MVSFORUMS.com

[sabarisavi]

Hi,

Sorry, if already this problem is resolved by Forums. Here the problem is,

I want to filter the Editing numeric filed to implicit numeric form. We got the pricing records thru Excel Spreadsheet, FTP transfer these records to Mainframe dataset. Problem is, when the customer sends the fields by preceding and succeeding zeros, our program is treat as new pricing. Say eg:

[CraigG]

What language is your program written in that is doing this compare? I think you have a serious programming error.

[Nic Clouston]

This probably should be in the Programming section or Utilities section but certainly not the JCL section.

All the same - what are your field definitions for this data. It looks as though you may be using a character field rather than a numeric field. If you want an assumed decimal point and the input has a real decimal point then you need to read it into an edited numeric field then move it to an unedited numeric field.
_________________
Utility and Program control cards are NOT, repeat NOT, JCL.

[sabarisavi]

Thanks, i suppose to post this conversation in the utilities Page. Well.

Our JOB flow is FTP send the dataset to EZTRIVE job to compare the price. But this EZT, have problem with comparing the 12.10 and new pricing 0012.10. Since both are same, but our EZT step treat as differnt records.

This point we dont want to change the EZTRIVE Program, instead write a sort, utility - reformat logic to supress the preceding and succeeding zeros in the price field.

for Instance, 00012.990 will become 12.99.

[kolusu]

sabarisavi,

Your posts are quite confusing. I am not sure what you are trying to do. However I am guessing that the excel file is ftp'd with tab delimiter and the postion of the numeric field varies . If that is the case then we need to use PARSE function to edit the numeric fields. Let us know and we will show you an example as to how to do it

Just show us how the input file Mainframe looks like

[sabarisavi]

Great, thanks Kolusu. Sorry for that if my statements are really confused. Well.

My Input file are below (Key field is - Unit Dosage is 11 Byte alphanumeric)

[CraigG]

In your input file unit dosage appears to be a 7 byte alphanumeric field not 11. And you are not supressing the leading or trailing zeros but adding them. That may be what is confusing people.

[sabarisavi]

Well,

Intially we thought to supress but future reference we have to keep leading and trailing zeros.

From my Input, Since text file made up with balnks. so FTP retreive the same.
It is sure, Unit Dosage is a 11 byte made up with blank spaces, it is not 7 byte. Here it shows bytes 7 but with preceeding and trailing blank ( spaces).

Let me know, if still i'm confussed.

[kolusu]

sabarisavi,

The following DFSORT JCL will give you desired results. I assumed that your input is FB recfm and 80 bytes LRECL

[sabarisavi]

Great Kolusu, you always prove that you are a Mainframe Encyclopedia. Thank you again.

It would be helpfull to every one that if you could you please brief us this
SYSIN CARD- Parse, UFF, M11 and %00 ie., simply this logic.

[kolusu]

sabarisavi,

Check this link which explains in detail about PARSE with examples

http://www-304.ibm.com/systems/support/storage/software/sort/mvs/peug/pdf/sortpeug.pdf

The first %00 operates starts scanning the input record from position 8 in the input file and it stops when it encounters the decimal dot , and the next %01 grabs the 3 bytes after the decimal dot

Check this link which explains in detail about UFF and SFF

http://www-304.ibm.com/systems/support/storage/software/sort/mvs/pdug/pdf/sortpdug.pdf

M11 is just an edit mask which makes the field to have leading zeroes. Here are the different edit masks available with DFSORT.

M11 Edit mask


Hope this helps...
_________________
Kolusu
www.linkedin.com/in/kolusu

[sabarisavi]

Thanks again, Well. Here, we had a simple problem, yet to resolve by our selves. Looking for your help.

First record in the Input file

XXXX 0012.10 20087


Current output file is

XXXX 0000012.010 20087

The Unit Dosage should not 0000012.010 instead it would be 0000012.100.

The Actual Output would be

XXXX 0000012.100 20087

[kolusu]

sabarisavi,

That is a weird way of representing the decimals. why would '10 ' considered as 100 ?. He already gave a zero why not give the other zero also ? Anyway change your control cards to the following and it will give you the desired results. We basically use ALTSEQ to convert any spaces to '0' for decimals

[sabarisavi]

Thanks. It is working fine. I hope, there is some miscommunication some where. Well,

FYI - We expect the output becomes .100 not for .010 thats what i meant, after decimal if any zero come followed by any number, the value is different rather than zero preceed the decimal point.

How ever, your new approch is working fine, but i dont know what is wrong with the old approch, ie.,

[kolusu]