I’m not an expert, but I heard that all those polterGates problems with Word documents were due to Word’s using service symbols right inside the text. May be or may be not, but trust limit to Word effectively doing word processing job for long manuscripts is over for those ages.
BTW, the last newsletter from Mellel, arrived lately and saying that their Mellel Mobile for iPad works fine now with mobile version of Bookends. Didn’t check it out, but that’s what they say. If so, this is the great news for many.
Thanks, this looks really useful! I think Obsidian seems like a case where indexing makes more sense, since my Obsidian notes are living documents that I want to be able to continue to edit from Obsidian. In contrast, I only care about Zotero’s ability to easily capture citation data via the Web Connector and to produce .bib files so I can make use of CWYW features in LaTeX (which is where I write all of my own journal articles, classroom handouts, etc.). However, I don’t actually care that much about Zotero as a PDF reader/annotator (though perhaps I’m underrating it in this respect), so I don’t mind importing the actual PDFs into DT3 as long as Zotero can still perform the other functions well.
I appreciate the info about BE and I should probably give the free trial a spin to see how well suited it is to my needs and whether the better integration with DT3 is worth it for me. But to go back to what JJW said above, I don’t think this thread is the right place to discuss the merits of ZT vs BE
If you want to see a really comprehensively thought-out method for using Zotero and Obsidian you might want to watch Bryan Jenks on YouTube. He has obviously put a lot of time into it.
Many, many smart people on here with smart strategies for seasoning and eating spaghetti…
In DT, I index both my Zotero articles (which I actually download and organize in my own folder structure, synced via Box) and my Obsidian vaults. I do most of my PDF reading and highlighting in Preview on both Mac and iOS, and write in Word. I’m getting closer to finishing my second book (and I promise you, DT devs, there will be a shout out in the acknowledgements), and after more years of work than I care to admit publicly, I’ve never had any issues with data loss or file corruption.
I suppose my one spaghetti quirk is that I prefer to manually insert citations into footnotes, so use Alfred and zothero.
Just my pennyworth, but I (now) use Zotero to get the ref and the PDF (I used to use Mendeley to watch the DTP pdfs which worked great until the Mendeley files got corrupted, including all the back ups and that was that). Advantage here over DTP is that Zotero can rename the file by Creator - Year - Title. Then I right click on the file in Zotero to reveal location and then drag (import) into DTP. So I get the citation ref in Zotero, a renamed file that makes sense, and it’s in DTP.
Here is a a fairly typical article (Open Access) that gets renamed by Zotero. I did have a look to see if renaming academic PDFs like this is something that DTP does, but didn’t find anything, though some seemed close e.g. Auto-renaming PDF after OCR based on content
@Silverstone Lovely stuff. Can I ask how you managed it?
Here it is. And some posts earlier. But this script works with Bookends, I didn’t use Zotero and don’t know if it is scriptable.
I wouldn’t rely on the OCR data, or data you can get automatically from PDF or PDF properties. There is often just garbage, and there is no sure regex to find what you need inside the PDF, unless it is some “hard format”. The best way I found - is to find the ref and click the script.
Interesting script. I remember I had a look on it earlier.
It’s worth to note, that the name of the PDF does not have to contain exact title and author, you can select it in the text and run the script. It uses the name of the PDF if you didn’t do that.
But. I also remember that I couldn’t adopt this script in my workflow. And that’s why.
This script would be a great automation, if it removed the manual text selecting and/or choosing from the list. This would be great for quick batch operations. But since I need to do it finally by hand (high probability of inconsistent results otherwise), it is almost doesn’t differ from my regular workflow: just to look in, say, google scholar for that selected text, choose adequate ref and download it. It is opened automatically in Reference Management software (RMS) and connects to this PDF in DT (with all needed metadata), so that you can keep it in sync afterwards. Name PDF and populate other custom metadata, do not go and choose a format at GitHub - citation-style-language/styles: Official repository for Citation Style Language (CSL) citation styles. from 1000 of existing, but make it quickly yourself in RMS as you like.
If you don’t need RMS sync, and just want to rename PDF, this script will make it a bit more quick and nice.
Zotero fetches metadata from a variety of web sources. Apparently they have made a considerable effort to expand coverage beyond PDFs with DOI.
Scripting in DT can handle DOIs. If you need more than that, IMO it’s not very practical to write your own script to inquire multiple catalogs at once, especially when there is a working alternative. Anyway, don’t expect a script to get good-quality metadata consistently from within a PDF, because there are so many different formats.
While this is true, if your script can get hold of any publication identifier or the publication’s title, then services such as OpenAlex are of great help to implement a custom workflow (e.g., just within DEVONthink).
For small uses like this one, the OpenAlex API is free and requires no authentication. Its data sources cover important providers such as CrossRef, PubMed, PubMed Central & the Internet Archive. One can run queries not only by ID but also, say, by publication title. The returned data (example JSON output) contain a lot of metadata, including basic bibliographic info, foreign identifiers, the PDF download URL, and references to cited/citing/related publications.
So a script could just work with a publication’s title (e.g. retrieved from the clipboard, or the text selected in DEVONthink or a web browser), perform a query and download & rename the PDF itself. In addition, it could also set a host of DEVONthink metadata for the PDF.
That said, reference managers such as Zotero & Bookends perform these tasks very well, so a good integration with these tools may be the better choice. But I wanted to mention OpenAlex as an example service that could facilitate a powerful custom workflow.
OpenAlex, or that Github script - all of them require manual steps: find a search string (exact author, title, doi etc), select manually one of the results found. You cannot automate it, so it doesn’t add much to the value of existing workflows. And as @meowky said, it’s hardly a good idea to re-invent RMS functionality with custom scripting
