feat(spanner): add spanner::Value support for TypeCode::UUID
/gcbrun
Codecov Report
:x: Patch coverage is 96.29630% with 2 lines in your changes missing coverage. Please review.
:white_check_mark: Project coverage is 92.92%. Comparing base (7595793) to head (967c7cb).
:warning: Report is 161 commits behind head on main.
| Files with missing lines | Patch % | Lines |
|---|---|---|
| google/cloud/spanner/value.cc | 87.50% | 2 Missing :warning: |
Additional details and impacted files
@@ Coverage Diff @@
## main #15076 +/- ##
=======================================
Coverage 92.92% 92.92%
=======================================
Files 2389 2389
Lines 214832 214886 +54
=======================================
+ Hits 199631 199693 +62
+ Misses 15201 15193 -8
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
:rocket: New features to boost your workflow:
- :snowflake: Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
Thanks again for the PR.
While the protos defining the UUID field are public, service support for the type isn't live yet. We'll merge this PR and write some integration tests when UUID is available in production.
While the protos defining the UUID field are public, service support for the type isn't live yet. We'll merge this PR and write some integration tests when UUID is available in production.
Okie dokie.
/gcbrun
While the protos defining the UUID field are public, service support for the type isn't live yet. We'll merge this PR and write some integration tests when UUID is available in production.
I would proffer that this should be merged irrespective of the rollout of service support. It is the server support that actually matters, not the conversion between types within the client.
Also, merging now would mean that there is no need to wait for a new client release when server support is enabled. And that means client-side development can start early, including writing the integration tests, which can also actually be run against the staging service.
Finally, it avoids the possibility of future merge conflicts.