Ready to Test: Improvements to File Cache / PDF Point Snapping in Background

We just deployed the latest to the TEST server:

1. Large files are properly caching - this is a big deal for speed. Most large PDFs were not caching properly before until we force the browser to cache. I guess browsers just decide "That file is too big, we won't cache it." So we had to write some work-arounds to force it to cache.
2. When uploading plans, we force cache client side after we verify that it uploaded to the server. We're intercepting HTTP requests and it doesn't have to re-download (at least for the user that uploaded the plans). Big speed improvement and bandwidth saved.
3. PDF point data for snapping is extracted using web workers (in the background). This should be a significant speed difference (at least avoiding browser hanging) especially for large plans.
4. Smart Select uses background web workers to extract text and vector data before sending to AI. Probably not much speed gain in terms of time, but offloading to a web worker will allow the main window to allow zoom/pan more smoothly while smart select is sending to AI. Overall, should just feel smoother when using smart select.

If you get a chance, please test on the TEST server and let us know if you run into any issues.

Regarding testing, it should just be overall smoother switching pages and navigating while it's processing PDF data, especially on large plans. One large set of plans was previously hanging for 10 seconds while processing PDF point snapping data for 3+ million points. Now it loads much more smoothly and processes that data in the background.