Add WalletRead::get_seed_account

[nuttycom]

Builds atop #1258

[codecov]

Codecov Report

Attention: Patch coverage is 12.00000% with 22 lines in your changes are missing coverage. Please review.

Project coverage is 63.22%. Comparing base (5e810d3) to head (79f5bb4).

Files	Patch %	Lines
zcash_client_sqlite/src/wallet.rs	0.00%	18 Missing ⚠️
zcash_client_backend/src/data_api.rs	0.00%	2 Missing ⚠️
zcash_client_sqlite/src/lib.rs	0.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1259      +/-   ##
==========================================
- Coverage   63.32%   63.22%   -0.11%     
==========================================
  Files         121      121              
  Lines       13496    13519      +23     
==========================================
+ Hits         8546     8547       +1     
- Misses       4950     4972      +22     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[str4d]

Reviewed 1271f98

[str4d]

utACK 79f5bb4

[str4d]

I don't understand the full subtleties of UNIQUE INDEX vs CONSTRAIN UNIQUE and why you'd choose one over the other. AFAICT:

• They both produce an index.
• UNIQUE INDEX can be removed and added without altering the table (and I guess CREATE UNIQUE INDEX fails if the table contains non-unique data?)
• UNIQUE INDEX can include expressions
• CONSTRAIN UNIQUE has greater support in ON CONFLICT logic (whereas UNIQUE INDEX only supports INSERT ... ON CONFLICT I think).

[nuttycom]

Yeah, I was looking at this today. I think that the ON CONFLICT capabilities are actually the same, from what I was reading, at least for recent releases? Given that the separate indices can be dropped and added independently, I think I have a slight preference for them, but I don't think it makes much difference one way or the other.

[str4d]

Thanks for adding this; I was unaware that uniqueness constraints were starting to be embedded in the indices (which prior to now I only thought were present for performance). I rely heavily on changes to init.rs to track changes to the database structure and determine what constraints are applied.