Complicated Query Help (for a relative beginner)

  • I'm more than sure that this is a question that's been addressed in the past on this forum and possibly others, but I'm not even what to call this sort of thing that I need help with. Again, forgive my ignorance, I'm SO not a SQL ninja 😛

    I have a table that houses information on employees:

    Name, email, NT ID, Manager, e-mail, account type, etc..

    I end up doing a very basic query to return all rows where account type is set to 'contractor' and the manager field for these users comes back with a group (ie Contractor Group 123). All of these contractor groups have entries in the same table, including manager information, which is what I need to drill down to.

    Essentially, I need to write one query for finding every contractor in my table and then listing their manager not as the contractor group, but as the manager of the contractor group. I wish I could give you a more technical term for what this type of operation is called, but like I said, I don't know.

    If anyone could point me to another forum thread that answers this question already and/or tell me what this sort of operation is called, I'd be much obliged and would be happy to get this stupid post off your board =)

    Thank you!

    Mitch

  • Please post table definitions, sample data and desired output. Read this to see the best way to post this to get quick responses.

    http://www.sqlservercentral.com/articles/Best+Practices/61537/

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Sorry about that, first time posting.

    The table I'm querying looks like this (just a copy/paste of the create statement):

    CREATE TABLE [dbo].[BasicAccountInfo](

    [UserObjectSID] [varchar](255) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,

    [UserSourceDomain] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserFullName] [varchar](125) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [NTUserID] [varchar](200) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserEmpID] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserDN] [varchar](175) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserManagerName] [varchar](255) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserCompany] [varchar](100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserAccountType] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserAccountStatus] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserDescriptionFromAD] [varchar](1000) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [Notes] [varchar](1000) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [UserCreationDateFromAD] [datetime] NULL,

    [UserLastLoginFromAD] [datetime] NULL,

    [NEPAccountFlag] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [FirstLogonFlag] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,

    [InfoGoodAsOf] [datetime] NULL,

    CONSTRAINT [PK_BasicAccountInfo] PRIMARY KEY CLUSTERED

    (

    [UserObjectSID] ASC

    )WITH (IGNORE_DUP_KEY = OFF) ON [PRIMARY]

    ) ON [PRIMARY]

    If I do a basic query like this:

    select top 1 UserFullName, UserManagerName

    from basicaccountinfo

    where useraccounttype = 'contractor'

    I get this result set:

    UserFullName

    Cxxxxx, Joanne

    UserManagerName

    CN=Contractors 1009340,OU=Contractor Group Accounts,DC=xxx,DC=com

    Instead of seeing that "Contractors 1009340" as the manager name, I want to see the manager name of the contractors group, as seen by the result of this query:

    select UserFullName, UserManagerName

    from basicaccountinfo

    where userfullname = 'Contractors 1009340'

    Result:

    UserFullName

    Contractors 1009340

    UserManagerName

    CN=Vise\, Jeff,OU=Users,OU=Axxx,DC=xxx,DC=com

    So the desired query/result would look like this:

    select UserFullName, UserManagerName

    from basicaccountinfo

    where useraccounttype = 'contractor'

    UserFullName

    Cxxxxx, Joanne

    UserManagerName

    CN=Vxxx\, Jeff,OU=Users,OU=Axxx,DC=xxx,DC=com

    Obviously, this info is all coming out of AD, but I just need help with the SQL side of things. Again, sorry for the noobie question.

    Thanks in advance,

    Mitch

  • Did you read the article?

    What I'm asking is sample data in a usable format, something I don't have to spend hours on to get into a SQL table.

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Hi Mitch

    Looks like your problem could be solved by using a recursive CTE.

    If you set up sample table scripts and data insertion scripts as Gail suggests, you will have a few solutions in no time. Read the article in her link, it will show you how to do this. The little time this will take, will save you oodles later.

    Cheers

    ChrisM@home


    [font="Arial"]Low-hanging fruit picker and defender of the moggies[/font]

    For better assistance in answering your questions, please read this[/url].


    Understanding and using APPLY, (I)[/url] and (II)[/url] Paul White[/url]

    Hidden RBAR: Triangular Joins[/url] / The "Numbers" or "Tally" Table: What it is and how it replaces a loop[/url] Jeff Moden[/url]

  • Hey Mitch,

    Here's some code to demonstrate one approach to the problem.

