Merge branch 'master' into py36escseq

Author: willkg

.travis.yml

@@ -5,6 +5,7 @@ python:
   - "3.3"
   - "3.4"
   - "3.5"
+  - "3.6"
   - "pypy"
 
 sudo: false

AUTHORS.rst

@@ -43,3 +43,4 @@ Patches and suggestions
 - Marc Abramowitz
 - Jon Dufresne
 - Ville Skyttä
+- Jonathan Vanasco

CHANGES.rst

@@ -1,6 +1,12 @@
 Change Log
 ----------
 
+unreleased
+~~~~~~~~~~~~~~~~~~
+
+* Added `itemscope` as boolean attribute 
+  https://github.com/html5lib/html5lib-python/issues/194
+
 0.999999999/1.0b10
 ~~~~~~~~~~~~~~~~~~
 

html5lib/constants.py

@@ -588,7 +588,7 @@
 ])
 
 booleanAttributes = {
-    "": frozenset(["irrelevant"]),
+    "": frozenset(["irrelevant", "itemscope"]),
     "style": frozenset(["scoped"]),
     "img": frozenset(["ismap"]),
     "audio": frozenset(["autoplay", "controls"]),

html5lib/filters/alphabeticalattributes.py

@@ -8,13 +8,24 @@
     from ordereddict import OrderedDict
 
 
+def _attr_key(attr):
+    """Return an appropriate key for an attribute for sorting
+
+    Attributes have a namespace that can be either ``None`` or a string. We
+    can't compare the two because they're different types, so we convert
+    ``None`` to an empty string first.
+
+    """
+    return (attr[0][0] or ''), attr[0][1]
+
+
 class Filter(base.Filter):
     def __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(),
-                                          key=lambda x: x[0]):
+                                          key=_attr_key):
                     attrs[name] = value
                 token["data"] = attrs
             yield token

html5lib/tests/serializer-testdata/options.test

@@ -46,6 +46,29 @@
                 "quote_attr_values": "always"
             }
         },
+        {
+            "expected": [
+                "<div itemscope>"
+            ],
+            "input": [
+                [
+                    "StartTag",
+                    "http://www.w3.org/1999/xhtml",
+                    "div",
+                    [
+                        {
+                            "namespace": null,
+                            "name": "itemscope",
+                            "value": "itemscope"
+                        }
+                    ]
+                ]
+            ],
+            "description": "quote_attr_values='always' with itemscope",
+            "options": {
+                "quote_attr_values": "always"
+            }
+        },
         {
             "expected": [
                 "<div irrelevant>"
@@ -171,6 +194,29 @@
                 "use_trailing_solidus": true
             }
         },
+        {
+            "expected": [
+                "<div itemscope=itemscope>"
+            ],
+            "input": [
+                [
+                    "StartTag",
+                    "http://www.w3.org/1999/xhtml",
+                    "div",
+                    [
+                        {
+                            "namespace": null,
+                            "name": "itemscope",
+                            "value": "itemscope"
+                        }
+                    ]
+                ]
+            ],
+            "description": "minimize_boolean_attributes=false",
+            "options": {
+                "minimize_boolean_attributes": false
+            }
+        },
         {
             "expected": [
                 "<div irrelevant=irrelevant>"
@@ -194,6 +240,29 @@
                 "minimize_boolean_attributes": false
             }
         },
+        {
+            "expected": [
+                "<div itemscope=\"\">"
+            ],
+            "input": [
+                [
+                    "StartTag",
+                    "http://www.w3.org/1999/xhtml",
+                    "div",
+                    [
+                        {
+                            "namespace": null,
+                            "name": "itemscope",
+                            "value": ""
+                        }
+                    ]
+                ]
+            ],
