A language-specific search engine for the living web.
https://ohana.nicholasgarcia.com
- Go 80.8%
- TypeScript 16.5%
- Nix 1.3%
- JavaScript 0.8%
- Shell 0.3%
- Other 0.3%
| docs | ||
| services | ||
| .env.example | ||
| .gitignore | ||
| .gitmodules | ||
| docker-compose.yaml | ||
| flake.lock | ||
| flake.nix | ||
| LICENSE | ||
| README.md | ||
Ohana 🌿
A language-specific search engine for the living web.
Motivation
This search engine is designed to do two things:
- Facilitate search beyond mainstream platforms (Reddit, Weibo, etc) into the re-fledgling indie web (e.g. personal websites, Neocities pages, niche forums, etc).
- Allow for language-specific filtering of search results.
In other words, this tool helps you browse the living web in your target language.
Timeline
v 1.0
- Fully functional web crawler
- Non-pornographic domain blocklist
- Upper bounds on site quantity based on arbitrary target
- Embedded SQLite queue for BFS
- Give embedded SQLite queue a Docker volume for data persistence across restarts
- Move domain-based crawler-specific enqueing from RAM to SQLite (perhaps add a column to the queue with the crawler ID the domain belongs to for tracking purposes?)
- Implement lexical search via pg_textsearch
- Implement a minimal client-facing Spring Boot service to query the database and display search results
v 2.0
- Basic ranking engine functionality
- Crawler upgrades
- Pornographic domain blocklist
- Source domain blacklist from v2fly/domain-list-community
- Upper bounds on site quantity based on "indie-ness" ranking
- Client
- Divide search results into multiple pages like Google
- Add CSS styling
- Add "recently added" section to homepage
- Add "most visited" section to webpage
- Add lightmode-darkmode
- Add "featured website" section
- Add privacy policy
- Add manifesto/about page
v 3.0
TBD