Comment by incanus77
9 hours ago
It's not trivial, but roughly: use AppleScript/osascript to get the URLs, but mostly pass them to a ~50 line Bash script which:
- Brings in the date path components for the dumped-to folder
- Makes a hash of the URL for an Obsidian doc (each tab gets their own doc)
- Uses Chrome command line (--headless --disable-gpu --dump-dom) to save a snapshot of the page contents
- Uses it again with --screenshot to make a thumbnail
- Create an Obsidian doc from a template
- If it's a single tab dump, pass -o to the script, which opens it in Obsidian for review
Lastly, I use the relatively-new Bases feature in Obsidian to make a nice "cards" view of the docs with their thumbnails.
I'm hoping to clean it up at some point and maybe release it, but it's one of those classic one-shot systems that just works for me for now.
> - Uses Chrome command line (--headless --disable-gpu --dump-dom) to save a snapshot of the page contents > - Uses it again with --screenshot to make a thumbnail
You could combine both of those into "run Archivebox somewhere and pass the URLs into that" (which is what I do for "URLs I save to Instapaper" - they go to my Linkhut, Pinboard, my Archivebox, and once I've fixed my code, to archive.org as well.)