Revert "Revert "Update selenium-webdriver gem to 4.2.0""
#51960
Conversation
Hamms
added
the
Ruby Update
| @@ -41,7 +41,7 @@ def create_response(code, body, content_type) | |||
| if (msg = exception.message.match(/unexpected response, code=(?<code>\d+).*\n(?<error>.*)/)) | |||
| error = msg[:error] | |||
| error = JSON.parse(error)['value']['error'] rescue error | |||
| exception.message.replace("Error #{msg[:code]}: #{error}") | |||
| raise exception, "Error #{msg[:code]}: #{error}", exception.backtrace | |||
There was a problem hiding this comment.
I definitely like this approach better than mutating the exception.
There was a problem hiding this comment.
Likewise! Shout out to @snickell, who actually came up with this before I copied it out of his WIP PR. I assume it was in response to the fact that the message string is now frozen internally by Selenium and throws an error when we try to modify it, but I agree that overall it's much cleaner this way. Also, I learned that you can pass the backtrace manually as an optional third argument to Kernel#raise!
There was a problem hiding this comment.
Looks good!
Wish [test all browsers] passed on drone :-P, good workaround. For my future ref, here's the drone run for test all browsers: https://drone.cdn-code.org/code-dot-org/code-dot-org/29313/2/5
Reverts #51958, reapplying #51635. The original attempt consistently failed UI tests on iPhone and iPad when trying to interact with dropdowns, because we were trying to apply a desktop-Safari-specific workaround when executing in mobile Safari.
In Selenium 3,
@browser.browseris:Safarion desktop Safari and:safarion mobile. In Selenium 4, it's:safarion both, so in addition to browser detection I explicitly inspect browser capabilities (as we do here) to determine whether or not to apply the workaround.Testing Story
Verified locally that the tests pass with the new modifications; also ran with
[test all browsers]and verified that the only tests which failed in that run did not fail during the actual DTT.