I am trying to query posts and use 'count' to get the total amount of comments and likes to display. My query looks like this
const posts = await knex
.from("posts")
.select("posts.id as id", "posts.text", "posts.user_id")
.leftJoin("comments", "comments.post_id", "posts.id")
.count("comments.post_id as comments")
.leftJoin("likes", "likes.post_id", "posts.id")
.count("likes.post_id as likes")
.groupBy("posts.id");
res.send(posts);
However, I get different results if I exclude comments or likes and do something like this:
const posts = await knex
.from("posts")
.select("posts.id as id", "posts.text", "posts.user_id")
.leftJoin("comments", "comments.post_id", "posts.id")
.count("comments.post_id as comments")
.groupBy("posts.id");
res.send(posts);
I feel like I am doing something wrong. What is the correct way to chain multiple 'count' and 'leftJoins'?
Firstly start with SQL query and then convert it to Knex.
As #nbk said, when you joining the comments at the final result you will receive a row for each comment.
One option is using sub-query in select, the query will look like:
Select posts.id as id, posts.text, posts.user_id,
(Select count(*) from comments where comments.post_id=posts.id) as comments,
(Select count(*) from likes where likes.post_id=posts.id) as likes,
From posts;
This query can be converted to Knex:
const posts = await knex
.from('posts')
.select(
'posts.id as id',
'posts.text',
'posts.user_id',
knex('comments')
.count('*')
.whereRaw('?? = ??', ['comments.post_id', 'posts.id'])
.as('comments'),
knex('likes').count('*').whereRaw('?? = ??', ['likes.post_id', 'posts.id']).as('likes')
);
Related
! have three table
inventories
enter image description here
warehouses
enter image description here
inventory_has_warehouses
enter image description here
I have use laravel yajra datatable. i need sum and search of starting_balance this field in inventory_has_warehouses pivot table
my code:
$id = Auth::user()->id;
$row = Inventory::with('contact')->with('warehouse')
->select(
'inventories.*',
DB::raw('SUM(inventory_has_warehouses.starting_balance) as total')
)
->leftJoin('inventory_has_warehouses', 'inventory_has_warehouses.inventory_id', '=', 'inventories.id')
->leftJoin('warehouses', 'warehouses.id', '=', 'inventory_has_warehouses.warehouse_id')
->where('inventories.subscriber_id',$id)
->groupBy('inventories.id');
$datatable = DataTables::of($row)
->filterColumn('total', function($query, $keyword) {
$query->whereRaw('sum(inventory_has_warehouses.starting_balance) like ?', ['%'.$keyword.'%']);
})
return $datatable->make(true);
but i fount this type of error
Exception Message:↵↵SQLSTATE[HY000]: General error: 1111 Invalid use
of group function (SQL: select count() as aggregate from (select
inventories., SUM(inventory_has_warehouses.starting_balance) as
total from inventories left join inventory_has_warehouses on
inventory_has_warehouses.inventory_id = inventories.id left
join warehouses on warehouses.id =
inventory_has_warehouses.warehouse_id where
inventories.subscriber_id = 2 and inventories.status = 1 and
(LOWER(inventories.itemcode) LIKE %1% or
LOWER(inventories.purchasedescription) LIKE %1% or exists (select
* from contacts where inventories.supplier = contacts.id and LOWER(contacts.name) LIKE %1%) or
(sum(inventory_has_warehouses.starting_balance) like %1%)) group by
inventories.id) count_row_table)
mysql query
select inventories., SUM(inventory_has_warehouses.starting_balance)
as total from inventories left join inventory_has_warehouses on
inventory_has_warehouses.inventory_id = inventories.id left join
warehouses on warehouses.id = inventory_has_warehouses.warehouse_id
where inventories.subscriber_id = 2 and inventories.status = 1 and
(LOWER(inventories.itemcode) LIKE %1% or
LOWER(inventories.purchasedescription) LIKE %1% or exists (select *
