serverless: let reject in callbackWrapper call captureExceptionAsync with the error

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!

[leemhenson]

Fixes#2956

I modified fakeCallback to not return the supplied error. I don't think you can assume that callbacks will always do that. That exposes the bug in the existing test suites if we add a check:

expect(Sentry.captureException).toBeCalledTimes(1);

Second commit fixes the issue by discarding the callback result and rejecting with the error instead.

• If you've added code that should be tested, please add tests.
• Ensure your code lints and the test suite passes (yarn lint) & (yarn test).

[marshall-lee]

I also have a kinda fix for this issue in #2945.
I was forced to take a some different approach but basically it looks similar. The key difference is that there's no need to invoke callback at all since we already coerce the handler to async.

[kamilogorek]

Merged #2945 as it included a lot of other serverless related work. Nonetheless, Really appreciate your contribution @leemhenson. Thanks!