+            "description": "minimize_boolean_attributes=false with empty value",
+            "options": {
+                "minimize_boolean_attributes": false
+            }
+        },
         {
             "expected": [
                 "<div irrelevant=\"\">"

html5lib/tests/test_alphabeticalattributes.py

@@ -0,0 +1,81 @@
+from __future__ import absolute_import, division, unicode_literals
+
+try:
+    from collections import OrderedDict
+except ImportError:
+    from ordereddict import OrderedDict
+
+import pytest
+
+import html5lib
+from html5lib.filters.alphabeticalattributes import Filter
+from html5lib.serializer import HTMLSerializer
+
+
+@pytest.mark.parametrize('msg, attrs, expected_attrs', [
+    (
+        'no attrs',
+        {},
+        {}
+    ),
+    (
+        'one attr',
+        {(None, 'alt'): 'image'},
+        OrderedDict([((None, 'alt'), 'image')])
+    ),
+    (
+        'multiple attrs',
+        {
+            (None, 'src'): 'foo',
+            (None, 'alt'): 'image',
+            (None, 'style'): 'border: 1px solid black;'
+        },
+        OrderedDict([
+            ((None, 'alt'), 'image'),
+            ((None, 'src'), 'foo'),
+            ((None, 'style'), 'border: 1px solid black;')
+        ])
+    ),
+])
+def test_alphabetizing(msg, attrs, expected_attrs):
+    tokens = [{'type': 'StartTag', 'name': 'img', 'data': attrs}]
+    output_tokens = list(Filter(tokens))
+
+    attrs = output_tokens[0]['data']
+    assert attrs == expected_attrs
+
+
+def test_with_different_namespaces():
+    tokens = [{
+        'type': 'StartTag',
+        'name': 'pattern',
+        'data': {
+            (None, 'id'): 'patt1',
+            ('http://www.w3.org/1999/xlink', 'href'): '#patt2'
+        }
+    }]
+    output_tokens = list(Filter(tokens))
+
+    attrs = output_tokens[0]['data']
+    assert attrs == OrderedDict([
+        ((None, 'id'), 'patt1'),
+        (('http://www.w3.org/1999/xlink', 'href'), '#patt2')
+    ])
+
+
+def test_with_serializer():
+    """Verify filter works in the context of everything else"""
+    parser = html5lib.HTMLParser()
+    dom = parser.parseFragment('<svg><pattern xlink:href="#patt2" id="patt1"></svg>')
+    walker = html5lib.getTreeWalker('etree')
+    ser = HTMLSerializer(
+        alphabetical_attributes=True,
+        quote_attr_values='always'
+    )
+
+    # FIXME(willkg): The "xlink" namespace gets dropped by the serializer. When
+    # that gets fixed, we can fix this expected result.
+    assert (
+        ser.render(walker(dom)) ==
+        '<svg><pattern id="patt1" href="#patt2"></pattern></svg>'
+    )

requirements-install.sh

@@ -1,17 +1,15 @@
 #!/bin/bash -ex
 
-pip install pip==6.1.0
+if [[ $SIX_VERSION ]]; then
+  pip install six==$SIX_VERSION
+fi
 
-pip install -U -r requirements-test.txt
+pip install -r requirements-test.txt
 
 if [[ $USE_OPTIONAL == "true" ]]; then
-  pip install -U -r requirements-optional.txt
-fi
-
-if [[ $SIX_VERSION ]]; then
-  pip install six==$SIX_VERSION
+  pip install -r requirements-optional.txt
 fi
 
 if [[ $CI == "true" ]]; then
-  pip install -U codecov
+  pip install codecov
 fi

setup.cfg

@@ -9,3 +9,6 @@ exclude = .git,__pycache__,.tox,doc
 [flake8]
 ignore = N
 max-line-length = 139
+
+[metadata]
+license_file = LICENSE

setup.py

@@ -70,6 +70,7 @@ def default_environment():
     'Programming Language :: Python :: 3.3',
     'Programming Language :: Python :: 3.4',
     'Programming Language :: Python :: 3.5',
+    'Programming Language :: Python :: 3.6',
     'Topic :: Software Development :: Libraries :: Python Modules',
     'Topic :: Text Processing :: Markup :: HTML'
 ]

tox.ini

@@ -1,5 +1,5 @@
 [tox]
-envlist = {py26,py27,py33,py34,py35,pypy}-{base,optional}
+envlist = {py26,py27,py33,py34,py35,py36,pypy}-{base,optional}
 
 [testenv]
 deps =
@@ -13,5 +13,5 @@ deps =
   optional: -r{toxinidir}/requirements-optional.txt
 
 commands =
-  {envbindir}/py.test
+  {envbindir}/py.test {posargs}
   {toxinidir}/flake8-run.sh