Confluence publishing uses correct text types (pantsbuild#7746)

Without this change, actually publishing to Confluence fails with encoding errors.

Specifically:
user needs converting because xmlrpclib hard-codes what kinds it can encode, and doesn't count future's newstr.
url needs converting because it's detected as a newbytes which has its `encode` disabled. Maybe this should instead be calling `binary_type(...).encode("utf-8")`?

contrib/confluence/src/python/pants/contrib/confluence/tasks/confluence_publish.py

@@ -44,11 +44,11 @@ def prepare(cls, options, round_manager):
   def __init__(self, *args, **kwargs):
     super(ConfluencePublish, self).__init__(*args, **kwargs)
 
-    self.url = self.get_options().url
+    self.url = str(self.get_options().url)
     self.force = self.get_options().force
     self.open = self.get_options().open
     self._wiki = None
-    self.user = self.get_options().user
+    self.user = str(self.get_options().user)
 
   def wiki(self):
     raise NotImplementedError('Subclasses must provide the wiki target they are associated with')