Limit amount of results using group by without affecting JSON output? - laravel

I'm trying to build a threaded comments system by grouping the parent_ids together and limit the results using taken.
Comment table
$table->increments('id');
$table->text('content');
$table->integer('post_id');
$table->integer('parent_id')->index()->nullable();
$table->string('username');
$table->string('user_image')->default('http://lorempixel.com/60/60/people/');
$table->timestamps();
When I don't use take() to limit the result the JSON outputs as expected by being groupedBy('parent_id')
Query without take()
$post->comments->groupBy('parent_id');
JSON Output example
Query with take()
$post->comments->take(5)->groupBy('parent_id')
When I use take() it changes the JSON output to no longer include the grouped by parent_id as keys.
JSON Output example
How do I limit results without having an effect on the JSON output?
edited
Post Controller
public function index(Post $post)
{
$comments = $post->comments->groupBy('parent_id');
return $comments;
}
Edit
Why were the other replies deleted from this thread?
Edit 2
Oddly enough this query works based on the limit I set. So if I set the limit to a lower limit like 5 I get the grouped by keys outputting normally and JSON. However, if I set the limit by 5 I don't get those keys. See the JSON outputs above:
$comments = DB::table('comments')
->where('post_id',1)
->orderByDesc('created_at')
->limit(7)
->get();
return collect($comments)->groupBy('parent_id');

This is caused by the way json_encode() handles arrays with numeric keys.
In the case without ->take(5), the array keys are [0, 1, 6, 9]. They are not sequential and so get encoded as a JSON object.
In the case with ->take(5), the array keys are [0, 1]. These are sequential and so get encoded as a JSON array.
You can solve this by using null instead of 0 for parentless comments. Using null to represent non-existent data is also a better solution in general.

Related

How I get data like 0 to 50 and 51 to 100 in laravel

I want to get data from database with range like 0 to 50 and 51 to 100, this is possible? if possible please talk me how to do that.
$users = User::all();
I am not able to get clearly what do you want. But I have provided some examples below that might work for you.
$users = DB::table('users')
->whereBetween('id', [1, 50])
->get();
Also try skip and take.
You may use the skip and take methods to limit the number of results returned from the query or to skip a given number of results in the query:
$users = DB::table('users')->skip(50)->take(50)->get();
Alternatively, you may use the limit and offset methods. These methods are functionally equivalent to the take and skip methods, respectively:
$users = DB::table('users')
->offset(50)
->limit(50)
->get();
Hope this helps!
There are a number of options available to you. If you're looking to obtain a working data set of x results at a time and not concerned about the id of the results, then one of the following might be of use:
Pagination
Chunking results
Both of the above will allow you to define the size/number of records returned for your data set.
If you're looking to return records between specific id values, then you'll want to use something like whereBetween on your database queries. The whereBetween clause can be used in conjunction with the pagination and chunking.

Get last few rows in ascending order in Laravel

What is the best way (Laravel way) to get last few rows of a database table in ascending order? I want the result to be an array of objects as like:
[{"name1":"value1"}, {"name2":"value2"}]
My recent code is like:
$users = User::take(2)->latest()->get()->reverse();
return $users;
Output looks like:
{"1":{"id":582, "name":"name1"}, "0":{"id":583, "name":"name2"}}
But I want the output to be like:
[{"id":582, "name":"name1"}, {"id":583, "name":"name2"}]
When you use the reverse() method, it modifies the Collection, but doesn't re-index it, meaning the Collection is now [1, 0]. JS/JSON treats this as an Object instead of an array, hence the difference:
{"1":{"id":582, "name":"name1"}, "0":{"id":583, "name":"name2"}}
// VS
[{"id":582, "name":"name1"}, {"id":583, "name":"name2"}]
To solve this, use the values() method to re-index the array before return:
$users = User::take(2)->latest()->get()->reverse();
return $users->values();
// [{"id":582, "name":"name1"}, {"id":583, "name":"name2"}]
If you want them in the other order, don't use reverse(), and you shouldn't need values() either:
$users = User::take(2)->latest()->get();
return $users;
// [{"id":583, "name":"name2"}, {"id":582, "name":"name1"}]

Is Laravel sortBy slower or heavier to run than orderBy?

My concern is that while orderBy is applied to the query, I'm not sure how the sortBy is applied?
The reason for using sortBy in my case is because I get the collection via the model (i.e. $user->houses->sortBy('created_at')).
I'm just concerned about the performance: is sortBy simply looping each object and sorting them?, or is Laravel smart enough to simply transform the sortBy into an orderBy executed within the original query?
You need orderBy in order to perform a SQL order.
$user->houses()->orderBy('created_at')->get()
You can also eager load the houses in the right order to avoid N+1 queries.
$users = User::with(['houses' => function ($query) {
return $query->orderBy('created_at');
}])->get();
$orderedHouses = $users->first()->houses;
The sortBy method is applied to the Collection so indeed, it will looping each objects.
The orderBy() method is much more efficient than the sortBy() method when querying databases of a non-trivial size / at least 1000+ rows. This is because the orderBy() method is essentially planning out an SQL query that has not yet run whereas the sortBy() method will sort the result of a query.
For reference, it is important to understand the difference between a Collection object and a Builder object in Laravel.
A builder object is, essentially, an SQL query that has not been run. In contrast, a collection is essentially an array with some extra functionality/methods added. Sorting an array is much less efficient than pulling the data from the DB in the correct format on the actual query.
example code :
<?php
// Plan out a query to retrieve the posts alphabetized Z-A
// This is still a query and has not actually run
$posts = Posts::select('id', 'created_at', 'title')->orderBy('title', 'desc');
// Now the query has actually run. $posts is now a collection.
$posts = $posts->get();
// If you want to then sort this collection object to be ordered by the created_at
timestamp, you *could* do this.
// This will run quickly with a small number or rows in the result,
// but will be essentially unusable/so slow that your server will throw 500 errors
// if the collection contains hundreds or thousands or objects.
$posts = $posts->sortBy('created_at');

laravel database query Does `where` always need `first()`?

