cmd/compile: use bufio.Reader directly in lexer

Removes an intermediate layer of functions that was clogging up a
corner of the compiler's profile graph.

I can't measure a performance improvement running a large build
like jujud, but the profile reports less total time spent in
gc.(*lexer).getr.

Change-Id: I3000585cfcb0f9729d3a3859e9023690a6528591
Reviewed-on: https://go-review.googlesource.com/20565
Reviewed-by: Robert Griesemer <gri@golang.org>
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

src/cmd/compile/internal/gc/bimport.go

@@ -8,8 +8,8 @@
 package gc
 
 import (
+	"bufio"
 	"cmd/compile/internal/big"
-	"cmd/internal/obj"
 	"encoding/binary"
 	"fmt"
 )
@@ -20,7 +20,7 @@ import (
 // changes to bimport.go and bexport.go.
 
 // Import populates importpkg from the serialized package data.
-func Import(in *obj.Biobuf) {
+func Import(in *bufio.Reader) {
 	p := importer{in: in}
 	p.buf = p.bufarray[:]
 
@@ -137,7 +137,7 @@ func idealType(typ *Type) *Type {
 }
 
 type importer struct {
-	in       *obj.Biobuf
+	in       *bufio.Reader
 	buf      []byte   // for reading strings
 	bufarray [64]byte // initial underlying array for buf, large enough to avoid allocation when compiling std lib
 	pkgList  []*Pkg
@@ -855,16 +855,16 @@ func (p *importer) ReadByte() (byte, error) {
 // byte is the bottleneck interface for reading from p.in.
 // It unescapes '|' 'S' to '$' and '|' '|' to '|'.
 func (p *importer) byte() byte {
-	c := obj.Bgetc(p.in)
+	c, err := p.in.ReadByte()
 	p.read++
-	if c < 0 {
+	if err != nil {
-		Fatalf("importer: read error")
+		Fatalf("importer: read error: %v", err)
 	}
 	if c == '|' {
-		c = obj.Bgetc(p.in)
+		c, err = p.in.ReadByte()
 		p.read++
-		if c < 0 {
+		if err != nil {
-			Fatalf("importer: read error")
+			Fatalf("importer: read error: %v", err)
 		}
 		switch c {
 		case 'S':
@@ -875,5 +875,5 @@ func (p *importer) byte() byte {
 			Fatalf("importer: unexpected escape sequence in export data")
 		}
 	}
-	return byte(c)
+	return c
 }

src/cmd/compile/internal/gc/export.go

@@ -5,6 +5,7 @@
 package gc
 
 import (
+	"bufio"
 	"bytes"
 	"cmd/internal/obj"
 	"fmt"
@@ -387,7 +388,7 @@ func dumpexport() {
 			pkgMap = make(map[string]*Pkg)
 			pkgs = nil
 			importpkg = mkpkg("")
-			Import(obj.Binitr(&copy)) // must not die
+			Import(bufio.NewReader(&copy)) // must not die
 			importpkg = nil
 			pkgs = savedPkgs
 			pkgMap = savedPkgMap

src/cmd/compile/internal/gc/lex.go

@@ -7,6 +7,7 @@
 package gc
 
 import (
+	"bufio"
 	"cmd/compile/internal/ssa"
 	"cmd/internal/obj"
 	"flag"
@@ -335,15 +336,16 @@ func Main() {
 
 		linehistpush(infile)
 
-		bin, err := obj.Bopenr(infile)
+		f, err := os.Open(infile)
 		if err != nil {
 			fmt.Printf("open %s: %v\n", infile, err)
 			errorexit()
 		}
+		bin := bufio.NewReader(f)
 
 		// Skip initial BOM if present.
-		if obj.Bgetrune(bin) != BOM {
+		if r, _, _ := bin.ReadRune(); r != BOM {
-			obj.Bungetrune(bin)
+			bin.UnreadRune()
 		}
 
 		block = 1
@@ -362,7 +364,7 @@ func Main() {
 		lexlineno++
 
 		linehistpop()
-		obj.Bterm(bin)
+		f.Close()
 	}
 
 	testdclstack()
@@ -541,7 +543,7 @@ func saveerrors() {
 	nerrors = 0
 }
 
-func arsize(b *obj.Biobuf, name string) int {
+func arsize(b *bufio.Reader, name string) int {
 	var buf [ArhdrSize]byte
 	if _, err := io.ReadFull(b, buf[:]); err != nil {
 		return -1
@@ -555,14 +557,11 @@ func arsize(b *obj.Biobuf, name string) int {
 	return i
 }
 
