src/title_embd/preprocessing.py

import pandas as pd
from src.configs.config import METADATA_FILE

def preprocess_titles():
    metadata = pd.read_csv(METADATA_FILE)
    titles = metadata["Nom du document"].tolist()
    links = metadata["Lien"].tolist()
    return [{"title": t, "link": l} for t, l in zip(titles, links)]