os/user: add non-cgo versions of Lookup, LookupId

If you cross compile for a Unix target and call user.Lookup("root")
or user.LookupId("0"), we'll try to read the answer out of
/etc/passwd instead of returning an "unimplemented" error.

The equivalent cgo function calls getpwuid_r in glibc, which
may reach out to the NSS database or allow callers to register
extensions. The pure Go implementation only reads from /etc/passwd.

Change-Id: I56a302d634b15ba5097f9f0d6a758c68e486ba6d
Reviewed-on: https://go-review.googlesource.com/37664
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

src/os/user/lookup_stubs.go

@@ -15,7 +15,6 @@ import (
 )
 
 func init() {
-	userImplemented = false
 	groupImplemented = false
 }
 
@@ -46,14 +45,6 @@ func current() (*User, error) {
 	return u, fmt.Errorf("user: Current not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
 }
 
-func lookupUser(username string) (*User, error) {
-	return nil, errors.New("user: Lookup requires cgo")
-}
-
-func lookupUserId(uid string) (*User, error) {
-	return nil, errors.New("user: LookupId requires cgo")
-}
-
 func listGroups(*User) ([]string, error) {
 	return nil, errors.New("user: GroupIds requires cgo")
 }

src/os/user/lookup_unix.go

@@ -10,12 +10,15 @@ package user
 import (
 	"bufio"
 	"bytes"
+	"errors"
 	"io"
 	"os"
+	"strconv"
 	"strings"
 )
 
 const groupFile = "/etc/group"
+const userFile = "/etc/passwd"
 
 var colon = []byte{':'}
 
@@ -23,54 +26,140 @@ func init() {
 	groupImplemented = false
 }
 
-func findGroupId(id string, r io.Reader) (*Group, error) {
+// lineFunc returns a value, an error, or (nil, nil) to skip the row.
+type lineFunc func(line []byte) (v interface{}, err error)
+
+// readColonFile parses r as an /etc/group or /etc/passwd style file, running
+// fn for each row. readColonFile returns a value, an error, or (nil, nil) if
+// the end of the file is reached without a match.
+func readColonFile(r io.Reader, fn lineFunc) (v interface{}, err error) {
 	bs := bufio.NewScanner(r)
-	substr := []byte(":" + id)
 	for bs.Scan() {
-		lineBytes := bs.Bytes()
-		if !bytes.Contains(lineBytes, substr) || bytes.Count(lineBytes, colon) < 3 {
+		line := bs.Bytes()
+		// There's no spec for /etc/passwd or /etc/group, but we try to follow
+		// the same rules as the glibc parser, which allows comments and blank
+		// space at the beginning of a line.
+		line = bytes.TrimSpace(line)
+		if len(line) == 0 || line[0] == '#' {
 			continue
 		}
-		text := strings.TrimSpace(removeComment(string(lineBytes)))
+		v, err = fn(line)
+		if v != nil || err != nil {
+			return
+		}
+	}
+	return nil, bs.Err()
+}
+
+func matchGroupIndexValue(value string, idx int) lineFunc {
+	var leadColon string
+	if idx > 0 {
+		leadColon = ":"
+	}
+	substr := []byte(leadColon + value + ":")
+	return func(line []byte) (v interface{}, err error) {
+		if !bytes.Contains(line, substr) || bytes.Count(line, colon) < 3 {
+			return
+		}
 		// wheel:*:0:root
-		parts := strings.SplitN(text, ":", 4)
-		if len(parts) < 4 {
-			continue
+		parts := strings.SplitN(string(line), ":", 4)
+		if len(parts) < 4 || parts[0] == "" || parts[idx] != value ||
+			// If the file contains +foo and you search for "foo", glibc
+			// returns an "invalid argument" error. Similarly, if you search
+			// for a gid for a row where the group name starts with "+" or "-",
+			// glibc fails to find the record.
+			parts[0][0] == '+' || parts[0][0] == '-' {
+			return
+		}
+		if _, err := strconv.Atoi(parts[2]); err != nil {
+			return nil, nil
 		}
-		if parts[2] == id {
 		return &Group{Name: parts[0], Gid: parts[2]}, nil
 	}
-	}
-	if err := bs.Err(); err != nil {
+}
+
+func findGroupId(id string, r io.Reader) (*Group, error) {
+	if v, err := readColonFile(r, matchGroupIndexValue(id, 2)); err != nil {
 		return nil, err
+	} else if v != nil {
+		return v.(*Group), nil
 	}
 	return nil, UnknownGroupIdError(id)
 }
 
