DSXchange

Hi all,
I have to read a txt file and load it to the database.This is daily cycle.
The file name is in the format say,SAMPLE_20091020.txt,SAMPLE_20091021.txt ..
I cant change the file name manually ...
How can i handle this situation of handling the file and getting the date from it.

Thanks in advance,
Raj

I think this is a common requirementYou can Parameterise the datepart of the file name.

If you are aware that what dates will come in files you can assign the value for date param from Sequencer and pass to the sequential file stage.

your File name in seq file property will look like

SAMPLE_#dataparam#.txt

Use the File Pattern option in the stage.

I searched for any topics regarding FILE PATTERN,
but i could find in dsxchange..
Can you please shre me the idea of using it or the thread that has the detail about it..

Thanks,
Raj

I searched for any topics regarding FILE PATTERN,
but i could find in dsxchange..
Can you please shre me the idea of using it or the thread that has the detail about it..

Thanks,
Raj

But, In that case how can it read yesterdays date's file? i think he had mentioned like daily files comes with the dates

I searched for any topics regarding FILE PATTERN,
but i could find in dsxchange..
Can you please shre me the idea of using it or the thread that has the detail about it..

Thanks,
Raj

Look into your online PDF "Parallel Job Developer Guide" and search for "file pattern".

Raj,

Do you want to disconsider the date part or you need to check the dates when reading from the file?

i have the date in the records itself, So there is no need to consider the date part...

But i should not run the yesterdays files again and again..?

Thanks,

But i should not run the yesterdays files again and again..?

Thanks,

Then you either need to rename the file once processed so it no longer matches your pattern, or you need to move it to another directory.

Is there any option other than renamiming the Production files....

Ummm... moving them to another directory? An 'archive' one would be typical. Harder would be to keep a list of filenames processed and skip those filenames when you see them again.