24 lines
1.0 KiB
Python
24 lines
1.0 KiB
Python
import re
|
|
|
|
from difflib import get_close_matches
|
|
from configs import AccountConfig
|
|
|
|
|
|
def word_straighten(word, ref_list, threshold=0.8):
|
|
matches = get_close_matches(word, ref_list, n=1, cutoff=threshold)
|
|
return matches[0] if matches else word
|
|
|
|
|
|
def category_finder(text, output_template="{kategori} {numara}"):
|
|
categories = AccountConfig.CATEGORIES
|
|
result = {category: [] for category in categories}
|
|
for category, patterns in categories.items():
|
|
words = re.split(r"\W+", text)
|
|
straighten_words = [word_straighten(word, patterns) for word in words]
|
|
straighten_text = " ".join(straighten_words)
|
|
pattern = r"(?:\b|\s|^)(?:" + "|".join(map(re.escape, patterns)) + r")(?:\s*|:|\-|\#)*(\d+)(?:\b|$)"
|
|
if founds_list := re.findall(pattern, straighten_text, re.IGNORECASE):
|
|
list_of_output = [output_template.format(kategori=category, numara=num) for num in founds_list]
|
|
result[category].extend([i for i in list_of_output if str(i).replace(" ", "")])
|
|
return result
|