 func findGroupName(name string, r io.Reader) (*Group, error) {
-	bs := bufio.NewScanner(r)
-	substr := []byte(name + ":")
-	for bs.Scan() {
-		lineBytes := bs.Bytes()
-		if !bytes.Contains(lineBytes, substr) || bytes.Count(lineBytes, colon) < 3 {
-			continue
-		}
-		text := strings.TrimSpace(removeComment(string(lineBytes)))
-		// wheel:*:0:root
-		parts := strings.SplitN(text, ":", 4)
-		if len(parts) < 4 {
-			continue
-		}
-		if parts[0] == name && parts[2] != "" {
-			return &Group{Name: parts[0], Gid: parts[2]}, nil
-		}
-	}
-	if err := bs.Err(); err != nil {
+	if v, err := readColonFile(r, matchGroupIndexValue(name, 0)); err != nil {
 		return nil, err
+	} else if v != nil {
+		return v.(*Group), nil
 	}
 	return nil, UnknownGroupError(name)
 }
 
+// returns a *User for a row if that row's has the given value at the
+// given index.
+func matchUserIndexValue(value string, idx int) lineFunc {
+	var leadColon string
+	if idx > 0 {
+		leadColon = ":"
+	}
+	substr := []byte(leadColon + value + ":")
+	return func(line []byte) (v interface{}, err error) {
+		if !bytes.Contains(line, substr) || bytes.Count(line, colon) < 6 {
+			return
+		}
+		// kevin:x:1005:1006::/home/kevin:/usr/bin/zsh
+		parts := strings.SplitN(string(line), ":", 7)
+		if len(parts) < 6 || parts[idx] != value || parts[0] == "" ||
+			parts[0][0] == '+' || parts[0][0] == '-' {
+			return
+		}
+		if _, err := strconv.Atoi(parts[2]); err != nil {
+			return nil, nil
+		}
+		if _, err := strconv.Atoi(parts[3]); err != nil {
+			return nil, nil
+		}
+		u := &User{
+			Username: parts[0],
+			Uid:      parts[2],
+			Gid:      parts[3],
+			Name:     parts[4],
+			HomeDir:  parts[5],
+		}
+		// The pw_gecos field isn't quite standardized. Some docs
+		// say: "It is expected to be a comma separated list of
+		// personal data where the first item is the full name of the
+		// user."
+		if i := strings.Index(u.Name, ","); i >= 0 {
+			u.Name = u.Name[:i]
+		}
+		return u, nil
+	}
+}
+
+func findUserId(uid string, r io.Reader) (*User, error) {
+	i, e := strconv.Atoi(uid)
+	if e != nil {
+		return nil, errors.New("user: invalid userid " + uid)
+	}
+	if v, err := readColonFile(r, matchUserIndexValue(uid, 2)); err != nil {
+		return nil, err
+	} else if v != nil {
+		return v.(*User), nil
+	}
+	return nil, UnknownUserIdError(i)
+}
+
+func findUsername(name string, r io.Reader) (*User, error) {
+	if v, err := readColonFile(r, matchUserIndexValue(name, 0)); err != nil {
+		return nil, err
+	} else if v != nil {
+		return v.(*User), nil
+	}
+	return nil, UnknownUserError(name)
+}
+
 func lookupGroup(groupname string) (*Group, error) {
 	f, err := os.Open(groupFile)
 	if err != nil {
@@ -89,11 +178,20 @@ func lookupGroupId(id string) (*Group, error) {
 	return findGroupId(id, f)
 }
 
