Handle unicode in entity search

Fixes #1934

endpoints/api/search.py

@@ -13,6 +13,7 @@ from operator import itemgetter
 from stringscore import liquidmetal
 from util.names import parse_robot_username
 
+import anunidecode # Don't listen to pylint's lies. This import is required.
 import math
 
 @resource('/v1/entities/<prefix>')
@@ -27,6 +28,11 @@ class EntitySearch(ApiResource):
   @nickname('getMatchingEntities')
   def get(self, prefix, parsed_args):
     """ Get a list of entities that match the specified prefix. """
+
+    # Ensure we don't have any unicode characters in the search, as it breaks the search. Nothing
+    # being searched can have unicode in it anyway, so this is a safe operation.
+    prefix = prefix.encode('unidecode', 'ignore').replace(' ', '').lower()
+
     teams = []
     org_data = []