Read Line feeds in seqential file stage

dkaranth · Post by **dkaranth** » Thu Jul 03, 2008 6:05 am

Hi guys,

I have a txt file which contains many records. Many of these records span multiple lines(i.e there can be line feed char).

eg:
111, "I work on
Datastage"

How can read one such record using the sequential file stage in Parallel job?

Another issue is that these multi-line fields can also contain comma (,). This comma also needs to be rejected by datastage and not considered as a field delimiter.

Can you please help me with this?

Regards,
Deepak

keshav0307 · Post by **keshav0307** » Thu Jul 03, 2008 6:15 am

by defining the records delimiter correctly.

DSDexter · Post by **DSDexter** » Thu Jul 03, 2008 6:33 am

It doesnt matter how much lines you record is running into. What matters is where record delimiter is located. So giving the exact column and records delimiters should serve the purpose. :D

Thinking outside the box....
Your monitor can display only those many characters with the current screen resolution

dkaranth · Post by **dkaranth** » Thu Jul 03, 2008 7:45 am

Hi guys,

I just tried the folowing input:

Col0|val0,Col1|Val
ue1Z

Here Z is the record de-limiter and comma is the value delimeter.

I get the error:
Sequential_File_1,0: Delimiter for field "Value" not found; input: {Col1|Val}

But if I give the folowing input, it works. i.e all in the same line
Col0|val0,Col1|Value1Z

So, does that mean that whatever record de-limiter we use needs to be in the same line as the rest of the record?

Regards,
Deeapk

DSDexter · Post by **DSDexter** » Thu Jul 03, 2008 7:54 am

dkaranth wrote:Hi guys,

I just tried the folowing input:

Col0|val0,Col1|Val
ue1Z

Here Z is the record de-limiter and comma is the value delimeter.

I get the error:
Sequential_File_1,0: Delimiter for field "Value" not found; input: {Col1|Val}

But if I give the folowing input, it works. i.e all in the same line
Col0|val0,Col1|Value1Z

So, does that mean that whatever record de-limiter we use needs to be in the same line as the rest of the record?

Regards,
Deeapk

Have you entered a line fed after Val?

There might be something after Val that is not visible to the naked eye. Using the od utility should explain you in detail. :D

dkaranth · Post by **dkaranth** » Thu Jul 03, 2008 8:37 am

Hi,

I checked with the od utility. There are no characters after Val except line feed. Still getting the error

ray.wurlod · Post by **ray.wurlod** » Thu Jul 03, 2008 4:17 pm

Is the record delimiter (that you have shown as "Z") also a linefeed?

dkaranth · Post by **dkaranth** » Fri Jul 04, 2008 12:06 am

Hi Ray,

There is a line feed after Z. But shouldn't DS ignore that and consider Z as the delimiter?

[quote="ray.wurlod"]Is the record delimiter (that you have shown as "Z") also a linefeed? ...[/quote]

ray.wurlod · Post by **ray.wurlod** » Fri Jul 04, 2008 4:37 am

Try pre-processing the file to eliminate the linefeeds that follow "Z" characters, and rely on the definition of "Z" as the record delimiter.

You should also be able to edit the table definition to indicate that the column in question contains line terminator characters (linefeeds).

dkaranth · Post by **dkaranth** » Fri Jul 04, 2008 4:45 am

Hi Ray,

My requirement is as follows:

Some of the attributes/columns in my input file can span multiple lines(i.e. it has a line feed). And I should read them as is and transfer to another system.

If I remove the line feeds, then I wont be serving the purpose.

Any other suggestions?

Regards,
Deepak

ray.wurlod · Post by **ray.wurlod** » Fri Jul 04, 2008 4:51 am

Read again what I wrote.

Remove those linefeeds that follow "Z".