fix: fall back to old behavior if newer value is nil in core data tracker

[armcknight]

Fixes a possible crash, because NSEntityDescription.name is actually declared as nullable, so if it is nil and we try to use it as a key in a dictionary, that will throw an exception. In case it is nil, we'll fall back to the prior behavior for this operation.

[github-actions]

Performance metrics 🚀

	Plain	With Sentry	Diff
Startup time	1219.51 ms	1239.44 ms	19.93 ms
Size	20.76 KiB	433.23 KiB	412.47 KiB

Baseline results on branch: main

Startup times

Revision	Plain	With Sentry	Diff
ea73af6	1230.96 ms	1244.98 ms	14.02 ms
5b6694b	1221.71 ms	1259.06 ms	37.35 ms
f715499	1234.26 ms	1259.40 ms	25.14 ms
20163bb	1248.92 ms	1258.48 ms	9.56 ms
4c00f8c	1231.62 ms	1237.76 ms	6.14 ms
89d72e7	1222.60 ms	1252.78 ms	30.18 ms
3e7aa41	1214.02 ms	1238.84 ms	24.82 ms
c50d363	1215.10 ms	1231.82 ms	16.72 ms
c00eafe	1253.04 ms	1256.41 ms	3.37 ms
45d3ca5	1238.53 ms	1263.09 ms	24.55 ms

App size

Revision	Plain	With Sentry	Diff
ea73af6	20.76 KiB	425.75 KiB	404.99 KiB
5b6694b	20.76 KiB	426.11 KiB	405.34 KiB
f715499	20.76 KiB	427.23 KiB	406.47 KiB
20163bb	20.76 KiB	426.82 KiB	406.06 KiB
4c00f8c	20.76 KiB	419.62 KiB	398.86 KiB
89d72e7	20.76 KiB	431.87 KiB	411.11 KiB
3e7aa41	20.76 KiB	432.31 KiB	411.55 KiB
c50d363	20.76 KiB	432.68 KiB	411.92 KiB
c00eafe	20.76 KiB	432.88 KiB	412.12 KiB
45d3ca5	20.76 KiB	427.54 KiB	406.78 KiB

Previous results on branch: armcknight/fix/possible-nil-crash

Startup times

Revision	Plain	With Sentry	Diff
c23c255	1229.71 ms	1248.02 ms	18.31 ms
fa8b1a9	1219.48 ms	1248.94 ms	29.46 ms
172904f	1224.24 ms	1251.98 ms	27.73 ms
f393202	1247.47 ms	1263.53 ms	16.06 ms

App size

Revision	Plain	With Sentry	Diff
c23c255	20.76 KiB	432.30 KiB	411.54 KiB
fa8b1a9	20.76 KiB	432.25 KiB	411.49 KiB
172904f	20.76 KiB	433.23 KiB	412.47 KiB
f393202	20.76 KiB	432.25 KiB	411.49 KiB

[armcknight]

tests should pass after merging #2867

[brustolin]

Good catch.
I never saw entity name being nil, but since is nullable I guess is possible.

Looks good, I think we need a test with nil entity name in order to finish this.

[armcknight]

Good catch. I never saw entity name being nil, but since is nullable I guess is possible.

Looks good, I think we need a test with nil entity name in order to finish this.

So when I try to use the current core data tracker tests to write one like that, it throws the following exception:

/Users/andrewmcknight/Code/organization/getsentry/repos/public/sentry-cocoa/Tests/SentryTests/Integrations/Performance/CoreData/TestCoreDataStack.swift:10: error: -[SentryTests.SentryCoreDataTrackerTests testEntityWithNilName] : Entity name must not be nil. (NSInvalidArgumentException)

[brustolin]

So when I try to use the current core data tracker tests to write one like that, it throws the following exception

Ow, now I remember seen something like this.
Even if non nullability is enforced at run time, I don't mind keeping the changing in this PR. I guess users can do some crazy thing that would make a nil name.

[brustolin]

LGTM