Remove support for end-of-life Python 2.7

.appveyor.yml

@@ -4,8 +4,6 @@ environment:
     PATH: "C:\\Python27\\Scripts\\;%PATH%"
     PYTEST_COMMAND: "coverage run -m pytest"
   matrix:
-    - TOXENV: py27-base
-    - TOXENV: py27-optional
     - TOXENV: py35-base
     - TOXENV: py35-optional
     - TOXENV: py36-base
@@ -21,7 +19,7 @@ test_script:
   - tox
 
 after_test:
-  - python debug-info.py
+  - C:\\Python35\\python.exe debug-info.py
 
 on_success:
   - codecov

.travis.yml

@@ -1,10 +1,8 @@
 language: python
 python:
   - "pypy3"
-  - "pypy"
   - "3.6"
   - "3.5"
-  - "2.7"
 
 cache: pip
 
@@ -14,7 +12,6 @@ env:
   matrix:
     - TOXENV=optional
     - TOXENV=base
-    - TOXENV=six19-optional
 
 matrix:
   include:
@@ -24,9 +21,6 @@ matrix:
     - python: "3.7"
       dist: xenial  # required for Python >= 3.7
       env: TOXENV=base
-    - python: "3.7"
-      dist: xenial  # required for Python >= 3.7
-      env: TOXENV=six19-optional
 
