publicsuffix: spruce up code gen and speed up PublicSuffix

jub0bs · gopherbot · commit fe7f0391aa99 · 2025-02-28T12:42:09.000-08:00
Rely on functions from the slices package where convenient. Drop custom max functions in favor of max builtin. Remove unused non-exported functions. Reduce the number of bounds checks. Replace calls to strings.LastIndex by calls to strings.LastIndexByte. goos: darwin goarch: amd64 pkg: golang.org/x/net/publicsuffix cpu: Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz │ old │ new │ │ sec/op │ sec/op vs base │ PublicSuffix-8 13.46µ ± 0% 13.23µ ± 0% -1.67% (p=0.000 n=20) │ old │ new │ │ B/op │ B/op vs base │ PublicSuffix-8 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=20) ¹ ¹ all samples are equal │ old │ new │ │ allocs/op │ allocs/op vs base │ PublicSuffix-8 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=20) ¹ ¹ all samples are equal Change-Id: Id72967560884d98a5c0791ccea73dbb27d120c2c GitHub-Last-Rev: 87567e7 GitHub-Pull-Request: #233 Reviewed-on: https://go-review.googlesource.com/c/net/+/652236 Reviewed-by: Damien Neil <dneil@google.com> Commit-Queue: Ian Lance Taylor <iant@golang.org> Auto-Submit: Ian Lance Taylor <iant@golang.org> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> Reviewed-by: Michael Pratt <mpratt@google.com>
diff --git a/publicsuffix/gen.go b/publicsuffix/gen.go
@@ -21,6 +21,7 @@ package main
 import (
 	"bufio"
 	"bytes"
+	"cmp"
 	"encoding/binary"
 	"flag"
 	"fmt"
@@ -29,7 +30,7 @@ import (
 	"net/http"
 	"os"
 	"regexp"
-	"sort"
+	"slices"
 	"strings"
 
 	"golang.org/x/net/idna"
@@ -62,39 +63,13 @@ var (
 	maxLo         uint32
 )
 
-func max(a, b int) int {
-	if a < b {
-		return b
-	}
-	return a
-}
-
-func u32max(a, b uint32) uint32 {
-	if a < b {
-		return b
-	}
-	return a
-}
-
 const (
 	nodeTypeNormal     = 0
 	nodeTypeException  = 1
 	nodeTypeParentOnly = 2
 	numNodeType        = 3
 )
 
-func nodeTypeStr(n int) string {
-	switch n {
-	case nodeTypeNormal:
-		return "+"
-	case nodeTypeException:
-		return "!"
-	case nodeTypeParentOnly:
-		return "o"
-	}
-	panic("unreachable")
-}
-
 const (
 	defaultURL   = "https://publicsuffix.org/list/effective_tld_names.dat"
 	gitCommitURL = "https://api.github.com/repos/publicsuffix/list/commits?path=public_suffix_list.dat"
@@ -251,7 +226,7 @@ func main1() error {
 	for label := range labelsMap {
 		labelsList = append(labelsList, label)
 	}
-	sort.Strings(labelsList)
+	slices.Sort(labelsList)
 
 	combinedText = combineText(labelsList)
 	if combinedText == "" {
@@ -509,15 +484,13 @@ func (n *node) child(label string) *node {
 		icann:    true,
 	}
 	n.children = append(n.children, c)
-	sort.Sort(byLabel(n.children))
+	slices.SortFunc(n.children, byLabel)
 	return c
 }
 
-type byLabel []*node
-
-func (b byLabel) Len() int           { return len(b) }
-func (b byLabel) Swap(i, j int)      { b[i], b[j] = b[j], b[i] }
-func (b byLabel) Less(i, j int) bool { return b[i].label < b[j].label }
+func byLabel(a, b *node) int {
+	return strings.Compare(a.label, b.label)
+}
 
 var nextNodesIndex int
 
@@ -557,7 +530,7 @@ func assignIndexes(n *node) error {
 		n.childrenIndex = len(childrenEncoding)
 		lo := uint32(n.firstChild)
 		hi := lo + uint32(len(n.children))
-		maxLo, maxHi = u32max(maxLo, lo), u32max(maxHi, hi)
+		maxLo, maxHi = max(maxLo, lo), max(maxHi, hi)
 		if lo >= 1<<childrenBitsLo {
 			return fmt.Errorf("children lo %d is too large, or childrenBitsLo is too small", lo)
 		}
@@ -586,20 +559,6 @@ func printNodeLabel(w io.Writer, n *node) error {
 	return nil
 }
 
