Spaces:
Running
Running
Yurii Paniv
commited on
Commit
·
17873e4
1
Parent(s):
6180416
Fix formatter
Browse files- tests/test_formatter.py +2 -0
- ukrainian_tts/formatter.py +4 -1
tests/test_formatter.py
CHANGED
|
@@ -1,6 +1,7 @@
|
|
| 1 |
from ukrainian_tts.formatter import preprocess_text
|
| 2 |
import pytest
|
| 3 |
|
|
|
|
| 4 |
@pytest.mark.parametrize(
|
| 5 |
"text,expected",
|
| 6 |
[
|
|
@@ -40,6 +41,7 @@ import pytest
|
|
| 40 |
def test_formatter(text, expected):
|
| 41 |
assert preprocess_text(text) == expected
|
| 42 |
|
|
|
|
| 43 |
# Purspose of these tests, to have clearly separate list of issues
|
| 44 |
# in the conversion. Once fixed, these cases should move to test_formatter
|
| 45 |
# We still want make sure that no changes happens there, as any regressions
|
|
|
|
| 1 |
from ukrainian_tts.formatter import preprocess_text
|
| 2 |
import pytest
|
| 3 |
|
| 4 |
+
|
| 5 |
@pytest.mark.parametrize(
|
| 6 |
"text,expected",
|
| 7 |
[
|
|
|
|
| 41 |
def test_formatter(text, expected):
|
| 42 |
assert preprocess_text(text) == expected
|
| 43 |
|
| 44 |
+
|
| 45 |
# Purspose of these tests, to have clearly separate list of issues
|
| 46 |
# in the conversion. Once fixed, these cases should move to test_formatter
|
| 47 |
# We still want make sure that no changes happens there, as any regressions
|
ukrainian_tts/formatter.py
CHANGED
|
@@ -29,6 +29,7 @@ def replace_currency_with_words(text, currency, num_form):
|
|
| 29 |
text = text.replace("€", CURRENCY[currency][num_form])
|
| 30 |
return text
|
| 31 |
|
|
|
|
| 32 |
def find_any_char(text: str, find: str, start: int):
|
| 33 |
result = -1
|
| 34 |
for c in find:
|
|
@@ -38,11 +39,12 @@ def find_any_char(text: str, find: str, start: int):
|
|
| 38 |
|
| 39 |
return result
|
| 40 |
|
|
|
|
| 41 |
# Have to check if I can use https://github.com/lang-uk/tokenize-uk
|
| 42 |
def simple_tokenizer(text: str):
|
| 43 |
start = 0
|
| 44 |
index = find_any_char(text, " ,", start)
|
| 45 |
-
while
|
| 46 |
word = text[start:index]
|
| 47 |
yield word
|
| 48 |
separator = text[index]
|
|
@@ -52,6 +54,7 @@ def simple_tokenizer(text: str):
|
|
| 52 |
|
| 53 |
yield text[start:]
|
| 54 |
|
|
|
|
| 55 |
def preprocess_text(text):
|
| 56 |
text = text.lower()
|
| 57 |
# currencies
|
|
|
|
| 29 |
text = text.replace("€", CURRENCY[currency][num_form])
|
| 30 |
return text
|
| 31 |
|
| 32 |
+
|
| 33 |
def find_any_char(text: str, find: str, start: int):
|
| 34 |
result = -1
|
| 35 |
for c in find:
|
|
|
|
| 39 |
|
| 40 |
return result
|
| 41 |
|
| 42 |
+
|
| 43 |
# Have to check if I can use https://github.com/lang-uk/tokenize-uk
|
| 44 |
def simple_tokenizer(text: str):
|
| 45 |
start = 0
|
| 46 |
index = find_any_char(text, " ,", start)
|
| 47 |
+
while index >= 0:
|
| 48 |
word = text[start:index]
|
| 49 |
yield word
|
| 50 |
separator = text[index]
|
|
|
|
| 54 |
|
| 55 |
yield text[start:]
|
| 56 |
|
| 57 |
+
|
| 58 |
def preprocess_text(text):
|
| 59 |
text = text.lower()
|
| 60 |
# currencies
|