    DECLARE @data TABLE

    (

    primary_key INTEGER NOT NULL PRIMARY KEY,

    account_type NVARCHAR(30) NOT NULL,

    full_name NVARCHAR(125) NOT NULL,

    manager_name NVARCHAR(125) NOT NULL

    );

    INSERT@data (primary_key, account_type, full_name, manager_name)

    VALUES(1, N'contractor', N'Cxxxxx, Joanne', N'CN=Contractors 1009340,OU=Contractor Group Accounts,DC=xxx,DC=com'),

    (2, N'manager', N'Contractors 1009340', N'CN=Vise\, Jeff,OU=Users,OU=Axxx,DC=xxx,DC=com');

    SELECTD.full_name,

    M.manager_name

    FROM@data D

    OUTER

    APPLY(

    SELECT*

    FROM@data D2

    WHERED2.full_name = SUBSTRING(D.manager_name, 4, CHARINDEX(N',', D.manager_name) - 4)

    ANDD2.account_type = N'manager'

    ) M

    WHERED.account_type = N'contractor';

  • Hey Paul,

    Thanks for the code. I, unfortunately, can't add an "Account Type" of manager as you've suggested, as that field in my db is imported from another source.

    I am going to just post some sample data from scratch, as has been suggested:

    --Create the basic table

    CREATE TABLE Users

    (

    primary_key INT NOT_NULL PRIMARY KEY,

    FullName VARCHAR(100) NOT_NULL,

    AccountType VARCHAR(50) NOT_NULL,

    Manager VARCAHR(100) NOT_NULL

    );

    --Insert some sample data

    INSERT Users (primary_key, FullName, AccountType, Manager)

    VALUES

    (1, 'John Smith', 'Employee', 'Sally Jones'),

    (2, 'Jake Ryan', 'Contractor', 'Contractor Group 1'),

    (3, 'Contractor Group 1', 'Other', 'Tom Lumberg');

    --A simple select statement for all contractors and their managers will only render me a result of a "Contractor Group"

    SELECT FullName, Manager

    FROM Users

    WHERE AccountType = 'Contractor'

    --Results:

    -- FullName Manager

    -- Jake Ryan Contractor Group 1

    --I'd like to see a result set of:

    -- FullName Manager

    -- Jake Ryan Tom Lumberg

    It's that second result set that I'm unsure how to produce in a single query. Does that make better sense with sample data? Sorry I messed that up a couple times =/

    Thanks,

    Mitch

  • Mitch,

    The code I posted was to demonstrate technique - it was not intended as a complete solution.

    To be fair, I think you could easily have found the solution for yourself by looking at it and understanding how it works. Nevertheless, here is a complete solution, based on your sample code. Please take the time to understand how it works so that you are able to refine it as required.

    --Create the basic table

    DECLARE @Users TABLE

    (

    primary_key INT NOT NULL PRIMARY KEY,

    FullName VARCHAR(100) NOT NULL,

    AccountType VARCHAR(50) NOT NULL,

    Manager VARCHAR(100) NOT NULL

    );

    --Insert some sample data

    INSERT @Users (primary_key, FullName, AccountType, Manager)

    VALUES

    (1, 'John Smith', 'Employee', 'Sally Jones'),

    (2, 'Jake Ryan', 'Contractor', 'Contractor Group 1'),

    (3, 'Contractor Group 1', 'Other', 'Tom Lumberg');

    --A simple select statement for all contractors and their managers will only render me a result of a "Contractor Group"

    SELECT FullName, Manager

    FROM @Users

    WHERE AccountType = 'Contractor'

    --Results:

    -- FullName Manager

    -- Jake Ryan Contractor Group 1

    --I'd like to see a result set of:

    -- FullName Manager

    -- Jake Ryan Tom Lumberg

    SELECT U.FullName,

    M.Manager

    FROM @Users U

    OUTER

    APPLY (

    SELECT *

    FROM @Users U2

    WHERE U2.FullName = U.Manager

    AND U2.AccountType = 'Other'

    ) M

    WHERE U.AccountType = 'Contractor';

  • Why don't use SELF JOIN in this case? Is it simple than outer apply?

    SELECT U.FullName,U1.Manager

    FROM

    @Users U

    join @Users U1 on U.Manager=U1.FullName and U1.AccountType = 'Other'

    WHERE

    U.AccountType = 'Contractor';

  • nguyennd (1/11/2010)


    Why don't use SELF JOIN in this case? Is it simple than outer apply?

    Why don't you try both and let us know what differences you find?

    It is certainly possible to write the query in a number of ways - which was why I previously said that I was posting code to demonstrate one possible approach.

Viewing 10 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic. Login to reply