Uploader: Check for existing files before beginning a new upload

See [discussion 28](https://github.com/orgs/CulverLab/discussions/28) for background information.

File hashes need to be checked against all SPARCd buckets (collections) on S3 to see if there's a match before uploading files. If a match is found, the upload destination will be the upload folder containing the file(s) already on S3.

If all files are uploaded, nothing changes and the upload is considered a success.

If a partial upload is detected then only the remaining files are uploaded (this is History resume if the History is found, otherwise a new History is created for this upload).

If there are no matches, the upload doesn't change and is placed in a new upload folder.

Start by locating the uploader entry point and the code responsible for S3 upload folders and History records; read discussion 28 for the background. Trace how file hashes and all SPARCd buckets can be accessed. Done means existing files are detected before upload, complete uploads are treated as successful, partial uploads resume correctly, and unmatched files follow the current new-upload path.

Written by the indexing model from the issue text.