def translate(input_phrase):
translator = Translator()
# detect language type of input phrase
detect_result = translator.detect(input_phrase)
language_type = detect_result.lang
# define data frame
trans_phrase_array = [''] * 5
colmuns_list = ['English', 'Japanese', 'Spanish', 'Chinese', 'Ch_Prns']
print('Language type: %s' % language_type)
# switch translation process depend on language type
if 'ja' in language_type: # Japanese
print('Input phrase is Japanese')
phrase_in_zh = to_chinese_simplified(translator, input_phrase)
trans_phrase_array[0] = to_english(translator, input_phrase).text
trans_phrase_array[1] = input_phrase
trans_phrase_array[2] = to_spanish(translator, input_phrase).text
trans_phrase_array[3] = phrase_in_zh.text
trans_phrase_array[4] = phrase_in_zh.pronunciation
elif 'zh-CN' in language_type: # Chinese(Simplified)
print('Input phrase is Simplified Chinese')
phrase_in_zh = to_chinese_simplified(translator, input_phrase)
trans_phrase_array[0] = to_english(translator, input_phrase).text
trans_phrase_array[1] = to_japanese(translator, input_phrase).text
trans_phrase_array[2] = to_spanish(translator, input_phrase).text
trans_phrase_array[3] = phrase_in_zh.text
trans_phrase_array[4] = phrase_in_zh.pronunciation
elif 'en' in language_type: # English
print('Input phrase is English')
phrase_in_zh = to_chinese_simplified(translator, input_phrase)
trans_phrase_array[0] = input_phrase
trans_phrase_array[1] = to_japanese(translator, input_phrase).text
trans_phrase_array[2] = to_spanish(translator, input_phrase).text
trans_phrase_array[3] = phrase_in_zh.text
trans_phrase_array[4] = phrase_in_zh.pronunciation
elif 'es' in language_type: # Spanish
print('Input phrase is Spanish')
phrase_in_zh = to_chinese_simplified(translator, input_phrase)
trans_phrase_array[0] = to_english(translator, input_phrase).text
trans_phrase_array[1] = to_japanese(translator, input_phrase).text
trans_phrase_array[2] = input_phrase
trans_phrase_array[3] = phrase_in_zh.text
trans_phrase_array[4] = phrase_in_zh.pronunciation
else:
print('Input phrase should be Japanese, English, Chinese or Spanish')
# write in csv
# df_trans_phrase = pd.DataFrame(index=[], columns=colmuns_list)
df_trans_phrase = pd.read_csv('phrase_translation_database.csv',
encoding='utf-8-sig',
index_col=0)
sr_trans_phrase = pd.Series(trans_phrase_array,
index=colmuns_list,
name=input_phrase)
df_trans_phrase = df_trans_phrase.append(sr_trans_phrase)
df_trans_phrase.to_csv('phrase_translation_database.csv',
encoding='utf-8-sig')
def detect_language_google(self, sample):
'''
EN: It uses Google Translate to detect the language of a given sample.
SP: Utiliza el Traductor de Google para detectar el idioma de una muestra dada.
:param sample: sample of text from which the language is detected / muestra de texto a partir de la cual detectar el idioma
:return: the detected language / el idioma detectado
'''
translator = Translator()
det = translator.detect(sample)
language_google = self.LANGUAGES_GOOGLE[det.lang]
return language_google
# Thai to English
translations = translator.translate(
text='ลมอ่อนพัดโชยมาน้ำตาก็ไหลรินเหลือเพียงกลิ่นหัวใจฟุ้งไปกับความเหงา',
src='th',
dest='en')
print('Original Text :')
print(translations.origin)
print('Translated Text :')
print(translations.text)
#%% language detection
print('----------- Language Detection --------------')
LangDetectObj = translator.detect('이 문장은 한글로 쓰여졌습니다.')
LangDetectObj = translator.detect('ไปกับพี่มั๊ยจ๊ะน้องสาว')
LangDetect = LangDetectObj.lang
LangConf = LangDetectObj.confidence
print('Detect language is : %s with confident : %.3f' % (LangDetect, LangConf))
#%% Translate from list
print('-----------Translator from list --------------')
translations = translator.translate(
['The quick brown fox', 'jumps over', 'the lazy dog'], dest='ko')
for translation in translations: # since the text is list, translations become list (iterable object)
print(translation.origin, ' -> ', translation.text)
class DataTranslation:
def __init__(self) -> None:
self.translator = Translator()
def make_translation(self, word: str) -> Union[bool, str]:
"""
Translate a word using GoogleTrans library
The language might be indicate in the 2 first character
:param word: string
:param language: string
:return: word string, or False if the word can't be translated
"""
if word == 'en:spreads':
# better translation for the word spreads
return 'Pate à tartiner'
elif word == 'en:sweets-spreads':
return 'Pâte à tartiner sucrée'
elif word == 'en:plant-based-spreads':
return 'Pâte à tartiner végétal'
elif word == ' ':
return False
else:
# take the language indication
language = self.get_language(word)
if not language:
return False
else:
# remove language indicator
word = self.slice_language(word)
# translate word to french
word = self.translator.translate(word, src=language,
dest='fr').text
# remove any type of punctuation
return word.replace('-', ' ')
def detect_lang(self, word: str) -> Union[bool, str]:
"""Detect language with googletrans"""
# detect language
tr = self.translator.detect(word)
if tr.confidence < 0.5:
return False
else:
for l in LANGCODES.values():
if tr.lang == l:
return tr.lang
return False
def slice_language(self, word: str) -> str:
if len(word) > 3 and word[2] == ':':
return word[3:]
return word
def get_language(self, word: str) -> str:
"""
Return the language of the word if its has a language indicator.
If its has not googletrans library will detect its language
"""
if len(self.slice_language(word)) == len(word):
# the word has no language indicator
lang = self.detect_lang(word)
else:
lang = word[:2]
tr = self.translator.detect(word)
# if language detected is different
# the confidence of the translator is above 90 %
# take the language from the translator
if tr.confidence > 0.9 and tr.lang != lang:
lang = tr.lang
return lang