[BUG] GQLgen Tracer - Introspection query is not always being filtered

[khwong-altium]

We are using GQL Tracer for our application.
We want to reduce the amount of noise so have enabled the WithoutTraceIntrospectionQuery Option.

However, the tracer sometimes collects fields under the Introspection query.
We have confirm:

• Playground comes with github.com/99designs/gqlgen: Filtered properly
• Retool introspection (https://www.retool.com) : Does not filtered

We have configured the tracer as follow:

func createGQLHandler(	schema graphql.ExecutableSchema) http.Handler {
	h := handler.New(schema)
	h.AddTransport(transport.Options{})
	h.AddTransport(transport.POST{})
	h.Use(extension.Introspection{})
	h.SetRecoverFunc(func(ctx context.Context, err interface{}) error {
		logger.Error("graphql panic", "err", err)
		debug.PrintStack()
		return gqlerror.Errorf("internal system error")
	})
	h.SetErrorPresenter(func(ctx context.Context, err error) *gqlerror.Error {
		logger.Error("graphql error", "err", err)
		return graphql.DefaultErrorPresenter(ctx, err)
	})
	if config.DD_TRACE_ENABLED {
		t := gqlgentrace.NewTracer(
			gqlgentrace.WithAnalytics(true),
			gqlgentrace.WithServiceName(logging.DDServiceName),
			gqlgentrace.WithoutTraceIntrospectionQuery(),
			gqlgentrace.WithoutTraceTrivialResolvedFields(),
		)
		h.Use(t)
	}
	httpAuthware, gqlAuthware := authenticator.NewAuthMiddleware(skipAudCheck)
	h.AroundOperations(gqlAuthware)
	return httpAuthware(h)
}

Version of dd-trace-go
1.70.3

Describe what happened:
Query comes from Retool traced all the fields under it.

Describe what you expected:

• All fields under any query operation named IntrospectionQuery shall not be traced.

Steps to reproduce the issue:

• Login Retool
• Connecting any resource with above tracer.
• Creating a query to the GQL resource and trigger the Introspection Query.

Additional environment details (Version of Go, Operating System, etc.):

[mtoffl01]

Hello @khwong-altium ,

Thanks for opening this issue. We are looking into it and will follow up once we have more information. In the meantime, if you would like to contribute, we'll be glad to review and merge your contribution!

In case this is time sensitive for you, please reach out to Datadog Support so we can best prioritize.