Merge branch 'hotfix/unicode_title'

Author: ikirudennis

CHANGELOG.textile

@@ -1,5 +1,9 @@
 h1. Textile Changelog
 
+h2. Version 2.3.3
+* Bugfix: Unicode in URL titles no longer break everything ("#30":https://github.com/textile/python-textile/issues/30)
+* Display DeprecationWarning when using textile on Python 2.6.
+
 h2. Version 2.3.2
 * Bugfix: properly handle @":"@ as text, not a link.
 

tests/test_github_issues.py

@@ -79,3 +79,9 @@ def parseWapProfile(self, url):
 
 \t<p>Of course there&#8217;s a lot more error handling to do (and useful data to glean off the <a href="XML"><span class="caps">XML</span></a>), but being able to cut through all the usual parsing crap is immensely gratifying.</p>""")
     assert result == expect
+
+def test_github_issue_30():
+    text ='"Tëxtíle (Tëxtíle)":http://lala.com'
+    result = textile.textile(text)
+    expect = '\t<p><a href="http://lala.com" title="Tëxtíle">Tëxtíle</a></p>'
+    assert result == expect

textile/__init__.py

@@ -1,8 +1,20 @@
 from __future__ import unicode_literals
 
+import sys
+import warnings
+
 from .core import textile, textile_restricted, Textile
 from .version import VERSION
 
 __all__ = ['textile', 'textile_restricted']
 
 __version__ = VERSION
+
+
+if sys.version_info[:2] == (2, 6):
+    warnings.warn(
+        "Python 2.6 is no longer supported by the Python core team, please "
+        "upgrade your Python. A future version of textile will drop support "
+        "for Python 2.6",
+        DeprecationWarning
+    )

textile/core.py

@@ -874,7 +874,10 @@ def _casesdefault(c, pop, popped, url_chars, counts, pre):
         url = self.shelveURL(self.encode_url(urlunsplit(uri_parts)))
         attributes = parse_attributes(atts)
         if title:
-            attributes['title'] = title
+            # if the title contains unicode data, it is annoying to get Python
+            # 2.6 and all the latter versions working properly.  But shelving
+            # the title is a quick and dirty solution.
+            attributes['title'] = self.shelve(title)
         attributes['href'] = url
         if self.rel:
             attributes['rel'] = self.rel

textile/utils.py

@@ -50,24 +50,24 @@ def generate_tag(tag, content, attributes=None):
     content are strings, the attributes argument is a dictionary.  As
     a convenience, if the content is ' /', a self-closing tag is generated."""
     content = six.text_type(content)
-    element = ElementTree.Element(tag, attrib=attributes)
     enc = 'unicode'
     if six.PY2:
         enc = 'UTF-8'
     if not tag:
         return content
+    element = ElementTree.Element(tag, attrib=attributes)
     # FIXME: Kind of an ugly hack.  There *must* be a cleaner way.  I tried
-    # adding text by assigning it to a.text.  That results in non-ascii text
-    # being html-entity encoded.  Not bad, but not entirely matching
-    # php-textile either.
+    # adding text by assigning it to element_tag.text.  That results in
+    # non-ascii text being html-entity encoded.  Not bad, but not entirely
+    # matching php-textile either.
     try:
         element_tag = ElementTree.tostringlist(element, encoding=enc,
                 method='html')
         element_tag.insert(len(element_tag) - 1, content)
         element_text = ''.join(element_tag)
     except AttributeError:
         # Python 2.6 doesn't have the tostringlist method, so we have to treat
-        # it different.
+        # it differently.
         element_tag = ElementTree.tostring(element, encoding=enc)
         element_text = re.sub(r"<\?xml version='1.0' encoding='UTF-8'\?>\n",
                 '', element_tag)

textile/version.py

@@ -1 +1 @@
-VERSION = '2.3.2'
+VERSION = '2.3.3'