def __init__(self,
dict_path=None,
name='Spelling_Aug',
aug_min=1,
aug_max=10,
aug_p=0.3,
stopwords=None,
tokenizer=None,
reverse_tokenizer=None,
include_reverse=True,
stopwords_regex=None,
verbose=0):
super().__init__(action=Action.SUBSTITUTE,
name=name,
aug_p=aug_p,
aug_min=aug_min,
aug_max=aug_max,
stopwords=stopwords,
tokenizer=tokenizer,
reverse_tokenizer=reverse_tokenizer,
device='cpu',
verbose=verbose,
stopwords_regex=stopwords_regex,
include_detail=False)
self.dict_path = dict_path if dict_path else os.path.join(
LibraryUtil.get_res_dir(), 'word', 'spelling', 'spelling_en.txt')
self.include_reverse = include_reverse
self.model = self.get_model(force_reload=False)
def __init__(self,
name='Keyboard_Aug',
aug_char_min=1,
aug_char_max=10,
aug_char_p=0.3,
aug_word_p=0.3,
aug_word_min=1,
aug_word_max=10,
stopwords=None,
tokenizer=None,
reverse_tokenizer=None,
include_special_char=True,
include_numeric=True,
include_upper_case=True,
lang="en",
verbose=0,
stopwords_regex=None,
model_path=None,
min_char=4,
include_detail=False):
super().__init__(action=Action.SUBSTITUTE,
name=name,
min_char=min_char,
aug_char_min=aug_char_min,
aug_char_max=aug_char_max,
aug_char_p=aug_char_p,
aug_word_min=aug_word_min,
aug_word_max=aug_word_max,
aug_word_p=aug_word_p,
tokenizer=tokenizer,
reverse_tokenizer=reverse_tokenizer,
stopwords=stopwords,
device='cpu',
verbose=verbose,
stopwords_regex=stopwords_regex,
include_special_char=include_special_char,
include_detail=include_detail)
# TODO: support other type of keyboard
self.keyboard_type = 'qwerty'
self.include_special_char = include_special_char
self.include_numeric = include_numeric
self.include_upper_case = include_upper_case
self.include_lower_case = True
self.lang = lang
if model_path is None:
if lang not in ['en', 'th']:
raise ValueError(
'Only support en and th now. You may provide the keyboard mapping '
'such that we can support "{}"'.format(lang))
self.model_path = os.path.join(LibraryUtil.get_res_dir(), 'char',
'keyboard', lang + '.json')
else:
self.model_path = model_path
self.model = self.get_model(include_special_char, include_numeric,
include_upper_case, lang, self.model_path)
def __init__(self, name='Keyboard_Aug', aug_char_min=1, aug_char_max=10, aug_char_p=0.3,
aug_word_p=0.3, aug_word_min=1, aug_word_max=10, stopwords=None,
tokenizer=None, reverse_tokenizer=None, include_special_char=True, include_numeric=True,
include_upper_case=True, lang="en", verbose=0, stopwords_regex=None, model_path=None,
min_char=4):
super().__init__(
action=Action.SUBSTITUTE, name=name, min_char=min_char, aug_char_min=aug_char_min, aug_char_max=aug_char_max,
aug_char_p=aug_char_p, aug_word_min=aug_word_min, aug_word_max=aug_word_max, aug_word_p=aug_word_p,
tokenizer=tokenizer, reverse_tokenizer=reverse_tokenizer, stopwords=stopwords, device='cpu',
verbose=verbose, stopwords_regex=stopwords_regex, include_special_char=include_special_char,
include_detail=False)
# TODO: support other type of keyboard
self.keyboard_type = 'qwerty'
self.include_special_char = include_special_char
self.include_numeric = include_numeric
self.include_upper_case = include_upper_case
self.include_lower_case = True
self.lang = lang
if model_path is None:
lang_list = set(
map(
lambda file_name: file_name.replace(".json", ""),
os.listdir(
os.path.join(LibraryUtil.get_res_dir(), "char", "keyboard")
),
)
)
if lang not in lang_list:
raise ValueError(
"Only support en and th now. You may provide the keyboard mapping "
'such that we can support "{}"'.format(lang)
)
self.model_path = os.path.join(
LibraryUtil.get_res_dir(), "char", "keyboard", lang + ".json"
)
else:
self.model_path = model_path
self.model = self.get_model(include_special_char, include_numeric, include_upper_case, lang, self.model_path)
def get_model(cls, dict_of_path):
# Use default
if not dict_of_path:
default_path = os.path.join(LibraryUtil.get_res_dir(), 'char', 'ocr', 'en.json')
model = ReadUtil.read_json(default_path)
return nmc.Ocr(model=model)
# Use dict
if type(dict_of_path) is dict:
return nmc.Ocr(model=dict_of_path)
# Use json from file
model = ReadUtil.read_json(dict_of_path)
if not model:
raise ValueError('The dict_of_path does not exist. Please check "{}"'.format(dict_of_path))
return nmc.Ocr(model=model)