#!/usr/bin/env python
#
# Create vector embeddings of articles.

import argparse
import pandas as pd
from wikivector import vector


def main(map_file, text_dir, h5_file):
    df = pd.read_csv(map_file)
    items = df['item'].tolist()
    vectors = vector.embed_items_use(items, text_dir)
    vector.save_vectors(vectors, items, h5_file)


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('map_file', help="Path to item map file")
    parser.add_argument('text_dir', help="Path to directory with text files")
    parser.add_argument('h5_file', help="Path to output hdf5 file")
    args = parser.parse_args()
    main(args.map_file, args.text_dir, args.h5_file)
