Show HN: Bogrep – Grep Your Bookmarks
github.comOccasionally, I need full-text search on my bookmarks to find a specific bookmark I added some time ago.
Usually, browsers only support search of bookmarks in the website's title tag which is often not sufficient to find a bookmark.
Some browsers, like Safari, provide built-in support for full-text search, but can only be used for content that's already in the browser's history.
Bogrep imports your bookmarks from multiple browsers (removing duplicates), and then downloads and caches them in plaintext (without images or videos). These cached bookmarks are subsequently used for full-text search. Simple but fine idea. Wonder why nobody in the "PKM-scene" came up with this before. Small suggestion: could you make it available for other package managers like homebrew, since not everyone has the rust tool chain installed? This is a bit late.. but a lot of PKM-scene are using bookmarkers and readers with offline webpage saving and search already. I suppose this is enough support as a lot of scene seems more interested in optimizing the information organization tooling experience that comes with the separate apps and bookmarking formats. I think these sorts of automated personal knowledge gathering systems are an under developed concept. There is things going on though with projects like Spyglass which does local website indexing and search. Spyglass looks interesting, providing a unified search GUI for which I use a bunch of CLI search tools (ripgrep, ripgrep-all, and now bogrep) as well as some simple bash scripts. Somehow I still prefer some CLI-based tooling though which I can configure as needed instead of a fully-fledged GUI-based solution. Thanks, yep I'm currently working on proper support for Bogrep on macOS. > Wonder why nobody in the "PKM-scene" came up with this before. They have. Many, many, many, many, many times. could you post some examples? The only implementations I remember were quite hard to setup, buggy, and/or user-unfriendly UIs. No thanks. You have moved the goalposts from "nobody[...] came up with this before" to "implementations I remember were quite hard to setup, buggy, and/or user-unfriendly", and I generally have a policy to disengage at that point. "Moving the goalpost", "disengage at this point"? I was simply interested to find out if there are decent apps for that. Oo First post meant "nobody came up with something fulfilling that goal in such a clean and simple way, at least not that I know off". Phrased a bit short cause I was on mobile, sorry.
Not everyone on the internet is trying to win an argument. Was just curious to learn about apps I might not be familiar with. > Not everyone on the internet is trying to win an argument. You're right. Not everyone is. You definitely were, though, when you wrote this comment. Does anyone remember the name of the service or similar software that takes a screenshot every N seconds and runs OCR on it, so that you can search everything you saw on your screen at any time? I remember HN balked at the cost. A combination of that and this would make my life so much better, i think. That or giving away all of my computers. Are you talking about rewind.ai? (no affiliation) I have been eyeing the service, I tried the free version and it was pretty amazing. But I can't bring myself to pay the $30/mo. Hoping somehow they offer this for $10/mo someday. "Rewind compresses, transcribes, encrypts, and stores your data locally so only you have access." sounds good. Was your CPU usage ok when running the background service? CPU usage was definitely noticeable, but not more than 8-10% IIRC (on M1). I remember lots of tiny 0->10% spikes rather than a steady ~5% when viewing the CPU graphs. it says on m1 and m2 macs it uses 20-40% of one core, or up to 5% of all CPU. Local transcription is expensive (in my experience). It appears i am misremembering, because i thought it was for windows. Oh well. Using Bogrep, I could only find https://tenderowl.com/work/frog/ in my bookmarks which is a bit related, but doesn't seem to be what you're looking for. Sorry for the derail! bogrep + everything (windows filesystem search) + the screenshot "service" - i'm about 80% sure it was a service and not self-hosted - would completely obviate the need to remember anything ever again! It would also eliminate the hassle of firefox or chrome "forgetting" my bookmarks every few years in an apparently random fashion. I've been having HN "dropbox" moments about this... how hard can it be? I've hacked together OCR to use my phone to control my icom ic-7100 - the person who makes/sells the bluetooth serial ports compatible with icom Ci-V was out during the pandemic, and before, when i needed one, so i wired a raspberry pi to the Ci-V port, enabled BT file transfers on the pi, and using repeaterbook on the phone with GPS i could find the info page for a nearby repeater, do the 2 finger screenshot, hit share, and the pi would OCR it and get the frequency and offset. I never bothered to extract the PL tone frequency, because my radio can find that quickly on an active repeater. So there's a workflow, and i have spare machines to do the OCR and inserting into solr (or whatever). Interesting setup! Personally, I would be cautious to use such a screenshot service though. Even if self-hosted, it could screenshot credentials, environment files, and my password manager for example, which would then be stored in plaintext somewhere? So I would rather prefer to "keep control" by using a search utility when needed instead of scanning the background permanently.