failsafe parsing and more logging

readability/readability.py

@@ -28,17 +28,17 @@ class Document:
 	TEXT_LENGTH_THRESHOLD = 25
 	RETRY_LENGTH = 250
 
-	def __init__(self, input, **options):
-		self.input = inpuunicodear
+	def __init__(self, input, notify=None, **options):
+		self.input = input
 		self.options = defaultdict(lambda: None)
 		for k, v in options.items():
 			self.options[k] = v
+		self.notify = notify or logging.info
 		self.html = None
 
 	def _html(self, force=False):
 		if force or self.html is None:
-			notify = self.options['notify'] or (lambda x: None)
-			self.html = parse(self.input, self.options['url'], notify=notify)
+			self.html = parse(self.input, self.options['url'], notify=self.notify)
 		return self.html
 	
 	def content(self):
@@ -48,6 +48,7 @@ class Document:
 		return get_title(self._html())
 
 	def summary(self):
+		try:
 			ruthless = True
 			while True:
 				self._html(True)
@@ -74,6 +75,9 @@ class Document:
 					continue # try again
 				else:
 					return cleaned_article
+		except StandardError, e:
+			logging.exception('error getting summary:')
+			raise Unparseable(str(e))
 
 	def get_article(self, candidates, best_candidate):
 		# Now that we have the top candidate, look through its siblings for content that might also be related.
@@ -322,6 +326,7 @@ def main():
 	if not (len(args) == 1 or options.url):
 		parser.print_help()
 		sys.exit(1)
+	logging.basicConfig(level=logging.DEBUG)
 
 	file = None
 	if options.url: