chore: Duplicate ANRTracker classes

[philipphofmann]

This is the first step for #3492, which is part of the EPIC AppHang improvements #4261.

The changes for adding the option to ignore non-fully blocking app hangs are bloating up, and the PR would be huge. Instead, we first duplicate the AppHangTracker classes, and then we make smaller digestible PRs. Once we're happy with the AppHang changes, we delete the old V1 version and replace it with the new one. We could also run a beta with the V2 first before changing the algorithm for everyone.

#skip-changelog

[codecov]

Codecov Report

Attention: Patch coverage is 97.35577% with 11 lines in your changes missing coverage. Please review.

Project coverage is 91.568%. Comparing base (d8cc6ae) to head (252e6a6).

Files	Patch %	Lines
...ions/ANR/SentryANRTrackingIntegrationV2Tests.swift	95.620%	6 Missing ⚠️
Sources/Sentry/SentryBaseIntegration.m	71.428%	2 Missing ⚠️
...sts/Integrations/ANR/SentryANRTrackerV2Tests.swift	98.518%	2 Missing ⚠️
Sources/Sentry/SentryANRTrackingIntegrationV2.m	97.500%	1 Missing ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##              main     #4262       +/-   ##
=============================================
+ Coverage   91.542%   91.568%   +0.026%     
=============================================
  Files          612       616        +4     
  Lines        49470     49886      +416     
  Branches     17916     18044      +128     
=============================================
+ Hits         45286     45680      +394     
- Misses        4091      4113       +22     
  Partials        93        93               

Files	Coverage Δ
Sources/Sentry/SentryANRTrackerV2.m	100.000% <100.000%> (ø)
Sources/Sentry/SentryDependencyContainer.m	96.153% <100.000%> (+0.235%)	⬆️
Sources/Sentry/SentryOptions.m	99.264% <100.000%> (+0.005%)	⬆️
...s/SentryTests/Helper/SentryTestThreadWrapper.swift	100.000% <100.000%> (ø)
Tests/SentryTests/SentryOptionsTest.m	97.902% <100.000%> (+0.002%)	⬆️
Sources/Sentry/SentryANRTrackingIntegrationV2.m	97.500% <97.500%> (ø)
Sources/Sentry/SentryBaseIntegration.m	95.238% <71.428%> (-1.701%)	⬇️
...sts/Integrations/ANR/SentryANRTrackerV2Tests.swift	98.518% <98.518%> (ø)
...ions/ANR/SentryANRTrackingIntegrationV2Tests.swift	95.620% <95.620%> (ø)

... and 4 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d8cc6ae...252e6a6. Read the comment docs.

[github-actions]

Performance metrics 🚀

	Plain	With Sentry	Diff
Startup time	1210.81 ms	1229.29 ms	18.47 ms
Size	21.58 KiB	703.94 KiB	682.36 KiB

Baseline results on branch: main

Startup times

Revision	Plain	With Sentry	Diff
e998fd0	1241.49 ms	1262.63 ms	21.14 ms
b15521e	1221.69 ms	1237.51 ms	15.82 ms
ef5821b	1253.18 ms	1265.46 ms	12.28 ms
533c68f	1236.54 ms	1256.68 ms	20.14 ms
4ae9b7a	1223.94 ms	1243.90 ms	19.96 ms
5230990	1222.86 ms	1243.59 ms	20.73 ms
bd2afa6	1241.37 ms	1246.20 ms	4.83 ms
189b629	1211.16 ms	1224.30 ms	13.14 ms
3f1be0f	1208.12 ms	1225.72 ms	17.60 ms
3de9971	1224.38 ms	1245.33 ms	20.95 ms

App size

Revision	Plain	With Sentry	Diff
e998fd0	21.58 KiB	414.59 KiB	393.01 KiB
b15521e	21.58 KiB	573.17 KiB	551.59 KiB
ef5821b	21.58 KiB	414.96 KiB	393.37 KiB
533c68f	21.58 KiB	630.28 KiB	608.70 KiB
4ae9b7a	21.58 KiB	632.04 KiB	610.46 KiB
5230990	21.58 KiB	682.39 KiB	660.81 KiB
bd2afa6	20.76 KiB	420.55 KiB	399.79 KiB
189b629	20.76 KiB	399.69 KiB	378.93 KiB
3f1be0f	20.76 KiB	414.44 KiB	393.69 KiB
3de9971	21.58 KiB	574.17 KiB	552.59 KiB

Previous results on branch: feat/non-fatal-app-hangs

Startup times

Revision	Plain	With Sentry	Diff
ec40ccd	1228.87 ms	1245.30 ms	16.43 ms

App size

Revision	Plain	With Sentry	Diff
ec40ccd	21.58 KiB	703.94 KiB	682.36 KiB

[brustolin]

Looks good. There's not much to review other than the duplication. However, I would keep this PR open and merge the other changes related to ANR into this one, so we don't make a release with this duplication.

Something that I believe fits this PR is a protection to avoid enabling both integrations at the same time. Users should either use V1 or V2

[philipphofmann]

However, I would keep this PR open and merge the other changes related to ANR into this one, so we don't make a release with this duplication.

The downside with that approach is that we could run into possible merge conflicts, and I need to merge main multiple times into this branch. I wouldn't worry too much about the duplication. It's around 500 lines of code.

Something that I believe fits this PR is a protection to avoid enabling both integrations at the same time. Users should either use V1 or V2

The option V2 is currently in SentryOptions+Private.h, so nobody can enable it. I will ensure that this can't happen in an upcoming PR. Then it's also easier to review. Furthermore, we might just swap V1 with V2 and not offer the possibility of having both.

[brustolin]

I would prefer that some changes be made to the classes already. I will leave this to your discretion.