From: Alan Donovan <adonovan@google.com>
Date: Fri, 23 May 2025 15:35:41 +0000 (-0400)
Subject: go/token: benchmark FileSet.{Position,AddExistingFiles}
X-Git-Tag: go1.25rc1~63
X-Git-Url: http://www.git.cypherpunks.su/?a=commitdiff_plain;h=c146a61d4c9f76484206a83fa0aac64547f8acdf;p=gostls13.git

go/token: benchmark FileSet.{Position,AddExistingFiles}

This CL adds a benchmark of FileSet.Position, the lookup
operation, and the new AddExistingFiles. It is evident
that its behavior is quadratic in important cases:

(Apple M1)
BenchmarkFileSet_AddExistingFiles/sequence-8         	       3	 362768139 ns/op

Change-Id: I256fdc776135e1924666d127afb37dacbefc860f
Reviewed-on: https://go-review.googlesource.com/c/go/+/675875
Reviewed-by: Robert Findley <rfindley@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
---

diff --git a/src/go/token/export_test.go b/src/go/token/export_test.go
new file mode 100644
index 0000000000..b1bd26958d
--- /dev/null
+++ b/src/go/token/export_test.go
@@ -0,0 +1,9 @@
+// Copyright 2025 The Go 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 token
+
+// exports for tests
+
+func SearchInts(a []int, x int) int { return searchInts(a, x) }
diff --git a/src/go/token/position_bench_test.go b/src/go/token/position_bench_test.go
index 41be7285b7..7bb9de8946 100644
--- a/src/go/token/position_bench_test.go
+++ b/src/go/token/position_bench_test.go
@@ -2,9 +2,14 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package token
+package token_test
 
 import (
+	"go/build"
+	"go/token"
+	"math/rand/v2"
+	"os"
+	"path/filepath"
 	"testing"
 )
 
@@ -14,11 +19,103 @@ func BenchmarkSearchInts(b *testing.B) {
 		data[i] = i
 	}
 	const x = 8
-	if r := searchInts(data, x); r != x {
+	if r := token.SearchInts(data, x); r != x {
 		b.Errorf("got index = %d; want %d", r, x)
 	}
 	b.ResetTimer()
 	for i := 0; i < b.N; i++ {
-		searchInts(data, x)
+		token.SearchInts(data, x)
 	}
 }
+
+func BenchmarkFileSet_Position(b *testing.B) {
+	rng := rand.New(rand.NewPCG(rand.Uint64(), rand.Uint64()))
+
+	// Create a FileSet based on the files of net/http,
+	// a single large package.
+	netHTTPFset := token.NewFileSet()
+	pkg, err := build.Default.Import("net/http", "", 0)
+	if err != nil {
+		b.Fatal(err)
+	}
+	for _, filename := range pkg.GoFiles {
+		filename = filepath.Join(pkg.Dir, filename)
+		fi, err := os.Stat(filename)
+		if err != nil {
+			b.Fatal(err)
+		}
+		netHTTPFset.AddFile(filename, -1, int(fi.Size()))
+	}
+
+	// Measure randomly distributed Pos values across net/http.
+	b.Run("random", func(b *testing.B) {
+		base := netHTTPFset.Base()
+		for b.Loop() {
+			pos := token.Pos(rng.IntN(base))
+			_ = netHTTPFset.Position(pos)
+		}
+	})
+
+	// Measure random lookups within the same file of net/http.
+	// (It's faster because of the "last file" cache.)
+	b.Run("file", func(b *testing.B) {
+		var file *token.File
+		for file = range netHTTPFset.Iterate {
+			break
+		}
+		base, size := file.Base(), file.Size()
+		for b.Loop() {
+			_ = netHTTPFset.Position(token.Pos(base + rng.IntN(size)))
+		}
+	})
+
+	// Measure random lookups on a FileSet with a great many files.
+	b.Run("manyfiles", func(b *testing.B) {
+		fset := token.NewFileSet()
+		for range 25000 {
+			fset.AddFile("", -1, 10000)
+		}
+		base := fset.Base()
+		for b.Loop() {
+			pos := token.Pos(rng.IntN(base))
+			_ = fset.Position(pos)
+		}
+	})
+}
+
+func BenchmarkFileSet_AddExistingFiles(b *testing.B) {
+	// Create the "universe" of files.
+	fset := token.NewFileSet()
+	var files []*token.File
+	for range 25000 {
+		files = append(files, fset.AddFile("", -1, 10000))
+	}
+	rand.Shuffle(len(files), func(i, j int) {
+		files[i], files[j] = files[j], files[i]
+	})
+
+	// choose returns n random files.
+	choose := func(n int) []*token.File {
+		res := make([]*token.File, n)
+		for i := range res {
+			res[i] = files[rand.IntN(n)]
+		}
+		return files[:n]
+	}
+
+	// Measure the cost of	creating a FileSet with a large number
+	// of files added in small handfuls, with some overlap.
+	// This case is critical to gopls.
+	b.Run("sequence", func(b *testing.B) {
+		for b.Loop() {
+			b.StopTimer()
+			fset2 := token.NewFileSet()
+			fset2.AddExistingFiles(files[:10000]...)
+			b.StartTimer()
+
+			for range 1000 {
+				fset2.AddExistingFiles(choose(10)...) // about one package
+			}
+		}
+	})
+}