Cache optimized regexp matchers (#465)

* Cache optimized regexp matchers

Signed-off-by: Marco Pracucci <marco@pracucci.com>

* Added BenchmarkNewFastRegexMatcher_CacheMisses

Signed-off-by: Marco Pracucci <marco@pracucci.com>

* Improved benchmark

Signed-off-by: Marco Pracucci <marco@pracucci.com>

* Improved benchmark

Signed-off-by: Marco Pracucci <marco@pracucci.com>

* Use LRU cache v2

Signed-off-by: Marco Pracucci <marco@pracucci.com>

* Run gofumpt

Signed-off-by: Marco Pracucci <marco@pracucci.com>

---------

Signed-off-by: Marco Pracucci <marco@pracucci.com>

go.mod

@@ -28,6 +28,7 @@ require (
 	github.com/grafana/regexp v0.0.0-20221122212121-6b5c0a4cb7fd
 	github.com/grpc-ecosystem/grpc-gateway v1.16.0
 	github.com/hashicorp/consul/api v1.20.0
+	github.com/hashicorp/golang-lru/v2 v2.0.2
 	github.com/hashicorp/nomad/api v0.0.0-20230308192510-48e7d70fcd4b
 	github.com/hetznercloud/hcloud-go v1.41.0
 	github.com/ionos-cloud/sdk-go/v6 v6.1.4

go.sum

@@ -444,6 +444,8 @@ github.com/hashicorp/golang-lru v0.5.0/go.mod h1:/m3WP610KZHVQ1SGc6re/UDhFvYD7pJ
 github.com/hashicorp/golang-lru v0.5.1/go.mod h1:/m3WP610KZHVQ1SGc6re/UDhFvYD7pJ4Ao+sR/qLZy8=
 github.com/hashicorp/golang-lru v0.6.0 h1:uL2shRDx7RTrOrTCUZEGP/wJUFiUI8QT6E7z5o8jga4=
 github.com/hashicorp/golang-lru v0.6.0/go.mod h1:iADmTwqILo4mZ8BN3D2Q6+9jd8WM5uGBxy+E8yxSoD4=
+github.com/hashicorp/golang-lru/v2 v2.0.2 h1:Dwmkdr5Nc/oBiXgJS3CDHNhJtIHkuZ3DZF5twqnfBdU=
+github.com/hashicorp/golang-lru/v2 v2.0.2/go.mod h1:QeFd9opnmA6QUJc5vARoKUSoFhyfM2/ZepoAG6RGpeM=
 github.com/hashicorp/logutils v1.0.0/go.mod h1:QIAnNjmIWmVIIkWDTG1z5v++HQmx9WQRO+LraFDTW64=
 github.com/hashicorp/mdns v1.0.0/go.mod h1:tL+uN++7HEJ6SQLQ2/p+z2pH24WQKWjBPkE0mNTz8vQ=
 github.com/hashicorp/mdns v1.0.4/go.mod h1:mtBihi+LeNXGtG8L9dX59gAEa12BDtBQSp4v/YAJqrc=

model/labels/regexp.go

@@ -18,6 +18,7 @@ import (
 
 	"github.com/grafana/regexp"
 	"github.com/grafana/regexp/syntax"
+	lru "github.com/hashicorp/golang-lru/v2"
 )
 
 const (
@@ -29,6 +30,14 @@ const (
 	optimizeEqualStringMatchersThreshold = 16
 )
 
+var fastRegexMatcherCache *lru.Cache[string, *FastRegexMatcher]
+
+func init() {
+	// Ignore error because it can only return error if size is invalid,
+	// but we're using an hardcoded size here.
+	fastRegexMatcherCache, _ = lru.New[string, *FastRegexMatcher](10000)
+}
+
 type FastRegexMatcher struct {
 	re *regexp.Regexp
 
@@ -43,6 +52,24 @@ type FastRegexMatcher struct {
 }
 
 func NewFastRegexMatcher(v string) (*FastRegexMatcher, error) {
+	// Check the cache.
+	if matcher, ok := fastRegexMatcherCache.Get(v); ok {
+		return matcher, nil
+	}
+
+	// Create a new matcher.
+	matcher, err := newFastRegexMatcherWithoutCache(v)
+	if err != nil {
+		return nil, err
+	}
+
+	// Cache it.
+	fastRegexMatcherCache.Add(v, matcher)
+
+	return matcher, nil
+}
+
+func newFastRegexMatcherWithoutCache(v string) (*FastRegexMatcher, error) {
 	parsed, err := syntax.Parse(v, syntax.Perl)
 	if err != nil {
 		return nil, err

model/labels/regexp_test.go

@@ -18,6 +18,7 @@ import (
 	"fmt"
 	"math/rand"
 	"os"
+	"strconv"
 	"strings"
 	"testing"
 	"time"
@@ -89,10 +90,46 @@ func TestNewFastRegexMatcher(t *testing.T) {
 }
 
 func BenchmarkNewFastRegexMatcher(b *testing.B) {
-	for _, r := range regexes {
-		b.Run(getTestNameFromRegexp(r), func(b *testing.B) {
+	runBenchmark := func(newFunc func(v string) (*FastRegexMatcher, error)) func(b *testing.B) {
+		return func(b *testing.B) {
+			for _, r := range regexes {
+				b.Run(getTestNameFromRegexp(r), func(b *testing.B) {
+					for n := 0; n < b.N; n++ {
+						_, err := newFunc(r)
+						if err != nil {
+							b.Fatal(err)
+						}
+					}
+				})
+			}
+		}
+	}
+
+	b.Run("with cache", runBenchmark(NewFastRegexMatcher))
+	b.Run("without cache", runBenchmark(newFastRegexMatcherWithoutCache))
+}
+
+func BenchmarkNewFastRegexMatcher_CacheMisses(b *testing.B) {
+	// Init the random seed with a constant, so that it doesn't change between runs.
+	randGenerator := rand.New(rand.NewSource(1))
+
+	tests := map[string]string{
+		"simple regexp":  randString(randGenerator, 10),
+		"complex regexp": strings.Join(randStrings(randGenerator, 100, 10), "|"),
+	}
+
+	for testName, regexpPrefix := range tests {
+		b.Run(testName, func(b *testing.B) {
+			// Ensure the cache is empty.
+			fastRegexMatcherCache.Purge()
+
+			b.ResetTimer()
+
 			for n := 0; n < b.N; n++ {
-				_, err := NewFastRegexMatcher(r)
+				// Unique regexp to emulate 100% cache misses.
+				regexp := regexpPrefix + strconv.Itoa(n)
+
+				_, err := NewFastRegexMatcher(regexp)
 				if err != nil {
 					b.Fatal(err)
 				}