Tinderbox Training Video 58 - Annotating with Highlights App, Tinderbox, DEVONThink, and Zotero

Ok. I reactivated my Highlights 1 app. Will only use it to extract the markdown files and annotate in bookends. This way, luckily, I don’t have to pay the subscription. Is there a possibility that you could share your stamp code with us? Will have to rewatch the video to get closer to what you do.

Regarding your problem with exploding at the *** delimiter, why don’t you export from highlights every citation in an individual file as this is an option?

take care,

J.

All the sample code is in the attached file.

Thank you so much. Totally overlooked that. Sorry for the inconvenience!

No problem. Let me know if you see any need for modifications. I can’t remember if I’ve uploaded the refinements I’ve been making.

what should this OnAdd action do:
action($Text("cnSource"));

This action first looks for a note named “cnSource”. Having found that note, we get its text. Finally, we treat that text as an action to be performed.

In essence, this lets us write the action in the text of a note, which might be more convenient for lengthy actions than the inspector window.

2 Likes

Hi @satikusala , I recently logged back into the TB forum and saw that you asked a question about my work process as it relates to OmniOutliner.

Basically, I’ve been annotating my PDF files with MarginNote, a PDF annotation application that can export the just annotations (i.e., the notes generated from highlights, along with a MarginNote-created URL for each annotation, and the notes and hashtags attached to each annotation/note). MarginNote allows users to export these annotations in different file formats, namely: .rtf , docx , iThoughtsX , and OmniOutliner export files.

I found that it could: (1) Export MarginNote-produced annotation in an OmniOutliner export file format, in which the annotated notes are segmented by the _note , Tags , Book Title , URL columns (2) then I use a script to reformat the OmniOutliner file – in tandem with a TaskPaper file that’s has a set of categories (first level), and their corresponding tag (second level) – so that file is populated with new columns (categories) whenever I’ve annotated notes with their corresponding hashtags (I can further explain this for you if you’d like), and then (3) I’m able to open the reformed file as an OPLM in Tinderbox.

I should add that I wouldn’t have been able to do any of this without the incredible and generous assistance of @mwra

Unfortunately, MarginNote’s developers haven’t been updating the app, and so it has become much more difficult to use. So… I’m now seeking a PDF annotation app that can:

  • Allow me to highlight and tag annotated notes (that’s crucially important for my work process);

  • Export just the annotated notes in OmniOutliner – or , at the very least, in a format that I can open in OmniOutliner, and will mirror the MarginNote - OmniOutliner file output (i.e., in which the annotated notes are segmented by the _note , Tags , Book Title , URL columns);

  • Also export the annotated files in PDFs that don’t have a prohibitively large file size and (if possible) have URL generated links for the PDF annotations that can function offline;

I’ve looked at Highlight long ago, and it doesn’t look like it exports files as OmniOutliner files (though maybe there’s a conversion process that can work), and I’m not sure if it can incorporate tags into its annotated notes.

2 Likes

same problem for me. I like the user interface of MarginNote more then those of any other annotation app I found. The sync between my Mac and the iPad works most of the time without a problem. My problem starts with DevonThink. This is my final repository for the annotated PDFs and the MarginNote format is not usable at all. Same with Bookends and MarginNote.
I tried Highlights and a few others - most of them offer a flat PDF format usable in other apps, but the UI or the sync… I still use MarginNote and this is a dead end until the developer will react - seems as likely as not :frowning:

2 Likes

Same experience. It has been excruciatingly frustrating. I’m sad to hear that Highlights doesn’t provide a viable alternative, especially since I was hoping it would provide one hopeful solution! Re: Devonthink, I simply make a few different output versions, and dump one in there for archival purposes. Of course, I sure wish MarginNote could make a PDF output size that wasn’t so bloated – and could work off-line.

What is the file format that OmniOutliner wants? I’ve found that Highlights has met all my needs

The step from my annotation app to TBX is the transformation of annotations (text written by someone else) into my own excerpts. I may add a reference to the original text but to write the ideas down in my own words is an important step for me. The map view in MarginNote helps a lot with this task - nothing like that in Highlights. Highlights is close to what DevonThink already offers - no advantages for my work.

1 Like

When I was helping @jprint714 with MarginNote to OmniOutliner ISTR we were using MarginNote’s oo3 (OmniOutliner v3) export which was arriving in OmniOutliner with some problems. The root cause seems to be that MarginNote’s designers appear to see no real need for (ROI in) decent export. MarginNote’s export is also dumping discrete fields of source data into a single text block, which again shows no understanding of why people export. So, even if Markdown can export data, careful post-processing is needed before the original data is back in addressable fields.

Still, I’ve seen many knowledge apps go through this myopic phase of “why do you need to export”, the implication being he tool is so great you don’t need to use anything else. Thus export (and import) are viewed as unwanted enhancement cost rather than, as they should be, a core feature. Also odd how people want to offer a (usually buggy) less common format rather than simple CSV/TSV. The latter has its issues but they are well known and community help is thus much easier to find.