from contacts where inventories.supplier = contacts.id and
LOWER(contacts.name) LIKE %1%) or
(sum(inventory_has_warehouses.starting_balance) like %1%)) group by
inventories.id
$id = Auth::user()->id;
$row = DB::table('inventories')->select('inventories.*','contacts.name',DB::raw('SUM(inventory_has_warehouses.starting_balance) as total'))
->leftJoin('contacts', 'inventories.supplier', '=', 'contacts.id')
->leftJoin('inventory_has_warehouses', 'inventories.id', '=', 'inventory_has_warehouses.inventory_id')
->where('inventories.subscriber_id',$id)
->groupBy('inventory_has_warehouses.inventory_id');
if ($keyword = $request->get('search')['value']) {
$row->having(DB::raw('SUM(inventory_has_warehouses.starting_balance)'), 'like', '%'.$keyword.'%');
$row->orHaving('inventories.itemcode', 'like', '%'.$keyword.'%');
$row->orHaving('inventories.purchasedescription', 'like', '%'.$keyword.'%');
$row->orHaving('contacts.name', 'like', '%'.$keyword.'%');
}
$datatable = DataTables::of($row)
->filterColumn('total', function($query, $keyword) {
})
return $datatable->make(true);
try to group by inventory_has_warehouses.inventory_id
Edited answer:
Your really giving hardtime for someone to read your query and give you help but to correct your query here is a reformatted query with corrections:
SELECT
inventories.*,
SUM(inventory_has_warehouses.starting_balance) as total
FROM
inventories LEFT JOIN inventory_has_warehouses
ON
inventory_has_warehouses.inventory_id = inventories.id LEFT JOIN warehouses
ON
warehouses.id = inventory_has_warehouses.warehouse_id
WHERE
inventories.subscriber_id = 2
AND
inventories.status = 1
AND
(LOWER(inventories.itemcode) LIKE '%1%' or LOWER(inventories.purchasedescription)
LIKE '%1%' OR EXISTS
(SELECT
*
FROM
contacts
WHERE
inventories.supplier = contacts.id
AND
LOWER(contacts.name) LIKE '%1%')
OR
(SUM(inventory_has_warehouses.starting_balance) LIKE '%1%')) GROUP BY inventories.id
the problem with query you sent is your like statement only have this:
LIKE %1%
this statement expects strings and it only says:
inventories.
this should be specific to a column you need or just use inventories.* to display all columns of that table but the error still doesn't make sense because it says and:
select count() as aggregate
maybe one of those would solve but upon reformatting your code I notice firstly the syntax error but this is very basic maybe you can start on just running a very simple query for the moment this might be the query that would work for you:
SELECT
inventories.id AS inventory_id,
warehouses.id,
SUM(inventory_has_warehouses.starting_balance) AS total
FROM
inventories LEFT JOIN inventory_has_warehouses
ON inventories.id = inventory_has_warehouses.inventory_id
LEFT JOIN
warehouses
ON warehouses.id = inventory_has_warehouses.warehouse_id
GROUP BY
inventory_has_warehouses.inventory_id
from this start adding the conditions one by one until the error appears again (do this on mysql query window not through laravel code) not yet sure how laravel handles sql queries but the format you sent will really cause an error and also if your going to post a question here make sure to make it reader friendly or someone might slam you cause its hard to read a code that is not properly formatted. ;)
also one thing I forgot make sure that the inventories.id is the primary key of that table or this will still cause you an error refer to this link for more details https://dev.mysql.com/doc/refman/5.7/en/group-by-handling.html
I am using Laravel 5.5.
I have a database which contains users. The problem is that some users exist more than one time because of a bug. I want to query my database and select all "unique" users.
By using the word "unique" I mean the below :
If a user with email "test#test.com" exists 50 times I want the row that created_at is closest to now.
