cmd/autogardener/changed_files.go

// Copyright 2022 The Fuchsia Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

package main

import (
	"strings"
)

// scoreChangedFileProximity computes a 0-100 score of the proximity of a Gerrit
// change's affected files to the given test GN label.
func scoreChangedFileProximity(changedFiles []string, gnLabel string) int {
	// If there's no GN label and we can't do this analysis, that significantly
	// lowers the confidence.
	if gnLabel == "" {
		return 0
	}

	strippedLabel := strings.Trim(strings.Split(gnLabel, ":")[0], "/")
	labelParts := strings.Split(strippedLabel, "/")

	var proximityScores []float64

	for _, file := range changedFiles {
		pl := sharedPrefixLength(labelParts, strings.Split(file, "/"))

		// Proximity is determined by the ratio of the shared prefix length to
		// the length of the GN label directory.
		//
		// If a test is declared in directory `src/foo/bar`, any files within
		// `src/foo/bar` or any subdirectory thereof will have a proximity ratio
		// of 1.
		proximityScore := 100 * float64(pl) / float64(len(labelParts))

		proximityScores = append(proximityScores, proximityScore)

		// Add extra weighting for higher proximity scores. If a change touches
		// a couple files close to the test, along with a large number of
		// unrelated files, it should still be scored higher than a change that
		// touches only files that are moderately close to the test.
		for i := 0; i <= int(proximityScore/25); i++ {
			proximityScores = append(proximityScores, proximityScore)
		}
	}

	return int(average(proximityScores))
}

// sharedPrefixLength computes the length of the longest shared prefix of the
// two slices, i.e. the maximum number N such that p1[:N] == p2[:N].
func sharedPrefixLength(p1, p2 []string) int {
	var i int
	for ; i < len(p1) && i < len(p2); i++ {
		if p1[i] != p2[i] {
			break
		}
	}
	return i
}