| # nuse_modules/keyword_extractor.py | |
| import os | |
| import requests | |
| import json | |
| from models_initialization.mistral_registry import mistral_generate | |
| def extract_last_keywords(raw: str, max_keywords: int = 8) -> list[str]: | |
| segments = raw.strip().split("\n") | |
| for line in reversed(segments): | |
| line = line.strip() | |
| if line.lower().startswith("extract") or not line or len(line) < 10: | |
| continue | |
| if line.count(",") >= 2: | |
| parts = [kw.strip().strip('"') for kw in line.split(",") if kw.strip()] | |
| if all(len(p.split()) <= 3 for p in parts) and 1 <= len(parts) <= max_keywords: | |
| return parts | |
| return [] | |
| def keywords_extractor(question: str) -> list[str]: | |
| prompt = ( | |
| f"Extract the 3–6 most important keywords from the following question. " | |
| f"Return only the keywords, comma-separated (no explanations):\n\n" | |
| f"{question}" | |
| ) | |
| raw_output = mistral_generate(prompt, max_new_tokens=32) | |
| keywords = extract_last_keywords(raw_output) | |
| print("Raw extracted keywords:", raw_output) | |
| print("Parsed keywords:", keywords) | |
| return keywords | |