Exclude download, redirects, reload, embed and framed link visit types from the get_top_frecent_site_infos query

[gabrielluong]

Fixes #3503

Pull Request checklist

• Quality: This PR builds and tests run cleanly
  • automation/all_tests.sh runs to completion and produces no failures
  • Note: For changes that need extra cross-platform testing, consider adding [ci full] to the PR title.
• Tests: This PR includes thorough tests or an explanation of why it does not
• Changelog: This PR includes a changelog entry in CHANGES_UNRELEASED.md or an explanation of why it does not need one
  • Any breaking changes to Swift or Kotlin binding APIs are noted explicitly
• Dependencies: This PR follows our dependency management guidelines
  • Any new dependencies are accompanied by a summary of the due dilligence applied in selecting them.

[gabrielluong]

This follows the allowed types (complement of the excluded types) used in the "History" screen of Fenix https://github.com/mozilla-mobile/fenix/blob/8d1bd10e6b00d270cfde13719fa91a8d8f6da7a3/app/src/main/java/org/mozilla/fenix/components/history/PagedHistoryProvider.kt#L36.

[codecov-commenter]

Codecov Report

Merging #3505 into main will decrease coverage by 0.00%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##             main    #3505      +/-   ##
==========================================
- Coverage   76.53%   76.52%   -0.01%     
==========================================
  Files         230      230              
  Lines       30753    30754       +1     
==========================================
  Hits        23536    23536              
- Misses       7217     7218       +1     

Impacted Files	Coverage Δ
components/places/src/storage/history.rs	95.49% <0.00%> (-0.07%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b45bb56...b6376ef. Read the comment docs.

[linabutler]

Great, thank you so much @gabrielluong! I left a comment about performance concerns (one table scan to two), but I think we should get this landed now to unblock you, and file a follow-up to figure out an overarching plan for top sites storage in Places.

[linabutler]

So, I'm wondering about the performance implications for large history sets. We're already doing a full table scan on moz_places, since we don't index on anything except frecency; now we'll need another table scan on moz_historyvisits.

I don't want to block your PR on that, but it's something to think about if we want to push top site storage down into Places. Having a top_site tri-state column in moz_places (always show, never show, or show based on frecency) would let us do that without the second scan—we'd pick the "always show" sites first, and then fill the rest based on frecency up to the limit. "Never show" would be for sites that the user explicitly removes from top sites, as well as ones that have forbidden types—and we can determine those cheaply at insert time! 🎉