def Pic_Book(x):
im1 = Image.open(r'C:\\Users\\user\\Desktop\\darlusail\\book.jpg')
Pic_cell=sheet.cell(x,17).value
Title_cell=str(sheet.cell(x,4).value)
draw = ImageDraw.Draw(im1)
font= ImageFont.truetype('arial.ttf', size=10) # desired size
txt = arabic_reshaper.reshape(Title_cell)# starting position of the message
message=get_display(txt)
color = 'rgb(0, 0, 0)' # black color
wrapper = textwrap3.TextWrapper(width=20) # draw the message on the background with text wrapper
word_list = wrapper.wrap(text=message)
caption_new = ''
ad = AlphabetDetector()
if (ad.is_arabic(txt)==True or ad.is_latin(txt)!=True):#checks if the text whatever it is has Arabic words
print(" the name contains arabic words")
for ii in reversed(word_list):
caption_new = caption_new + ii + '\n'
if(ad.is_latin(txt)==True):
print(" the name doesn't contains arabic words")
for ii in word_list:
caption_new = caption_new + ii + '\n'
print(caption_new)
font= ImageFont.truetype('arial.ttf', size=10)#define the font for the text
color = 'rgb(0, 0, 0)'
#w,h = draw.textsize(message, font=font)
w,h = draw.textsize(caption_new, font=font)
W,H = im1.size
#x,y = 0.5*(W-w),0.90*H-h
x,y=0.5*(W-w),0.5*H
draw.text((x,y), caption_new, fill=color,font=font)
size=563,285 #define the dimensions to crop in the next step
im1=im1.resize(size,Image.ANTIALIAS)
im2=im1.save(r"C:"+Pic_cell) #save the image with the new configurations
def process_word(word):
ad = AlphabetDetector()
if not ad.is_latin(word):
return ''
banned_fragments = [
'.html', 'http://', 'https://', '.jpg', '.svg', '.png'
]
for banned in banned_fragments:
if banned in word:
return ''
word = word.lower()
word = PorterStemmer().stem(word)
chars = list(punctuation)
chars = chars + ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
chars = chars + ['®']
for char in chars:
word = word.replace(char, '')
stop_words = list(stopwords.words('english'))
stop_words += ['aa', 'aaa', 'bb', 'bbb']
for stop in stop_words:
if stop == word:
return ''
if len(set(word)) == 1:
return ''
return word
def parse_text(bot, update):
input = update.message.text
ad = AlphabetDetector()
if ad.is_latin(input):
update.message.reply_text(
"Right now only cyrillic characters are supported :( \n"
"Enter a Kyrgyz text:")
else:
match_list = run_apertium_tagger(input, mode="text")
output_list, error_list = [], []
for word_match in match_list:
if not word_match:
continue
else:
word_match_1 = str(word_match[1]).replace("J", "Й")
word_match_1 = word_match_1.replace("j", "й")
if word_match[2] is not None:
output_list.append(word_match_1 + "*" + word_match[2] +
"*")
else:
output_list.append("_" + word_match_1 + "_")
error_list.append(word_match_1)
update.message.reply_text(" ".join(output_list),
parse_mode=ParseMode.MARKDOWN)
if error_list:
update.message.reply_text(
"⚠️ These words were not recognized by the parser ⚠️")
for word in error_list:
update.message.reply_text(word)
update.message.reply_text("*-----*", parse_mode=ParseMode.MARKDOWN)
def parse_input(bot, update, user_data):
input = update.message.text
ad = AlphabetDetector()
if ad.is_latin(input):
update.message.reply_text(
"Right now only cyrillic characters are supported :( \n"
"Enter a Kyrgyz word:")
else:
result = run_apertium_tagger(input)
if user_data.setdefault("stem", []):
user_data["stem"] = []
result_2 = str(result[2]).replace("J", "Й")
result_2 = result_2.replace("j", "й")
user_data['stem'].append(result_2.lower())
if result[3] is not None:
user_data['stem'].append(1)
user_data['stem'].append(result[3])
reply = result_2 + "*" + result[3] + "*"
update.message.reply_text(reply, parse_mode=ParseMode.MARKDOWN)
update.message.reply_text(
"Look up stem in the dictionary? \n"
"Press /find or enter next word to continue.")
else:
user_data['stem'].append(0)
update.message.reply_text(result_2)
update.message.reply_text(
"Parsing didn't work :( \n"
"Look up stem in the dictionary? \n"
"Press /find or enter next word to continue.")
def check_case(file):
lists = file.readlines()
ad = AlphabetDetector()
for word in lists:
if word.isalpha() and ad.is_latin(word) and word.islower() == False:
print(word)
return "Please repair the case to lowercase in " + file.name + "."
return "PASS"
def drop_cols_not_in_mapping(df, mapper):
ad = AlphabetDetector()
for header in list(df.columns):
if ad.is_hebrew(header) and clean_text(header) not in list(
mapper.keys()):
df = drop_col_if_exists(df, header)
if ad.is_latin(header) and header not in list(mapper.values()):
df = drop_col_if_exists(df, header)
return df
def _test_caps_to_camel_case(a, b):
ad = AlphabetDetector()
if type(a) is not str:
assert a == b
elif type(a) is str and not ad.is_latin(a):
assert a == b
elif type(a) is str and len(a) < 4 or a.upper() != a:
assert a == b
else:
assert len(a) == len(b)
assert a != b
def isLatin(string):
try:
ad = AlphabetDetector()
return ad.is_latin(string)
except:
return False
pre, ext = os.path.splitext(file)
with open(pre+".pdfdata", 'w') as target_file:
sentences = []
for paragraph in txt2paragraph(file):
paragraph=paragraph.replace('"',' ')
paragraph=paragraph.replace('”',' ')
paragraph=paragraph.replace('“',' ')
paragraph=paragraph.replace('•',' ')
paragraph=paragraph.replace('?','.')
paragraph=paragraph.replace('!','.')
paragraph=paragraph.replace(';',' ')
paragraph=paragraph.replace(':',' ')
paragraph=paragraph.replace("’","'")
paragraph=' '.join(paragraph.split())
temp = nltk.sent_tokenize(paragraph)
sentences.extend(temp)
if len(temp)>1 and temp[-1].endswith('.'):
paragraph = ""
for sentence in sentences:
sentence = filter_non_printable(sentence)
if ad.is_latin(sentence) and not re.search(r"([\s.!#?^@-\\\|\*,]+\w[\s.!#?^@-\\\|\*,]+)+\w[\s.!#?^@-\\\|\*,]+",sentence) and not re.search(r"\/\d+|\d+\/",sentence):
if paragraph!="":
paragraph+=" "
paragraph += sentence
if paragraph is not "":
target_file.write(paragraph + "\n")
# else:
# print(sentences)
sentences.clear()
os.remove(file)