-// removeComment returns line, removing any '#' byte and any following
-// bytes.
-func removeComment(line string) string {
-	if i := strings.Index(line, "#"); i != -1 {
-		return line[:i]
+func lookupUser(username string) (*User, error) {
+	f, err := os.Open(userFile)
+	if err != nil {
+		return nil, err
 	}
-	return line
+	defer f.Close()
+	return findUsername(username, f)
+}
+
+func lookupUserId(uid string) (*User, error) {
+	f, err := os.Open(userFile)
+	if err != nil {
+		return nil, err
+	}
+	defer f.Close()
+	return findUserId(uid, f)
 }

src/os/user/lookup_unix_test.go

@@ -8,6 +8,7 @@
 package user
 
 import (
+	"reflect"
 	"strings"
 	"testing"
 )
@@ -19,6 +20,9 @@ nobody:*:-2:
 nogroup:*:-1:
 wheel:*:0:root
 emptyid:*::root
+invalidgid:*:notanumber:root
++plussign:*:20:root
+-minussign:*:21:root
       
 daemon:*:1:root
     indented:*:7:
@@ -36,7 +40,12 @@ var groupTests = []struct {
 	{testGroupFile, "kmem", "2"},
 	{testGroupFile, "notinthefile", ""},
 	{testGroupFile, "comment", ""},
+	{testGroupFile, "plussign", ""},
+	{testGroupFile, "+plussign", ""},
+	{testGroupFile, "-minussign", ""},
+	{testGroupFile, "minussign", ""},
 	{testGroupFile, "emptyid", ""},
+	{testGroupFile, "invalidgid", ""},
 	{testGroupFile, "indented", "7"},
 	{testGroupFile, "# comment", ""},
 	{"", "emptyfile", ""},
@@ -83,6 +92,8 @@ var groupIdTests = []struct {
 	{testGroupFile, "comment", ""},
 	{testGroupFile, "7", "indented"},
 	{testGroupFile, "4", ""},
+	{testGroupFile, "20", ""}, // row starts with a plus
+	{testGroupFile, "21", ""}, // row starts with a minus
 	{"", "emptyfile", ""},
 }
 
