feat(spanner): add samples for proto columns #13759
Conversation
Also add "singer_cc_proto" dependencies to the tests that use it directly, as would be required if strict checking was enabled.
product-auto-label
bot
added
api: spanner
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #13759 +/- ##
==========================================
- Coverage 93.23% 93.15% -0.09%
==========================================
Files 2233 2233
Lines 193351 193517 +166
==========================================
- Hits 180265 180263 -2
- Misses 13086 13254 +168 ☔ View full report in Codecov by Sentry. |
|
Here is the summary of changes. You are about to add 7 region tags.
This comment is generated by snippet-bot.
|
| @@ -989,6 +990,22 @@ void AddTimestampColumn( | |||
| } | |||
| // [END spanner_add_timestamp_column] | |||
|
|
|||
| //! [drop-column] | |||
There was a problem hiding this comment.
comment: It looked suspicious that this tag is not referenced by a @snippet. I assume the tag is there in case we Ctrl+F the drop-column banner.
There was a problem hiding this comment.
I think(?) we have a few such non-devrel, non-snippet tags that are indeed only there to match their sub-command name, or otherwise delineate a block of code.
In this particular case it was only introduced so that I could adapt the pre-existing table-creation sample to the prerequisites of the proto-column samples. If I recall correctly, the Java samples have "setup" and "teardown" code for each sample so that they are all independent, but we've evolved a model where we try to execute them in (some) sequence, so occasionally some between-sample adaptation is necessary.
| google::protobuf::FileDescriptorSet fds; | ||
| google::cloud::spanner::testing::SingerInfo::default_instance() | ||
| .GetMetadata() | ||
| .descriptor->file() |
There was a problem hiding this comment.
I assume we can assume this is never nullptr, or that if it is, something else has gone wrong.
There was a problem hiding this comment.
My recollection is that it can indeed never be null.
| if (!commit_result) throw std::move(commit_result).status(); | ||
| } | ||
| std::cout << "Update was successful " | ||
| << "[spanner_update_data_with_proto_message_column]\n"; |
There was a problem hiding this comment.
Here and below, is printing the [spanner_*] intentional?
Is there logic to when we print it and when we don't?
There was a problem hiding this comment.
Intentional, yes. As you can see, many do, so I usually try to keep doing that (although now that we have SampleBanner("tag"), it is probably less important). So, I'm afraid the only logic is: Is the "[tag]" even slightly helpful, and do you remember to add it.
| void UpdateDataWithProtoMessageColumn(google::cloud::spanner::Client client) { | ||
| google::cloud::spanner::testing::SingerInfo singer_proto; | ||
| singer_proto.set_singer_id(2); | ||
| singer_proto.set_birth_date("1942-06-18"); |
Also add "singer_cc_proto" dependencies to the tests that use it directly, as would be required if strict checking was enabled.
This change is