reading fixed width file

[ajithaselvan]

Hi
I'm receiving input(fixedwidth) in below format
PCR2D11 D 93 G08010CR201
PCR2D2 DC REP G08010CR2 03-01-2011 DC115509435
PCR2DD SOURCE: DEPOSITORY TRUST CO.

I'm reading it as char(1)-Col 1,char(4)-Col 2,char(1)-Col 3,char(50)-Col 4 into sequential file.
when i view the data, i'm getting it as below
Col1 Col2 Col3 Col4
P CR2D 1 1
P CR2D 2
P CR2D D

i'm able to view first 3 columns. But for the 4th column, only first value of first row is printed. But i need all the values to be printed in the 4th filed.

Kindly help me to resolve it


Thanks in advance

[BI-RMA]

Hi ajithaselvan,

check the Hex-values of characters in position 8 (row 1) and 7 (row 2-3). Are these blanks or possibly something like your record-delimiter?

[jwiles]

Based on your description, I assume that your input record is fixed-width (56 bytes) with no delimiters between your columns, and that the last (fourth) column contains free-form text. Sounds like a file that was generated by a mainframe source.

What is your input schema?

Have you checked the format tab for the output link options to make sure that the field delimiters option is set correctly?

Regards,

[ajithaselvan]

I found there is a junk value(special character) is placed at the poition 8.
I believe, that is the reason for not able to read by datastage.
Can we handle junk value in datastage while reading file?

jwiles: yeah. file is generated from mainframe.
checked the format tab. set fields defaults (delimiter and quote) as none


Regards,

[jwiles]

As BI-RMA suggests, view the records in hex mode to see what the junk data actually is. I'm going to guess that it's either binary zero (x'00') or maybe a tab character (x'09').

If possible, ask the data supplier if they can remove non-printable characters from the data prior to transmission to your server. If this isn't possible, you may be able to do the same using an external filter in the sequential file stage, calling a sed or awk script. DSGuru28 could suggest an appropriate script, I bet.

Regards,

[BI-RMA]

I found there is a junk value(special character) is placed at the poition 8.
I believe, that is the reason for not able to read by datastage.
Can we handle junk value in datastage while reading file?

Hello ajithaselvan,

Yes, we can!

You might, for example, use a filter-command on the sequential-file stage. sed might be a good idea. Of course, You will have to make sure that You replace just the correct values...