Add support for returning file list and multi-thread download

[Nik-V9]

Made minor changes to the download module to support return of entire file list.

Added an example multi-thread download script to enable quick download of entire dataset. The file also has a retry logic to enable download of failed zip files.

[Amigoshan]

Thanks for the PR! I found a few problems. Could you modify the interfaces so that they are consistent with other interfaces in the examples? I suggest that we have an interface like this:
ta.multi_thread_download(env = env,
difficulty = ['easy', 'hard'],
modality = ['image', 'depth'],
camera_name = ['lcam_front', 'lcam_right', 'lcam_back', 'lcam_left', 'lcam_top', 'lcam_bottom'],
unzip = True,
num_workers = 8)
In the current package, the data_root is already defined in the initialization function. Defining another one using argparse is confusing.
I don't understand why you added one return value targetfilelist to the download function. It is not used in the example file.

I think one better way is to add the multi-thread function into the downloader.py, using a similar interface as the existing download function. We can keep track of the failed files and restart the download for them.