Angular org selector / fm-editor required field validation issues

[Bill Erickson]

Evergreen 3.3

First noted in bug #1825851. There are 2 related issues:

1. There is no way to mark an <eg-org-select> component as required and the fielmapper editor should be taught to pass the value.

2. <eg-org-select> components within a fieldmapper editor show as invalid when marked as required with no value present (assuming code for #1), but fail to invalidate the editor form. It's as if the editor form is not aware the org-select input is one of its inputs.

This means it's possible to click Save in the editor even though a required org-select contains no value and is visibly styled as invalid.

[Bill Erickson (berick)]

Branch implementing a fix for issue #1 above:

https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/berick/lp1838784-org-select-required

The fix for #2 will require building atop this branch.

[Remington Steed (rjs7)]

Another example of this: In Local Admin -> Address Alerts, you can click "Save" on a new alert with no org unit selected, and you still see a "Success" toast message. (The alert isn't created.) Seeing a false "Success" message is bad.

Bill, I tested your branch, and it does show the Org Unit field as required now (with a red left-border). But I'm still able to save without choosing an Org Unit (and get the "Success" message). So it still needs to be considered when enabling the Save button (wherever that happens).

[Remington Steed (rjs7)]

And now I see that is probably what you meant by #2. Okay, as you said, it looks like the fix for #1 is working.

[Tiffany Little (tslittle)]

Going to add the pullrequest tag so this can be tested for Feedback Fest next week.

[Tiffany Little (tslittle)]

Hi Bill, can you give this a rebase? Thanks!

[Bill Erickson (berick)]

Rebased:

https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/berick/lp1838784-org-select-required-v2

[Tiffany Little (tslittle)]

Well, on the plus side I can confirm that the fix for #1 does still work great. Unfortunately, #2 still seems to be a problem.

[Terran McCanna (tmccanna)]

Removing pullrequest as per comment 7

[Tiffany Little (tslittle)]

I think this is useful enough that I'm going to mark it for pullrequest for issue #1, and then we can open a new bug for issue #2.

[Galen Charlton (gmc)]

Noting that I've opened bug 2054233, which if successful would resolve both issues.

I do not share the opinion that a fix for #1 is enough as a stop-gap, as it could create a question in the user's mind regarding whether form validation is working at all.

[Terran McCanna (tmccanna)]

Marking needsdiscussion as per Galen's comment #10