slotrans/whos_hiring

review tool for Hacker News "Who's Hiring?" threads

Concept

Each month, a new "Who's Hiring?" thread pops up on Hacker News. At least in theory, these are a good place to look for tech jobs. But, the threads quickly accumulate hundreds of comments spanning several pages, and reviewing them in a web browser is tedious and fragile. You can collapse threads to keep track of which ones you've dismissed, but if your browser ever crashes or restarts, or if you click between pages, you lose that state. There are some existing tools that ingest the posts in these threads and make them searchable, but I find they don't really give me the workflow that I want.

My goal was to be able to quickly flip through all the posts in a thread, and for each one mark it as rejected (irrelevant tech, bad company, distant location, etc) or save it for further consideration.

With an emphasis on "quickly": the review tool itself is built using dearpygui, which in turn is built on Dear ImGui, the immediate-mode GUI framework for C++. The UI is very fast, far faster than 99.9% of web interfaces. Though it is in theory possible to make a suitably fast web UI using the right techniques, I hate web UI stuff (can't CSS my way out of a wet paper bag), and I wanted an excuse to play with dearpygui.

Usage

(Quick caveat: I made this entirely for myself, it definitely works on my machine, so these instructions are a bit half-assed...)

This project uses Poetry, and includes an appropriate pyproject.toml. The declared Python version is 3.9 because that's what I used, but 3.7+ ought to work. You will need the SQLite .dll/.so, which is not always included with Python.

First, get scraped

python scrape.py --url "https://news.ycombinator.com/item?id=28719320" --output-dir foo

Then extract the data

python parse.py --input-dir foo --output-file bar.json

Now get clickin'

python review.py --json-file bar.json --db-file whatever.db

Your progress is saved automatically, exit any time. When resuming later, the JSON file is no longer necessary.

python review.py --db-file whatever.db

The db file is a SQLite database with a single table and can be manipulated with any appropriate tools.

It's ugly!

Beauty is in the eye of the beholder.

The code is ugly!

Without a doubt.