hooks.add_entity: ASCIIfy stroke characters.

import2ledger/hooks/add_entity.py

@@ -19,6 +19,16 @@ class AddEntityHook:
                 s = re.sub(open_c + self.NO_PARENS + close_c, '', s)
         return s if s else last_s
 
+    def _destroke_chr(self, c):
+        name = unicodedata.name(c, '')
+        if name.endswith(' WITH STROKE'):
+            return unicodedata.lookup(name[:-12])
+        else:
+            return c
+
+    def _destroke(self, s):
+        return ''.join(self._destroke_chr(c) for c in s)
+
     def _entity_parts(self, s, trim_re):
         for word in s.split():
             word = unicodedata.normalize('NFKD', word)
@@ -36,6 +46,7 @@ class AddEntityHook:
     def run(self, data):
         if ('payee' in data) and ('entity' not in data):
             payee = self._remove_parens(data['payee'])
+            payee = self._destroke(payee)
             entity = self._str2entity(payee, self.NONASCII_RE)
             if not entity:
                 entity = self._str2entity(payee, self.NONALNUM_RE)

tests/test_hooks.py

@@ -18,6 +18,7 @@ def test_load_all():
     ('Fran Doe-Smith', 'Doe-Smith-Fran'),
     ('Alex(Nickname) Smith', 'Smith-Alex'),
     ('稲荷', '稲荷'),
+    ('Pøweł', 'Powel'),
 ])
 def test_add_entity(payee, expected):
     data = {'payee': payee}