We Keep Coding

MySQL IN clause: How can I get multiple rows when I use a same value multiple times?

I'm programming a C# Windows Forms Application in Visual Studio and I'm trying to get data about prices of products and the amount a user has added a product to its shopping list from my local MySQL-database into a List(int).
What I do is following:
If a user has added a product 4 times to their shopping list, I'm adding the barcode of the product 4 times to my List(int).
This is working but when I'm reading out all items of the List with the String.Join()-method into the IN-clause of my query and execute it, it only returns a row one time altough the IN-operator has the same barcode multiple times.
The following is how I'm adding barcodes to my List(int)
int count = 0;
List<int> barcodes = new List<int>();
MySqlCommand cmd = new MySqlCommand("SELECT product_barcode, amount FROM shopping_list_items WHERE shopping_list_id = " + current_shoppingListID + ";", db.connection);
cmd.ExecuteNonQuery();
var reader = cmd.ExecuteReader();
while (reader.Read())
{
do
{
barcodes.Add(Int32.Parse(reader["product_barcode"].ToString()));
count++;
} while (count < Int32.Parse(reader["amount"].ToString()));
}
reader.Close();
This is how I'm executing my query and assign the values to variables:
MySqlCommand cmdSum = new MySqlCommand("SELECT sum(price) AS 'total', supermarket_id FROM prices WHERE barcode IN (" + String.Join(", ", barcodes) + ") GROUP BY supermarket_id;", db.connection);
cmdSum.ExecuteNonQuery();
var readerSum = cmdSum.ExecuteReader();
while (readerSum.Read())
{
switch (double.Parse(readerSum["supermarket_id"].ToString()))
{
case 1:
sumSupermarket1 = double.Parse(readerSum["total"].ToString());
break;
case 2:
sumSupermarket2 = double.Parse(readerSum["total"].ToString());
break;
case 3:
sumSupermarket3 = double.Parse(readerSum["total"].ToString());
break;
}
}
A simplified query just to make it simple may look like this:
SELECT name FROM products WHERE barcode IN (13495, 13495, 13495);
If the above one is my query then I want it to return 3 the same rows.
So my question now is, how can I get multiple rows altough I use a same value multiple times in the IN-clause of a MySQL-query?

Q: how can I get multiple rows altough I use a same value multiple times in the IN-clause of a MySQL-query?
A: We don't. That's not how IN () works.
Note that
WHERE foo IN ('fee','fi','fi','fi')`
Is shorthand for
WHERE ( foo = 'fee'
OR foo = 'fi'
OR foo = 'fi'
OR foo = 'fi'
)
Understand what's happening here. MySQL is going to examine each row, and for each row it checks to see if this condition returns TRUE or not. If the row satisfies the condition, the row gets returned. Otherwise the row is not returned.
It doesn't matter that a row with foo value of 'fi' satisfies multiple conditions. All MySQL cares about is that the condition inside the parens ultimately evaluates to TRUE.
As an illustration, consider:
WHERE ( t.picked_by = 'peter piper'
OR t.picked_amount = 'peck'
OR t.name LIKE '%pickled%'
OR t.name LIKE '%pepper%'
)
There could be a row that satisfies every one of these conditions. But the WHERE clause is only asking if the entire condition evaluates to TRUE. If it does, return the row. If it doesn't, then exclude the row. We don't get four copies of a row because more than one of the conditions is satisfied.
So how do we get a set with multiple copies of a row?
As one possible option, we could use separate SELECT statements and combine the results with UNION ALL set operator. Something like this:
SELECT p1.name FROM product p1 WHERE p1.barcode IN (13495)
UNION ALL
SELECT p2.name FROM product p2 WHERE p2.barcode IN (13495)
UNION ALL
SELECT p3.name FROM product p3 WHERE p3.barcode IN (13495)
Note that the result from this query is significantly different than the result from the original query.
There are other query patterns that can return an equivalent set.
FOLLOWUP
Without an understanding of the use case, the specification, I'm just guessing at what we are attempting to achieve. Based on the two queries shown in the code (which follows a common pattern we see in code that is vulnerable to SQL Injection),
The shopping list:
SELECT i.product_barcode
, i.amount
FROM shopping_list_item i
WHERE i.shopping_list_id = :id
What is amount? Is that the quantity ordered? We want two cans of this, or three pounds of that? Seems like we would want to multiply the unit price by the quantity ordered to get the cost. (Two cans is going to cost twice as much as one can.)
If what we are after is the total cost of the items on the shopping list from multiple stores, we could do something like this:
SELECT SUM(p.price * s.amount) AS `total`
, p.supermarket_id
FROM ( SELECT i.product_barcode
, i.amount
FROM shopping_list_item i
WHERE i.shopping_list_id = :id
) s
JOIN price p
ON p.barcode = s.product_barcode
GROUP
BY p.supermarket_id
Note that if a particular product_barcode is not available for particular supermarket_id, that item on the list will be excluded from the total, i.e. we could get a lower total for a supermarket that doesn't have everything on our list.
For performance, we can eliminate the inline view, and write the query like this:
SELECT SUM(p.price * i.amount) AS `total`
, p.supermarket_id
FROM shopping_list_item i
JOIN price p
ON p.barcode = i.product_barcode
WHERE i.shopping_list_id = :id
GROUP
BY p.supermarket_id
If we absolutely have to rip through the shopping list query, and then use the rows from that to create a second query, we could form a query that looks something like this:
SELECT SUM(p.price * i.amount) AS `total`
, p.supermarket_id
FROM ( -- shopping_list here
SELECT '13495' AS product_barcode, '1'+0 AS amount
UNION ALL SELECT '13495', '1'+0
UNION ALL SELECT '13495', '1'+0
UNION ALL SELECT '12222', '2'+0
UNION ALL SELECT '15555', '5'+0
-- end shopping_list
) i
JOIN price p
ON p.barcode = i.product_barcode
WHERE i.shopping_list_id = :id
GROUP
BY p.supermarket_id

