fix: prevent SystemExit(0) from being reported as Sentry error

[sentry]

This PR addresses an issue where normal program termination (indicated by SystemExit: 0) was being captured and reported as an error by Sentry.

The root cause was that the _before_send function in codecov_cli/opentelemetry.py, responsible for filtering Sentry events, did not explicitly handle SystemExit with a code of 0 or None. Consequently, the Sentry SDK would capture these normal exits as unhandled exceptions.

The fix modifies the _before_send function to check for SystemExit exceptions. If the exception type is SystemExit and its code is 0 or None (which signifies a clean exit), the event is now suppressed by returning None. This ensures that Sentry only receives reports for actual errors, reducing noise and improving the signal-to-noise ratio in error monitoring.

Fixes CLI-G7

[codecov]

❌ 1 Tests Failed:

Tests completed	Failed	Passed	Skipped
723	1	722	1

View the full list of 1 ❄️ flaky test(s)

api.temp.calculator.test_calculator::test_divide

Flake rate in main: 100.00% (Passed 0 times, Failed 1625 times)

Stack Traces | 0.001s run time

def
                test_divide():
                > assert Calculator.divide(1, 2) == 0.5
                E assert 1.0 == 0.5
                E + where 1.0 = <function Calculator.divide at 0x104c9eb90>(1, 2)
                E + where <function Calculator.divide at 0x104c9eb90> = Calculator.divide
                .../temp/calculator/test_calculator.py:30: AssertionError

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}