I have a system_name field that is nullable that I want to validate for uniqueness if the input is not null.
As based on the recommended solutions given by others, it can just be
$this->validate($request, [
'system_name' => 'nullable|unique:systems'
]);
Why is it that this solution work? Given that when a null value is entered for system_name, shouldn't it pass the nullable rule but fail the unique rule as the null value is not unique since it has been used before?
Can I also enquire what is the difference between encapsulating the rule with '' vs [ ] ? As shown by the validation done on username and email
public function rules()
{
return [
'username' => 'required|min:8',
'email' => ['required', Rule::unique('users')]
];
}
CodePudding user response:
The documentation for the nullable
rule simply states:
The field under validation may be
null
.
What it doesn't state that if the field is null
, the subsequent validation is ignored. So in your case, the unique:systems
rule is not called. And from a logical standpoint, this makes sense, as a nullable
column/index should not treat null
as a unique value.1
Additionally, the difference between 'nullable|unique:systems'
and ['nullable', 'unique:systems']
(or other variants) is: nothing, really. In older versions of Laravel, the 'rule1|rule2'
syntax was the only method available. Compare the documentation for Laravel 5.4 and Laravel 6.x:
https://laravel.com/docs/5.4/validation#basic-usage https://laravel.com/docs/6.x/validation#basic-usage
The 6.x
version includes:
Alternatively, validation rules may be specified as arrays of rules instead of a single
|
delimited string
Nowadays, both syntaxes are acceptable, mainly to support older versions for backwards compatibility, and syntax preference. I like the |
delimited string, but the []
method is a bit more flexible (since you can use Rules and Functions as shown in some of the examples in the Documentation.)
1 Typically; that might not be true for all DB types, but from my experience this is the case.