Importing and updating

First, a little background. I don’t personally subscribe to all the technical journals I use in my work, but my university library does. If I come across an online article that I can’t access, I’ll clip the abstract page to DevonThink, flag it with a label, and send it to a smart group based on that label. Every so often, I’ll export the smart group to my laptop, trundle off to the library, and download the underlying articles.

(It would be nice to automate the download process, but since every journal is a little different, I suspect that requires more complex scripting than I’m willing to do. Or maybe DevonAgent? But that’s another post.)

(Due to the vagaries of licensing, the article retrieval must be done while physically at the library. Since I’m not a student, I can’t get off-campus access.)

(I don’t want to just copy the entire database to the laptop, as it’s too big to do so efficiently and I’m worried about data corruption.)

So far this works great. The problem comes when I update my main DevonThink database. I would like to automatically identify and delete (or at least un-label) the abstract pages for the articles I retrieved, keeping in mind that these are scattered across numerous groups in several different databases.

The fully automatic approach would have me flag the abstracts with a different label as I retrieve them, then Boolean AND the pre- and post-retrieval abstract groups. Except I can’t figure out how to do that, and even if I could DTP won’t seem to let me delete items from a smart group.

Any suggestions?


Maybe it’s possible to search for all pages/articles via a global smart group in the sidebar of DEVONthink Pro (Office) 2? Assuming that the pages/articles have a common attribute of course.

It’s possible but you have (at least at the moment) to use Data > Move All Instances To Trash as smart groups return only one instance (like searching does).

Hmmm… Let’s be a little more precise about what I’m doing.

  1. Capture abstract for interesting article. Label it “Important” and create a smart group for that label.
  2. Export the “Important” smart group for remote use. Re-import it to an instance of DTP on my laptop.
  3. Download some, but not all, of the articles in the Important group. Label the abstracts for the captured articles as “Done.”
  4. Re-import the “Done” abstracts to the main DTP Inbox.
    (Labels do seem to survive the export/re-import process, fortunately.)

What I would like to do, but haven’t figured out how, is

  1. Find all items for which there is both an “Important” and a “Done” instance.
  2. Delete these items.

But as I write this, it occurs to me that the Duplicates smart group might do the trick nicely. Time to do some experimenting…


Try using State as a searchable characteristic of your abstracts, as well as the Important Label.

When you have downloaded a complete paper, check the State box of the abstract (Done).

Now you can search for Important and Done in the full Search window. The results are your candidates for deletion.

Ah ha! I never use State, so I’d forgotten it existed. This should do the trick nicely.