setup_data.py

"""
setup_data.py
-------------
Downloads and extracts the Kaggle dataset automatically.
Requires kaggle API credentials (~/.kaggle/kaggle.json).

Get your token from: https://www.kaggle.com/settings -> API -> Create New Token
Place kaggle.json at C:\\Users\\<you>\\.kaggle\\kaggle.json
"""

import os
import zipfile
import subprocess
import sys


DATASET = "clmentbisaillon/fake-and-real-news-dataset"
DATA_DIR = os.path.join(os.path.dirname(__file__), "data")


def download():
    os.makedirs(DATA_DIR, exist_ok=True)

    print("Downloading dataset from Kaggle…")
    result = subprocess.run(
        [sys.executable, "-m", "kaggle", "datasets", "download",
         "-d", DATASET, "-p", DATA_DIR],
        capture_output=True, text=True
    )

    if result.returncode != 0:
        print("ERROR:", result.stderr)
        print("\nMake sure kaggle is installed and your API key is set up.")
        print("  pip install kaggle")
        print("  Place kaggle.json at ~/.kaggle/kaggle.json")
        sys.exit(1)

    print(result.stdout)

    # Unzip
    zip_path = os.path.join(DATA_DIR, "fake-and-real-news-dataset.zip")
    if os.path.exists(zip_path):
        print("Extracting…")
        with zipfile.ZipFile(zip_path, "r") as z:
            z.extractall(DATA_DIR)
        os.remove(zip_path)
        print(f"Done. Files saved to {DATA_DIR}/")
    else:
        print("Zip not found — check if download succeeded.")


if __name__ == "__main__":
    download()