import mlcroissant as mlc
import pandas as pd
import os

# These values are used by mlcroissant
# to perform the necessary auth to fetch the data
# Provide your Kaggle username and API key
os.environ['CROISSANT_BASIC_AUTH_USERNAME'] = 'enzetao'
os.environ['CROISSANT_BASIC_AUTH_PASSWORD'] = "a97ddf556baf4db4723d78aab014b599"


# Fetch the Croissant JSON-LD
# croissant_dataset = mlc.Dataset('https://www.kaggle.com/datasets/yueyin27/refref/croissant/download')
# croissant_dataset = mlc.Dataset('https://www.kaggle.com/datasets/nguyenhung1903/nerf-synthetic-dataset/croissant/download')
# croissant_dataset = mlc.Dataset('https://www.kaggle.com/datasets/enzetao/refref/croissant/download')
# croissant_dataset = mlc.Dataset('https://www.kaggle.com/datasets/muratkokludataset/rice-image-dataset/croissant/download')
url = 'https://huggingface.co/api/datasets/fashion_mnist/croissant'

print(mlc.Dataset(url).metadata.to_json())
import tensorflow_datasets as tfds
builder = tfds.core.dataset_builders.CroissantBuilder(
    jsonld=url,
    record_set_ids=["fashion_mnist"],
    file_format='array_record',
)
builder.download_and_prepare()

# # Check what record sets are in the dataset
# record_sets = croissant_dataset.metadata.record_sets
# print(record_sets)

# # Fetch the records and put them in a DataFrame
# record_set_df = pd.DataFrame(croissant_dataset.records(record_set=record_sets[0].uuid))
# record_set_df.head()