-func icannStr(icann bool) string {
-	if icann {
-		return "I"
-	}
-	return " "
-}
-
-func wildcardStr(wildcard bool) string {
-	if wildcard {
-		return "*"
-	}
-	return " "
-}
-
 // combineText combines all the strings in labelsList to form one giant string.
 // Overlapping strings will be merged: "arpa" and "parliament" could yield
 // "arparliament".
@@ -616,18 +575,15 @@ func combineText(labelsList []string) string {
 	return text
 }
 
-type byLength []string
-
-func (s byLength) Len() int           { return len(s) }
-func (s byLength) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
-func (s byLength) Less(i, j int) bool { return len(s[i]) < len(s[j]) }
+func byLength(a, b string) int {
+	return cmp.Compare(len(a), len(b))
+}
 
 // removeSubstrings returns a copy of its input with any strings removed
 // that are substrings of other provided strings.
 func removeSubstrings(input []string) []string {
-	// Make a copy of input.
-	ss := append(make([]string, 0, len(input)), input...)
-	sort.Sort(byLength(ss))
+	ss := slices.Clone(input)
+	slices.SortFunc(ss, byLength)
 
 	for i, shortString := range ss {
 		// For each string, only consider strings higher than it in sort order, i.e.
@@ -641,7 +597,7 @@ func removeSubstrings(input []string) []string {
 	}
 
 	// Remove the empty strings.
-	sort.Strings(ss)
+	slices.Sort(ss)
 	for len(ss) > 0 && ss[0] == "" {
 		ss = ss[1:]
 	}
diff --git a/publicsuffix/list.go b/publicsuffix/list.go
@@ -88,7 +88,7 @@ func PublicSuffix(domain string) (publicSuffix string, icann bool) {
 	s, suffix, icannNode, wildcard := domain, len(domain), false, false
 loop:
 	for {
-		dot := strings.LastIndex(s, ".")
+		dot := strings.LastIndexByte(s, '.')
 		if wildcard {
 			icann = icannNode
 			suffix = 1 + dot
@@ -129,7 +129,7 @@ loop:
 	}
 	if suffix == len(domain) {
 		// If no rules match, the prevailing rule is "*".
-		return domain[1+strings.LastIndex(domain, "."):], icann
+		return domain[1+strings.LastIndexByte(domain, '.'):], icann
 	}
 	return domain[suffix:], icann
 }
@@ -178,26 +178,28 @@ func EffectiveTLDPlusOne(domain string) (string, error) {
 	if domain[i] != '.' {
 		return "", fmt.Errorf("publicsuffix: invalid public suffix %q for domain %q", suffix, domain)
 	}
-	return domain[1+strings.LastIndex(domain[:i], "."):], nil
+	return domain[1+strings.LastIndexByte(domain[:i], '.'):], nil
 }
 
 type uint32String string
 
 func (u uint32String) get(i uint32) uint32 {
 	off := i * 4
-	return (uint32(u[off])<<24 |
-		uint32(u[off+1])<<16 |
-		uint32(u[off+2])<<8 |
-		uint32(u[off+3]))
+	u = u[off:] // help the compiler reduce bounds checks
+	return uint32(u[3]) |
+		uint32(u[2])<<8 |
+		uint32(u[1])<<16 |
+		uint32(u[0])<<24
 }
 
 type uint40String string
 
 func (u uint40String) get(i uint32) uint64 {
 	off := uint64(i * (nodesBits / 8))
-	return uint64(u[off])<<32 |
-		uint64(u[off+1])<<24 |
-		uint64(u[off+2])<<16 |
-		uint64(u[off+3])<<8 |
-		uint64(u[off+4])
+	u = u[off:] // help the compiler reduce bounds checks
+	return uint64(u[4]) |
+		uint64(u[3])<<8 |
+		uint64(u[2])<<16 |
+		uint64(u[1])<<24 |
+		uint64(u[0])<<32
 }
