gocognit lintering warnings for Range support

[atc0005]

Overview

Output from a recent CI run against the development branch:

golangci-lint has version 1.61.0 built with go1.23.1 from a1d6c560 on 2024-09-09T17:44:42Z
level=warning msg="The linter 'exportloopref' is deprecated (since v1.60.2) due to: Since Go1.22 (loopvar) this linter is no longer relevant. Replaced by copyloopvar."
range.go:147:1: cognitive complexity 17 of func `(*Plugin).EvaluateThreshold` is high (> 10) (gocognit)
func (p *Plugin) EvaluateThreshold(perfData ...PerformanceData) error {
^
range.go:49:1: cognitive complexity 11 of func `(Range).checkOutsideRange` is high (> 10) (gocognit)
func (r Range) checkOutsideRange(valueAsAFloat float64) bool {
^
range.go:79:1: cognitive complexity 11 of func `ParseRangeString` is high (> 10) (gocognit)
func ParseRangeString(input string) *Range {
^
Error: Process completed with exit code 1.

Relevant bits:

range.go:147:1: cognitive complexity 17 of func `(*Plugin).EvaluateThreshold` is high (> 10) (gocognit)
func (p *Plugin) EvaluateThreshold(perfData ...PerformanceData) error {
^

range.go:49:1: cognitive complexity 11 of func `(Range).checkOutsideRange` is high (> 10) (gocognit)
func (r Range) checkOutsideRange(valueAsAFloat float64) bool {
^

range.go:79:1: cognitive complexity 11 of func `ParseRangeString` is high (> 10) (gocognit)
func ParseRangeString(input string) *Range {
^

While these warnings have been emitted for a while now they don't impact functionality so haven't been a priority. I'd like to resolve them as part of the next release.

References

• Warning and Critical Threshold Handling #178
• Refactor newly added Range support #181
• Fix panic if ParseRangeCheck returns nil #236

[atc0005]

Original implementation references for review to better understand current implementation:

• https://github.com/nagios-plugins/nagios-plugin-perl/blob/master/lib/Nagios/Monitoring/Plugin/Range.pm
• https://github.com/nagios-plugins/nagios-plugin-perl/blob/master/lib/Nagios/Monitoring/Plugin/Threshold.pm

[atc0005]

range.go:49:1: cognitive complexity 11 of func `(Range).checkOutsideRange` is high (> 10) (gocognit)
func (r Range) checkOutsideRange(valueAsAFloat float64) bool {
^

I worked with this method for a bit, then turned to popular LLMs for alternatives. Those suggestions were much better than mine and appear to improve on the original. The Google Gemini suggestion appears to be the easiest to grok, so will plan to use it to replace the original.

Google Gemini (model 'Gemini 1.5 Flash')

func (r Range) checkOutsideRange(valueAsAFloat float64) bool {
	// Explanation of the Simplification:
	//
	// Infinite Bounds:
	//
	// If the start is infinite, the value is outside the range only if it's
	// greater than the end. If the end is infinite, the value is outside the
	// range only if it's less than the start.
	//
	// Finite Bounds:
	//
	// The value is outside the range if it's either less than the start or
	// greater than the end.

	// Handle infinite bounds first
	if r.StartInfinity {
		return valueAsAFloat > r.End
	} else if r.EndInfinity {
		return valueAsAFloat < r.Start
	}

	// Handle finite bounds
	return valueAsAFloat < r.Start || valueAsAFloat > r.End
}

I moved the explanation notes within the function for reference.

ChatGPT (model 'GPT-4o')

func (r Range) checkOutsideRange(value float64) bool {
	// Explanation of the Simplification:
	//
	// Each case is now focused only on the conditions that make the value
	// outside the range. The final default case covers the fully infinite
	// range, simplifying the logic to just return false since no bounds
	// restrict the range.

	switch {
	case !r.StartInfinity && value < r.Start:
		return true
	case !r.EndInfinity && value > r.End:
		return true
	case r.StartInfinity && r.EndInfinity:
		return false
	default:
		return false
	}
}

As before, I placed the explanation for the changes within the function for reference.