Spec compliance: OTEL_PROPAGATORS should still work when OTEL_SDK_DISABLED

[mikeblum]

No description provided.

[trask]

can you see if you can add a test for this?

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.83%. Comparing base (2fcd5f5) to head (90a58ef).
Report is 2 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #7062      +/-   ##
============================================
- Coverage     89.84%   89.83%   -0.01%     
- Complexity     6610     6612       +2     
============================================
  Files           740      740              
  Lines         19981    19983       +2     
  Branches       1966     1966              
============================================
+ Hits          17951    17952       +1     
+ Misses         1440     1439       -1     
- Partials        590      592       +2     

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

[mikeblum]

Interesting. I feel like historically class loaders are pretty hard to test properly. This test was an attempt in vain to get the code coverage over 90%. Should I just strike it as it’s quite out of scope for this PR?

…

On Thu, Feb 6, 2025 at 15:15 jack-berg ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In sdk-extensions/autoconfigure/src/test/java/io/opentelemetry/sdk/autoconfigure/AutoConfiguredOpenTelemetrySdkTest.java <#7062 (comment)> : > + () -> + AutoConfiguredOpenTelemetrySdk.builder() + .addPropertiesSupplier(() -> singletonMap("otel.experimental.config.file", "")) + .addPropertiesSupplier(() -> singletonMap("otel.sdk.disabled", "true")) + .build()); + } + + @test + @SuppressLogger(AutoConfiguredOpenTelemetrySdkBuilder.class) + void serviceClassLoader() { + ClassLoader classLoader = null; + assertThatThrownBy( + () -> AutoConfiguredOpenTelemetrySdk.builder().setServiceClassLoader(classLoader)) + .isInstanceOf(NullPointerException.class) + .hasMessageContaining("serviceClassLoader"); + ClassLoader mockClassLoader = mock(ClassLoader.class); I think this is what's causing the test to fail: mockito/mockito#1696 (comment) <mockito/mockito#1696 (comment)> I can recreate the test failure locally on my machine. — Reply to this email directly, view it on GitHub <#7062 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA5ZPN2VN5N3FFVKPR45P2D2OPGINAVCNFSM6AAAAABWJNNAG2VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKOJZHE3TENZQGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jack-berg]

Should I just strike it as it’s quite out of scope for this PR?

Yeah I think that's the simplest path.