cmd/autogardener/commit_tag.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 (
	"regexp"
	"slices"
	"strings"
)

var (
	commitTagRegex = regexp.MustCompile(`\[([-\w]+)\]`)

	// Commit tags that will be ignored, because they often appear as substrings
	// of test names but aren't correlated with the area of the codebase that
	// the test applies to with.
	// TODO(olivernewman): Generate this list on-the-fly by querying all test
	// names and ignoring any tag that appears in more than ~10% of test names.
	ignoreList = []string{
		"cm",
		"component",
		"cmx",
		"fuchsia",
		"pkg",
		"roll",
		"test",
		"tests",
	}
)

// hasMatchingTag determines whether the test name contains any of the given
// commit message's bracket-delimited tags as a substring.
func hasMatchingTag(commitMessage, testName string) bool {
	separator := "_"

	// Convert all hyphens to underscores so that we don't ignore matches just
	// because of differing separators.
	testName = strings.ReplaceAll(testName, "-", separator)

	// Ignore tefmocheck failures, which are generally log signatures that only
	// rarely contain keywords that meaningfully correspond to commit tags.
	// This is a bit of a hack, we should ideally have some more general test
	// tag in ResultDB to indicate that this tefmocheck failures are not regular
	// tests.
	if strings.HasPrefix(testName, "testing_failure_mode") {
		return false
	}
	firstLine := strings.Split(commitMessage, "\n")[0]
	tagMatches := commitTagRegex.FindAllStringSubmatch(firstLine, -1)
	for _, match := range tagMatches {
		tag := match[1]
		// Standardize separators.
		for _, c := range []string{" ", "-"} {
			tag = strings.ReplaceAll(tag, c, separator)
		}
		if slices.Contains(ignoreList, tag) {
			continue
		}
		if strings.Contains(testName, tag) {
			return true
		}
	}
	return false
}