@@ -115,3 +126,151 @@ func TestFindGroupId(t *testing.T) {
 		}
 	}
 }
+
+const testUserFile = `   # Example user file
+root:x:0:0:root:/root:/bin/bash
+daemon:x:1:1:daemon:/usr/sbin:/usr/sbin/nologin
+bin:x:2:3:bin:/bin:/usr/sbin/nologin
+     indented:x:3:3:indented:/dev:/usr/sbin/nologin
+sync:x:4:65534:sync:/bin:/bin/sync
+negative:x:-5:60:games:/usr/games:/usr/sbin/nologin
+man:x:6:12:man:/var/cache/man:/usr/sbin/nologin
+allfields:x:6:12:mansplit,man2,man3,man4:/home/allfields:/usr/sbin/nologin
++plussign:x:8:10:man:/var/cache/man:/usr/sbin/nologin
+-minussign:x:9:10:man:/var/cache/man:/usr/sbin/nologin
+
+malformed:x:27:12 # more:colons:after:comment
+
+struid:x:notanumber:12 # more:colons:after:comment
+
+# commented:x:28:12:commented:/var/cache/man:/usr/sbin/nologin
+      # commentindented:x:29:12:commentindented:/var/cache/man:/usr/sbin/nologin
+
+struid2:x:30:badgid:struid2name:/home/struid:/usr/sbin/nologin
+`
+
+var userIdTests = []struct {
+	in   string
+	uid  string
+	name string
+}{
+	{testUserFile, "-5", "negative"},
+	{testUserFile, "2", "bin"},
+	{testUserFile, "100", ""}, // not in the file
+	{testUserFile, "8", ""},   // plus sign, glibc doesn't find it
+	{testUserFile, "9", ""},   // minus sign, glibc doesn't find it
+	{testUserFile, "27", ""},  // malformed
+	{testUserFile, "28", ""},  // commented out
+	{testUserFile, "29", ""},  // commented out, indented
+	{testUserFile, "3", "indented"},
+	{testUserFile, "30", ""}, // the Gid is not valid, shouldn't match
+	{"", "1", ""},
+}
+
+func TestInvalidUserId(t *testing.T) {
+	_, err := findUserId("notanumber", strings.NewReader(""))
+	if err == nil {
+		t.Fatalf("findUserId('notanumber'): got nil error")
+	}
+	if want := "user: invalid userid notanumber"; err.Error() != want {
+		t.Errorf("findUserId('notanumber'): got %v, want %s", err, want)
+	}
+}
+
+func TestLookupUserId(t *testing.T) {
+	for _, tt := range userIdTests {
+		got, err := findUserId(tt.uid, strings.NewReader(tt.in))
+		if tt.name == "" {
+			if err == nil {
+				t.Errorf("findUserId(%s): got nil error, expected err", tt.uid)
+				continue
+			}
+			switch terr := err.(type) {
+			case UnknownUserIdError:
+				if want := "user: unknown userid " + tt.uid; terr.Error() != want {
+					t.Errorf("findUserId(%s): got %v, want %v", tt.name, terr, want)
+				}
+			default:
+				t.Errorf("findUserId(%s): got unexpected error %v", tt.name, terr)
+			}
+		} else {
+			if err != nil {
+				t.Fatalf("findUserId(%s): got unexpected error %v", tt.name, err)
+			}
+			if got.Uid != tt.uid {
+				t.Errorf("findUserId(%s): got uid %v, want %s", tt.name, got.Uid, tt.uid)
+			}
+			if got.Username != tt.name {
+				t.Errorf("findUserId(%s): got name %s, want %s", tt.name, got.Username, tt.name)
+			}
+		}
+	}
+}
+
+func TestLookupUserPopulatesAllFields(t *testing.T) {
+	u, err := findUsername("allfields", strings.NewReader(testUserFile))
+	if err != nil {
+		t.Fatal(err)
+	}
+	want := &User{
+		Username: "allfields",
+		Uid:      "6",
+		Gid:      "12",
+		Name:     "mansplit",
+		HomeDir:  "/home/allfields",
+	}
+	if !reflect.DeepEqual(u, want) {
+		t.Errorf("findUsername: got %#v, want %#v", u, want)
+	}
+}
+
+var userTests = []struct {
+	in   string
+	name string
+	uid  string
+}{
+	{testUserFile, "negative", "-5"},
+	{testUserFile, "bin", "2"},
+	{testUserFile, "notinthefile", ""},
+	{testUserFile, "indented", "3"},
+	{testUserFile, "plussign", ""},
+	{testUserFile, "+plussign", ""},
+	{testUserFile, "minussign", ""},
+	{testUserFile, "-minussign", ""},
+	{testUserFile, "   indented", ""},
+	{testUserFile, "commented", ""},
+	{testUserFile, "commentindented", ""},
+	{testUserFile, "malformed", ""},
+	{testUserFile, "# commented", ""},
+	{"", "emptyfile", ""},
+}
+
+func TestLookupUser(t *testing.T) {
+	for _, tt := range userTests {
+		got, err := findUsername(tt.name, strings.NewReader(tt.in))
+		if tt.uid == "" {
+			if err == nil {
+				t.Errorf("lookupUser(%s): got nil error, expected err", tt.uid)
+				continue
+			}
+			switch terr := err.(type) {
+			case UnknownUserError:
+				if want := "user: unknown user " + tt.name; terr.Error() != want {
+					t.Errorf("lookupUser(%s): got %v, want %v", tt.name, terr, want)
+				}
+			default:
+				t.Errorf("lookupUser(%s): got unexpected error %v", tt.name, terr)
+			}
+		} else {
+			if err != nil {
+				t.Fatalf("lookupUser(%s): got unexpected error %v", tt.name, err)
+			}
+			if got.Uid != tt.uid {
+				t.Errorf("lookupUser(%s): got uid %v, want %s", tt.name, got.Uid, tt.uid)
+			}
+			if got.Username != tt.name {
+				t.Errorf("lookupUser(%s): got name %s, want %s", tt.name, got.Username, tt.name)
+			}
+		}
+	}
+}

src/os/user/user_test.go

@@ -70,6 +70,9 @@ func TestLookup(t *testing.T) {
 	if err != nil {
 		t.Fatalf("Current: %v", err)
 	}
+	// TODO: Lookup() has a fast path that calls Current() and returns if the
+	// usernames match, so this test does not exercise very much. It would be
+	// good to try and test finding a different user than the current user.
 	got, err := Lookup(want.Username)
 	if err != nil {
 		t.Fatalf("Lookup: %v", err)