When looking at a cool utility, it is often useful to start one’s evaluation in the manual and by looking at import/export. Unless you know you’ll only do work in that app and the result will never leave the app, the value of the app falls somewhat, regardless of any cool features. Caveat emptor.

2 Likes

Thanks for that explanation, @mwra.

I wish there was another annotation app out there that enabled me to highlight and tag annotated notes, and produced an output that would better interface with my work process (which eventually flows into DEVONthink and Tinderbox). So far I haven’t found one, but I’m eager to find an alternative solution. Really.

@satikusala - I’m glad to hear that Highlights has met all of your needs! I’ve written to the developer to ask if it can incorporate tag - annotated notes – and to see if Highlights can produce an OmniOutliner export. (I’m happy to walk you through how and why I use this process, if you’re interested.) So far, I haven’t heard back from him, and don’t see that there’s a user forum.

Have you ever tried Skim? https://sourceforge.net/projects/skim-app/ It is a long while since I used it, but it is free and is quite capable, though I don’t remember it having the particular capabilities you seem to desire. But it might be another to add to the list of options. Personally, my preferred PDF reader is PDF Expert https://pdfexpert.com, but again, it doesn’t have the particular features you mention.

2 Likes

Thanks. I actually do have Skim but, as you say, it doesn’t have the features I’m seeking, alas… Ditto that for PDF Expert. But thanks so much for the suggestions!

I’m running a few months behind here Michael, but congrats and thanks for making such an impressive (family of!) video. This one here really struck a chord. While no longer an academic, I still have need to create academic-like research papers. I’ve struggled for years with Tinderbox: some successes then frustration, but I kept at it, sensing that it really IS the tool I need. Over the last 1-2 weeks, I finally broke thru with the help of this “Annotating” video.

I’ve been using Highlights for a few years now and collecting info in DT, but until digging thru your Video 58 couldn’t make the connections I sought with TBX. I’ve successfully used your Parse Highlights App Annotations script. But I’ve run into a few snags on which I hope you can set me straight.

(1) I can parse out the article title, authors, the URL and annotation type and color, but haven’t figured out how to resolve the full journal reference. The line comes into TBX as, e.g.,

Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2019, Volume 234, Issue 4, Page 936–949

I am able to peel off the "(bookends://sonnysoftware …) into its KA ($URLBE - I use Bookends rather than Zotero), but haven’t figured out how to resolve the [Proceedings …] – my hunch is that there’s a wrinkle I’m missing - am I incorrectly using special characters? - in your parsing script:

if($Text.contains("### \[(.*?)\n")){
   $MyString=$1
};
$URLBE=$MyString.following("](").replace("\)","");
$MyString="";

(2) Images: I have the captured image (from Highlights) in DT, but when I drag it into TBX and apply the script, I get the blue question mark image. I think that usually means the app can’t resolve image location or type. Got any idea on what I’m missing? When I drag the image-containing MD file into TBX, that last line reads (pre-Script), e.g.,

Page 937, Orange, Square:

I modified (for my DT directory) the last line in your “Square” script to read:

if($AnnotationType=="Square"){
   $ReadOnly=false;
   $Prototype="pImage";
   $DisplayedAttributes=;
   if($Text.contains("!.*")){$MyString=$0};
   $Text=$MyString.replace("assets/","/Users/garyholland/Library/Application Support/DEVONthink 3/Inbox.dtBase2/Files.noindex/jpg/^value($DT3ImgDir)^/");

Have I missed a definition of $DT3ImgDir?

I’d appreciate any insights you can provide. Thanks

Looks like I lost the directory information for the image …

From DT the line reads
Ater a 1 and paired brackets: (assets/Page937Image304.7483887381276_383.4196014249501-236.18080054274083_123.63189026458713.jpg)

Path to that image per Finder is /Users/garyholland/Library/Application Support/DEVONthink 3/Inbox.dtBase2/Files.noindex/jpg/1/Page937Image304.7483887381276_383.4196014249501-236.18080054274083_123.63189026458713.jpg

Ah, yes…I must admit that I’ve refined my process quite a bit since I produced this video–I’m moving at the speed of data, i.e. fast.

As for getting the full-citation into Tinderbox, I actually have another process for this, a script I’ve developed to process an RIS file. As for images, yes, this is a total pain, one that I’ve not fully grokked yet because I’ve not been able to resolve a Devonthink link to a path that the HTML src tag will read. As a result, I have a different approach now for handling images–I put them in a separate file direction and have a dedicated prototype.

If you’re up for it, let’s hop on a call and we can compare notes. I’ll DM you.

1 Like

A belated thank you, Michael: that was a generous chunk of time you gave me in our chat. It helped to see what you’ve done with your industry map - inspiring stuff. I’m working to build out my own map, taking advantage of some of the insights you’ve been providing.

I see that some recent posting on the forum takes up the file path vs DT’s links: I’ll need to review them to try to resolve some of my image incorporation challenges.

1 Like

Did you see my latest:

From an industry analysis perspective, this one is pretty cool!!!

1 Like