update setup.py and fix #95: don't add slashes to void elements

mf2py/dom_helpers.py

@@ -4,7 +4,7 @@
 import bs4
 import re
 
-from bs4.element import Tag, NavigableString, Comment
+from bs4.element import Tag, NavigableString, Comment, MinimalHTMLFormatter
 
 if sys.version < '3':
     from urlparse import urljoin
@@ -21,6 +21,11 @@
 _reduce_spaces_regex = re.compile(r" {2,}")
 
 
+class MinimalHTML5Formatter(MinimalHTMLFormatter):
+    """An HTML formatter that omits the slash in void tags and othewise does minimal replacement"""
+    void_element_close_prefix = None
+
+
 def try_urljoin(base, url, allow_fragments=True):
     """attempts urljoin, on ValueError passes through url. Shortcuts http(s):// urls"""
     if url.startswith(("https://", "http://")):

mf2py/parse_property.py

@@ -1,7 +1,7 @@
 """functions to parse the properties of elements"""
 from __future__ import unicode_literals, print_function
 
-from .dom_helpers import get_attr, get_img_src_alt, get_textContent, try_urljoin
+from .dom_helpers import get_attr, get_img_src_alt, get_textContent, try_urljoin, MinimalHTML5Formatter
 from .datetime_helpers import normalize_datetime, DATETIME_RE, TIME_RE
 from . import value_class_pattern
 
@@ -104,6 +104,6 @@ def datetime(el, default_date=None):
 def embedded(el, base_url=''):
     """Process e-* properties"""
     return {
-        'html': el.decode_contents().strip(),    # secret bs4 method to get innerHTML
+        'html': el.decode_contents(formatter=MinimalHTML5Formatter()).strip(),    # secret bs4 method to get innerHTML
         'value': get_textContent(el, replace_img=True, base_url=base_url)
     }

setup.py

@@ -20,8 +20,8 @@
       install_requires=[
           # Keep in sync with requirements.txt!
           'html5lib>=1.0.1',
-          'requests>=2.18.4',
-          'BeautifulSoup4>=4.6.0',
+          'requests>=2.19.1',
+          'BeautifulSoup4>=4.6.3',
       ],
       tests_require=[
           'lxml',