• The Four Hundred
  • Subscribe
  • Media Kit
  • Contributors
  • About Us
  • Contact
Menu
  • The Four Hundred
  • Subscribe
  • Media Kit
  • Contributors
  • About Us
  • Contact
  • SQL Can Return One or a Few Records

    February 23, 2005 Hey Ted

    SQL’s SELECT INTO statement works fine as long as only one record matches the criteria in the WHERE clause. If the query returns more than one record, the host variables are loaded with data, but the SQL State variable, SQLSTT, has a value of 21000, which according to the manual indicates a “Cardinality Violation.” Can I assume that the record loaded into the host variables really is the first record in the returned dataset? Or do you have a better suggestion?

    –Brad

    I suspect that the data in the host variables comes from the record you’re looking for, but being the ultra-conservative I am (compared to me, Ronald Reagan is in the same league with Karl Marx), I wouldn’t count on it. Let me give you another suggestion.

    If you’re at V5R1 or above, add the FETCH FIRST n ROWS ONLY clause to the SELECT statement. The number n can be omitted and defaults to the value one, and either ROW and ROWS is acceptable. Here’s an example:

    select * into :SomeDataStructure
      from SomeTable
     Where Something = SomethingElse
     order by SomeField
     fetch first row only
    

    If you’re running an earlier release, you’ll have to go to the trouble of declaring a cursor, opening it, and fetching from it one time.

    You can use the FETCH FIRST n ROWS ONLY clause with the regular SELECT statement, not just SELECT INTO. I threw together the following short program to illustrate this usage. It lists the first five customers in alphabetical order by name. That is to say, it sorts the records in alphabetical order, then returns the first five records from the sorted data.

    Fqsysprt   o    f  132        printer
    
    D SqlNormal       c                   const('00000')
    D SqlEOF          c                   const('02000')
    D Error           s             12a
    D CustRec       e ds                  extname(QCUSTCDT)
    
    C/exec sql
    C+ set option closqlcsr=*endmod
    C/end-exec
    
    C/exec sql
    C+ declare input cursor for
    C+ select * from qcustcdt
    C+ order by lstnam,init
    C+ fetch first 5 rows only
    C/end-exec
    
    C                   eval      *inlr = *on
    
    C/exec sql
    C+ open input
    C/end-exec
    
    C                   if        SqlStt <> SqlNormal
    C                   eval      Error = 'Open'
    C                   return
    C                   endif
    
    C                   dow       '1'
    C/exec sql
    C+ fetch input into :CustRec
    C/end-exec
    C                   if        SqlStt = SqlEOF
    C                   leave
    C                   endif
    C                   if        SqlStt <> SqlNormal
    C                   eval      Error = 'Fetch'
    C                   except    ErrorLine
    C                   return
    C                   endif
    C                   except    DtlLine
    C                   enddo
    
    Oqsysprt   e            ErrorLine   1
    O                       Error
    O                       SqlStt           +0001
    Oqsysprt   e            DtlLine     1
    O                       CusNum
    O                       LstNam           +0001
    O                       Init             +0001
    

    I sometimes use FETCH FIRST n ROWS only in interactive SQL to create a small result set. It’s similar to using the Number of Records to Copy (NBRRCDS) parameter of the Copy File (CPYF) command.


    I have read that some database management systems permit you to use a host variable to specify the number of rows to return. DB2/400 does not support this feature, at least not in V5R2. I tried it to make sure. I understand all databases in the DB2 family require you to hardcode a literal number of records.

    –Ted

    Click here to contact Ted Holt by e-mail.

    Share this:

    • Share on Reddit (Opens in new window) Reddit
    • Share on Facebook (Opens in new window) Facebook
    • Share on LinkedIn (Opens in new window) LinkedIn
    • Share on X (Opens in new window) X
    • Email a link to a friend (Opens in new window) Email

    Tags:

    Sponsored by
    ARCAD Software

    [IBM i Auditing Tool] Discover the Health of Your IBM i Applications in Minutes with ARCAD iCheck-up

    How well do you really know the health of your IBM i applications? Hidden complexity, unused code, and synchronization issues can make maintenance more challenging and slow down modernization efforts.

    What if you could identify these issues in just a few clicks, without installing anything on your IBM i system?

    Meet ARCAD iCheck-up, an auditing tool designed to give IBM i teams a clear, actionable overview of their applications in minutes.

    With iCheck-up, you can:

    • Assess code complexity: Identify complex source files and prioritize refactoring efforts.
    • Detect dead code: Uncover unused code that may be cluttering your applications.
    • Spot synchronization issues: Identify inconsistencies between source files and compiled objects.
    • Track code changes: Explore modification history and better understand your application’s evolution.
    • Generate a shareable report: Get a comprehensive, easy-to-read HTML report with key metrics and insights.

    Fast, simple, and secure. iCheck-up runs from Windows, requires no installation on your IBM i system, and leaves no files or code behind. Your generated report contains metrics only, keeping your source code confidential.

    Whether you’re maintaining existing applications, planning modernization, or looking to improve code quality, iCheck-up helps you make more informed decisions.

    Ready to gain deeper insights into your IBM i applications? Discover ARCAD iCheck-up and take the first step toward better code quality today!

    Share this:

    • Share on Reddit (Opens in new window) Reddit
    • Share on Facebook (Opens in new window) Facebook
    • Share on LinkedIn (Opens in new window) LinkedIn
    • Share on X (Opens in new window) X
    • Email a link to a friend (Opens in new window) Email

    CPYSPLF and AFPDS Admin Alert: Configuring Output Queues to Print Multi-Copy Spooled Files on Remote Printers

    Leave a ReplyCancel reply

Volume 5, Number 8 -- February 23, 2005
THIS ISSUE
SPONSORED BY:

T.L. Ashford
WorksRight Software
COMMON

Table of Contents

  • CPYSPLF and AFPDS
  • SQL Can Return One or a Few Records
  • Admin Alert: Configuring Output Queues to Print Multi-Copy Spooled Files on Remote Printers

Content archive

  • The Four Hundred
  • Four Hundred Stuff
  • Four Hundred Guru

Recent Posts

  • Bob Code Assistant Gets Better Understanding Of IBM i Applications
  • IBM i In The Cloud And The Case For Reversing Replication
  • Guru: From Hiking Boots To Long-Term Memory
  • Inside IBM’s New PowerHA Cloud Offering, Full System Manager
  • IBM i PTF Guide, Volume 28, Number 39
  • Power Systems Turns In Another Great Quarter
  • Big Blue Is Right On Schedule With Fall 2026 IBM i Tech Refreshes
  • Guru: Environment Variables – Global Variables Wearing Hiking Boots
  • A New IBM i Entity Takes Shape: Power By Fortra
  • IBM i PTF Guide, Volume 28, Number 38

Subscribe

To get news from IT Jungle sent to your inbox every week, subscribe to our newsletter.

Pages

  • About Us
  • Contact
  • Contributors
  • Four Hundred Monitor
  • IBM i PTF Guide
  • Media Kit
  • Subscribe

Search

Copyright © 2025 IT Jungle