only consider partials in multi-words for initial count

This ensures that it is less likely that we exclude meaningful
words like 'hauptstrasse' just because they are frequent.
This commit is contained in:
Sarah Hoffmann
2021-06-26 11:57:09 +02:00
parent 5dd24b3ef0
commit b9fbfeff67
2 changed files with 3 additions and 3 deletions

View File

@@ -150,9 +150,8 @@ def test_init_word_table(tokenizer_factory, test_config, place_row, word_table):
tok = tokenizer_factory()
tok.init_new_db(test_config)
assert word_table.get_partial_words() == {('test', 1), ('52', 1),
assert word_table.get_partial_words() == {('test', 1),
('no', 1), ('area', 2),
('holzstrasse', 1), ('holzstr', 1),
('holz', 1), ('strasse', 1),
('str', 1)}