We do a bunch of work flipping an enabled check box to be stored as a disabled flag. Let's align the MenuLinkContent entity with others in core and just use a status field that's 1 for enabled, and 0 for disabled.
| Comment | File | Size | Author |
|---|---|---|---|
| #38 | 2305707-37.patch | 37.36 KB | dawehner |
| #34 | 2305707-32.patch | 37.35 KB | pwolanin |
| #32 | 2305707-32.patch | 37.35 KB | linl |
| #29 | increment.txt | 1.48 KB | pwolanin |
| #29 | 2305707-29.patch | 38.07 KB | pwolanin |
Comments
Comment #1
pwolanin commentedComment #2
pwolanin commentedComment #3
pwolanin commentedper YesCT, let's make the field name make sense with a boolean value, and using enabled means fewer form changes
Comment #4
pwolanin commentedStarting conversion - still some fails.
Comment #6
cilefen commentedThis should fix MenuTreeStorageTest.
Comment #7
cilefen commentedComment #8
cilefen commentedA few more fixes... but I have to stop for now.
Comment #11
cilefen commentedComment #13
cilefen commentedComment #15
pwolanin commentedre:
I think the test code wants you to use TRUE or FALSE.
Comment #16
cilefen commentedComment #17
dawehnerI wonder whether we considered to use status? At least this is what config entities and node use.
Comment #19
pwolanin commented@dawehner - YesCT argued that "status" is a confusing field name if we do not define constants like NODE_PUBLISHED
In contrast, "enabled" being 1 or 0 is evident without any constant - as is "hidden" currently.
Comment #20
cilefen commentedFrom at least #4 on there is a regression that displays the filter tips menu to anonymous users. It is this path:
Comment #21
cilefen commentedRenamed MenuTreeParameters::onyEnabledLinks to onlyEnabledLinks. I assume that was a typo.
Comment #23
pwolanin commented@cilefen - I'm confused about #20 - nothing in the patch should be affecting actual access
Comment #24
cilefen commented@pwolanin - yet there it is. Test it manually and see what I mean.
Comment #25
dawehnerFixed the failures as well as the problem mentioned by @cliefen.
On top of that I had a quick scan through the patch and could not find anything problematic.
Comment #26
pwolanin commentedre-roll for a moved file and fix the standard profile too (not sure why that's not causing a test fail - it used to)
Comment #29
pwolanin commentedSmall fixes, solve the test fails locally
Comment #30
dawehnerI just fixed a tiny little detail of the tests.
This is a huge improvement because having to revert the logic all over the place is just a bad idea in general and could lead
easily to problems.
Comment #31
alexpottNeeds a reroll
Comment #32
linl commentedRerolled.
Comment #34
pwolanin commentedHere's my reroll. fixed conflicts for 2 different core commits - form state object and the test change.
Comment #35
pwolanin commentedNo changes since #30 except resolving small conflicts, so I think this can be back to rtbc
Comment #37
cilefen commentedComment #38
dawehnerRerolled. FieldDefinition => baseFieldDefinition.
Comment #39
alexpottCommitted 4ae3ee3 and pushed to 8.0.x. Thanks!
Change test method name on commit.