You would probably be better off investigating LINQ to SQL rather than using direct SQL and injection.
You can use an inline table join to accomplish what you want:
"SELECT sum(price) AS 'total', supermarket_id
FROM (select "+barcodes[0]+"as bc union all select "+String.Join(" union all select ", barcodes.Skip(1).ToArray())+") w
JOIN prices p ON p.barcode = w.bc
GROUP BY supermarket_id;"
Note: If you can name the column with the inline table alias (I couldn't test that) you could simplify the inline table generation.

LINQ Null Join with Pivot Table

I'm trying to get a list of servers thay may or may not belong to 1 or more groups to display in a grid.
Example
ServerID IP GroupID
1 192.168.1.44 1
1 192.168.1.44 10
2 192.168.1.45 1
3 192.168.1.46 2
4 192.168.1.47 null
5 192.168.1.48 null
If I have no records In the GroupServer Table. (Since there is no groups or groups exist but they are not assigned) I expect to get something like this:
ServerID IP GroupID
1 192.168.1.44 null
2 192.168.1.45 null
3 192.168.1.46 null
4 192.168.1.47 null
5 192.168.1.48 null
Since is a Many-to-Many relationship. I have
Group Table
Server Table
GroupServer Table
I could not find a LINQ Pivot Table example.
So I tried to buid my own.
var query = (from sg in context.ServerGroups
join servers in context.Servers on sg.ServerID equals servers.ID
join groups in context.Groups on sg.GroupID equals groups.ID
into serverxgroup
from gAddrBilling in serverxgroup.DefaultIfEmpty()
select new
{
ServerID = sg.ServerID,
ServerIP = server.IP,
GroupID = sg.GroupID
});
The Query above does not retrieve anything
And I quiet dont understand what the "from gAddrBilling" is for. Since I modify a snippet I was trying to make work. So I wonder if someone has already faced a problem like this and give me some hint, snippet or advice about what is what I'm missing.
Thank you.

