Attempt to reuse previously materialized datasets

[mvdbeek]

The conditions are:

• the replacement dataset exists in the same object store
• the replacement dataset exists in a history that the user owns
• the replacement dataset has a HDA of the same datatype
• the dataset hash or dataset source hash matches
• the replacement dataset is not purged or deleted and in OK state
• the same transform is applied

Providing hashes is currently a bit of a niche thing, but we do use this
in BRC analytics (in particular all fastq files contain hashes).
This should help a lot with demo'ing things.

We should also be able to make use of this in planemo, where this would be a realistic path to enable "invocation resume" functionality.

We might eventually allow this for public datasets as well, but perhaps this should be a little more explicit. We could for instance include cache hints in the dataset request syntax (maybe something like cache_strategy: own, cache_strategy: public, cache_strategy: never) and a top level setting for the workflow request.

How to test the changes?

(Select all options that apply)

• I've included appropriate automated tests.
• This is a refactoring of components with existing test coverage.
• Instructions for manual testing are as follows:
  1. [add testing steps and prerequisites here if you didn't write automated tests covering all your changes]

License

• I agree to license these and all my past contributions to the core galaxy codebase under the MIT license.