Ask HN: Getting a single file from a public GitHub repo?

1 point·999900000999··8 comments

Let's say I have a public GitHub repo called Notes

The entire repo is only a Readme file, and a folder with 20 text files.

Is it possible to list the 20 text files and download one. If possible I'd love to do this via a Rest API, and without cloning the whole repo.

View on HN

For downloading a single file, I do it manually by clicking on the "Raw" button the using Copy/Paste. The raw button leads to a url starting with "https://raw.githubusercontent.com/" If that url is stable, then presumably I could've used wget instead of manual copy/paste.

You can get a single file, but you cannot list them, afaik, maybe I'm wrong? Cloning is not that expensive, consider how many deps you fetch in a typical day.

I actually do this to fetch initial / bootstrap content stored in a repo, and hold it in memory without the need to write it. I am considering changing this and exec'n out to git as a more secure / standard practice

I'm thinking you could probably configure GitHub pages to return JSON.

And then you'd have a call which would basically return a listing of all the files, along with the URLs to get them.

At that point, why not just put the listing as a file in the repo and just fetch that directly?

I'd have to generate that file and commit it though.

ah, you mean GHP producing something like a sitemap.json for your repo, hmmmmm

Exactly, I want to use the Git repo as a read only REST API.

You could always use something like S3 or GCS if GHP doesn't work

Reply on news.ycombinator.com