| Summary: | Update the output format for run-api-tests | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Sam Weinig <sam> | ||||||
| Component: | New Bugs | Assignee: | Sam Weinig <sam> | ||||||
| Status: | RESOLVED FIXED | ||||||||
| Severity: | Normal | ||||||||
| Priority: | P2 | ||||||||
| Version: | 528+ (Nightly build) | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Attachments: |
|
||||||||
|
Description
Sam Weinig
2015-01-09 21:23:48 PST
Created attachment 244403 [details]
Patch
Created attachment 244404 [details]
Sample Output
Attaching a screenshot of what the new output looks like, including one simulated failure.
Comment on attachment 244403 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=244403&action=review > Tools/ChangeLog:9 > + indented based on suite, as we were no really using suite very well) but Typo: “no really” > Tools/ChangeLog:20 > + Augment the custom gtest printer by replacing the tokens "**PASS**" and "**FAIL**" > + with colorized variants and strip out leaks spew (the leaks can be added back using > + the new --show-leaks argument). Are the colors going to be a problem when looking at things like buildbot output? > Tools/Scripts/run-api-tests:185 > + Why? (In reply to comment #3) > Comment on attachment 244403 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=244403&action=review > > > Tools/ChangeLog:9 > > + indented based on suite, as we were no really using suite very well) but > > Typo: “no really” > > > Tools/ChangeLog:20 > > + Augment the custom gtest printer by replacing the tokens "**PASS**" and "**FAIL**" > > + with colorized variants and strip out leaks spew (the leaks can be added back using > > + the new --show-leaks argument). > > Are the colors going to be a problem when looking at things like buildbot > output? I don't know. But if they are, I will add an option to remove them. > > > Tools/Scripts/run-api-tests:185 > > + > > Why? open3 is weird. If you pass 0 handle to the stderr parameter it will send stderr to the stdout handle, which is what I want in the verbose or leak printing case. In the non-verbose case, we can send stderr to devnull Committed r178236: <http://trac.webkit.org/changeset/178236> |