refactor: use idiomatic snafu in prover_db_indexer

[JayWhite2357]

Rationale for this change

We should use snafu in the manner it was intended.

What changes are included in this PR?

• context is used instead of map_err
• fields of the error varients are named source rather than error

Are these changes tested?

This is a pure refactoring PR.

[github-actions]

1.70.0

Bug Fixes

• make on chain indexer keys saturate rather than default to (5f794de
• use finalized block in prover-db-indexer (50438ab

Features

• add helper that queries the client for s. (4fc3952
• add client native interface and extenality (d9c6c4d
• add generic config key-value store for offchain workers (117b845
• Document Snapshot Download Process (4c2608b
• register client externality to offchain worker (338c213
• use generic key value config for prover_db_indexing instead of storage/CLI based config (faff564

[tlovell-sxt]

Not against this. If we ever transition back to thiserror I suppose keeping map_err would be a slightly smaller diff? maybe? not a big deal

[JayWhite2357]

Not against this. If we ever transition back to thiserror I suppose keeping map_err would be a slightly smaller diff? maybe? not a big deal

I've thought about that. But, I don't know if we'll do that any time soon.