feat: Change type of DecryptedRoomEvent::event to Raw<AnyTimelineEvent>

[kaylendog]

• Change DecryptedRoomEvent::event to Raw<AnyTimelineEvent>
• Update usages to pattern match on AnyTimelineEvent::MessageLike where necessary

[codecov]

Codecov Report

❌ Patch coverage is 50.00000% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 88.57%. Comparing base (6c944a9) to head (b1ab5c4).
⚠️ Report is 33 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
...dk-base/src/event_cache/store/integration_tests.rs	0.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5512   +/-   ##
=======================================
  Coverage   88.57%   88.57%           
=======================================
  Files         339      339           
  Lines       93627    93627           
  Branches    93627    93627           
=======================================
+ Hits        82927    82934    +7     
+ Misses       6565     6558    -7     
  Partials     4135     4135           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[andybalaam]

Test failure seems to be an internal problem at codspeed

[andybalaam]

Looks good, and kudos for breaking into small commits. I like each commit to compile and pass tests, however, because it makes git bisect work nicely, so I'd prefer all in one commit if these don't compile separately :-)

No need to change that now though: we can squash merge this if we agree it should all be combined.

For the changelog: please can you list the pub functions that are affected and what users should do to if they use these methods?

[codspeed-hq]

CodSpeed Performance Report

Merging #5512 will not alter performance

_{Comparing kaylendog:kaylendog/decrypted-room-event (864fbeb) with main (6c944a9)}

Summary

✅ 31 untouched benchmarks

[codspeed-hq]

Unable to generate the performance report

There was an internal error while processing the run's data. We're working on fixing the issue. Feel free to contact us on Discord or at [email protected] if the issue persists.

[Hywan]

Thanks for the patch. Can you explain me why you need that? The changes are good and nice, kudos for that, but you forgot to explain the motivations for that ^^.

[kaylendog]

Thanks for the patch. Can you explain me why you need that? The changes are good and nice, kudos for that, but you forgot to explain the motivations for that ^^.

It's fundamentally about being able to avoid casting when implementing encrypted state events. At the moment, using Raw<AnyMessageLikeEvent> requires that a decrypted event can never be a state event (unless it is unchecked-casted). There are a couple of places in my older #5456 PR where I needed to cast between Any(Sync)StateEvent and AnyMessageLikeEvent and vice versa. This change would allow me to avoid the casting when decrypting state events.

I would have liked to feature gate this behind experimental-encrypted-state-events but that would require conditional compilation flags all over the place.