tests for #135 #241
tests for #135 #241
Conversation
src/protocol/context.rs
Outdated
| } | ||
|
|
||
| #[tokio::test] | ||
| pub async fn narrow() -> Result<(), BoxError> { |
There was a problem hiding this comment.
it seems that you don't need those tests to be async?
There was a problem hiding this comment.
I tried changing them to regular tests, but there is a call to tokio::spawn within InMemoryNetwork::new within make_world.
andyleiserson
force-pushed
the
issue135-tests
branch
from
471dd9c
to
273ce34
Compare
| // // do something *without* narrowing context by `i` or using it as RecordId | ||
| // } | ||
| // ``` | ||
| #[ignore] // broken, see #135 |
There was a problem hiding this comment.
I'm not sure what the material difference is between this test and the previous one. What am I missing?
There was a problem hiding this comment.
The idea is that, in the specific case where i is a record ID, narrowing a context twice in this way should not be an error. It is okay to construct the same context path multiple times, as long as you're doing it for different record IDs.
In the test case it would be possible to avoid the issue by moving the call to narrow outside the loop, but in general the body of the for record_id loop can be spread across multiple functions and it may not be desirable to pull the calls to narrow all the way up. (main...andyleiserson:ipa:issue135 shows what that might look like.)
|
No longer applicable after #253. |
Some tests I wrote for #135. Haven't settled on a strategy yet.