Improve no search results page by caching "search inside" response

[RayBB]

Problem

The video explains it well. In short, we don't cache the "search inside" request so when people go to different search types with no result, it loads each time for 1-2 seconds. A small improvement would be to cache the search inside partial endpoint for a while. Maybe 1 hour? Seems unlikely to change often but also don't want to make a cache last a long time for no reason.

Video (with audio)

https://github.com/user-attachments/assets/cfd6c47d-8dc2-4436-aa54-553956aa4c10

Reproducing the bug

1. Go to https://openlibrary.org/search/authors?q=%22cory+doctorow+1234%22
2. Click the different tabs, see the "search inside" loading bars

• Expected behavior: Caches and therefore loads fast on subsequent requests
• Actual behavior: Loads slowly on each reaquest

Context

n/a

Breakdown

Requirements Checklist

• [ ]

Related files

*

Stakeholders

• @cdrini do you have any thoughts on this one?

  ---

Instructions for Contributors

• Please run these commands to ensure your repository is up to date before creating a new branch to work on this issue and each time after pushing code to Github, because the pre-commit bot may add commits to your PRs upstream.

[cdrini]

Nice good idea! I think we should cache this by setting the HTTP headers to cache this response on the client side for a short time (maybe like 5min), but only for the FullTextSearch partial and only if there's no error.

(Note shorter time, since it's rather unlikely that a user will hit the search inside multiple times, so it's not worth taking up valuable storage space on their machine for this cache for any extended period of time!)

[asharan2buff]

Can I take this up?

[weifanwu]

Hi @RayBB, I am new to this project, i wonder if I can take this issue? thanks!