chifra chunks - could have a --generate option

[tjayrush]

for --pins that would simply generate the CIDs from the existing files -- not upload them to either Pinata or IPFS. Just generate the CID.

This would make re-generation easier if there is an error uploading.

Also, there's no way currently to know which pins on Pinata are being used, nor if hte entire dataset is even present on Pinata.