use key presence for stacktrace filter instead of value

[cleverchuk]

Context:

Span stacktrace filter now checks key presence instead of limiting to only string-key values. This type agnostic approach is consistent with current documentation.

Test Plan:

New unit test and regression.

Test services data

1. e-1712644058766987264
2. e-1712643928659124224
3. e-1742334541200846848
4. e-1777406072376840192

[Copilot]

Pull request overview

This PR changes the SpanStacktraceFilter to check for attribute key presence instead of checking for string-typed attribute values. This makes the filter type-agnostic and aligns with the documented behavior.

Key changes:

• Modified filter logic to check if attribute keys exist regardless of their value types
• Changed filterAttributes from static to instance field to support proper testing
• Enhanced test coverage with multiple attribute types (string, long, string arrays, long arrays)

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
SpanStacktraceFilter.java	Refactored the test method to check for attribute key presence instead of string attribute values, and changed filterAttributes from static to instance field
SpanStacktraceFilterTest.java	Replaced mock-based tests with comprehensive tests covering multiple attribute types and proper setup/teardown

[Copilot]

Changing filterAttributes from static to instance field may cause thread safety issues if SpanStacktraceFilter instances are shared across threads. Each call to test() modifies this field. Consider making this field static and using thread-safe operations, or document that instances should not be shared across threads.

[Copilot]

This implementation creates a new Set and copies all attribute keys on every invocation. For spans with many attributes, this could be inefficient. Consider using a stream-based approach that short-circuits on the first match: filterAttributes.stream().anyMatch(attr -> readableSpan.getAttributes().asMap().keySet().stream().anyMatch(key -> key.getKey().equals(attr)))

Suggested change

	Set<String> attributes =
	readableSpan.getAttributes().asMap().keySet().stream()
	.map(AttributeKey::getKey)
	.collect(Collectors.toSet());
	attributes.retainAll(filterAttributes);
	return !attributes.isEmpty();
	return filterAttributes.stream().anyMatch(
	attr -> readableSpan.getAttributes().asMap().keySet().stream()
	.anyMatch(key -> key.getKey().equals(attr))
	);

[cheempz]

Any merit to this suggestion?

[cleverchuk]

yeah. compute-wise superior and memory-wise inferior in worst case scenario. all round superior actually.

[cheempz]

LGTM pending question.

[cheempz]

Any merit to this suggestion?

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

[cheempz]

Thanks for the revisit!

[cleverchuk]

merged because lambda-release-test failing is not actual test failure. it's failure to write test results file.