Re: Help improve performance of RPG program - Long Post

Rich,

I had a quick scan on your code and some suggestions:

I see you use embedded SQL and to make it more generic i would use an SQLDA DS 
to use on the different file (the F-specs can be removed)

Create an SQL statement with the selected fields (point 5). Prepare and 
describe the SQL statement using the SQLDA, then you have the field names used 
in the result set (point 3: second record).

Fetch one record at the time using the SQLDA, format your date fields and 
output string (point 4).

Regards,
Carel Teijgeler

*********** REPLY SEPARATOR  ***********

On 24-5-05 at 20:34 Rich Dotson wrote:

>I have a program that we use to create export files that are ftp*d to our
>   data warehouse application.  The program works great functionally but we
>   would like to improve the performance on files that have a lot of fields
>   and a lot of records (500,000+).
>    
>   The program specs were:
>
>   1) Must be *generic* enough to process, without modification, any file on 
> our iSeries.
>   2) First row of file must contain:
>      a) File Name
>      b) Number of fields in file
>      c) Number of Records in the file 
>      d) Last G/L posting date
>   3) Second row of the file must contain the field names
>   4) Subsequent rows will be the data
>
>   A sample of the first three rows would look like:
>   "XAT90090",5,1297,05/22/2005
>   "EFFDATE","APPLCD","ACCTNO","ACCTTYP","TRANAMT"
>   05/19/2005,20,1234567890,1,-190.78

>   4) All dates must be in MM/DD/CCYY format.  For this I created a file
>   (XAP10005L1) that contains the fields that contain dates and the format
>   that they are in: a) File Name  b) Date Field Name  c)Stored Date
>   Format  (*LongJul, *MDY, *JUL, *YMD, etc..)  When processing a field I
>   check this file to see if it is a date field and reformat it if it is.
>
>   5) Only the fields that the user selects should appear in the export
>   file.    The way I solved this issue is if the user does not want all the
>   fields in the data file, I created another PF DDS containing only the
>   desired fields.   I pull the data from one file based on the fields this
>   *template* file.
>