pipeline/vectorizer.py

import joblib
from sklearn.feature_extraction.text import TfidfVectorizer

VECTORIZER_PATH = "models/tfidf_vectorizer.pkl"

def fit_and_save_vectorizer(texts, max_features=5000):
    """Fits TF-IDF vectorizer and saves it to disk."""
    vectorizer = TfidfVectorizer(max_features=max_features)
    X_tfidf = vectorizer.fit_transform(texts)
    joblib.dump(vectorizer, VECTORIZER_PATH)
    return X_tfidf

def load_vectorizer():
    """Loads the TF-IDF vectorizer from disk."""
    return joblib.load(VECTORIZER_PATH)