- Issue created by @larowlan
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
Had a shot at using PATCH for this, but not sure what impact it will have on subsequent AJAX requests.
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
Yep, doesn't work with AJAX forms because the form needs to be rebuilt and that requires the params - and they are no longer in the body of the request because now we're POSTing the Drupal form
That was far too easy
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
Ok, added a hidden field with those values to ensure it works with AJAX.
Yes these can be tampered, but so could the URL so π€·
Maybe we need an anti-tamper hash?
- πΊπΈUnited States effulgentsia
Is this for URLs where we're saving the submission to autosave storage? If so, GET is incorrect HTTP semantics, not only length constrained.
- π¦πΊAustralia larowlan π¦πΊπ.au GMT+10
This is for when we fetch the component inputs form
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
wim leers β made their first commit to this issueβs fork.
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
LGTM, just a minor remark about the changes in
inputBehaviors.tsx
(https://git.drupalcode.org/project/experience_builder/-/merge_requests/6...) being AFAICT avoidable and in a way that makes XB slightly more robust.I self-addressed that with a trivial commit, resulting in a tighter scope that make me comfortable merging this: https://git.drupalcode.org/project/experience_builder/-/merge_requests/6....
- π¬π§United Kingdom longwave UK
I was also concerned about this a while back but this is a neat solution and much simpler than I expected it to be! Two minor points, one that can wait until later, the other is just a naming issue, otherwise this looks great to me so leaving at RTBC.
-
wim leers β
committed f990f565 on 0.x authored by
larowlan β
Issue #3491215 by larowlan, wim leers, longwave: Pass model props to...
-
wim leers β
committed f990f565 on 0.x authored by
larowlan β
- π§πͺBelgium wim leers Ghent π§πͺπͺπΊ
Merged per https://git.drupalcode.org/project/experience_builder/-/merge_requests/6...
@longwave's concerns about special casing the
form_
-prefixed key-value pairs are accurate, but it's pre-existing. This doesn't make that future problem harder, but keeps it identical. Automatically closed - issue fixed for 2 weeks with no activity.