[forms] disabled FormGroups include disabled controls in `.value`

Which @angular/* package(s) are the source of the bug?

forms

Is this a regression?

No

Description

It was working fine in Angular 11, Now it showing up disabled control in form.value

https://github.com/angular/angular/issues/36894 As per this disabled control get excluded from form.value if you want disabled control value then you should use getRawValue()

Please provide a link to a minimal reproduction of the bug

https://stackblitz.com/edit/angular-ws73gc-1fgccc?file=src%2Fapp%2Fparent%2Fchildren%2Fchildren.component.ts

Please provide the exception or error you saw

No response

Please provide the environment you discovered this bug in

Angular: 12.2.7

Anything else?

No response

[dylhunn]

As I described above, .value excludes disabled controls unless the entire group is disabled. I think we probably are not going to change this. It’s really weird, but at least it’s consistent, and I can understand the motivation – it would be even more confusing if .value returned the empty object! Plus, changing it would break the universe.

I think the “fix” for this is to use .getRawValue() instead of .value. This is a radical idea, but maybe we should deprecate .value. The exclusion of disabled controls isn’t terribly useful anyway. Closing for now though, since that’s a bigger question.

[dylhunn]

This is happening because when the entire FormGroup is disabled, calling .value does include disabled children.

Consider the following:

myForm = new FormGroup({
  a: new FormControl({ value: 'e', disabled: false }),
  b: new FormControl({ value: 'd', disabled: true }),
});
console.log(this.myForm.value);

This will log only the value of a. However, setting a to disabled: true will cause the entire containing FormGroup to become disabled:

myForm = new FormGroup({
  a: new FormControl({ value: 'e', disabled: true }),
  b: new FormControl({ value: 'd', disabled: true }),
});
console.log(this.myForm.value);

and now this prints both a and b.

This is quite counter-intuitive, and perhaps we should change it, but I don’t think it’s a regression.