feat(gnovm): Test Coverage Support #2616
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #2616 +/- ##
==========================================
- Coverage 60.90% 60.49% -0.42%
==========================================
Files 564 563 -1
Lines 75273 75068 -205
==========================================
- Hits 45848 45409 -439
- Misses 26055 26272 +217
- Partials 3370 3387 +17
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
….go and op_eval.go
|
For simplicity for reviewing I'll change this PR as ready for review. |
notJoon
requested review from
moul,
piux2,
deelawn and
thehowl
as code owners
|
@notJoon |
|
@zivkovicmilos It seems the problem is occurring when retrieving and updating the line information executed in the |
|
Hey @notJoon, thanks for covering this. If I get a chance, I'll try to get a review in. I just wanted to comment because I just added a bounty on the original issue: #1121 This has been a long time coming; we had set out the requirements for a small while before you published this PR; but we have been slow in getting the bounties published 🫠 The requirements are slightly different to what you implemented here in general, and reflect the general comment I'd make on the PR right now: I like the terminal visualization, but ultimately it doesn't matter. Terminal visualization should be the responsibility of other tools based on the eventual coverprofile. If we use coverprofiles properly, it means we can re-use the tools Go already has; and that's more important. So, I'd recommend you move forward more in the direction of ensuring compatibility in that direction and as specified in the requirements. Note that the bounties exclude AiB/NT employees from working on the bounty, but not grantees; so at your discretion, you can also claim the bounty. Thanks! |
|
@thehowl Looks like the bounty program is gonna start 😮 The CLI output was temporarily added to view during development, so it doesn't matter actually. Ultimately, the goal is to make it as similar as possible to Go's test coverage functionality. Originally, before the bounty was added, I was planning to split the work to reduce the PR size, but in this case, I wonder if all the work needs to be completed in a single PR? Or should I submit a separate work plan? There seems to be more information in the attached link, but I don't have permission to access it. Thank you! |
|
separate please! smaller PRs are better in general :) once the issue is fully addressed, we'll propose how we intend to distribute the bounty (mostly in case you're not the only once who contributed to it) :) |
|
apologies, misclicked. |
Good. I'll get this done as quickly as possible 👍 |
Description
closes #1121
Support testing coverage functionality into previous testing framework. This feature tracks and reports which parts of the code were executed during test runs.
Key Changes
CoverageDataandFileCoveragestructsPopOpinMachine.Run()methodrecordCoveragefunction for each statement to obtaining accurate coverage tracking resultAddHitmethodCLI Options
CLI Options
-covergno test <path> -cover-view=<pattern>gno test <path> -cover -view=demo/foo-show-hitsgno test <path> -cover -view=demo/foo -show-hits-html=<file>gno test <path> -cover -html=coverage.htmlAll coverage-related flags must be preceded by the
-coverflag to be active.Notes:
-viewoption supports partial file path patterns and can match multiple files.-view, the output is color-coded: green for executed lines, yellow for executable but not executed lines, and white for non-executable lines.-show-hitsoption adds an orange hit count next to each executed line when used with-view.