Associated fields question

[Jaime Perry]

I have considered that option. Like you, I generally do not like temp 
files. I understand the associated fields logic but when I check @af in 
the selection processing, it is always "1" whether it is the first time 
I hit the record or the ninth (using the index built on the associated 
field). This makes me wonder if @af works in a selection process or if 
there is some environmental variable that needs to be set/changed to 
make it work right. And I do need it in  the selection as  my sorts are 
based of of it.

Jaime Perry

Nancy Palmquist wrote:
> Jaime Perry wrote:
>> Hello,
>>
>> I am having a problem with a report that is based on a file that 
>> contains several grouping of associated fields that can contain up to 
>> 10 elements each.  I have included a screen shot of a record to use 
>> as an example.   My selection is based on a date range and the last 
>> three characters of the department field (which is a union code).  
>> The date criteria is generic to the whole record (not associated, one 
>> field for this).  My sorts need to be union code, job number and 
>> employee code.
>>  
>>  RECORD NUMBER: 5
>>  EMPLOYEE CODE:  789
>>
>>  DEPT          JOB NO.   COMP    COST CODE    RATE   TYPE       
>> HOURS      EXTENSION
>>  000D10       000000       DE953    5200 000           39.50       
>> R            8.00             316.00
>>  228D10       228000       DE953    5200 000           39.50       
>> R             8.00             316.00
>>  243D10       243000       DE953    5200 000           39.50       
>> R             8.00             316.00
>>  266D20       266000       DE953    5200 000           39.50       
>> R             8.00             316.00
>>  274D10       274000       DE953    5200 000           39.50       
>> R             8.00             316.00
>>
>> My problem is that I need to select and sort the record for each 
>> instance where the union code matches the selection criteria.  In the 
>> above example, 4 of the 5 lines would pass (the D10 union codes).   I 
>> built an index on the job number associated field (J1) and this works 
>> fine as far as letting me select the record multiple times, but the 
>> real problem I am having is that I cannot determine which instance I 
>> am dealing with.  The index gives me this record 5 times but I only 
>> need 4.  I need a way to determine the instance so I can filter out 
>> unneeded instances and pass the correct job number to the sort.  I 
>> thought @af would work, but it reports a "1" each time I hit the 
>> record.  I am not sure if @af is supposed to work in a selection 
>> process or not or if some environmental variable is set wrong/not set 
>> that affects @af.  If there is another way to select a record more 
>> than once, that would work as well.
>> Thank you
>> Jaime Perry
>>
>> _______________________________________________
>> Filepro-list mailing list
>> Filepro-list at lists.celestial.com
>> http://mailman.celestial.com/mailman/listinfo/filepro-list
>>
>>
> Jamie,
>
> I am not usually a proponent of a work file, but for sanity sake this 
> might be a good example where it is very helpful and saves a bunch of 
> programming time.
>
> Set up a file with the data fields you need to generate the sorts, 
> selections and report output you need.
>
> Clear the file.  Then add a record for each associated occurance in 
> the data that you want to consider.
>
> Once that is done, the selections, sorts and reports are very simple.
>
> Your complex sort and selection can make this solution very 
> attractive.  Of course, you can do it with the associated field logic 
> but it can be tricky.
>
> Since the work file is cleared and loaded each time, it is important 
> that two users are not running the same report at the same time.  It 
> would be possible to make a work file for each user to eliminate any 
> possible overlap if that could be an issue.  ("key"{@id on Unix, who 
> knows on windows.)
>
> Nancy
>