My query, which returns all users is written below :
DB::table('users')
->select('name', 'surname', 'email', 'phone', 'answers', 'newsletter', 'created_at')
->get();
I got confused and I'm not sure if i should use limit combining it with order by created_at column.
Any ideas?
Ok, this is what you need to do: First off, you get a table with the users in their last created_at 'version'. Now you have a list of emails and dates. Then you perform a left join of all the users with that temporary table.
TL;DR:
$users = DB::select('select t1.* from users t1 right join (SELECT email, MAX(created_at) as created_at from users group by email) as t2 on t1.email=t2.email and t1.created_at=t2.created_at');
I hate raw SQL, and I hate subqueries, but this is the only way I know using generic SQL (I mean, you could do a better MySQL or MSSQL native queries, but this should do for you.)
You can use
DB::table('users')->select('name', 'surname', 'email','phone','answers','newsletter','created_at')->orderBy('created_at', 'desc')->groupBy('email')->get();
For more help refer Order By before Group By using Eloquent (Laravel)
What you need is groupby and orderby
try this code
DB::table('users')->select('name', 'surname', 'email','phone','answers','newsletter','created_at')
->orderBy('created_at', 'desc')
->groupBy('email')
->get();
hope it will help you if you need further info try above link!
To get latest user record among duplicates you can use a self join
DB::table('users as u')
->select('u.*')
->leftJoin('users as u1', function ($join) {
$join->on('u.email', '=', 'u1.email')
->whereRaw(DB::raw('u.created_at < u1.created_at'));
})
->whereNull('u1.id')
->get();
In plain SQL it would be something like
select u.*
from users u
left join users u1 on u.email = u1.email
and u.created_at < u1.created_at
where u1.id is null
I'm trying to perform what I assume is a very simple query on a MySQL DB. Here's my table setup;
Table 1 - CMS_AccessLevels
accessLevel
titleColor
Table 2 - CMS_Users
userID
username
userEmail
userAvatar
userSignature
accessLevel
I've already got this query;
SELECT `titleColor` FROM `CMS_AccessLevels` WHERE `accessLevel` = (SELECT `accessLevel` FROM `CMS_Users` WHERE `userID` = 3)
This works correctly and returns the correct titleColor value based on the accessLevel matching across both tables.
Now, what I want to do is also grab some of the values from CMS_Users as well. For the sake of simplicity, let's assume I want to grab only a few of the values, so my result set might look something like this;
userID|username|userAvatar|accessLevel|titleColor
-------------------------------------------------
0 |Scott |image.png | 6 |#FFFFFF
or as a PHP Array (shown just so you can see the logical layout if the above table didn't make sense);
array('userID' => $result['userID'],
'username' => $result['username'],
'userAvatar' => $result['userAvatar'],
'accessLevel' => $result['accessLevel'],
'titleColor' => $result['titleColor'];
Let's say I want to get userID, userName, userAvatar and accessLevel from CMS_Users, and titleColor from CMS_AccessLevels where CMS_Users.userID is equal to '3', remembering that CMS_AccessLevels.accessLevel and CMS_Users.accessLevel MUST match.
Realistically, the only piece of data I know before running the query is userID.
Is it possible to do this with a single query?
Try this:
SELECT u.userID, u.username, u.userAvatar, u.accessLevel, al.titleColor
FROM CMS_AccessLevels al
INNER JOIN CMS_Users u
ON u.accessLevel = al.accessLevel
WHERE u.userID = 3
You are using subqueries whereas joins will be the right choice. You might try something like
SELECT a.titleColor AS titleColor, u.username AS username FROM CMS_users u INNER JOIN CMS_AccessLevels a ON u.accessLevel = a.accessLevel WHERE u.userID = '3'
I want to translate query like this:
SELECT * FROM Product WHERE Product.ID in (SELECT Product.ID FROM other_table)
into LINQ. I read about using the contains method but the problem is that it generates a lot of parameters for each id passed in like this:
WHERE [Product].[ID] IN (#p0, #p1)
If I had for example one bilion parameters I want to pass into my query the server won't be able to execute such a long query. Is it possible to create LINQ query in such a way that the generated SQL will be close to the original?
Thanks,
Romek
If you are using large tables then IN statments are a bad idea, they are very slow. You should be doing joins.
Anyway, here is what you want;
using(dbDataContext db = new dbDataContext())
{
var result = from p in db.products
join o in db.other_table
on p.ID equals o.ID
select p;
}
You should be able to use join for this.
other_Table.Join(product, ot => ot.Id, pd => pd.Id, (pd, ot) => pd);
MySQL Server Version: Server version: 4.1.14
MySQL client version: 3.23.49
Tables under discussion: ads_list and ads_cate.
Table Relationship: ads_cate has many ads_list.
Keyed by: ads_cate.id = ads_list.Category.
I am not sure what is going on here, but I am trying to use COUNT() in a simple agreggate query, and I get blank output.
Here is a simple example, this returns expected results:
$queryCats = "SELECT id, cateName FROM ads_cate ORDER BY cateName";
But if I modify it to add the COUNT() and the other query data I get no array return w/ print_r() (no results)?
$queryCats = "SELECT ads_cate.cateName, ads_list.COUNT(ads_cate.id),
FROM ads_cate INNER JOIN ads_list
ON ads_cate.id = ads_list.category
GROUP BY cateName ORDER BY cateName";
Ultimately, I am trying to get a count of ad_list items in each category.
Is there a MySQL version conflict on what I am trying to do here?
NOTE: I spent some time breaking this down, item by item and the COUNT() seems to cause the array() to disappear. And the the JOIN seemed to do the same thing... It does not help I am developing this on a Yahoo server with no access to the php or mysql error settings.
I think your COUNT syntax is wrong. It should be:
COUNT(ads_cate.id)
or
COUNT(ads_list.id)
depending on what you are counting.
Count is an aggregate. means ever return result set at least one
here you be try count ads_list.id not null but that wrong. how say Myke Count(ads_cate.id) or Count(ads_list.id) is better approach
you have inner join ads_cate.id = ads_list.category so Count(ads_cate.id) or COUNT(ads_list.id) is not necessary just count(*)
now if you dont want null add having
only match
SELECT ads_cate.cateName, COUNT(*),
FROM ads_cate INNER JOIN ads_list
ON ads_cate.id = ads_list.category
GROUP BY cateName
having not count(*) is null
ORDER BY cateName
all
SELECT ads_cate.cateName, IFNULL(COUNT(*),0),
FROM ads_cate LEFT JOIN ads_list
ON ads_cate.id = ads_list.category
GROUP BY cateName
ORDER BY cateName
Did you try:
$queryCats = "SELECT ads_cate.cateName, COUNT(ads_cate.id)
FROM ads_cate
JOIN ads_list ON ads_cate.id = ads_list.category
GROUP BY ads_cate.cateName";
I am guessing that you need the category to be in the list, in that case the query here should work. Try it without the ORDER BY first.
You were probably getting errors. Check your server logs.
Also, see what happens when you try this:
SELECT COUNT(*), category
FROM ads_list
GROUP BY category
Your array is empty or disappear because your query has errors:
there should be no comma before the FROM
the "ads_list." prefix before COUNT is incorrect
Please try running that query directly in MySQL and you'll see the errors. Or try echoing the output using mysql_error().
Now, some other points related to your query:
there is no need to do ORDER BY because GROUP BY by default sorts on the grouped column
you are doing a count on the wrong column that will always give you 1
Perhaps you are trying to retrieve the count of ads_list per ads_cate? This might be your query then:
SELECT `ads_cate`.`cateName`, COUNT(`ads_list`.`category`) `cnt_ads_list`
FROM `ads_cate`
INNER JOIN `ads_list` ON `ads_cate`.`id` = `ads_list`.`category`
GROUP BY `cateName`;
Hope it helps?