Closed (fixed)
Project:
Drupal Canvas
Version:
1.x-dev
Component:
Shape matching
Priority:
Normal
Category:
Feature request
Assigned:
Unassigned
Reporter:
Created:
9 Oct 2025 at 19:33 UTC
Updated:
26 Nov 2025 at 15:09 UTC
Jump to comment: Most recent, Most recent file


Comments
Comment #2
hooroomooComment #3
wim leersI think we should gather a list of things we think that should NEVER be available to map.
I think a clear example is all revision information, because thats' basically internal information. Or perhaps "author-only" information, if you will.
Comment #4
wim leersPinged @lauriii to confirm 😊
Comment #5
lauriiiLet's just omit
revision_log_message. The two other examples have valid use cases i.e. displaying the author who had edited the node and when it was edited. I checked again and didn't see any other values we would have to omit.Comment #6
wim leersThanks!
Comment #8
wim leersComment #9
wim leersI think @phenaproxima is well-positioned to review this given his recent involvement and his strong opinions on #3555300: Distinguish between prop source matchers (comprehensive, objective) and suggesters (ordered by importance, filtered by relevance: subjective) 👍
Comment #10
wim leersOops!
Comment #11
phenaproximaThe nature of the change here is straightforward but the code could maybe use a touch of a documentation and a small refactor for clarity.
Comment #12
phenaproximaComment #13
wim leersResponded to @phenaproxima's feedback.
But more importantly: @lauriii promised me he'll check tomorrow for other cases, such as "Default translation" for a user, and "User status" for a user (or a revision user):
(Those are
type: booleanmatches. "Default translation" is a boolean field on ALL translatable content entity types. "User status" is to convey whether the user is active or blocked.)Comment #14
lauriiiI've added couple more fields that we should consider as irrelevant.
User Status field you may want to use if you're building a content template for the user entity – so let's not get rid of that.
Comment #15
wim leersComment #16
wim leersPer
created a core feature request: #3557353: Allow specifying additional metadata on Typed Data field definitions (or perhaps data definitions).
Comment #17
phenaproximaWe could use the existing
DataDefinition::setSetting(). Despite not being an interface method, core relies on it in many, many places.Comment #18
wim leers#17: but only for actual settings for the given field type?
Comment #19
wim leersImplemented everything proposed by @lauriii 👍
Pivoting the implementation towards metadata-in-Typed-Data (asked by @lauriiii, explained by me on the MR why not feasible, disputed by @phenaproxima in #17, and that itself disputed by me in #18, for which I created a core feature request in #16) can happen in a follow-up issue.
This solves the problem at hand without presuming that we have sufficient data to bake in a formal API. I think we should first gain more real-world experience. IOW: hardcoded heuristics now, evolve to metadata-in-Typed Data later. With or without #3557353.
Comment #20
wim leersComment #21
phenaproxima+1 RTBC. The changes make sense to me and it's wise that these subjective decisions are confined to PropSourceSuggester.
Comment #22
wim leersComment #24
wim leersCypress job #1 is a known failure, see #3539693-24: inputBehaviorsCommon mutates the passed props, breaking the rules of React. So went ahead and merged.