Hi ports@,
While looking at py-lxml's deps I came across an issue in py-html5lib.
py-lxml 3.6.0 broke "make test" but it turns out it's html5lib's
fault: it was using lxml incorrectly, the issue was noticed and fixed
already, but not yet released:
https://github.com/lxml/lxml/pull/172#issuecomment-169119391
This update backports a patch from head of line to address the issue:
make test wins now with py-lxml 3.6.0. It also has the happy property
of not breaking "make test" with the current py-lxml (3.4.3). It also
addresses a portcheck nit (line > 80 cols).
If nobody cares I'll also take MAINTAINER.
Pax, -A
--
http://haqistan.net/~attila | [email protected] | 0x62A729CF
Index: Makefile
===================================================================
RCS file: /cvs/ports/devel/py-html5lib/Makefile,v
retrieving revision 1.11
diff -u -p -r1.11 Makefile
--- Makefile 1 Jan 2016 20:22:09 -0000 1.11
+++ Makefile 27 Apr 2016 00:08:27 -0000
@@ -6,6 +6,7 @@ MODPY_EGG_VERSION= 0.9999999
DISTNAME= html5lib-${MODPY_EGG_VERSION}
PKGNAME= py-${DISTNAME}
CATEGORIES= devel www textproc
+REVISION= 1
HOMEPAGE= https://github.com/html5lib/html5lib-python/
@@ -25,6 +26,7 @@ FLAVORS = python3
FLAVOR ?=
do-test:
- cd ${WRKSRC} && ${LOCALBASE}/bin/nosetests${MODPY_BIN_SUFFIX} --first-package-wins
+ cd ${WRKSRC} && ${LOCALBASE}/bin/nosetests${MODPY_BIN_SUFFIX} \
+ --first-package-wins
.include <bsd.port.mk>
Index: patches/patch-html5lib_ihatexml_py
===================================================================
RCS file: patches/patch-html5lib_ihatexml_py
diff -N patches/patch-html5lib_ihatexml_py
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ patches/patch-html5lib_ihatexml_py 27 Apr 2016 00:08:27 -0000
@@ -0,0 +1,15 @@
+$OpenBSD$
+
+Backport of patch for comment issue
+--- html5lib/ihatexml.py.orig Fri Apr 22 18:44:14 2016
++++ html5lib/ihatexml.py Fri Apr 22 18:45:49 2016
+@@ -225,6 +225,9 @@ class InfosetFilter(object):
+ while "--" in data:
+ warnings.warn("Comments cannot contain adjacent dashes", DataLossWarning)
+ data = data.replace("--", "- -")
++ if data.endswith("-"):
++ warnings.warn("Comments cannot end in a dash", DataLossWarning)
++ data += " "
+ return data
+
+ def coerceCharacters(self, data):
Index: patches/patch-html5lib_treebuilders_etree_lxml_py
===================================================================
RCS file: patches/patch-html5lib_treebuilders_etree_lxml_py
diff -N patches/patch-html5lib_treebuilders_etree_lxml_py
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ patches/patch-html5lib_treebuilders_etree_lxml_py 27 Apr 2016 00:08:27 -0000
@@ -0,0 +1,43 @@
+$OpenBSD$
+
+Backport patch from HEAD to stop test fails due to py-lxml update to 3.6.0
+https://github.com/lxml/lxml/pull/172#issuecomment-169119391
+--- html5lib/treebuilders/etree_lxml.py.orig Mon Sep 7 18:36:25 2015
++++ html5lib/treebuilders/etree_lxml.py Fri Apr 22 18:50:50 2016
+@@ -54,7 +54,7 @@ class Document(object):
+ def testSerializer(element):
+ rv = []
+ finalText = None
+- infosetFilter = ihatexml.InfosetFilter()
++ infosetFilter = ihatexml.InfosetFilter(preventDoubleDashComments=True)
+
+ def serializeElement(element, indent=0):
+ if not hasattr(element, "tag"):
+@@ -189,7 +189,7 @@ class TreeBuilder(_base.TreeBuilder):
+
+ def __init__(self, namespaceHTMLElements, fullTree=False):
+ builder = etree_builders.getETreeModule(etree, fullTree=fullTree)
+- infosetFilter = self.infosetFilter = ihatexml.InfosetFilter()
++ infosetFilter = self.infosetFilter = ihatexml.InfosetFilter(preventDoubleDashComments=True)
+ self.namespaceHTMLElements = namespaceHTMLElements
+
+ class Attributes(dict):
+@@ -257,7 +257,7 @@ class TreeBuilder(_base.TreeBuilder):
+ data = property(_getData, _setData)
+
+ self.elementClass = Element
+- self.commentClass = builder.Comment
++ self.commentClass = Comment
+ # self.fragmentClass = builder.DocumentFragment
+ _base.TreeBuilder.__init__(self, namespaceHTMLElements)
+
+@@ -344,7 +344,8 @@ class TreeBuilder(_base.TreeBuilder):
+
+ # Append the initial comments:
+ for comment_token in self.initial_comments:
+- root.addprevious(etree.Comment(comment_token["data"]))
++ comment = self.commentClass(comment_token["data"])
++ root.addprevious(comment._element)
+
+ # Create the root document and add the ElementTree to it
+ self.document = self.documentClass()
