Tags on annotations?

I’m trying to figure out how to link ideas to PDF annotations. The idea of adding a tag to a PDF annotation would probably work for me, but it seems that tags are only for files.

I’m discovering the past discussions about TheBrain, and wondering if thoughts in TheBrain can use annotation links. Or is there some other feature of DT that does what I have in mind? Imagine being able to see all PDF annotations by tag, across all files. That’s the general idea. Any way to do that?

Also, I realize I could paste all my annotation links into a shared file, but that’s not really the approach I’m looking for, I think it wouldn’t be scalable.

That’s right, only database items can be tagged but not the parts or annotations of documents.

That’s a very desirable though advanced feature.

The only app I know of which can do that is LiquidText; if you are looking to tag annotations then LiquidText probably also has a whole lot of other features you would appreciate.

You can store your LiquidText files in Devonthink and then use “Open in” to launch LiquidText. And you can put Devonthink Item-Links inside LiquidText files.


It looks like the annotations I make in DT are not stored in the original PDF. Are all of my annotations lost outside of DT? How do I get them into LT? Manually recreate them?

It looks like I have to export to PDF to have my annotations available out of DT, is that the intended workflow? Export the PDF after making annotations and then open in LT?

The annotations are stored in the PDF and should be immediately visible after saving in other apps (e.g. Preview or Quick Look).

I am not sure about whether DT annotations are visible in LT - I never tried that.

When I have needs this complex I make LT my tool for all annotations on the document(s). DT simply becomes the repository for the LT file(s).

I tried out LT but at the moment it looks like there’s too much of a barrier to try to use this together with DT. I also have video and audio files that I link to, and so I have situations where I want to outline an idea or set of ideas with the supporting annotations and video and audio references.

I think DT could remedy this situation really easily by adding a new document type which is more free-form, in the way that LT is free-form. The way that LT allows me to draw a shape to cut out of the PDF and then place that into the workspace (annotation file) is really useful, if DT had this it would make annotation files a lot more useful. Having each link show a preview of what it links to is very helpful when coming back to review the annotation file. Imagine having your own writing and lots of little previews along the margins of the document so you can see a snippet of every reference. That’s what LT offers and DT could easily accomplish with a new document type. Maybe call it “canvas” and allow the user to drag elements to place them.

I think you’re underestimating what it takes to implement such a thing. :thinking:

Is there a reason you’re not just exporting your annotations to a separate file using the “summarise annotations” function?

Seems to me your problem stems from leaving your annotations in the PDF: you want to tag annotations separate to the PDF, but you’ve left your annotations in the PDF file and that isn’t currently possible. It is easily fixed by creating a separate annotations file and then using it however you like.

(It doesn’t have all the functionality you want, but it covers your actual problem = I want to tag my annotations and then manipulate them independently of the PDF).

1 Like

This could be done with an HTML document. The snippets could be img elements, absolutely positioned, and the user can drag them around to place them within the document.

You would need a way to gather up the annotations and then paste them. You could add an option to draw the shape within a PDF, and then the resulting element gets added to either the paste buffer, or give DT a storage area where these annotations can be reviewed, pasted, or deleted.

No, sorry. I think you’re not understanding me. I want the annotations to remain in the PDF’s. I simply want to have my annotation files show me a preview of what I’m linking to. Picture a letter written out, and for each citation, there is a little clip of paper which shows the thing I’m referencing. Think of it as a more insightful link. Like a thumbnail, it shows you a preview of what is behind the link.

The request is noted but as I said, this isn’t easily done, especially since you are talking about generating some kind of image of the annotations.

1 Like

I was able to make a successful test using an HTML file. I added width="50%" to make the image sizes more appropriate for the window. Then I wrapped each with an anchor tag and pasted the annotation link in for the href value. I’m really glad this works!

<html dir="auto"><head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<body style="color: white; margin: 1.5em; font-family: Times-Roman, Times; font-size: 12px; overflow-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;">

    <p>Laura Garwin said they didn't have a controls</p>
    <a href="x-devonthink-item://5274C665-FE3E-4754-97B1-930C443E2F11?page=108&amp;annotation=Highlight&amp;x=66&amp;y=213">
        <img width="50%" src="file:///Users/rob/Databases/CCA.dtBase2/Files.noindex/png/7/Screenshot%202023-11-13%20at%208.44.42%20AM.png">

    <p>FP listed Pt blanks in their paper</p>
    <a href="x-devonthink-item://0B7C8B66-04AA-4395-8B27-8BC558A7B8AC?page=4&amp;annotation=Highlight&amp;x=72&amp;y=159">
        <img width="50%" src="file:///Users/rob/Databases/CCA.dtBase2/Files.noindex/png/1f/Screenshot%202023-11-13%20at%208.23.26%20AM.png">

Here’s what it looks like.

You can see the image files next to the HTML file. They were just drag-and-dropped into the group directly from the macOS native screenshot utility.

This works pretty well but does require me to edit HTML manually. I’m looking into AppleScript and the DT AppleScript API to see what features are available to me to streamline the process. I will probably move the images to a different group to keep them stored safely but out of sight.


It sounds like I could use a smart rule to run a custom AppleScript when my screenshot is added to my designated group. The script would copy the annotation link for the currently selected annotation and populate the URL field with the annotation link for the newly added image file. Then I would probably use another AppleScript to automate the process of generating the correct HTML img tag and link in my HTML file.

Is that doable?

It’s not a wise idea to link to items in the internal structure of a database. The location can change, leaving you broken links.

And yes, it’s possible to use an Execute Script smart rule action to run an AppleScript when you drop a screen capture into a specific group.

Are you saying the images should be stored in the OS folder structure? Since I keep everything in Dropbox, that could work.

No, I’m not saying that. You can actually use DEVONthink’s item links as an img src in HTML in our app.

And I strongly encourage you to read Help > Tutorials > Indexing Files as well as the In & Out > Importing & Indexing section of the built-in Help and manual before going down that trail.

Thanks for the html code example
I use Formated Note files (html code), edited with an external text editor
and the DT file link instead of the OS file link in the src line

Could you elaborate for me on what you meant by this? Do you mean that it’s a bad idea to keep annotation links in an HTML annotation file?

@EthicalEgret wants to use the “annotation link”, sounds like its item link which is of course fine.

1 Like