-func skiptopkgdef(b *obj.Biobuf) bool {
+func skiptopkgdef(b *bufio.Reader) bool {
 	// archive header
-	p := obj.Brdline(b, '\n')
+	p, err := b.ReadString('\n')
-	if p == "" {
+	if err != nil {
-		return false
+		log.Fatalf("reading input: %v", err)
-	}
-	if obj.Blinelen(b) != 8 {
-		return false
 	}
 	if p != "!<arch>\n" {
 		return false
@@ -672,10 +671,10 @@ func loadsys() {
 	incannedimport = 1
 
 	importpkg = Runtimepkg
-	parse_import(obj.Binitr(strings.NewReader(runtimeimport)), nil)
+	parse_import(bufio.NewReader(strings.NewReader(runtimeimport)), nil)
 
 	importpkg = unsafepkg
-	parse_import(obj.Binitr(strings.NewReader(unsafeimport)), nil)
+	parse_import(bufio.NewReader(strings.NewReader(unsafeimport)), nil)
 
 	importpkg = nil
 	incannedimport = 0
@@ -761,12 +760,13 @@ func importfile(f *Val, indent []byte) {
 
 	importpkg.Imported = true
 
-	imp, err := obj.Bopenr(file)
+	impf, err := os.Open(file)
 	if err != nil {
 		Yyerror("can't open import: %q: %v", path_, err)
 		errorexit()
 	}
-	defer obj.Bterm(imp)
+	defer impf.Close()
+	imp := bufio.NewReader(impf)
 
 	if strings.HasSuffix(file, ".a") {
 		if !skiptopkgdef(imp) {
@@ -776,7 +776,13 @@ func importfile(f *Val, indent []byte) {
 	}
 
 	// check object header
-	p := obj.Brdstr(imp, '\n', 1)
+	p, err := imp.ReadString('\n')
+	if err != nil {
+		log.Fatalf("reading input: %v", err)
+	}
+	if len(p) > 0 {
+		p = p[:len(p)-1]
+	}
 
 	if p != "empty archive" {
 		if !strings.HasPrefix(p, "go object ") {
@@ -800,23 +806,23 @@ func importfile(f *Val, indent []byte) {
 	// $$B\n (new format): import directly, then feed the lexer a dummy statement
 
 	// look for $$
-	var c int
+	var c byte
 	for {
-		c = obj.Bgetc(imp)
+		c, err = imp.ReadByte()
-		if c < 0 {
+		if err != nil {
 			break
 		}
 		if c == '$' {
-			c = obj.Bgetc(imp)
+			c, err = imp.ReadByte()
-			if c == '$' || c < 0 {
+			if c == '$' || err != nil {
 				break
 			}
 		}
 	}
 
 	// get character after $$
-	if c >= 0 {
+	if err == nil {
-		c = obj.Bgetc(imp)
+		c, _ = imp.ReadByte()
 	}
 
 	switch c {
@@ -826,7 +832,7 @@ func importfile(f *Val, indent []byte) {
 
 	case 'B':
 		// new export format
-		obj.Bgetc(imp) // skip \n after $$B
+		imp.ReadByte() // skip \n after $$B
 		Import(imp)
 
 	default:
@@ -879,9 +885,7 @@ const (
 
 type lexer struct {
 	// source
-	bin    *obj.Biobuf
+	bin *bufio.Reader
-	peekr1 rune
-	peekr2 rune // second peekc for ...
 
 	nlsemi bool // if set, '\n' and EOF translate to ';'
 
@@ -1025,8 +1029,9 @@ l0:
 		}
 
 		if c1 == '.' {
-			c1 = l.getr()
+			p, err := l.bin.Peek(1)
-			if c1 == '.' {
+			if err == nil && p[0] == '.' {
+				l.getr()
 				c = LDDD
 				goto lx
 			}
@@ -1886,49 +1891,26 @@ func pragcgo(text string) {
 }
 
