src/cmd/pack/pack.go - go - Git at Google

 // Copyright 2014 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

 import (
 	"bufio"
 	"bytes"
 	"errors"
 	"fmt"
 	"io"
 	"log"
 	"os"
 	"strconv"
 	"strings"
 	"time"
 	"unicode/utf8"
 )

 /*
 The archive format is:

 First, on a line by itself
 	!<arch>

 Then zero or more file records. Each file record has a fixed-size one-line header
 followed by data bytes followed by an optional padding byte. The header is:

 	%-16s%-12d%-6d%-6d%-8o%-10d`
 	name mtime uid gid mode size

 (note the trailing backquote). The %-16s here means at most 16 *bytes* of
 the name, and if shorter, space padded on the right.
 */

 const usageMessage = `Usage: pack op file.a [name....]
 Where op is one of cprtx optionally followed by v for verbose output.
 For compatibility with old Go build environments the op string grc is
 accepted as a synonym for c.

 For more information, run
 	godoc cmd/pack`

 func usage() {
 	fmt.Fprintln(os.Stderr, usageMessage)
 	os.Exit(2)
 }

 func main() {
 	log.SetFlags(0)
 	log.SetPrefix("pack: ")
 	// need "pack op archive" at least.
 	if len(os.Args) < 3 {
 		log.Print("not enough arguments")
 		fmt.Fprintln(os.Stderr)
 		usage()
 	}
 	setOp(os.Args[1])
 	var ar *Archive
 	switch op {
 	case 'p':
 		ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
 		ar.scan(ar.printContents)
 	case 'r':
 		ar = archive(os.Args[2], os.O_RDWR, os.Args[3:])
 		ar.scan(ar.skipContents)
 		ar.addFiles()
 	case 'c':
 		ar = archive(os.Args[2], os.O_RDWR|os.O_TRUNC, os.Args[3:])
 		ar.addPkgdef()
 		ar.addFiles()
 	case 't':
 		ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
 		ar.scan(ar.tableOfContents)
 	case 'x':
 		ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
 		ar.scan(ar.extractContents)
 	default:
 		log.Printf("invalid operation %q", os.Args[1])
 		fmt.Fprintln(os.Stderr)
 		usage()
 	}
 	if len(ar.files) > 0 {
 		log.Fatalf("file %q not in archive", ar.files[0])
 	}
 }

 // The unusual ancestry means the arguments are not Go-standard.
 // These variables hold the decoded operation specified by the first argument.
 // op holds the operation we are doing (prtx).
 // verbose tells whether the 'v' option was specified.
 var (
 	op      rune
 	verbose bool
 )

 // setOp parses the operation string (first argument).
 func setOp(arg string) {
 	// Recognize 'go tool pack grc' because that was the
 	// formerly canonical way to build a new archive
 	// from a set of input files. Accepting it keeps old
 	// build systems working with both Go 1.2 and Go 1.3.
 	if arg == "grc" {
 		arg = "c"
 	}

 	for _, r := range arg {
 		switch r {
 		case 'c', 'p', 'r', 't', 'x':
 			if op != 0 {
 				// At most one can be set.
 				usage()
 			}
 			op = r
 		case 'v':
 			if verbose {
 				// Can be set only once.
 				usage()
 			}
 			verbose = true
 		default:
 			usage()
 		}
 	}
 }

 const (
 	arHeader    = "!<arch>\n"
 	entryHeader = "%s%-12d%-6d%-6d%-8o%-10d`\n"
 	// In entryHeader the first entry, the name, is always printed as 16 bytes right-padded.
 	entryLen   = 16 + 12 + 6 + 6 + 8 + 10 + 1 + 1
 	timeFormat = "Jan _2 15:04 2006"
 )

 // An Archive represents an open archive file. It is always scanned sequentially
 // from start to end, without backing up.
 type Archive struct {
 	fd       *os.File // Open file descriptor.
 	files    []string // Explicit list of files to be processed.
 	pad      int      // Padding bytes required at end of current archive file
 	matchAll bool     // match all files in archive
 }

