November 11, 2010 at 8:23 am
Could you have a task that runs first which renames the file, dropping the h:m:s bit and then feed that into your dataflow as normal?
November 11, 2010 at 8:37 am
Hi Phil,
Thanks for the reply. But this folder has files from every day, so there would be a lot. How would i rename the particular file that i want?
November 11, 2010 at 8:48 am
Accepted good practice is to move a file which has been processed to a separate archive folder - then you don't get a mixture of processed and unprocessed in the same place. Could you adopt this - becomes easier then.
Also, if you know that the final six characters of all of the filenames in the folder are hhmmss and can safely be removed, why not rename them all via a script? (You'd need to put something in the script to recognise those filenames which have already been processed - eg if the ninth character from the right (ignoring file extension) is non-numeric, do not rename...
November 11, 2010 at 10:14 am
Here is what i ended up doing.
Since there would atleast one file uploaded everyday, before i call my routine, i have another package that wipes out the files from this directory. So, when the file is dumped again, there will be only one file in that directory.
Using Foreachloop task, I am getting the file from that directory and storing that in a variable. This variable is then what the Flat File Manager looks at to import from.
For this project, it was okay for me to wipe those files out since someone was planning on cleaning that directory anyway. And this directory is dedicated to just one client who would be uploading only one file after midnight.
Thanks for giving me ideas.
Viewing 4 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply