Shell Programming and Scripting

I have an HP-UX server that runs a script each night. The script connects to an SFTP server and downloads all xml files (if any are present) from a certain folder, and then deletes the files from the SFTP server. So sometimes it will download a new file, sometimes it will download 2 or 3 new files, and sometimes it will download zero new files.

The files are all downloaded to a folder on my server that contains dozens of other similar files.

If/when a new file gets downloaded, I need to run other scripts against it, to processes the data. I'm doing it manually right now. Does anyone know of a better way to identify 1. Whether or not any new files are present. 2. What the file name is of the new files.

My thought is to capture the output of "ls", then run the sftp download script, then capture the output of "ls" again, and run a diff on the two. That seems a little clunky though, there must be a cleaner way.

Any ideas are appreciated!

Why not just drop the files in an initially empty directory?

Regards,
Alister

Code:

find /my/folder -ls | grep "`date +'%b %d'`" | grep xml | cut -c 69-

FYI I cobbled something together. This looks like it'll give me what I need. It will list the xml files in the directory that have today's date on them.

Wouldn't the time stamps reflect the download time? And, I'm not sure if verbosity can be increased so you can capture a complete transfer log.

Correct. Time stamps reflect the download time. So if no new files are downloaded, it will be an empty list. If 3 new files are downloaded, it will return the names of those 3 files. Exactly what I need. Files are downloaded at 3 am. The processing script runs at 7:30 am. So any new files present will have the current date.