How to append metadata to a grouped selection by primary key - sql-server

Question

I have a list of favorites like:
Sample Data
| key | item_id | list_name | customer_id |meta |
|-----|---------|-----------|--------------|---------------|
| 1 | A-11 | aa11 | 001 | unique-data-1 |
| 2 | A-11 | bb22 | 001 | unique-data-2 |
| 3 | A-26 | cc33 | 001 | unique-data-3 |
| 4 | A-28 | aa11 | 002 | unique-data-4 |
| 5 | J-52 | aa11 | 001 | unique-data-5 |
| 6 | X-53 | aa11 | 001 | unique-data-6 |
Desired Output
for #item_id nvarchar(20) = 'A-11'
| key | isFavorited | list_name | meta |
|-----|-------------|-----------|---------------|
| 1 | Y | aa11 | unique-data-1 |
| 2 | Y | bb22 | unique-data-2 |
| 3 | N | cc33 | unique-data-3 |
And would like to return a selection of all available lists, as well as whether or not a particular item is part of that list, with its meta data.
declare #item_id nvarchar(20) = 'A-11'
declare #customer_id nvarchar(20) = 001
select
[key],
[isFavorited] = max(case when [item_id] = #item_id then 'Y' else 'N' end)
[list_name]
[meta]
from favorites
where customer_id = #customer_id
group by [list_name], [key], [meta]
Issues when trying various methods:
The issue I'm having is that since the meta is unique the group by destroys the uniqueness of the select
A cross apply like the following doesn't apply the correct meta based on a matching key.
cross apply (
select top 1
[meta]
from favorites
where customer_id = #customer_id
)
When selecting by row number, the actual key to join back to is lost, so I'm unable to join the meta.
"noRow" = row_number() over(order by h_po_no asc)
I'd like to
Pass in an item_id and customer_id
Return all lists for that customer
Get favorite status for each list of passed in item_id
An item is flagged favorite if it matches both list_name and item_id for a given customer_id
Get row primary key and meta data
How can I return a distinct selection of list_name, isFavorite status, key, and it's meta?

Answer 1

To return the desired output you don't need any aggregation at all. A simple case expression and a where clause will accomplish this.
declare #Favorites table
(
MyKey int
, item_id varchar(10)
, list_name varchar(10)
, customer_id varchar(10)
, meta varchar(20)
)
insert #Favorites values
(1, 'A-1', 'list-1', '001', 'unique-data-1')
, (2, 'A-1', 'list-2', '001', 'unique-data-2')
, (3, 'A-2', 'list-3', '001', 'unique-data-3')
, (4, 'A-2', 'list-1', '002', 'unique-data-1')
select *
from #Favorites
declare #item_id nvarchar(20) = 'A-1'
, #customer_id nvarchar(20) = '001'
select f.MyKey
, isFavorited = case when f.item_id = #item_id then 'Y' else 'N' end
, listName = f.list_name
, f.meta
from favorites f
where f.customer_id = #customer_id
order by f.MyKey

Answer 2

Test Parameters
declare #item_id nvarchar(20) = 'A-11'
declare #customer_id nvarchar(20) = 001
Solution
drop table if exists #tmp
;with cte as (
select
[key],
[list_name],
[rn] = row_number() over (partition by list_name order by list_name desc)
from favorites
where customer_id = #customer_id
group by [list_name], [key], [meta]
)
select *
into #tmp
from cte
where [rn] = 1
select
[i],
[json],
#t.[tmp]
from #tmp #t
inner join (
select
[isFavorited] = max(case when [item_id] = #item_id then 'Y' else 'N' end)
[list_name]
from favorites
where customer_id = #customer_id
group by [list_name]
) j
on j.list_name = #t.list_name