start removal of six and improve type annotation (#607)

* fix unfinished Python2 -> 3 translation
* remove some six usage
* annotate
* fix regression in Six removal
* mypy: self.enf is always defined
* fix return type of cups.py
* Usb idVendor/idProduct are integers
* self.default_args is always defined
* tweak usb_args, PEP589 is better
* lp.py: reassure mypy
* correctly cast call to CashDrawerError()
* update CUPS test
* add missing close() method in metaclass
* document a bug in typeshed
* query_status() returns bytes as seen in constants.py
* remove more SIX usage
* test examples too
* remove type comment where type is annotated
* adapt behavior of cups printer to match other implementations

---------

Co-authored-by: Patrick Kanzler <dev@pkanzler.de>
Co-authored-by: Patrick Kanzler <4189642+patkan@users.noreply.github.com>

src/escpos/magicencode.py

@@ -23,7 +23,7 @@ from .constants import CODEPAGE_CHANGE
 from .exceptions import Error
 
 
-class Encoder(object):
+class Encoder:
     """Take available code spaces and pick the right one for a given character.
 
     Note: To determine the code page, it needs to do the conversion, and
@@ -202,7 +202,7 @@ def split_writable_text(encoder, text, encoding):
     return text, None
 
 
-class MagicEncode(object):
+class MagicEncode:
     """Help switching to the right code page.
 
     A helper that helps us to automatically switch to the right
@@ -292,7 +292,7 @@ class MagicEncode(object):
 
     def write_with_encoding(self, encoding, text):
         """Write the text and inject necessary codepage switches."""
-        if text is not None and type(text) is not six.text_type:
+        if text is not None and type(text) is not str:
             raise Error(
                 f"The supplied text has to be unicode, but is of type {type(text)}."
             )