My convoluted workflow for journal articles
Here’s what I’m currently doing for finding, organizing, and reading journal articles:
- Search via Google Scholar or PubMed.
- Add any vaguely interesting articles to Paperpile. Paperpile will helpfully download PDFs for most of the articles, and it makes it easy to find any PDFs it can’t download — its workflows.
- Once I’ve exhausted my search queries, go through Paperpile and label everything I just found with a “Not processed” label.
- Go through the “Not processed” label and star anything I want to read or keep for later.
- Use Paperpile’s Google Drive sync feature to grab all the PDFs for starred articles (it puts them in their own subfolder in Drive).
- Unstar all the selected references in Paperpile (just click on one star while all are selected). Then remove the “Not processed” label.
- Use a hacked-together script to clean up filenames of the downloaded PDFs into my preferred format:
$last-name-of-first-author $year - $title.pdf. Manually fix up any ones that aren’t automatically corrected.
- Put the PDFs into a folder hierarchy in Google Drive (my preferred way of syncing/sharing PDFs). Dropbox would work fine too.
- Use Bear to take notes in Markdown, referencing the properly-formatted filenames of PDFs (which I can search for to find the PDFs quickly if needed).
- Use Documents by Readdle to read/annotate on my iPad, which syncs back to Google Drive. Use Highlights on my Mac to pull out my highlights in Markdown format if needed.
I would love to use Paperpile (or any reference management database), but I’ve been burned too many times by disappearing or buggy reference management software. I like files in folders — my organization (the folder structure) and references (i.e. filenames) are future proof, and everything can be backed up easily.
The folder can also be indexed and searched by DEVONthink, which allows for a bunch of cool search features.
Note: I’ve written previously in much greater detail on reference managers.