antigravity-hackathon-backend / utils /language_detector.py
anique-1's picture
Deploy NadraGuard backend to Hugging Face Spaces
7f3e64e
Raw
History Blame Contribute Delete
696 Bytes
import re
def detect_language(text: str) -> str:
# Very basic language detection for the hackathon
# Checks for Urdu script characters
urdu_pattern = re.compile(r'[\u0600-\u06FF]')
if urdu_pattern.search(text):
return "Urdu"
# Check for common Roman Urdu patterns (very simplified)
roman_urdu_keywords = ["aap", "hai", "kya", "nhi", "ji", "ho", "raha", "kar"]
words = text.lower().split()
matches = [w for w in words if w in roman_urdu_keywords]
if len(matches) > 1:
return "Roman Urdu"
return "English"
def normalize_text(text: str) -> str:
# Normalize encoding and whitespace
return " ".join(text.split()).strip()