class EbookTranslator:
def __init__(self,
path,
source_language,
target_language,
engine="Google"):
self.source_language = source_language
self.target_language = target_language
self.ebook = Ebook(path)
self.set_out_path()
self.set_counter_path()
self.set_start_point()
self.translator = Translator(source_language, target_language, engine)
def translate(self):
for counter, original_sentence in enumerate(tqdm(
self.ebook.sentences)):
translated_sentence = self.translator.translate(original_sentence)
both_sentences = f"{translated_sentence}\n\n{original_sentence}\n\n"
self.write_to_file(both_sentences, self.out_path)
self.write_to_file(str(counter), self.counter_path, mode="w")
self.quit()
def set_out_path(self):
self.out_path = self.ebook.path.replace(
".txt", f"_translated_to_{self.target_language}.txt")
def set_counter_path(self):
self.counter_path = self.ebook.path.replace(".txt", f"_counter.txt")
def write_counter_to_file(self, counter):
self.write_to_file(counter, self.counter_path, mode="w")
def write_to_file(self, text, path, mode="a"):
with open(path, mode) as file:
file.write(text)
def read_counter(self):
with open(self.counter_path) as file:
return int(file.read())
def set_start_point(self):
if os.path.exists(self.counter_path):
start_point = self.read_counter() + 1
print(f"Continuing translation from sentence {start_point}")
self.ebook.sentences = self.ebook.sentences[start_point:]
else:
print("Starting translation from the beginning.")
def quit(self):
self.translator.quit()
class TranslatorTest(unittest.TestCase):
def setUp(self):
warnings.filterwarnings("ignore", category=ResourceWarning, message="unclosed.*<ssl.SSLSocket.*>")
self.translator = Translator(source_language="en", target_language="de", engine="Google")
def test_1_set_engine(self):
self.translator.set_engine("Google")
self.assertTrue(self.translator.engine)
def test_2_translate(self):
text = get_mock_text()
translation = self.translator.translate(text)
self.assertTrue(isinstance(translation, str))
def test_3_quit(self):
self.translator.quit()
tw["content"] = translation[j]
except IndexError:
pass
news += deepcopy(translated_tweets)
translated_tweets = []
return news + tweets
if __name__ == '__main__':
reader = InterTASSReader('intertass-ES-train-tagged.xml')
tweets = list(reader.tweets()) # iterador sobre los tweets
langs = list(LANGUAGES.keys())[:50]
try:
langs.remove("es")
langs.remove('zh-cn')
langs.remove('zh-tw')
except ValueError:
pass
client = Translator(headless_browser=True, bulk=True)
augmented_train = []
for lang in tqdm(langs):
try:
augmented_train += bulk_translate(tweets, lang)
except:
pass
client.quit()
with open("augmented_data.pkl", "wb") as f:
pickle.dump(augmented_train, f)
