Testing of the credential form clutters up the full migration test. Let's make it standalone.
| Comment | File | Size | Author |
|---|---|---|---|
| #46 | 3143720-45.patch | 8.81 KB | quietone |
| #45 | interdiff-38-45.txt | 769 bytes | quietone |
| #45 | interdiff-38-45.txt | 769 bytes | quietone |
| #40 | reroll_diff_32-38.txt | 2.73 KB | spokje |
| #38 | 3143720-38.patch | 9.19 KB | spokje |
Comments
Comment #2
quietone commentedComment #4
quietone commentedThis moves the declaration of $destinationSiteVersion to MigrateUpgradeTestBase::setUp().
Comment #5
quietone commentedReroll since #3143719: Add a getCredentials helper method to migrate_drupal_ui functional tests was committed.
This confirms that benjifisher was correct in 3143719#5 to return an array.
Comment #7
quietone commentedRestore the line to get the legacy Drupal version.
Comment #8
quietone commentedPatch no longer applied, so I rerolled it.
Comment #10
quietone commentedI missed that the strings for the field paths had changed.
Comment #11
quietone commentedReroll because #3143717: Add helper methods to the migrate_drupal_ui functional tests was committed.
Comment #12
quietone commentedI think this can be removed and assertUpgrade can get the version number
Comment #13
quietone commentedFixes for #13.
Comment #14
quietone commenteds/public/protected for $modules.
Comment #15
mikelutzI think we need to return the correct types here, empty arrays for most of them, I think.
Comment #16
quietone commentedThis is suitable for a novice. A similar fix was done at https://www.drupal.org/project/drupal/issues/3143721#comment-13811695
Comment #17
ayushmishra206 commentedComment #18
ayushmishra206 commentedMade the changes suggested in #15. Thankyou for the help @quietone!
Comment #19
mikelutzRTBC pending tests.
Comment #21
ayushmishra206 commentedComment #22
ayushmishra206 commentedComment #23
ayushmishra206 commentedComment #24
spokjeReran test on #18, previous failure seemed to be random.
Comment #25
spokje@ayushmishra206 Sorry, had an old window open, resetting status to RTBC on #18 as per #19 by Mike Lutz
Comment #26
quietone commented@ayushmishra206, thanks for the patches! Please remember to upload an interdiff.
@Spokje, since there was a new patch, in #23, I don't see why the tests were rerun on the patch in #18. And, of course, not rerunning tests will save on resources.
I downloaded the patch in #14 and #18. An interdiff on those shows they are identical, that is, it does not contain the changes requested in #15.
The patches in #22 and #23 are also identical.
I reviewed the patch in #23 and found that is contains 3 of the 4 changes requested in #15. Just one more to fix!
This should be returning an empty array as per #15.
Comment #27
ayushmishra206 commentedComment #28
spokje@quietone, I started the rerun of the test on #18 when your comment #19 was the latest entry and the first testbot run on it seemed to be a random failure.
When the test was finished (this time successful, so it was a false positive) I didn't reload my open browser-tab and didn't notice the (identical, so not so) new patches. Hence my comment #25 which is now completely out of sync.
Besides the above: You're completely right (as usual :)
Comment #29
ayushmishra206 commentedI am sorry for creating this much of confusion. I have made the changes suggested in #26. Please review
Comment #30
ayushmishra206 commentedComment #31
quietone commented@ayushmishra206, thanks. Now, all the return statements are there and tests are passing. Unfortunately, there are coding standard errors, see https://www.drupal.org/pift-ci-job/1821801.
@Spokje, ah, I see. That makes sense.
Comment #32
sarvjeetsingh commentedfixed the coding standards error.
Comment #33
mikelutzOkay, looks right now.
Comment #35
quietone commentedSigh, random failure from Drupal\Tests\quickedit\FunctionalJavascript\QuickEditIntegrationTest. Retesting
Comment #36
mikelutzComment #37
quietone commentedComment #38
spokjeReroll of patch #32
Comment #39
quietone commented@Spokje, please provide a diff. It helps the reviewer. There are instructions for creating an interdiff.
Comment #40
spokje@quietone Sorry, had actually created an interdiff, but forgot to upload it. Here it is now.
Comment #41
spokjeI think since it's a reroll this can be RTBC per #33, but I've been bitten before after rerolling and RTBC-ing it myself, so I'll leave this on NR for now.
Comment #42
quietone commented@Spokje, thanks for the interdiff. I was just about to upload my reroll of the patch when I refreshed the page and saw that you had already done it. That was fast of you! And, yes, it is best to let someone else review a reroll and set it back to RTBC. I have seen exceptions to that but they are very rare and only for minor changes.
The patch looks good to me, but I can't RTBC it.
Comment #43
benjifisherAs far as git is concerned, this is a very routine reroll: a few lines of context were removed from MigrateUpgradeExecuteTestBase.php in #3143721: Create a separate SourceProviderTest, and that is the only difference between the patches in #32 and #38.
I still want to look a little closer to make sure that the patch makes sense in that context.
Comment #44
benjifisherI am glad I decided to take a closer look. There is nothing wrong with the reroll, but this hunk is redundant:
Thanks for cleaning up the tests. They have to be maintainable, too!
Comment #45
quietone commentedGoodness, how did I miss that before. Thanks benjifisher!
Comment #46
quietone commentedUploading the patch helps. (In my defense, I had to run to the kitchen).
Comment #47
benjifisherMmm, what's cooking?
Thanks for the updated patch. Back to RTBC.
Comment #49
catchCommitted 1e03b1c and pushed to 9.1.x. Thanks!