Reputation: 99
The following code does not pluck the name column of the selected user
record. Rather, returns the entire row. Before I make a re-creatable example: Is this the expected behaviour here?
I want to explicitly select columns across joins to reduce my JSON payload size, and to return a nested model hierarchy to my clients.
I should add that I'm experiencing the same behaviour when using the pluck()
function as well, on the same line. Perhaps I've done something wrong.
There's tons of examples showing this approach with earlier versions of Laravel. Version 6 may have broken this.
$query = Post::whereHas('user.address', function ($query) use ($lat, $lon, $distance) {
$query->distance($lat, $lon, $distance);
})->with([
'user' => function ($query) {
$query->select('name'); // TODO: Report this bug. I've also tried pluck()
},
'user.address' => function ($query) use ($lat, $lon, $distance) {
$query->distance($lat, $lon, $distance);
},
'user.address.city',
'bids' => function ($query) {
$query->orderBy('amount', 'DESC');
},
'bids.user',
'images',
]);
Upvotes: 0
Views: 195
Reputation: 7637
pluck()
is a collection method, it executes the query and returns a simple Collection object of the field you specify.
Using pluck()
inside your subquery builder executes it (returning nothing, because you are assigning it to nothing) while the $query
variable is unmodified and behaves as normal returning all columns.
If you were to dump the value of the pluck()
inside this query, you would see it is an array of just names, and because of that, it has no affect on the query itself.
'user' => function ($query) {
dd($query->pluck('name'));
}
select()
should work fine in this case. You just need to also provide the relationship key or else it will just return a null object.
'user' => function ($query) {
$query->select(['id', 'name']);
},
Upvotes: 2