def remove_nikud(text):
# text_processor = TextProcessor(DEFAULT_CONFIG_PATH)
# # model2text = ModelInputToSentence(DEFAULT_CONFIG_PATH)
# # inputs, labels = text_processor.prepare_text(text)
# # empty_labels = [[]*len(labels)]
# # result = model2text.input_and_labels_to_sentence(inputs, empty_labels)
# # return result
cid = CharIdsSingleton.get_instance()
result = ""
for c in text:
if not cid.is_nikud(c):
result += c
return result
def get_prediction_for_text(text):
#TODO: think where to put this
path = "C:/nikud/model/resources/config.xlsx"
CharIdsSingleton(path)
text_processor = TextProcessor()
model2text = ModelInputToSentence()
result = ""
all_inputs, _ = text_processor.prepare_text(text)
predictions = []
for inputs in all_inputs:
labels = predict(inputs)
decoded = model2text.input_and_labels_to_sentence(inputs, labels)
result += decoded
return result
def setUpClass(cls):
path = "../resources/letters.xlsx"
CharIdsSingleton(path)
def __init__(self):
self.cid = CharIdsSingleton.get_instance()
def setUpClass(cls):
path = "../resources/test_charIdsConfig.xlsx"
CharIdsSingleton(path)
def setUp(self):
self.cid = CharIdsSingleton.get_instance()
def setUpClass(cls):
config_path = "../resources\letters.xlsx"
CharIdsSingleton(config_path)
def __init__(self):
self.cid = CharIdsSingleton.get_instance()
self.cp = ChunkToInputLabel()
def get_text(self, path):
ext = os.path.splitext(path)[-1].lower()
if ext == ".docx":
text = self.get_text_from_docx_file(path)
else:
raise ValueError("This file's extension is not supported: " +
path + "(supported extensions: .docx")
return text
def collect_from_file(self, path):
file_name = Path(path).stem
new_file_path = "C:/nikud/model/data/transcipts/" + file_name + "_arabic.txt"
new_file = open(new_file_path, "w", encoding="utf8")
text = self.get_text(path)
for word in text.split():
if self.isArabicInHebrewLetters(word):
new_file.write(" " + word)
new_file.close()
def collect_from_dir(self, dir_path):
directory = os.fsencode(dir_path)
for file in os.listdir(directory):
filename = os.fsdecode(file)
self.collect_from_file(dir_path + "/" + filename)
path = "../resources/config.xlsx"
CharIdsSingleton(path)
abuGal = ExtractArabicFromFiles()
abuGal.collect_from_dir("sources/חיאת עילתנא/docx")
def setUp(self):
self.cid = CharIdsSingleton.get_instance()
self.sentence2model = SentenceToModelInput()
def __init__(self, str):
self.str = str
self.cid = CharIdsSingleton.get_instance()
##self.verify_word(str)
self.letter_indices = self.find_letter_indices_in_word()