First, this is not a pivot query, but a regular query on many-to-may relationship via explicit junction table.
Second, looks like you are using Entity Framework, in which case you'd better define and use navigation properties rather than manual joins.
Third, and the most important, the structure of the query is wrong. If you want to get a list of servers that may or may not belong to 1 or more groups, then you should start your query from Servers (the table which records you want to be always included, not from link table where some ServerID are missing) and then use left outer joins to the other tables like this:
var query =
from s in servers in context.Servers
join sg in context.ServerGroups on s.ID equals sg.ServerID
into s_sg from sg in s_sg.DefaultIfEmpty() // make the above LEFT OUTER JOIN
// You can remove the next two lines if all you need is the GroupId
// and keep them if you need some other Group field in the select
join g in context.Groups on sg.GroupID equals g.ID
into sg_g from g in sg_g.DefaultIfEmpty() // make the above LEFT OUTER JOIN
select new
{
ServerID = s.ID,
ServerIP = s.IP, // or sg.IP?
GroupID = (int?)sg.GroupID
};

TSQL - Select from table with multiple join paths

That title is not very good, so consider the following. I have five tables:
User {
Id,
ProfileId // -> Profiles.Id
}
Profile {
Id
}
ProfilePermissionSets {
ProfileId // -> Profiles.Id
PermissionSetId // -> PermissionSets.Id
}
UserPermissionSets {
UserId // -> Users.Id
PermissionSetId // -> PermissionSets.Id
}
PermissionSets {
Id
}
Permissions {
Id,
PermissionSetId // -> PermissionSets.Id
}
And I want get all of the permissions for a user that are directly linked to it or indirectly through the profile. The not-quite-there SQL I've come up with so far is this:
SELECT [Pe].[Controller],
[Pe].[Action]
FROM [PermissionSets] AS [PS]
JOIN [UserPermissionSets] AS [UPS]
ON ([UPS].[PermissionSetId] = [PS].[Id])
JOIN [Users] AS [U]
ON ([U].[Id] = [UPS].[UserId])
JOIN [Profiles] AS [P]
ON ([P].[Id] = [U].[ProfileId])
JOIN [ProfilePermissionSets] AS [PPS]
ON ([PPS].[ProfileId] = [P].[Id])
JOIN [Permissions] AS [Pe]
ON ([Pe].[PermissionSetId] = [PS].[Id])
WHERE [U].[Id] = 4;
It returns back a correct count of rows, but it's repeating the controller or action over and over, so it's wrong. I'm hoping someone can help me correct it to show all of the distinct permission sets for the user. Ideally, I'd like to also change it so that it's all discovered starting at the user because that is what I have access to in the method I need to do this (the object is an Entity Framework class named User and will be browsed using LINQ).
UPDATED because I forgot that I really wanted the permissions not the permission sets.

Try this SQL
SELECT [Pe].[Controller],
[Pe].[Action]
FROM [Users] AS [U]
LEFT OUTER JOIN [UserPermissionSets] AS [UPS]
ON ([UPS].[UserId] = [U].[Id])
LEFT OUTER JOIN [ProfilePermissionSets] AS [PPS]
ON ([PPS].[ProfileId] = [U].[ProfileId])
LEFT OUTER JOIN [Permissions] AS [Pe]
ON ([Pe].[PermissionSetId] = [UPS].[PermissionSetId])
OR ([Pe].[PermissionSetId] = [PPS].[PermissionSetId])
WHERE [U].[Id] = 4;

So, messing around on LINQPad, I came up with this as far as the LINQ query:
user.PermissionSets.Union(user.Profile.PermissionSets).SelectMany(
ps =>
ps.Permissions.Select(
p =>
p.Controller + "." + p.Action));
And it produces what I want, BUT it does it by composing the results of a bunch of SQL queries. The biggest impact comes from profiles that have multiple permission sets, like say the Administrator. I don't think there's a way around it, and I only have a User object to work with, so I'm ok with the excess SQL queries, at least for now.

Duplicate field name problem in a nested multi-mapping Dapper pagination query

I've ran into an issue when trying to do multi-mapping using Dapper, for pagination queries.
Because I am using a nested query in this pagination scenario, there are multiple tables within the nested query that I must join to get my multi-mapped data, but some of these tables will share some fields of the same name which you can see in my example query below (e.g. id, displayname and email):
q = #"select * from (select p.id, p.title, p.etc...,
u1.id, u1.displayname, u1.email,
u2.id, u2.displayname, u2.email,
t.id, t.name,
row_number() over (order by " + sort.ToPostSortSqlClause() + ") as rownum" +
" from posts p" +
" join users u1 on p.owneruserid = u1.id" +
" join users u2 on p.lastediteduserid = u2.id" +
" join topics t on p.topicid = t.id" +
") seq where seq.rownum between #pLower and #pUpper";
In the example above you can see that within the nested query, there are going to be problems with the fields id (appears in the posts table, both users table joins and the topics table join), and also displayname and email (appear in both users table joins).
The only workaround I have thought of so far involves casting each of these 'problem' fields as a different name, but this then involves the very messy process of creating dummy properties in the affected models, so multimapping can map into these, and editing the 'real' properties in my models to also check the dummy property for a value if the real value has not been set.
Also, in the above scenario I would have to create x dummy properties where x is the number of joins I may have on the same table within a query (in this example, 2 joins on the same Users table, therefore requiring 2 uniquely named dummy properties just for Dapper mapping purposes).
This is obviously not ideal and I'm sure would have knock on problems and more untidyness as I created more of these multi-mapping pagination queries.
I'm hoping there is nice, clean solution to this problem?

