Use lxml to filter out html out of the title

Sometimes links were posted with random html in the title.
We make the code more robust by just always filtering out html

reddit_poster/__init__.py

@@ -12,6 +12,7 @@ from pelican.generators import Generator
 from functools import partial
 import logging
 import praw
+import lxml.html
 
 log = logging.getLogger(__name__)
 
@@ -31,22 +32,23 @@ def make_posts(generator, metadata, url):
     Make posts on reddit if it's not a draft, on whatever subs are specified
     """
     reddit = generator.get_reddit()
+    title =  lxml.html.fromstring(metadata['title']).text_content()
     if reddit is None:
         log.info("Reddit plugin not enabled")
         return
     if metadata.get('status') == "draft": # people don't want to post drafts
-        log.debug("ignoring draft %s" % metadata['title'])
+        log.debug("ignoring draft %s" % title)
         return
 
     collection = generator.settings['REDDIT_POSTER_COLLECT_SUB']
     sub = reddit.subreddit(collection)
-    results = sub.search(metadata['title'])
+    results = sub.search(title)
     if len([result for result in results]) > 0:
-        log.debug("ignoring %s because it is already on sub %s " % (metadata['title'], collection))
+        log.debug("ignoring %s because it is already on sub %s " % (title, collection))
         # post already was made to this sub
         return
     try:
-        submission = sub.submit(metadata['title'], url=url, resubmit=False)
+        submission = sub.submit(title, url=url, resubmit=False)
         cross_post(reddit, submission, metadata.get('subreddit'))
     except praw.exceptions.APIException as e:
         log.error("got an api exception: %s", e)