added support for preferred term

Author: soldni

quickumls.py

@@ -205,7 +205,7 @@ def _get_all_matches(self, ngrams):
             for match in ngram_cands:
                 cuisem_match = sorted(self.cuisem_db.get(match))
 
-                for cui, semtypes in cuisem_match:
+                for cui, semtypes, preferred in cuisem_match:
                     match_similarity = toolbox.get_similarity(
                         x=ngram_normalized.lower(),
                         y=match.lower(),
@@ -232,15 +232,16 @@ def _get_all_matches(self, ngrams):
                             'term': toolbox.safe_unicode(match),
                             'cui': cui,
                             'similarity': match_similarity,
-                            'semtypes': semtypes
+                            'semtypes': semtypes,
+                            'preferred': preferred
                         }
                     )
 
             if len(ngram_matches) > 0:
                 matches.append(
                     sorted(
                         ngram_matches,
-                        key=lambda m: m['similarity'],
+                        key=lambda m: m['similarity'] + m['preferred'],
                         reverse=True
                     )
                 )

toolbox.py

@@ -65,7 +65,7 @@ def prepare_string_for_db_input(s):
 
 
 def make_ngrams(s, n):
-    s = u'$${}$$'.format(safe_unicode(s))
+    s = u'{t}{s}{t}'.format(s=safe_unicode(s), t=('$' * (n - 1)))
     return (s[i:i+n] for i in xrange3(len(s) - n + 1))
 
 
@@ -236,7 +236,7 @@ def has_term(self, term):
         except KeyError:
             return
 
-    def insert(self, term, cui, semtypes):
+    def insert(self, term, cui, semtypes, is_preferred):
         term = prepare_string_for_db_input(safe_unicode(term))
         cui = prepare_string_for_db_input(safe_unicode(cui))
 
@@ -247,7 +247,7 @@ def insert(self, term, cui, semtypes):
         except KeyError:
             cuis = set()
 
-        cuis.add(cui)
+        cuis.add((cui, is_preferred))
         self.cui_db.Put(db_key_encode(term), pickle.dumps(cuis))
 
         try:
@@ -262,7 +262,11 @@ def get(self, term):
 
         cuis = pickle.loads(self.cui_db.Get(db_key_encode(term)))
         matches = (
-            (cui, pickle.loads(self.semtypes_db.Get(db_key_encode(cui))))
-            for cui in cuis
+            (
+                cui,
+                pickle.loads(self.semtypes_db.Get(db_key_encode(cui))),
+                is_preferred
+            )
+            for cui, is_preferred in cuis
         )
         return matches