internal/lazyregexp: add a lazy Regexp package

This was implemented as part of go/doc, but it's going to be useful in
other packages. In particular, many packages under cmd/go like web and
vcs make somewhat heavy use of global regexes, which add a non-trivial
amount of init work to the cmd/go program.

A lazy wrapper around regexp.Regexp will make it trivial to get rid of
the extra cost with a trivial refactor, so make it possible for other
packages in the repository to make use of it. While naming the package,
give the members better names, such as lazyregexp.New and
lazyregexp.Regexp.

We're also considering adding some form of a lazy API to the public
regexp package, so this internal package will allow us to get some
initial experience across std and cmd.

For #29382.

Change-Id: I30b0e72871d5267c309786f95f4cb15c68b2393d
Reviewed-on: https://go-review.googlesource.com/c/164040
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

src/go/build/deps_test.go

@@ -192,10 +192,11 @@ var pkgDeps = map[string][]string{
 	"runtime/trace":  {"L0", "context", "fmt"},
 	"text/tabwriter": {"L2"},
 
-	"testing":          {"L2", "flag", "fmt", "internal/race", "os", "runtime/debug", "runtime/pprof", "runtime/trace", "time"},
-	"testing/iotest":   {"L2", "log"},
-	"testing/quick":    {"L2", "flag", "fmt", "reflect", "time"},
-	"internal/testenv": {"L2", "OS", "flag", "testing", "syscall"},
+	"testing":             {"L2", "flag", "fmt", "internal/race", "os", "runtime/debug", "runtime/pprof", "runtime/trace", "time"},
+	"testing/iotest":      {"L2", "log"},
+	"testing/quick":       {"L2", "flag", "fmt", "reflect", "time"},
+	"internal/testenv":    {"L2", "OS", "flag", "testing", "syscall"},
+	"internal/lazyregexp": {"L2", "OS", "regexp"},
 
 	// L4 is defined as L3+fmt+log+time, because in general once
 	// you're using L3 packages, use of fmt, log, or time is not a big deal.
@@ -208,7 +209,7 @@ var pkgDeps = map[string][]string{
 
 	// Go parser.
 	"go/ast":     {"L4", "OS", "go/scanner", "go/token"},
-	"go/doc":     {"L4", "OS", "go/ast", "go/token", "regexp", "text/template"},
+	"go/doc":     {"L4", "OS", "go/ast", "go/token", "regexp", "internal/lazyregexp", "text/template"},
 	"go/parser":  {"L4", "OS", "go/ast", "go/scanner", "go/token"},
 	"go/printer": {"L4", "OS", "go/ast", "go/scanner", "go/token", "text/tabwriter"},
 	"go/scanner": {"L4", "OS", "go/token"},

src/go/doc/comment.go

@@ -8,6 +8,7 @@ package doc
 
 import (
 	"bytes"
+	"internal/lazyregexp"
 	"io"
 	"strings"
 	"text/template" // for HTMLEscape
@@ -69,7 +70,7 @@ const (
 	urlRx = protoPart + `://` + hostPart + pathPart
 )
 
-var matchRx = newLazyRE(`(` + urlRx + `)|(` + identRx + `)`)
+var matchRx = lazyregexp.New(`(` + urlRx + `)|(` + identRx + `)`)
 
 var (
 	html_a      = []byte(`<a href="`)
@@ -273,7 +274,7 @@ type block struct {
 	lines []string
 }
 
-var nonAlphaNumRx = newLazyRE(`[^a-zA-Z0-9]`)
+var nonAlphaNumRx = lazyregexp.New(`[^a-zA-Z0-9]`)
 
 func anchorID(line string) string {
 	// Add a "hdr-" prefix to avoid conflicting with IDs used for package symbols.

src/go/doc/reader.go

@@ -7,6 +7,7 @@ package doc
 import (
 	"go/ast"
 	"go/token"
+	"internal/lazyregexp"
 	"sort"
 	"strconv"
 )
@@ -439,9 +440,9 @@ func (r *reader) readFunc(fun *ast.FuncDecl) {
 }
 
 var (
-	noteMarker    = `([A-Z][A-Z]+)\(([^)]+)\):?`           // MARKER(uid), MARKER at least 2 chars, uid at least 1 char
-	noteMarkerRx  = newLazyRE(`^[ \t]*` + noteMarker)      // MARKER(uid) at text start
-	noteCommentRx = newLazyRE(`^/[/*][ \t]*` + noteMarker) // MARKER(uid) at comment start
+	noteMarker    = `([A-Z][A-Z]+)\(([^)]+)\):?`                // MARKER(uid), MARKER at least 2 chars, uid at least 1 char
+	noteMarkerRx  = lazyregexp.New(`^[ \t]*` + noteMarker)      // MARKER(uid) at text start
+	noteCommentRx = lazyregexp.New(`^/[/*][ \t]*` + noteMarker) // MARKER(uid) at comment start
 )
 
 // readNote collects a single note from a sequence of comments.

src/go/doc/lazyre.go → src/internal/lazyregexp/lazyre.go

@@ -2,7 +2,7 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package doc
+package lazyregexp
 
 import (
 	"os"
@@ -11,38 +11,38 @@ import (
 	"sync"
 )
 
-type lazyRE struct {
+type Regexp struct {
 	str  string
 	once sync.Once
 	rx   *regexp.Regexp
 }
 
-func (r *lazyRE) re() *regexp.Regexp {
+func (r *Regexp) re() *regexp.Regexp {
 	r.once.Do(r.build)
 	return r.rx
 }
 
-func (r *lazyRE) build() {
+func (r *Regexp) build() {
 	r.rx = regexp.MustCompile(r.str)
 	r.str = ""
 }
 
-func (r *lazyRE) FindStringSubmatchIndex(s string) []int {
+func (r *Regexp) FindStringSubmatchIndex(s string) []int {
 	return r.re().FindStringSubmatchIndex(s)
 }
 
-func (r *lazyRE) ReplaceAllString(src, repl string) string {
+func (r *Regexp) ReplaceAllString(src, repl string) string {
 	return r.re().ReplaceAllString(src, repl)
 }
 
-func (r *lazyRE) MatchString(s string) bool {
+func (r *Regexp) MatchString(s string) bool {
 	return r.re().MatchString(s)
 }
 
 var inTest = len(os.Args) > 0 && strings.HasSuffix(strings.TrimSuffix(os.Args[0], ".exe"), ".test")
 
-func newLazyRE(str string) *lazyRE {
-	lr := &lazyRE{str: str}
+func New(str string) *Regexp {
+	lr := &Regexp{str: str}
 	if inTest {
 		// In tests, always compile the regexps early.
 		lr.re()