I have a Laravel Controller method with the following code:
$listings = LikedListing::where('user_id', '=', $user->id)
->with('Listing.Photos')
->get();
This should return a collection of LikedListing records with Photos attached to each likedlisting record.
I have this SQL Query I need to inject into each record as well:
select u.id, l.address, u.first_name, u.last_name, ll.score
from listings l
left join liked_listings ll on ll.listing_id = l.id
left join users u on u.id = ll.user_id
where u.id in (
select secondary_user
from user_relationships
where primary_user = $primaryUser
)
and ll.listing_id = $listingId
Where $primaryUser = $user->id And $listingId is equal to the listingid inside each record in the collection.
I have absolutely no idea how to do this.
Maybe theres a model way of performing this? UserRelationship model has a primary_user column, which connects to a $user->id, and there is a secondary_user column, which acts like a follower userid, which is what we need in the final result (a list of all related users per listing)`
Can someone who has much far superior knowledge with Laravel please assist
My goal is to have the current collection of listing records with associated photos as well as following users (secondary_user) from the user_relationship table related to the primary_user (the logged in user) who have a record using user_id with the secondary_user value for that listing in the likedlisting table (obv assoicated with the listing_id). I already provided a raw sql query if thats the only option.
So in simple terms all related users who have liked a listing that the primary user has liked as well should be added to each listing record
Related
Trying to figure out how to fetch two related models(obviously united) of my Many To Many(Polymorphic) relationship.
What we have:
3 models: Bucket, Template and DesignPack.
Bucket has Many-To-Many(Polymorphic) relationship with Template and DesignPack(It means we have pivot table bucketables).In essence Bucket can have(be related with) both: Template and DesignPack.
Laravel 6.*
What I want to get:
I want to get a Bucket templates and design packs united in one collection and paginated!
Please check one of the solutions I've tried:
$templates = Bucket::find($bucket_id)->templates()->select(['id', 'file_name as name', 'size', 'preview']);
$design_packs = Bucket::find($bucket_id)->dps()->select(['id', 'name', 'size', 'preview']);
$all = $templates ->union($design_packs )->paginate(10);
Unfortunately that solution throws me the error(thought I checked what each request returns and it returns the same fields, not different):
"SQLSTATE[21000]: Cardinality violation: 1222 The used SELECT statements have a different number of columns (SQL: (select `id`, `size`, `preview`, `bucketables`.`bucket_id` as `pivot_bucket_id`, `bucketables`.`bucketable_id` as `pivot_bucketable_id`, `bucketables`.`bucketable_type` as `pivot_bucketable_type` from `design_packs` inner join `bucketables` on `design_packs`.`id` = `bucketables`.`bucketable_id` where `bucketables`.`bucket_id` = 3 and `bucketables`.`bucketable_type` = App\DesignPack and `design_packs`.`deleted_at` is null) union (select `id`, `size`, `preview` from `templates` inner join `bucketables` on `templates`.`id` = `bucketables`.`bucketable_id` where `bucketables`.`bucket_id` = 3 and `bucketables`.`bucketable_type` = App\Template and `templates`.`deleted_at` is null))"
Are there any different way to get what I want?
May be examples, documentation links or any helpful ideas?
Will be so grateful guys for any help!
Thank you!
You can pass closure to queries:
$templates = Bucket::whereHas('templates', function($query) use $bucket_id {
$query->where('bucket_id', $bucket_id);
})->get();
$designPacks = Bucket::whereHas('dps', function($query) use $bucket_id {
$query->where('bucket_id', $bucket_id);
})->get();
then merge 2 eloquent collections:
$mergedCollections = $templates->merge($designPacks);
now you have a collection of both results, you can select specific fields, limit the results or etc. you may want take a look at Laravel collection helpers.
also if you insist to use the union, you may want to take a look at this treat:
The used SELECT statements have a different number of columns (REDUX!!)
I have the following which brings back all users in a group along with their posts.
$group = Group::where('id', $id)->with('users.posts')->firstOrFail();
However, what I need is an additional join on the users to bring back additional (hasMany) information.
What I want is something like this (although this doesn't work)
$group = Group::where('id', $id)->with('users.posts,houses')->firstOrFail();
The sql would look something like
SELECT * FROM groups
JOIN group_users ON groups.id = group_users.group_id
JOIN users ON users.id = group_users.user_id
JOIN posts ON posts.user_id = users.id
JOIN house_users ON house_users.user_id = users.id
JOIN houses ON houses.id = house_users.house_id
WHERE groups.id = 123
If you pass a single argument to with(), it will look for a relationship with a matching name. Using a single string with a comma won't work as it won't parse and respect it. Since you're trying to use multiple relationships, this needs to be multiple signature, which there are a couple ways to accomplish.
First, array syntax:
->with(["users.posts", "houses"])
Second, multiple arguments:
->with("users.posts", "houses")
Either method will specify that you want multiple relationships loaded to your initial query; preference is given to whichever you find easier to read.
I've got a many to many relationship between a student and an institution_contact.
students should only ever have two institution_contacts and I have an attribute on the pivot table named type to be set as 1 or 2.
So, my pivot table looks like this:
institution_contact_student: id, institution_contact_id, student_id, type
I've run into difficulty in deciding how to approach the issue of adding/updating the pivot table. Let's say I have 100 students and I want to assign them a contact with the type of 1.
