Workflows for file import

adam1

I'm looking for some suggestions on how to solve a very specific problem.

When my users upload an image into TS, I want them to do it through a workflow so I can gather some very specific information as we go along.

Here's what I want to the workflow to do (roughly):

Now first of all I have to do this in 5.0 becuase my operations guys won't be able to do the upgrade to 5.5 for a little bit and I want to do this now.

It seems to me that one could trigger the file upload and new DCR using external tasks since "everything" in teamsite is now a URL.

From what my local TS specialist tells me, automatically setting the name of the new DCR is something that 5.5 does but not 5. I think we should be able to get around it still by using a cgi task that creates a dummy file in the correct location and sets the extended attributes (for which there are CLTs) before triggering the external task that edits the DCR. This might actually be a good way to go since it might allow me to sniff out things like the mimetype, and dimensions of the graphic (all data in my DCR) ahead of time.

Now I'm convinced this is possible but I know its probably a little out there on the scale of things. My TS specialist gets a little wild eyed when I walk into her quad and sit down so I'd like to present her with a basic outline of what to do for the harder parts.

Can anyone help me?
Please?
Adam

Migrateduser

That seems ok.

In 5.0 implementations creating a new DCR and then bringing it up for editing was the only way to do things.

You may want to consider offloading items to a custom menu item or using default TeamSite functionality.

A way that has been implemented before in 5.0.1
1) User imports files (via TeamSite GUI)
2) They select file, and run "new image" workflow menu item.
3) This menu item creates a DCR (possibly by copying a template DCR), and then adds the relevent fields culled from the file. It then immediately pops up a DCR edit screen, and starts a workflow.
4) output file is generated as the next step in workflow...

This approach provides a seemless user approach, however it requires a fair amount of custom coding.

What is in the output file? Is there a reason why data must be gathered in a DCR?

A few alternate approaches: (extra user step but easier to code)
1) A CGI task lets user select the file and uploads it to TeamSite. After uploading it reads the file and creates the DCR and adds it to the workflow.
2) The user can manually edit the DCR in the workflow.
3) As the next step the page is regenerated.

Or a metadata approach:
1) File imported to TeamSite.
2) External task writes extended attributes.
3) User brings up metadata screen and adds additional data
4) Output file is generated from perl script by reading data.

For creating the DCR, the easiest approach would be to first manually create a DCR. This DCR can be used as a 'template' for other DCRs that will be automatically created. If you are comfortable with XML parsers, you could simply parse and replace the fields. Otherwise, you could manually edit the DCR and balnk out all the unneaded fields. You could then use regular expressions to substitute in new values:
$dcr =~ s|<image_size></image_size>)|<image_size>24</image_size)



http://devnet.interwoven.com/site.fcgi/techlib/048807