There are 2 options I can think of:
option 1: join back to your extended properties outside of your nested query:
select s.*, t1.*, t2.* from
(
select s.*, ROW_NUMBER() OVER (order by somecol) AS RowNumber from Something s
) as X
left join Table t1 on Id = x.SomeId
left join Table t2 on Id = x.SomeOtherId
option 2: Extend SqlBuilder to handle column aliasing:
select s.*, /**unalias(Table,t1)**/, /**unalias(Table,t2)**/ from
(
select s.*, /**alias(Table,t1)**/, /**alias(Table,t2)**/ ROW_NUMBER() OVER (order by somecol) AS RowNumber from Something s
left join Table t1 on Id = x.SomeId
left join Table t2 on Id = x.SomeOtherId
) as X
Then define the alias macro to query and cache a list of columns from the db using INFORMATION_SCHEMA.COLUMNS and simply add a 'column as column_t1` string for each column.
Unalias can do the reverse quite simply.

TSQL Query Where All Records Must Exists to Return A Record

I am not sure even how to ask this question.
I have a table of tags:
TagId Tag
----- -----
1 Fruit
2 Meat
3 Grain
I have a table of events:
EventId Event
------- -----------
1 Eating Food
2 Buying Food
What I need to do is bring back only Events that have all selected tags associated with it.
If three tags are selected then only show event that have all three.
For example:
Mapping Table
EventId TagId
------- -----
1 1
1 3
2 1
If I write a query like this:
select * from MapTable where where tagId in (1,3)
This will return Eating Food and Buying Food.
But what I need to do is bring back events that have both tags 1 and 3. This means that the only event in this case I would return would be Eating Food as it has both selected tags.
I was wondering if this can be done in TSQL or if I will have to use the business layer to translate it into the object to return back to the GUI.
Thanks.

There was a very similar question yesterday: Query for exact match of users in a conversation in SQL Server
basically you can do this:
DECLARE #NumTags INT = 2
SELECT EventID
FROM EventTag
GROUP BY EventID
HAVING
Sum(CASE WHEN TagID IN (1, 3) THEN 1 ELSE 0 END) >= #NumTags
so this will find all events that both the tags exist in (this allows for instances where those two tags exist along with any additional tags)

Here is a solution for when you do not know the tags before hand.
Load the tags into a table variable and get the total count:
select #iCount = COUNT(*) from #Tags;
Then write your normal query and slam those results into a table variable:
insert into #EventTags(IsSet, EventId)
select distinct CASE WHEN TagID IN (select ID from #Tags) THEN 1 ELSE 0 END,
e.EventId
from Event_Tag e
inner join #Tags t on t.ID = e.TagId
Then to get back only Events that have ALL matching tags, not just ones that are in the selection, but ALL you do this:
select *
from Event_Tag e
inner join #Tags t on t.ID = e.TagId
where e.EventId in
( select EventId
from #EventTags
group by EventId
having count(EventId) = #iCount
)
Only bring back tags that have all tags associated.
Thank you again everyone for the ideas! Greatly appreciated all the feedback!

There's probably a better way to write it but this will give you what you are looking for:
select *
from event e
where exists(select * from maptable where eventid = e.eventid and tagid = 1) and exists(select * from maptable where eventid = e.eventid and tagid = 3)

You'll want to inner join the two tables, as follows
SELECT * FROM Events INNER JOIN MapTable ON MapTable.EventId=Events.EventID WHERE MapTable.TagID=1 AND MapTable.TagID=3