I am new to laravel and confused about some query methods.
find($id) is useful and returns a nice array, but sometimes I need to select by other fields rather than id.
The Laravel document said I could use where('field', '=', 'value') and return a bunch of data, which is fine.
What I can't understand is why I need to add ->first() every time, even if I am pretty sure there is only one single row matches the query.
It goes like this:
$query->where(..)->orderBy(..)->limit(..) etc.
// you can chain the methods as you like, and finally you need one of:
->get($columns); // returns Eloquent Collection of Models or array of stdObjects
->first($columns); // returns single row (Eloquent Model or stdClass)
->find($id); // returns single row (Eloquent Model or stdClass)
->find($ids); // returns Eloquent Collection
// those are examples, there are many more like firstOrFail, findMany etc, check the api
$columns is an array of fields to retrieve, default array('*')
$id is a single primary key value
$ids is an array of PKs, this works in find method only for Eloquent Builder
// or aggregate functions:
->count()
->avg()
->aggregate()
// just examples here too
So the method depends on what you want to retrieve (array/collection or single object)
Also the return objects depend on the builder you are using (Eloquent Builder or Query Builder):
User::get(); // Eloquent Colleciton
DB::table('users')->get(); // array of stdObjects
even if I am pretty sure there is only one single row matches the query.
Well Laravel cant read your mind - so you need to tell it what you want to do.
You can do either
User::where('field', '=', 'value')->get()
Which will return all objects that match that search. Sometimes it might be one, but sometimes it might be 2 or 3...
If you are sure there is only one (or you only want the first) you can do
User::where('field', '=', 'value')->first()
get() returns an array of objects (multiple rows)
while
first() returns a single object (a row)
You can of course use get() when you know it will return only one row, but you need to keep that in mind when addressing the result:
using get()
$rez = \DB::table('table')->where('sec_id','=','5')->get();
//will return one row in an array with one item, but will be addressed as:
$myfieldvalue = $rez[0]->fieldname;
using first()
$rez = \DB::table('table')->where('sec_id','=','5')->first();
// will also return one row but without the array, so
$myfieldvalue = $rez->fieldname;
So it depends on how you want to access the result of the query: as an object or as an array, and also depends on what "you know" the query will return.
first() is the equivalent of LIMIT 1 at the end of your SELECT statement. Even if your query would return multiple rows, if you use first() it will only return the first row

Laravel - When to use ->get()

I'm confused as to when ->get() in Laravel...
E.G. DB::table('users')->find(1) doesn't need ->get() to retrieve the results, neither does User::find(1)
The laravel docs say "...execute the query using the get or first method..."
I've read the Fluent Query Builder and Eloquent docs but don't understand when the usage of get() is required...
Thanks for the help
Since the find() function will always use the primary key for the table, the need for get() is not necessary. Because you can't narrow your selection down and that's why it will always just try to get that record and return it.
But when you're using the Fluent Query Builder you can nest conditions as such:
$userQuery = DB::table('users');
$userQuery->where('email', '=', 'foo#bar.com');
$userQuery->or_where('email', '=', 'bar#foo.com');
This allows you to add conditions throughout your code until you actually want to fetch them, and then you would call the get() function.
// Done with building the query
$users = $userQuery->get();
For find(n), you retrieve a row based on the primary key which is 'n'.
For first(), you retrieve the first row among all rows that fit the where clauses.
For get(), you retrieve all the rows that fit the where clauses. (Please note that loops are required to access all the rows or you will get some errors).
find returns one row from the database and represent it as a fluent / eloquent object. e.g. SELECT * FROM users WHERE id = 3 is equivalent to DB::table('users')->find(3);
get returns an array of objects. e.g. SELECT * FROM users WHERE created_at > '2014-10-12' is equivalent to DB::table('users')->where('created_at', '>', '2014-10-12')->get() will return an array of objects containing users where the created at field is newer than 4014-10-12.
The get() method will give you all the values from the database that meet your parameters where as first() gets you just the first result. You use find() and findOrFail() when you are searching for a key. This is how I use them:
When I want all data from a table I use the all() method
Model::all();
When I want to find by the primary key:
Model::find(1)->first();
Or
Model::findOrFail(1)->first();
This will work if there is a row with a primary key of one. It should only retrieve one row so I use first() instead of get(). Remember if you deleted the row that used key 1, or don't have data in your table, your find(1) will fail.
When I am looking for specific data as in a where clause:
Model::where('field', '=', 'value')->get();
When I want only the first value of the data in the where clause.
Model::where('field', '=', 'value')->first();
Basically what you need to understand is that get() return a collection(note that one object can be in the collection but it still a collection) why first() returns the first object from the result of the query(that is it returns an object)
#Take_away
Get() return a collection first() return an object
You can use get() method with latest() method to get the latest record that were recently added to your table
For example
$user=Student::latest()->get();
return all the data in descending order

Resources