Allow kagglehub.dataset_load to load specific file from nested directory from a dataset

[andrei-diaconescu]

Having a dataset with the following directory structure:

data
├── day_1
├── day_3
├── hour_1
├── hour_12
├── hour_4
├── hour_8
├── minute_15
├── minute_30
├── minute_5
├── monthly
└── weekly

Where each child directory has hundreds of .csv files, I want to be able to only download/load any single file from any single directory, without downloading/loading the whole dataset.

The approach would be:
weekly_candles = kagglehub.dataset_load(KaggleDatasetAdapter.PANDAS, "andreidiaconescu/binancepricedata", "weekly/BTCUSDT.csv")

to which you would get the following response:

KaggleApiHTTPError: 404 Client Error.
Resource not found at URL: https://www.kaggle.com/datasets/andreidiaconescu/binancepricedata/versions/3
The server reported the following issues: Not found
Please make sure you specified the correct resource identifiers.