diff --git a/html5lib/filters/alphabeticalattributes.py b/html5lib/filters/alphabeticalattributes.py
index fed6996c..4795baec 100644
--- a/html5lib/filters/alphabeticalattributes.py
+++ b/html5lib/filters/alphabeticalattributes.py
@@ -1,6 +1,6 @@
 from __future__ import absolute_import, division, unicode_literals
 
-from . import _base
+from . import base
 
 try:
     from collections import OrderedDict
@@ -8,9 +8,9 @@
     from ordereddict import OrderedDict
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
     def __iter__(self):
-        for token in _base.Filter.__iter__(self):
+        for token in base.Filter.__iter__(self):
             if token["type"] in ("StartTag", "EmptyTag"):
                 attrs = OrderedDict()
                 for name, value in sorted(token["data"].items(),
diff --git a/html5lib/filters/_base.py b/html5lib/filters/base.py
similarity index 100%
rename from html5lib/filters/_base.py
rename to html5lib/filters/base.py
diff --git a/html5lib/filters/inject_meta_charset.py b/html5lib/filters/inject_meta_charset.py
index ca33b70b..2059ec86 100644
--- a/html5lib/filters/inject_meta_charset.py
+++ b/html5lib/filters/inject_meta_charset.py
@@ -1,11 +1,11 @@
 from __future__ import absolute_import, division, unicode_literals
 
-from . import _base
+from . import base
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
     def __init__(self, source, encoding):
-        _base.Filter.__init__(self, source)
+        base.Filter.__init__(self, source)
         self.encoding = encoding
 
     def __iter__(self):
@@ -13,7 +13,7 @@ def __iter__(self):
         meta_found = (self.encoding is None)
         pending = []
 
-        for token in _base.Filter.__iter__(self):
+        for token in base.Filter.__iter__(self):
             type = token["type"]
             if type == "StartTag":
                 if token["name"].lower() == "head":
diff --git a/html5lib/filters/lint.py b/html5lib/filters/lint.py
index af231d8e..a9c0831a 100644
--- a/html5lib/filters/lint.py
+++ b/html5lib/filters/lint.py
@@ -2,21 +2,21 @@
 
 from six import text_type
 
-from . import _base
+from . import base
 from ..constants import namespaces, voidElements
 
 from ..constants import spaceCharacters
 spaceCharacters = "".join(spaceCharacters)
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
     def __init__(self, source, require_matching_tags=True):
         super(Filter, self).__init__(source)
         self.require_matching_tags = require_matching_tags
 
     def __iter__(self):
         open_elements = []
-        for token in _base.Filter.__iter__(self):
+        for token in base.Filter.__iter__(self):
             type = token["type"]
             if type in ("StartTag", "EmptyTag"):
                 namespace = token["namespace"]
diff --git a/html5lib/filters/optionaltags.py b/html5lib/filters/optionaltags.py
index 8f11fff4..f6edb734 100644
--- a/html5lib/filters/optionaltags.py
+++ b/html5lib/filters/optionaltags.py
@@ -1,9 +1,9 @@
 from __future__ import absolute_import, division, unicode_literals
 
-from . import _base
+from . import base
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
     def slider(self):
         previous1 = previous2 = None
         for token in self.source:
diff --git a/html5lib/filters/sanitizer.py b/html5lib/filters/sanitizer.py
index 7f81c0d1..47da50fe 100644
--- a/html5lib/filters/sanitizer.py
+++ b/html5lib/filters/sanitizer.py
@@ -5,7 +5,7 @@
 
 from six.moves import urllib_parse as urlparse
 
-from . import _base
+from . import base
 from ..constants import namespaces, prefixes
 
 __all__ = ["Filter"]
@@ -712,7 +712,7 @@
                                re.VERBOSE)
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
     """ sanitization of XHTML+MathML+SVG and of inline style attributes."""
     def __init__(self,
                  source,
@@ -739,7 +739,7 @@ def __init__(self,
         self.svg_allow_local_href = svg_allow_local_href
 
     def __iter__(self):
-        for token in _base.Filter.__iter__(self):
+        for token in base.Filter.__iter__(self):
             token = self.sanitize_token(token)
             if token:
                 yield token
diff --git a/html5lib/filters/whitespace.py b/html5lib/filters/whitespace.py
index dfc60eeb..89210528 100644
--- a/html5lib/filters/whitespace.py
+++ b/html5lib/filters/whitespace.py
@@ -2,20 +2,20 @@
 
 import re
 
-from . import _base
+from . import base
 from ..constants import rcdataElements, spaceCharacters
 spaceCharacters = "".join(spaceCharacters)
 
 SPACES_REGEX = re.compile("[%s]+" % spaceCharacters)
 
 
-class Filter(_base.Filter):
+class Filter(base.Filter):
 
     spacePreserveElements = frozenset(["pre", "textarea"] + list(rcdataElements))
 
     def __iter__(self):
         preserve = 0
-        for token in _base.Filter.__iter__(self):
+        for token in base.Filter.__iter__(self):
             type = token["type"]
             if type == "StartTag" \
                     and (preserve or token["name"] in self.spacePreserveElements):