 func (l *lexer) getr() rune {
-	// unread rune != 0 available
-	if r := l.peekr1; r != 0 {
-		l.peekr1 = l.peekr2
-		l.peekr2 = 0
-		if r == '\n' && importpkg == nil {
-			lexlineno++
-		}
-		return r
-	}
-
 redo:
-	// common case: 7bit ASCII
+	r, w, err := l.bin.ReadRune()
-	c := obj.Bgetc(l.bin)
+	if err != nil {
-	if c < utf8.RuneSelf {
+		if err != io.EOF {
-		if c == 0 {
+			Fatalf("io error: %v", err)
-			yyerrorl(lexlineno, "illegal NUL byte")
-			return 0
 		}
-		if c == '\n' && importpkg == nil {
+		return -1
+	}
+	switch r {
+	case 0:
+		yyerrorl(lexlineno, "illegal NUL byte")
+	case '\n':
+		if importpkg == nil {
 			lexlineno++
 		}
-		return rune(c)
+	case utf8.RuneError:
-	}
+		if w == 1 {
-	// c >= utf8.RuneSelf
+			yyerrorl(lexlineno, "illegal UTF-8 sequence")
-
+		}
-	// uncommon case: non-ASCII
+	case BOM:
-	var buf [utf8.UTFMax]byte
-	buf[0] = byte(c)
-	buf[1] = byte(obj.Bgetc(l.bin))
-	i := 2
-	for ; i < len(buf) && !utf8.FullRune(buf[:i]); i++ {
-		buf[i] = byte(obj.Bgetc(l.bin))
-	}
-
-	r, w := utf8.DecodeRune(buf[:i])
-	if r == utf8.RuneError && w == 1 {
-		// The string conversion here makes a copy for passing
-		// to fmt.Printf, so that buf itself does not escape and
-		// can be allocated on the stack.
-		yyerrorl(lexlineno, "illegal UTF-8 sequence % x", string(buf[:i]))
-	}
-
-	if r == BOM {
 		yyerrorl(lexlineno, "Unicode (UTF-8) BOM in middle of file")
 		goto redo
 	}
@@ -1937,8 +1919,7 @@ redo:
 }
 
 func (l *lexer) ungetr(r rune) {
-	l.peekr2 = l.peekr1
+	l.bin.UnreadRune()
-	l.peekr1 = r
 	if r == '\n' && importpkg == nil {
 		lexlineno--
 	}

src/cmd/compile/internal/gc/parser.go

@@ -13,7 +13,7 @@ package gc
 // to handle optional commas and semicolons before a closing ) or } .
 
 import (
-	"cmd/internal/obj"
+	"bufio"
 	"fmt"
 	"strconv"
 	"strings"
@@ -22,12 +22,12 @@ import (
 const trace = false // if set, parse tracing can be enabled with -x
 
 // parse_import parses the export data of a package that is imported.
-func parse_import(bin *obj.Biobuf, indent []byte) {
+func parse_import(bin *bufio.Reader, indent []byte) {
 	newparser(bin, indent).import_package()
 }
 
 // parse_file parses a single Go source file.
-func parse_file(bin *obj.Biobuf) {
+func parse_file(bin *bufio.Reader) {
 	newparser(bin, nil).file()
 }
 
@@ -40,7 +40,7 @@ type parser struct {
 
 // newparser returns a new parser ready to parse from src.
 // indent is the initial indentation for tracing output.
-func newparser(src *obj.Biobuf, indent []byte) *parser {
+func newparser(src *bufio.Reader, indent []byte) *parser {
 	var p parser
 	p.bin = src
 	p.indent = indent

src/cmd/internal/obj/util.go

@@ -129,18 +129,6 @@ func Bgetc(b *Biobuf) int {
 	return int(c)
 }
 
-func Bgetrune(b *Biobuf) int {
-	r, _, err := b.r.ReadRune()
-	if err != nil {
-		return -1
-	}
-	return int(r)
-}
-
-func Bungetrune(b *Biobuf) {
-	b.r.UnreadRune()
-}
-
 func (b *Biobuf) Read(p []byte) (int, error) {
 	return b.r.Read(p)
 }
@@ -158,17 +146,6 @@ func Brdline(b *Biobuf, delim int) string {
 	return string(s)
 }
 
-func Brdstr(b *Biobuf, delim int, cut int) string {
-	s, err := b.r.ReadString(byte(delim))
-	if err != nil {
-		log.Fatalf("reading input: %v", err)
-	}
-	if len(s) > 0 && cut > 0 {
-		s = s[:len(s)-1]
-	}
-	return s
-}
-
 func Blinelen(b *Biobuf) int {
 	return b.linelen
 }

test/syntax/ddd.go

@@ -0,0 +1,11 @@
+// errorcheck
+
+// Copyright 2016 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 main
+
+func f() {
+	g(f..3) // ERROR "unexpected literal \.3, expecting name or \("
+}