My current solution is to delete the contact then add it:
$students = Student::all(); // the 100 students
$contactId = InstitutionContact::first()->id; // the contact
foreach ($students as $student) {
// remove existing contact
$student
->institutionContacts()
->newPivotStatement()
->where('type', 1)
->delete();
// add new contact
$student
->institutionContacts()
->attach([$contactId => ['type' => 1]]);
}
However, I'm thinking that this is going to hit the database twice for each student, right? So would I be better off creating a model for the pivot table and removing all entries that matched the student id and the type then simply adding the new ones? Or would creating a model for the pivot table be considered bad practice and is there a better way of accomplishing this that I've missed?
Please note the reason I'm not using sync is because I'm relying on the type attribute to maintain only two contacts per student. I'm not aware of a way to modify an existing pivot without causing issues to my two contacts per student requirement.
Edit:
Instead of creating a model I could run the following code to perform the delete using DB.
DB::table('institution_contact_student') // the pivot table
->whereIn('student_id', $studentIds)
->where('type', 1)
->delete();
If I have understood your question correctly then you can use the updateExistingPivot method for updating your pivot table.But first of course you have to define the pivot in your relationship. For instance,
public function institutionContacts(){
return $this->belongsToMany('institutionContact')->withPivot('type');
}
after this, all you have to do is use the following code:
$student
->institutionContacts()
->updateExistingPivot($contactId, ["type" => 1]);
Hope this helps.
Some help with many to many relationships in Laravel:
Using the example for roles and users - basically:
a table for all the roles
a table for the users
and table with user_id and role_id.
I want to add to the third table, eg Year. basically the pivot table will have user_id, role_id and year_id.
I want to be able to make a query to pull for example all users assigned a specific role in a specific year. Eg All users with role_id = 2, and year_id = 1.
Any help will be appreciated
Before answering, I would like to suggest you not to put year on database like this.
All your tables should have created_at and updated_at which should be enough for that.
To filter users like you want. You could do this:
// This queries all users that were assigned to 'admin' role within 2013.
User::join('role_users', 'role_users.user_id', '=', 'users.id')
->join('roles', 'roles.id', '=', 'role_users.role_id')
->where('roles.name', '=', 'admin')
->where(DB::raw('YEAR(role_users.created_at)', '=', '2013')
->get();
This example may not be the precise query you are looking for, but should be enough for you to come up with it.
The best way to achieve a three way relation with Eloquent is to create a model for the table representing this relation. Pivot tables is meant to be used for two way relations.
You could have then a table called roles_users_year which could have data related to this 3 way relation like a timestamp or whatever...
A very late answer to a very old question, but Laravel has supported additional intermediate (pivot) table columns of at least Laravel 5.1 judging from the documentation, which hasn't changed at least through Laravel 6.x.
You can describe these extra columns when defining your many-to-many relationship:
return $this->belongsToMany(Role::class)->withPivot('column1', 'column2');
or in your case, the below would also do the job:
return $this->belongsToMany(Role::class)->withTimestamps();
which you can then access via the pivot attribute on your model:
$user = User::find(1);
foreach ($user->roles as $role) {
echo $role->pivot->created_at;
}
Note that the pivot attribute is on the distant relationship model (a single Role) and not on the relationship itself.
To get all the Roles assigned to Users in any given year, you might create a special relationship:
// User.php
public function rolesInYear($year) {
return $this->belongsToMany(Role::class)
->wherePivot('created_at', '>=', Carbon::create($year))
->wherePivot('created_at', '<', Carbon::create($year + 1));
}
My data structure is set up this way
A user takes a number of modules
A module contains a number of courses
Here's how the relationship looks like:
How do I get a list of courses the user takes?
The query I have now is:
var courses = (from ClassEnrollment enrolment in entities.ClassEnrollment
where enrolment.UserID == UserID
join Module module in entities.Module
on enrolment.ModuleID equals module.ID
select module.Course
).ToList();
However, this doesn't result in a list of courses, but rather a list of list of courses.
How can I flatten this query to a list of distinct courses?
According to your data structure screenshot, you have a one-to-many relationship between the ClassEnrollment and Module, as well as navigational property called Module. You also have a many-to-many relationship between Module and Course, but the navigational property should be called Courses. Given your code, you want something like this:
var courses = entities.
ClassEnrollment.
Where(e => e.UserID == UserID).
SelectMany(e => e.Module.Courses).
ToList();
Your question, however, mentions a user: A user takes a number of modules, How do I get a list of courses the user takes?. I don't see any User entity anywhere else, though, so it would be nice if you could clarify. Are you using LINQ-to-SQL, btw?
Something like this:
var courses = from ClassEnrollment enrolment in entities.ClassEnrollment
from module in entities.Module
where enrolment.ModuleID equals module.ID && enrolment.UserID equals UserID
select module.Course
Use SelectMany.
You can use
courses.SelectMany(c => c);
In your query you don't need explicitly specify the type for the range variables
Or you can join course to the query
var query = from enrolment in entities.ClassEnrollment
join module in entities.Module on enrolment.ModuleID equals module.ID
join course in entities.Course on module.CourseID equals course.ID
where enrolment.UserID == UserID
select course;
var course = query.ToList();