fix: Session replay redact view with transformation

[brustolin]

📜 Description

View transformations, used to interfere with the calculations of which areas to redact

Before	After

💚 How did you test it?

Unit Tests

📝 Checklist

You have to check all boxes before merging:

• I reviewed the submitted code.
• I added tests to verify the changes.
• No new PII added or SDK only sends newly added PII if sendDefaultPII is enabled.
• I updated the docs if needed.
• Review from the native team if needed.
• No breaking change or entry added to the changelog.
• No breaking change for hybrid SDKs or communicated to hybrid SDKs.

🔮 Next steps

[codecov]

Codecov Report

Attention: Patch coverage is 99.56897% with 1 line in your changes missing coverage. Please review.

Project coverage is 91.640%. Comparing base (7f720fb) to head (3bc34ab).
Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
...ests/SentryTests/SentryViewPhotographerTests.swift	99.248%	1 Missing ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##              main     #4308       +/-   ##
=============================================
+ Coverage   91.625%   91.640%   +0.015%     
=============================================
  Files          618       617        -1     
  Lines        50269     50264        -5     
  Branches     18138     18061       -77     
=============================================
+ Hits         46059     46062        +3     
+ Misses        4117      4110        -7     
+ Partials        93        92        -1     

Files with missing lines	Coverage Δ
...tegrations/SessionReplay/SentryReplayOptions.swift	93.103% <ø> (ø)
Sources/Swift/Tools/SentryViewPhotographer.swift	100.000% <100.000%> (+65.000%)	⬆️
Sources/Swift/Tools/UIRedactBuilder.swift	100.000% <100.000%> (+0.900%)	⬆️
Tests/SentryTests/UIRedactBuilderTests.swift	100.000% <100.000%> (ø)
...ests/SentryTests/SentryViewPhotographerTests.swift	99.248% <99.248%> (ø)

... and 15 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 7f720fb...3bc34ab. Read the comment docs.

[github-actions]

Performance metrics 🚀

	Plain	With Sentry	Diff
Startup time	1226.77 ms	1254.43 ms	27.66 ms
Size	21.58 KiB	713.49 KiB	691.91 KiB

Baseline results on branch: main

Startup times

Revision	Plain	With Sentry	Diff
dcec216	1238.94 ms	1261.06 ms	22.12 ms
27f970b	1223.48 ms	1239.51 ms	16.03 ms
c0f08e7	1230.67 ms	1246.31 ms	15.63 ms
5616e0a	1224.12 ms	1249.86 ms	25.74 ms
8aec30e	1244.71 ms	1262.20 ms	17.49 ms
45d3ca5	1238.53 ms	1263.09 ms	24.55 ms
1ae5768	1226.90 ms	1239.94 ms	13.04 ms
5f6f658	1221.08 ms	1241.84 ms	20.76 ms
075a044	1237.26 ms	1255.36 ms	18.10 ms
3cba0e8	1231.20 ms	1251.50 ms	20.30 ms

App size

Revision	Plain	With Sentry	Diff
dcec216	20.76 KiB	432.88 KiB	412.11 KiB
27f970b	21.58 KiB	706.97 KiB	685.39 KiB
c0f08e7	21.58 KiB	573.84 KiB	552.26 KiB
5616e0a	22.85 KiB	407.44 KiB	384.59 KiB
8aec30e	21.58 KiB	616.76 KiB	595.18 KiB
45d3ca5	20.76 KiB	427.54 KiB	406.78 KiB
1ae5768	21.58 KiB	655.72 KiB	634.14 KiB
5f6f658	21.58 KiB	699.25 KiB	677.67 KiB
075a044	20.76 KiB	420.41 KiB	399.65 KiB
3cba0e8	22.84 KiB	403.18 KiB	380.34 KiB

Previous results on branch: fix/redact-under-translucent

Startup times

Revision	Plain	With Sentry	Diff
a268d0b	1223.94 ms	1242.94 ms	19.00 ms
70ebda6	1238.09 ms	1247.11 ms	9.03 ms
b2f8339	1218.12 ms	1243.32 ms	25.20 ms
be34bdf	1228.30 ms	1244.78 ms	16.48 ms
11b37b6	1239.22 ms	1259.82 ms	20.59 ms
bf4bb59	1231.86 ms	1251.20 ms	19.35 ms
fda51b1	1229.89 ms	1237.02 ms	7.13 ms
1e7fd29	1242.35 ms	1260.59 ms	18.24 ms
8ee29d0	1227.21 ms	1242.84 ms	15.63 ms

App size

Revision	Plain	With Sentry	Diff
a268d0b	21.58 KiB	707.29 KiB	685.70 KiB
70ebda6	21.58 KiB	707.13 KiB	685.55 KiB
b2f8339	21.58 KiB	707.34 KiB	685.76 KiB
be34bdf	21.58 KiB	707.33 KiB	685.75 KiB
11b37b6	21.58 KiB	707.84 KiB	686.26 KiB
bf4bb59	21.58 KiB	707.34 KiB	685.76 KiB
fda51b1	21.58 KiB	707.34 KiB	685.76 KiB
1e7fd29	21.58 KiB	707.14 KiB	685.56 KiB
8ee29d0	21.58 KiB	707.40 KiB	685.82 KiB

[armcknight]

I did one pass on this but ran out of time for today. I wasn't able to manually test anything yet so I'd like to try that tomorrow if you could provide some instructions. I wasn't able to see any replays showing up on the dashboard.

[armcknight]

I've still got some questions but have been unable to get a replay displayed in the web app, so can't exercise the feature to answer them myself.

What I'd like answered in this PR description, in high-level terms:

• what was the bug we encountered
• what were we doing before that wasn't working, why wasn't it working
• what is the new strategy we're taking and why does it fix the bug

What i'd like to see written in code comments:

• how CGPathFillRule.evenOdd achieves the goal
• what does layerOriginalFrame calculate
• an explanation of the different scenarios covered in mapRedactRegion with some schematic diagrams for different scenarios

[brustolin]

what was the bug we encountered
what were we doing before that wasn't working, why wasn't it working
what is the new strategy we're taking and why does it fix the bug

We have text that isn’t being redacted behind a rotated element that shouldn’t be redacted.

Can we provide a detailed explanation of everything done in a PR?

Yes, but that could introduce bias into the review and consume a lot of time for something that should be clear from reading the code. Reviewing the code involves understanding what was done and identifying issues by reading it directly.

And It’s not about what changed or how the bug was fixed, but whether the current code meets the requirements.

What i'd like to see written in code comments:

• how CGPathFillRule.evenOdd achieves the goal

I don’t think it’s a good idea for the code to have a lot of comments just due to a lack of familiarity with the APIs. Of course, this is the type of question that can be raised during the review. The answer is that the even-odd fill rule allows me to add clip regions that will be ignored during a new draw in a easier way compared to CGPathFillRule.winding.

• what does layerOriginalFrame calculate

layerOriginalFrame returns the original frame of the layer (before transforms), but I've removed it since you reviewed it.

• an explanation of the different scenarios covered in mapRedactRegion with some schematic diagrams for different scenarios

I can explain any question you have in here, IMO, schematic diagrams within the code for this case would be too complex to be helpful in any way.

[armcknight]

Just a few more questions/comments here.

[romtsn]

I'm not gonna fight the code preferences fight 😅 only left one comment to check, otherwise LGTM, we'll probably need something similar on Android

[armcknight]

Thanks for sticking with me reviewing this. Please understand this was my first brush with any replay code, and the redaction logic here and elsewhere in the SDK is probably the most impactful feature we have from a privacy/legal standpoint. I wanted to make sure we get it right. Maintenance, performance and usability concerns aside, I haven't been able to find any functional flaws with the redaction feature, so let's do it.