Details
Identity compiles and runs. Commtest passes, also played with it manually.
Diff Detail
- Repository
- rCOMM Comm
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
In the future we might consider refactoring parse_{some_business_logic_name}_attribute() too, but it's not a direct purpose of this task, so I replaced only these "universal" functions
services/identity/src/database.rs | ||
---|---|---|
1697 ↗ | (On Diff #34976) | clippy fix |
services/identity/src/database/device_list.rs | ||
380–393 ↗ | (On Diff #34976) | Refactoring this one blew my mind 🤯 |
services/identity/src/error.rs | ||
33–34 ↗ | (On Diff #34976) | This one is mainly used in code that is going to be removed anyway as a part of ENG-5842 so for now I deprecated it. |
services/identity/src/database.rs | ||
---|---|---|
1352–1359 ↗ | (On Diff #34976) | Ahh I realized I can do even better |
services/identity/src/database/device_list.rs | ||
38–42 ↗ | (On Diff #34976) | This enum turns out to be unused |
In the future we might consider refactoring parse_{some_business_logic_name}_attribute() too, but it's not a direct purpose of this task, so I replaced only these "universal" functions
Can we make a task for it? (or possibly even just do it in another diff, seems like it would be quick)
services/identity/src/database.rs | ||
---|---|---|
1148–1151 ↗ | (On Diff #34976) | Nit: does this work? |
1574 ↗ | (On Diff #34976) | Why are we removing parse_string_attribute but not this one? |
services/identity/src/database.rs | ||
---|---|---|
1148–1151 ↗ | (On Diff #34976) | Yep, missed this, thanks |
1574 ↗ | (On Diff #34976) | Still used in yet-to-be-removed code |