 // archive opens (and if necessary creates) the named archive.
 func archive(name string, mode int, files []string) *Archive {
 	// If the file exists, it must be an archive. If it doesn't exist, or if
 	// we're doing the c command, indicated by O_TRUNC, truncate the archive.
 	if !existingArchive(name) || mode&os.O_TRUNC != 0 {
 		create(name)
 		mode &^= os.O_TRUNC
 	}
 	fd, err := os.OpenFile(name, mode, 0)
 	if err != nil {
 		log.Fatal(err)
 	}
 	checkHeader(fd)
 	return &Archive{
 		fd:       fd,
 		files:    files,
 		matchAll: len(files) == 0,
 	}
 }

 // create creates and initializes an archive that does not exist.
 func create(name string) {
 	fd, err := os.Create(name)
 	if err != nil {
 		log.Fatal(err)
 	}
 	_, err = fmt.Fprint(fd, arHeader)
 	if err != nil {
 		log.Fatal(err)
 	}
 	fd.Close()
 }

 // existingArchive reports whether the file exists and is a valid archive.
 // If it exists but is not an archive, existingArchive will exit.
 func existingArchive(name string) bool {
 	fd, err := os.Open(name)
 	if err != nil {
 		if os.IsNotExist(err) {
 			return false
 		}
 		log.Fatalf("cannot open file: %s", err)
 	}
 	checkHeader(fd)
 	fd.Close()
 	return true
 }

 // checkHeader verifies the header of the file. It assumes the file
 // is positioned at 0 and leaves it positioned at the end of the header.
 func checkHeader(fd *os.File) {
 	buf := make([]byte, len(arHeader))
 	_, err := io.ReadFull(fd, buf)
 	if err != nil || string(buf) != arHeader {
 		log.Fatalf("%s is not an archive: bad header", fd.Name())
 	}
 }

 // An Entry is the internal representation of the per-file header information of one entry in the archive.
 type Entry struct {
 	name  string
 	mtime int64
 	uid   int
 	gid   int
 	mode  os.FileMode
 	size  int64
 }

 func (e *Entry) String() string {
 	return fmt.Sprintf("%s %6d/%-6d %12d %s %s",
 		(e.mode & 0777).String(),
 		e.uid,
 		e.gid,
 		e.size,
 		time.Unix(e.mtime, 0).Format(timeFormat),
 		e.name)
 }

 // readMetadata reads and parses the metadata for the next entry in the archive.
 func (ar *Archive) readMetadata() *Entry {
 	buf := make([]byte, entryLen)
 	_, err := io.ReadFull(ar.fd, buf)
 	if err == io.EOF {
 		// No entries left.
 		return nil
 	}
 	if err != nil || buf[entryLen-2] != '`' || buf[entryLen-1] != '\n' {
 		log.Fatal("file is not an archive: bad entry")
 	}
 	entry := new(Entry)
 	entry.name = strings.TrimRight(string(buf[:16]), " ")
 	if len(entry.name) == 0 {
 		log.Fatal("file is not an archive: bad name")
 	}
 	buf = buf[16:]
 	str := string(buf)
 	get := func(width, base, bitsize int) int64 {
 		v, err := strconv.ParseInt(strings.TrimRight(str[:width], " "), base, bitsize)
 		if err != nil {
 			log.Fatal("file is not an archive: bad number in entry: ", err)
 		}
 		str = str[width:]
 		return v
 	}
 	// %-16s%-12d%-6d%-6d%-8o%-10d`
 	entry.mtime = get(12, 10, 64)
 	entry.uid = int(get(6, 10, 32))
 	entry.gid = int(get(6, 10, 32))
 	entry.mode = os.FileMode(get(8, 8, 32))
 	entry.size = get(10, 10, 64)
 	return entry
 }

 // scan scans the archive and executes the specified action on each entry.
 // When action returns, the file offset is at the start of the next entry.
 func (ar *Archive) scan(action func(*Entry)) {
 	for {
 		entry := ar.readMetadata()
 		if entry == nil {
 			break
 		}
 		action(entry)
 	}
 }

 // listEntry prints to standard output a line describing the entry.
 func listEntry(ar *Archive, entry *Entry, verbose bool) {
 	if verbose {
 		fmt.Fprintf(stdout, "%s\n", entry)
 	} else {
 		fmt.Fprintf(stdout, "%s\n", entry.name)
 	}
 }

 // output copies the entry to the specified writer.
 func (ar *Archive) output(entry *Entry, w io.Writer) {
 	n, err := io.Copy(w, io.LimitReader(ar.fd, entry.size))
 	if err != nil {
 		log.Fatal(err)
 	}
 	if n != entry.size {
 		log.Fatal("short file")
 	}
 	if entry.size&1 == 1 {
 		_, err := ar.fd.Seek(1, 1)
 		if err != nil {
 			log.Fatal(err)
 		}
 	}
 }

 // skip skips the entry without reading it.
 func (ar *Archive) skip(entry *Entry) {
 	size := entry.size
 	if size&1 == 1 {
 		size++
 	}
 	_, err := ar.fd.Seek(size, 1)
 	if err != nil {
 		log.Fatal(err)
 	}
 }

 // match reports whether the entry matches the argument list.
 // If it does, it also drops the file from the to-be-processed list.
 func (ar *Archive) match(entry *Entry) bool {
 	if ar.matchAll {
 		return true
 	}
 	for i, name := range ar.files {
 		if entry.name == name {
 			copy(ar.files[i:], ar.files[i+1:])
 			ar.files = ar.files[:len(ar.files)-1]
 			return true
 		}
 	}
 	return false
 }

 // addFiles adds files to the archive. The archive is known to be
 // sane and we are positioned at the end. No attempt is made
 // to check for existing files.
 func (ar *Archive) addFiles() {
 	if len(ar.files) == 0 {
 		usage()
 	}
 	for _, file := range ar.files {
 		if verbose {
 			fmt.Printf("%s\n", file)
 		}
 		fd, err := os.Open(file)
 		if err != nil {
 			log.Fatal(err)
 		}
 		ar.addFile(fd)
 	}
 	ar.files = nil
 }

 // FileLike abstracts the few methods we need, so we can test without needing real files.
 type FileLike interface {
 	Name() string
 	Stat() (os.FileInfo, error)
 	Read([]byte) (int, error)
 	Close() error
 }

 // addFile adds a single file to the archive
 func (ar *Archive) addFile(fd FileLike) {
 	defer fd.Close()
 	// Format the entry.
 	// First, get its info.
 	info, err := fd.Stat()
 	if err != nil {
 		log.Fatal(err)
 	}
 	// mtime, uid, gid are all zero so repeated builds produce identical output.
 	mtime := int64(0)
 	uid := 0
 	gid := 0
 	ar.startFile(info.Name(), mtime, uid, gid, info.Mode(), info.Size())
 	n64, err := io.Copy(ar.fd, fd)
 	if err != nil {
 		log.Fatal("writing file: ", err)
 	}
 	if n64 != info.Size() {
 		log.Fatalf("writing file: wrote %d bytes; file is size %d", n64, info.Size())
 	}
 	ar.endFile()
 }

 // startFile writes the archive entry header.
 func (ar *Archive) startFile(name string, mtime int64, uid, gid int, mode os.FileMode, size int64) {
 	n, err := fmt.Fprintf(ar.fd, entryHeader, exactly16Bytes(name), mtime, uid, gid, mode, size)
 	if err != nil || n != entryLen {
 		log.Fatal("writing entry header: ", err)
 	}
 	ar.pad = int(size & 1)
 }

 // endFile writes the archive entry tail (a single byte of padding, if the file size was odd).
 func (ar *Archive) endFile() {
 	if ar.pad != 0 {
 		_, err := ar.fd.Write([]byte{0})
 		if err != nil {
 			log.Fatal("writing archive: ", err)
 		}
 		ar.pad = 0
 	}
 }

 // addPkgdef adds the __.PKGDEF file to the archive, copied
 // from the first Go object file on the file list, if any.
 // The archive is known to be empty.
 func (ar *Archive) addPkgdef() {
 	for _, file := range ar.files {
 		pkgdef, err := readPkgdef(file)
 		if err != nil {
 			continue
 		}
 		if verbose {
 			fmt.Printf("__.PKGDEF # %s\n", file)
 		}
 		ar.startFile("__.PKGDEF", 0, 0, 0, 0644, int64(len(pkgdef)))
 		_, err = ar.fd.Write(pkgdef)
 		if err != nil {
 			log.Fatal("writing __.PKGDEF: ", err)
 		}
 		ar.endFile()
 		break
 	}
 }

 // readPkgdef extracts the __.PKGDEF data from a Go object file.
 func readPkgdef(file string) (data []byte, err error) {
 	f, err := os.Open(file)
 	if err != nil {
 		return nil, err
 	}
 	defer f.Close()

 	// Read from file, collecting header for __.PKGDEF.
 	// The header is from the beginning of the file until a line
 	// containing just "!". The first line must begin with "go object ".
 	rbuf := bufio.NewReader(f)
 	var wbuf bytes.Buffer
 	for {
 		line, err := rbuf.ReadBytes('\n')
 		if err != nil {
 			return nil, err
 		}
 		if wbuf.Len() == 0 && !bytes.HasPrefix(line, []byte("go object ")) {
 			return nil, errors.New("not a Go object file")
 		}
 		if bytes.Equal(line, []byte("!\n")) {
 			break
 		}
 		wbuf.Write(line)
 	}
 	return wbuf.Bytes(), nil
 }

 // exactly16Bytes truncates the string if necessary so it is at most 16 bytes long,
 // then pads the result with spaces to be exactly 16 bytes.
 // Fmt uses runes for its width calculation, but we need bytes in the entry header.
 func exactly16Bytes(s string) string {
 	for len(s) > 16 {
 		_, wid := utf8.DecodeLastRuneInString(s)
 		s = s[:len(s)-wid]
 	}
 	const sixteenSpaces = "                "
 	s += sixteenSpaces[:16-len(s)]
 	return s
 }

 // Finally, the actual commands. Each is an action.

 // can be modified for testing.
 var stdout io.Writer = os.Stdout

 // printContents implements the 'p' command.
 func (ar *Archive) printContents(entry *Entry) {
 	if ar.match(entry) {
 		if verbose {
 			listEntry(ar, entry, false)
 		}
 		ar.output(entry, stdout)
 	} else {
 		ar.skip(entry)
 	}
 }

 // skipContents implements the first part of the 'r' command.
 // It just scans the archive to make sure it's intact.
 func (ar *Archive) skipContents(entry *Entry) {
 	ar.skip(entry)
 }

 // tableOfContents implements the 't' command.
 func (ar *Archive) tableOfContents(entry *Entry) {
 	if ar.match(entry) {
 		listEntry(ar, entry, verbose)
 	}
 	ar.skip(entry)
 }

 // extractContents implements the 'x' command.
 func (ar *Archive) extractContents(entry *Entry) {
 	if ar.match(entry) {
 		if verbose {
 			listEntry(ar, entry, false)
 		}
 		fd, err := os.OpenFile(entry.name, os.O_WRONLY|os.O_CREATE|os.O_TRUNC, entry.mode)
 		if err != nil {
 			log.Fatal(err)
 		}
 		ar.output(entry, fd)
 		fd.Close()
 	} else {
 		ar.skip(entry)
 	}
 }
	// Copyright 2014 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

	import (
	"bufio"
	"bytes"
	"errors"
	"fmt"
	"io"
	"log"
	"os"
	"strconv"
	"strings"
	"time"
	"unicode/utf8"
	)

	/*
	The archive format is:

	First, on a line by itself
	!<arch>

	Then zero or more file records. Each file record has a fixed-size one-line header
	followed by data bytes followed by an optional padding byte. The header is:

	%-16s%-12d%-6d%-6d%-8o%-10d`
	name mtime uid gid mode size

	(note the trailing backquote). The %-16s here means at most 16 bytes of
	the name, and if shorter, space padded on the right.
	*/

	const usageMessage = `Usage: pack op file.a [name....]
	Where op is one of cprtx optionally followed by v for verbose output.
	For compatibility with old Go build environments the op string grc is
	accepted as a synonym for c.

	For more information, run
	godoc cmd/pack`

	func usage() {
	fmt.Fprintln(os.Stderr, usageMessage)
	os.Exit(2)
	}

	func main() {
	log.SetFlags(0)
	log.SetPrefix("pack: ")
	// need "pack op archive" at least.
	if len(os.Args) < 3 {
	log.Print("not enough arguments")
	fmt.Fprintln(os.Stderr)
	usage()
	}
	setOp(os.Args[1])
	var ar *Archive
	switch op {
	case 'p':
	ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
	ar.scan(ar.printContents)
	case 'r':
	ar = archive(os.Args[2], os.O_RDWR, os.Args[3:])
	ar.scan(ar.skipContents)
	ar.addFiles()
	case 'c':
	ar = archive(os.Args[2], os.O_RDWR\|os.O_TRUNC, os.Args[3:])
	ar.addPkgdef()
	ar.addFiles()
	case 't':
	ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
	ar.scan(ar.tableOfContents)
	case 'x':
	ar = archive(os.Args[2], os.O_RDONLY, os.Args[3:])
	ar.scan(ar.extractContents)
	default:
	log.Printf("invalid operation %q", os.Args[1])
	fmt.Fprintln(os.Stderr)
	usage()
	}
	if len(ar.files) > 0 {
	log.Fatalf("file %q not in archive", ar.files[0])
	}
	}

	// The unusual ancestry means the arguments are not Go-standard.
	// These variables hold the decoded operation specified by the first argument.
	// op holds the operation we are doing (prtx).
	// verbose tells whether the 'v' option was specified.
	var (
	op rune
	verbose bool
	)

	// setOp parses the operation string (first argument).
	func setOp(arg string) {
	// Recognize 'go tool pack grc' because that was the
	// formerly canonical way to build a new archive
	// from a set of input files. Accepting it keeps old
	// build systems working with both Go 1.2 and Go 1.3.
	if arg == "grc" {
	arg = "c"
	}

	for _, r := range arg {
	switch r {
	case 'c', 'p', 'r', 't', 'x':
	if op != 0 {
	// At most one can be set.
	usage()
	}
	op = r
	case 'v':
	if verbose {
	// Can be set only once.
	usage()
	}
	verbose = true
	default:
	usage()
	}
	}
	}

	const (
	arHeader = "!<arch>\n"
	entryHeader = "%s%-12d%-6d%-6d%-8o%-10d`\n"
	// In entryHeader the first entry, the name, is always printed as 16 bytes right-padded.
	entryLen = 16 + 12 + 6 + 6 + 8 + 10 + 1 + 1
	timeFormat = "Jan _2 15:04 2006"
	)

	// An Archive represents an open archive file. It is always scanned sequentially
	// from start to end, without backing up.
	type Archive struct {
	fd *os.File // Open file descriptor.
	files []string // Explicit list of files to be processed.
	pad int // Padding bytes required at end of current archive file
	matchAll bool // match all files in archive
	}

	// archive opens (and if necessary creates) the named archive.
	func archive(name string, mode int, files []string) *Archive {
	// If the file exists, it must be an archive. If it doesn't exist, or if
	// we're doing the c command, indicated by O_TRUNC, truncate the archive.
	if !existingArchive(name) \|\| mode&os.O_TRUNC != 0 {
	create(name)
	mode &^= os.O_TRUNC
	}
	fd, err := os.OpenFile(name, mode, 0)
	if err != nil {
	log.Fatal(err)
	}
	checkHeader(fd)
	return &Archive{
	fd: fd,
	files: files,
	matchAll: len(files) == 0,
	}
	}

	// create creates and initializes an archive that does not exist.
	func create(name string) {
	fd, err := os.Create(name)
	if err != nil {
	log.Fatal(err)
	}
	_, err = fmt.Fprint(fd, arHeader)
	if err != nil {
	log.Fatal(err)
	}
	fd.Close()
	}

	// existingArchive reports whether the file exists and is a valid archive.
	// If it exists but is not an archive, existingArchive will exit.
	func existingArchive(name string) bool {
	fd, err := os.Open(name)
	if err != nil {
	if os.IsNotExist(err) {
	return false
	}
	log.Fatalf("cannot open file: %s", err)
	}
	checkHeader(fd)
	fd.Close()
	return true
	}

	// checkHeader verifies the header of the file. It assumes the file
	// is positioned at 0 and leaves it positioned at the end of the header.
	func checkHeader(fd *os.File) {
	buf := make([]byte, len(arHeader))
	_, err := io.ReadFull(fd, buf)
	if err != nil \|\| string(buf) != arHeader {
	log.Fatalf("%s is not an archive: bad header", fd.Name())
	}
	}

	// An Entry is the internal representation of the per-file header information of one entry in the archive.
	type Entry struct {
	name string
	mtime int64
	uid int
	gid int
	mode os.FileMode
	size int64
	}

	func (e *Entry) String() string {
	return fmt.Sprintf("%s %6d/%-6d %12d %s %s",
	(e.mode & 0777).String(),
	e.uid,
	e.gid,
	e.size,
	time.Unix(e.mtime, 0).Format(timeFormat),
	e.name)
	}

	// readMetadata reads and parses the metadata for the next entry in the archive.
	func (ar Archive) readMetadata() Entry {
	buf := make([]byte, entryLen)
	_, err := io.ReadFull(ar.fd, buf)
	if err == io.EOF {
	// No entries left.
	return nil
	}
	if err != nil \|\| buf[entryLen-2] != '`' \|\| buf[entryLen-1] != '\n' {
	log.Fatal("file is not an archive: bad entry")
	}
	entry := new(Entry)
	entry.name = strings.TrimRight(string(buf[:16]), " ")
	if len(entry.name) == 0 {
	log.Fatal("file is not an archive: bad name")
	}
	buf = buf[16:]
	str := string(buf)
	get := func(width, base, bitsize int) int64 {
	v, err := strconv.ParseInt(strings.TrimRight(str[:width], " "), base, bitsize)
	if err != nil {
	log.Fatal("file is not an archive: bad number in entry: ", err)
	}
	str = str[width:]
	return v
	}
	// %-16s%-12d%-6d%-6d%-8o%-10d`
	entry.mtime = get(12, 10, 64)
	entry.uid = int(get(6, 10, 32))
	entry.gid = int(get(6, 10, 32))
	entry.mode = os.FileMode(get(8, 8, 32))
	entry.size = get(10, 10, 64)
	return entry
	}

	// scan scans the archive and executes the specified action on each entry.
	// When action returns, the file offset is at the start of the next entry.
	func (ar Archive) scan(action func(Entry)) {
	for {
	entry := ar.readMetadata()
	if entry == nil {
	break
	}
	action(entry)
	}
	}

	// listEntry prints to standard output a line describing the entry.
	func listEntry(ar Archive, entry Entry, verbose bool) {
	if verbose {
	fmt.Fprintf(stdout, "%s\n", entry)
	} else {
	fmt.Fprintf(stdout, "%s\n", entry.name)
	}
	}

	// output copies the entry to the specified writer.
	func (ar Archive) output(entry Entry, w io.Writer) {
	n, err := io.Copy(w, io.LimitReader(ar.fd, entry.size))
	if err != nil {
	log.Fatal(err)
	}
	if n != entry.size {
	log.Fatal("short file")
	}
	if entry.size&1 == 1 {
	_, err := ar.fd.Seek(1, 1)
	if err != nil {
	log.Fatal(err)
	}
	}
	}

	// skip skips the entry without reading it.
	func (ar Archive) skip(entry Entry) {
	size := entry.size
	if size&1 == 1 {
	size++
	}
	_, err := ar.fd.Seek(size, 1)
	if err != nil {
	log.Fatal(err)
	}
	}

	// match reports whether the entry matches the argument list.
	// If it does, it also drops the file from the to-be-processed list.
	func (ar Archive) match(entry Entry) bool {
	if ar.matchAll {
	return true
	}
	for i, name := range ar.files {
	if entry.name == name {
	copy(ar.files[i:], ar.files[i+1:])
	ar.files = ar.files[:len(ar.files)-1]
	return true
	}
	}
	return false
	}

	// addFiles adds files to the archive. The archive is known to be
	// sane and we are positioned at the end. No attempt is made
	// to check for existing files.
	func (ar *Archive) addFiles() {
	if len(ar.files) == 0 {
	usage()
	}
	for _, file := range ar.files {
	if verbose {
	fmt.Printf("%s\n", file)
	}
	fd, err := os.Open(file)
	if err != nil {
	log.Fatal(err)
	}
	ar.addFile(fd)
	}
	ar.files = nil
	}

	// FileLike abstracts the few methods we need, so we can test without needing real files.
	type FileLike interface {
	Name() string
	Stat() (os.FileInfo, error)
	Read([]byte) (int, error)
	Close() error
	}

	// addFile adds a single file to the archive
	func (ar *Archive) addFile(fd FileLike) {
	defer fd.Close()
	// Format the entry.
	// First, get its info.
	info, err := fd.Stat()
	if err != nil {
	log.Fatal(err)
	}
	// mtime, uid, gid are all zero so repeated builds produce identical output.
	mtime := int64(0)
	uid := 0
	gid := 0
	ar.startFile(info.Name(), mtime, uid, gid, info.Mode(), info.Size())
	n64, err := io.Copy(ar.fd, fd)
	if err != nil {
	log.Fatal("writing file: ", err)
	}
	if n64 != info.Size() {
	log.Fatalf("writing file: wrote %d bytes; file is size %d", n64, info.Size())
	}
	ar.endFile()
	}

	// startFile writes the archive entry header.
	func (ar *Archive) startFile(name string, mtime int64, uid, gid int, mode os.FileMode, size int64) {
	n, err := fmt.Fprintf(ar.fd, entryHeader, exactly16Bytes(name), mtime, uid, gid, mode, size)
	if err != nil \|\| n != entryLen {
	log.Fatal("writing entry header: ", err)
	}
	ar.pad = int(size & 1)
	}

	// endFile writes the archive entry tail (a single byte of padding, if the file size was odd).
	func (ar *Archive) endFile() {
	if ar.pad != 0 {
	_, err := ar.fd.Write([]byte{0})
	if err != nil {
	log.Fatal("writing archive: ", err)
	}
	ar.pad = 0
	}
	}

	// addPkgdef adds the __.PKGDEF file to the archive, copied
	// from the first Go object file on the file list, if any.
	// The archive is known to be empty.
	func (ar *Archive) addPkgdef() {
	for _, file := range ar.files {
	pkgdef, err := readPkgdef(file)
	if err != nil {
	continue
	}
	if verbose {
	fmt.Printf("__.PKGDEF # %s\n", file)
	}
	ar.startFile("__.PKGDEF", 0, 0, 0, 0644, int64(len(pkgdef)))
	_, err = ar.fd.Write(pkgdef)
	if err != nil {
	log.Fatal("writing __.PKGDEF: ", err)
	}
	ar.endFile()
	break
	}
	}

	// readPkgdef extracts the __.PKGDEF data from a Go object file.
	func readPkgdef(file string) (data []byte, err error) {
	f, err := os.Open(file)
	if err != nil {
	return nil, err
	}
	defer f.Close()

	// Read from file, collecting header for __.PKGDEF.
	// The header is from the beginning of the file until a line
	// containing just "!". The first line must begin with "go object ".
	rbuf := bufio.NewReader(f)
	var wbuf bytes.Buffer
	for {
	line, err := rbuf.ReadBytes('\n')
	if err != nil {
	return nil, err
	}
	if wbuf.Len() == 0 && !bytes.HasPrefix(line, []byte("go object ")) {
	return nil, errors.New("not a Go object file")
	}
	if bytes.Equal(line, []byte("!\n")) {
	break
	}
	wbuf.Write(line)
	}
	return wbuf.Bytes(), nil
	}

	// exactly16Bytes truncates the string if necessary so it is at most 16 bytes long,
	// then pads the result with spaces to be exactly 16 bytes.
	// Fmt uses runes for its width calculation, but we need bytes in the entry header.
	func exactly16Bytes(s string) string {
	for len(s) > 16 {
	_, wid := utf8.DecodeLastRuneInString(s)
	s = s[:len(s)-wid]
	}
	const sixteenSpaces = " "
	s += sixteenSpaces[:16-len(s)]
	return s
	}

	// Finally, the actual commands. Each is an action.

	// can be modified for testing.
	var stdout io.Writer = os.Stdout

	// printContents implements the 'p' command.
	func (ar Archive) printContents(entry Entry) {
	if ar.match(entry) {
	if verbose {
	listEntry(ar, entry, false)
	}
	ar.output(entry, stdout)
	} else {
	ar.skip(entry)
	}
	}

	// skipContents implements the first part of the 'r' command.
	// It just scans the archive to make sure it's intact.
	func (ar Archive) skipContents(entry Entry) {
	ar.skip(entry)
	}

	// tableOfContents implements the 't' command.
	func (ar Archive) tableOfContents(entry Entry) {
	if ar.match(entry) {
	listEntry(ar, entry, verbose)
	}
	ar.skip(entry)
	}

	// extractContents implements the 'x' command.
	func (ar Archive) extractContents(entry Entry) {
	if ar.match(entry) {
	if verbose {
	listEntry(ar, entry, false)
	}
	fd, err := os.OpenFile(entry.name, os.O_WRONLY\|os.O_CREATE\|os.O_TRUNC, entry.mode)
	if err != nil {
	log.Fatal(err)
	}
	ar.output(entry, fd)
	fd.Close()
	} else {
	ar.skip(entry)
	}
	}