `fsnative`
     Raises:
         TypeError: If no `bytes` path is passed
-        ValueError: If decoding fails or no encoding is given
+        ValueError: If decoding fails or the encoding is invalid
 
     Turns `bytes` to a `fsnative` path.
 
     The passed *encoding* is only used on platforms where paths are not
-    associated with an encoding (Windows for example). If you don't care about
-    Windows you can pass `None`.
+    associated with an encoding (Windows for example).
+
+    For Windows paths ``WTF-8`` is accepted if ``utf-8`` is used and
+    ``WTF-16`` accepted if ``utf-16-le`` is used.
     """
 
     if not isinstance(data, bytes):

tests/test_api.py

@@ -604,6 +604,8 @@ def test_fsn2bytes():
         with pytest.raises(TypeError):
             fsn2bytes(path)
 
+    assert fsn2bytes(fsnative(u"foo"), "utf-8") == fsn2bytes(fsnative(u"foo"))
+
 
 @pytest.mark.skipif(os.name != "nt", reason="win only")
 def test_fsn2bytes_surrogate_pairs():
@@ -737,6 +739,8 @@ def test_bytes2fsn():
         with pytest.raises(TypeError):
             bytes2fsn(b"data", object())
 
+    assert bytes2fsn(b"foo", "utf-8") == bytes2fsn(b"foo")
+
 
 def test_constants():
     assert isinstance(sep, fsnative)