Add: test-virtual (related issue #71) #72
Conversation
taqtiqa-mark
changed the title
This initial commit contains the state of play immediately after add the example crates: - test-project - test-procmacro-project Other than setting up the workspace references no changes to the example crates have been made. The integration test harness has not yet been implemented. Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Running:
cargo test workspace-tests
correctly produces:
Running src/main.rs (target/debug/deps/workspace_tests-8bdd65c7ce129a54)
Setup
Teardown
Signed-off-by: Mark Van de Vyver <[email protected]>
Failing tests remain unchanged:
failures:
pass
pass_args
pass_expect_expanded
pass_expect_expanded_args
pr61
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Correctly produces: Setup Running basic test Teardown Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
Signed-off-by: Mark Van de Vyver <[email protected]>
…rams Signed-off-by: Mark Van de Vyver <[email protected]>
|
Just to clarify the state of play here. Previously this was motivated by:
However, with some experience, that seemed an overly complicated approach (check out the current state of the fluvio integration test code...). Currently, one can get a generic and elegant template using DJB's redo build system, and following a couple of conventions. |
|
All related issues have been closed, but is there still a problem with your use case? |
|
Setting up a integration test env for a virtual workspace, using the hierarchy of crates that proc-macros necessitate, is a non trivial exercise. So I think there is some value adding this template to the two existing templates is worth while. As you have observed first hand, it exposed one bug, and its presence should guard against regressions and other issues. Understand if you disagree and wish to close this as out of scope. Otherwise, I'll ping you when I believe it is ready for review? |
|
Sorry for the late response. Most of the code being added here seems to be larger and more complex than necessary both as examples and as test cases, so I would prefer not to merge this PR as is, but I would be happy to accept a PR that adds smaller, simpler test cases. |
|
I'll have a look at what can be rationalized. I think you are correct, we are likely looking at different use cases. I'm not a FT programmer and find it more efficient to delete than add. Your managing a project for programmers. I'll see where I land and will ping you when there is something to look at. Or will close this if I can't make substantial simplifications. |
This initial commit contains the state of play immediately after
add the example crates:
Other than setting up the workspace references no changes
to the example crates have been made.
The integration test harness has not yet been implemented.
Signed-off-by: Mark Van de Vyver [email protected]