Cache pip dependencies using requirements.txt as cache key

[jovnc]

Summary

Fixes #68

• Hash your requirements.txt to generate a cache key
• On a cache hit: restore the pip download cache before pip install runs, so packages are installed from disk rather than re-downloaded
• On a cache miss (e.g. first run, or after requirements.txt changes): run normally and save the result to cache for the next run

Note: The arch-build job in publish.yml was intentionally left unchanged, it runs pip inside a Docker container (archlinux:base-devel), so setup-python's host-level cache doesn't reach it.

Verification

• CI action to download pip dependencies should be faster (from ~20s down to ~1s)

[Copilot]

Pull request overview

This PR aims to speed up CI runs by enabling pip dependency caching in GitHub Actions and by pinning dependencies in requirements.txt so the cache key changes only when dependencies change.

Changes:

• Pin all dependencies in requirements.txt (moving from loose specifiers to exact versions).
• Enable actions/setup-python pip caching in CI workflows (test.yml, test-pr.yml, publish.yml).

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 1 comment.

File	Description
requirements.txt	Converts dependencies to a fully pinned list to stabilize cache keys.
.github/workflows/test.yml	Enables pip caching for the main test workflow.
.github/workflows/test-pr.yml	Enables pip caching for PR-targeted E2E tests.
.github/workflows/publish.yml	Enables pip caching for publish/build jobs that run pip on the host runner.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.