 install:
   - pip install tox codecov

CHANGES.rst

@@ -1,6 +1,11 @@
 Change Log
 ----------
 
+UNRELEASED
+----------
+
+* Remove support for end-of-life Python 2.7.
+
 1.0.1
 ~~~~~
 

README.rst

@@ -91,8 +91,7 @@ More documentation is available at https://html5lib.readthedocs.io/.
 Installation
 ------------
 
-html5lib works on CPython 2.7+, CPython 3.4+ and PyPy.  To install it,
-use:
+html5lib works on CPython 3.5+ and PyPy. To install it, use:
 
 .. code-block:: bash
 
@@ -128,8 +127,8 @@ Please report any bugs on the `issue tracker
 Tests
 -----
 
-Unit tests require the ``pytest`` and ``mock`` libraries and can be
-run using the ``py.test`` command in the root directory.
+Unit tests require the ``pytest`` library and can be run using the ``py.test``
+command in the root directory.
 
 Test data are contained in a separate `html5lib-tests
 <https://github.com/html5lib/html5lib-tests>`_ repository and included

debug-info.py

@@ -1,5 +1,3 @@
-from __future__ import print_function, unicode_literals
-
 import platform
 import sys
 
@@ -12,7 +10,7 @@
     "maxsize": sys.maxsize
 }
 
-search_modules = ["chardet", "datrie", "genshi", "html5lib", "lxml", "six"]
+search_modules = ["chardet", "datrie", "genshi", "html5lib", "lxml"]
 found_modules = []
 
 for m in search_modules:

doc/conf.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python3
-# -*- coding: utf-8 -*-
 #
 # html5lib documentation build configuration file, created by
 # sphinx-quickstart on Wed May  8 00:04:49 2013.
@@ -250,7 +249,7 @@
 # If true, do not generate a @detailmenu in the "Top" node's menu.
 #texinfo_no_detailmenu = False
 
-class CExtMock(object):
+class CExtMock:
     """Required for autodoc on readthedocs.org where you cannot build C extensions."""
     def __init__(self, *args, **kwargs):
         pass

html5lib/__init__.py

@@ -20,7 +20,6 @@
 * :func:`~.serializer.serialize`
 """
 
-from __future__ import absolute_import, division, unicode_literals
 
 from .html5parser import HTMLParser, parse, parseFragment
 from .treebuilders import getTreeBuilder

html5lib/_ihatexml.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
 import re
 import warnings
 
@@ -183,7 +181,7 @@ def escapeRegexp(string):
 nonPubidCharRegexp = re.compile("[^\x20\x0D\x0Aa-zA-Z0-9\\-'()+,./:=?;!*#@$_%]")
 
 
-class InfosetFilter(object):
+class InfosetFilter:
     replacementRegexp = re.compile(r"U[\dA-F]{5,5}")
 
     def __init__(self,

html5lib/_inputstream.py

@@ -1,11 +1,8 @@
-from __future__ import absolute_import, division, unicode_literals
-
-from six import text_type
-from six.moves import http_client, urllib
-
+import urllib.response
 import codecs
 import re
 from io import BytesIO, StringIO
+import http.client
 
 import webencodings
 
@@ -48,7 +45,7 @@
 charsUntilRegEx = {}
 
 
-class BufferedStream(object):
+class BufferedStream:
     """Buffering for streams that do not have buffering of their own
 
     The buffer is implemented as a list of chunks on the assumption that
@@ -125,15 +122,15 @@ def _readFromBuffer(self, bytes):
 def HTMLInputStream(source, **kwargs):
     # Work around Python bug #20007: read(0) closes the connection.
     # http://bugs.python.org/issue20007
-    if (isinstance(source, http_client.HTTPResponse) or
+    if (isinstance(source, http.client.HTTPResponse) or
         # Also check for addinfourl wrapping HTTPResponse
         (isinstance(source, urllib.response.addbase) and
-         isinstance(source.fp, http_client.HTTPResponse))):
+         isinstance(source.fp, http.client.HTTPResponse))):
         isUnicode = False
     elif hasattr(source, "read"):
-        isUnicode = isinstance(source.read(0), text_type)
+        isUnicode = isinstance(source.read(0), str)
     else:
-        isUnicode = isinstance(source, text_type)
+        isUnicode = isinstance(source, str)
 
     if isUnicode:
         encodings = [x for x in kwargs if x.endswith("_encoding")]
@@ -145,7 +142,7 @@ def HTMLInputStream(source, **kwargs):
         return HTMLBinaryInputStream(source, **kwargs)
 
 
-class HTMLUnicodeInputStream(object):
+class HTMLUnicodeInputStream:
     """Provides a unicode stream of characters to the HTMLTokenizer.
 
     This class takes care of character encoding and removing or replacing
@@ -598,10 +595,6 @@ def __next__(self):
             raise TypeError
         return self[p:p + 1]
 
-    def next(self):
-        # Py2 compat
-        return self.__next__()
-
     def previous(self):
         p = self._position
         if p >= len(self):
@@ -679,7 +672,7 @@ def jumpTo(self, bytes):
             raise StopIteration
 
 
-class EncodingParser(object):
+class EncodingParser:
     """Mini parser for detecting character encoding from meta elements"""
 
     def __init__(self, data):
@@ -860,7 +853,7 @@ def getAttribute(self):
                 attrValue.append(c)
 
 
-class ContentAttrParser(object):
+class ContentAttrParser:
     def __init__(self, data):
         assert isinstance(data, bytes)
         self.data = data

html5lib/_tokenizer.py

@@ -1,7 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
-from six import unichr as chr
-
 from collections import deque
 
 from .constants import spaceCharacters
@@ -18,7 +14,7 @@
 entitiesTrie = Trie(entities)
 
 
-class HTMLTokenizer(object):
+class HTMLTokenizer:
     """ This class takes care of tokenizing HTML.
 
     * self.currentToken
@@ -44,7 +40,7 @@ def __init__(self, stream, parser=None, **kwargs):
 
         # The current token being created
         self.currentToken = None
-        super(HTMLTokenizer, self).__init__()
+        super().__init__()
 
     def __iter__(self):
         """ This is where the magic happens.

html5lib/_trie/__init__.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
 from .py import Trie as PyTrie
 
 Trie = PyTrie

html5lib/_trie/_base.py

@@ -1,17 +1,12 @@
-from __future__ import absolute_import, division, unicode_literals
-
-try:
-    from collections.abc import Mapping
-except ImportError:  # Python 2.7
-    from collections import Mapping
+from collections.abc import Mapping
 
 
 class Trie(Mapping):
     """Abstract base class for tries"""
 
     def keys(self, prefix=None):
         # pylint:disable=arguments-differ
-        keys = super(Trie, self).keys()
+        keys = super().keys()
 
         if prefix is None:
             return set(keys)

html5lib/_trie/datrie.py

@@ -1,7 +1,4 @@
-from __future__ import absolute_import, division, unicode_literals
-
 from datrie import Trie as DATrie
-from six import text_type
 
 from ._base import Trie as ABCTrie
 
@@ -10,7 +7,7 @@ class Trie(ABCTrie):
     def __init__(self, data):
         chars = set()
         for key in data.keys():
-            if not isinstance(key, text_type):
+            if not isinstance(key, str):
                 raise TypeError("All keys must be strings")
             for char in key:
                 chars.add(char)

html5lib/_trie/py.py

@@ -1,14 +1,11 @@
-from __future__ import absolute_import, division, unicode_literals
-from six import text_type
-
 from bisect import bisect_left
 
 from ._base import Trie as ABCTrie
 
 
 class Trie(ABCTrie):
     def __init__(self, data):
-        if not all(isinstance(x, text_type) for x in data.keys()):
+        if not all(isinstance(x, str) for x in data.keys()):
             raise TypeError("All keys must be strings")
 
         self._data = data

html5lib/_utils.py

@@ -1,9 +1,5 @@
-from __future__ import absolute_import, division, unicode_literals
-
 from types import ModuleType
 
-from six import text_type
-
 try:
     import xml.etree.cElementTree as default_etree
 except ImportError:
@@ -23,10 +19,10 @@
 # escapes.
 try:
     _x = eval('"\\uD800"')  # pylint:disable=eval-used
-    if not isinstance(_x, text_type):
+    if not isinstance(_x, str):
         # We need this with u"" because of http://bugs.jython.org/issue2039
         _x = eval('u"\\uD800"')  # pylint:disable=eval-used
-        assert isinstance(_x, text_type)
+        assert isinstance(_x, str)
 except:  # pylint:disable=bare-except
     supports_lone_surrogates = False
 else:

html5lib/constants.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
 import string
 
 EOF = None

html5lib/filters/alphabeticalattributes.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
 from . import base
 
 from collections import OrderedDict

html5lib/filters/base.py

@@ -1,7 +1,4 @@
-from __future__ import absolute_import, division, unicode_literals
-
-
-class Filter(object):
+class Filter:
     def __init__(self, source):
         self.source = source
 

html5lib/filters/inject_meta_charset.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, unicode_literals
-
 from . import base