*: update `-u` behavior

Fixes #16

In attempt to close https://github.com/vbatts/go-mtree/issues/16 I've
uncovered that the update was missing a function for symlink.
Additionally the update was not even opperating on the correct directory
hierarchy.

I've uncovered that os.Chtimes follows the symlink, and presumably only
Linux has an obscure way to set the mtime/atime on a symlink itself. So
I've made a custom lchtimes().

Also Mode follows through the symlink, and symlinks only ever have a
mode of 0777, so don't set them.

Lastly, directories need to have their mtime/atime set in a reverse
order after all other updates have been done. This is going to require
something like a `container/heap` to be unwound.

Also, each updateFunc will _only_ perform the update if it is needed. Much less
invasive this way.

Signed-off-by: Vincent Batts <vbatts@hashbangbash.com>

cmd/gomtree/main.go

@@ -254,18 +254,36 @@ func app() error {
 	if *flUpdateAttributes && stateDh != nil {
 		// -u
 		// this comes before the next case, intentionally.
-
-		// TODO brainstorm where to allow setting of xattrs. Maybe a new flag that allows a comma delimited list of keywords to update?
-		updateKeywords := []mtree.Keyword{"uid", "gid", "mode"}
-
-		result, err := mtree.Update(rootPath, stateDh, updateKeywords, nil)
+		result, err := mtree.Update(rootPath, specDh, mtree.DefaultUpdateKeywords, nil)
 		if err != nil {
 			return err
 		}
-
-		if result != nil {
+		if result != nil && len(result) > 0 {
 			fmt.Printf("%#v\n", result)
 		}
+
+		var res []mtree.InodeDelta
+		// only check the keywords that we just updated
+		res, err = mtree.Check(rootPath, specDh, mtree.DefaultUpdateKeywords, nil)
+		if err != nil {
+			return err
+		}
+		if res != nil {
+			out := formatFunc(res)
+			if _, err := os.Stdout.Write([]byte(out)); err != nil {
+				return err
+			}
+
+			// TODO: This should be a flag. Allowing files to be added and
+			//       removed and still returning "it's all good" is simply
+			//       unsafe IMO.
+			for _, diff := range res {
+				if diff.Type() == mtree.Modified {
+					return fmt.Errorf("mainfest validation failed")
+				}
+			}
+		}
+
 		return nil
 	}
 
@@ -300,7 +318,6 @@ func app() error {
 	// This is a validation.
 	if specDh != nil && stateDh != nil {
 		var res []mtree.InodeDelta
-
 		res, err = mtree.Compare(specDh, stateDh, currentKeywords)
 		if err != nil {
 			return err

entry.go

@@ -112,6 +112,16 @@ func (e Entry) AllKeys() []KeyVal {
 	return e.Keywords
 }
 
+// IsDir checks the type= value for this entry on whether it is a directory
+func (e Entry) IsDir() bool {
+	for _, kv := range e.AllKeys() {
+		if kv.Keyword().Prefix() == "type" {
+			return kv.Value() == "dir"
+		}
+	}
+	return false
+}
+
 // EntryType are the formats of lines in an mtree spec file
 type EntryType int
 

stat_unix.go

@@ -0,0 +1,18 @@
+// +build !windows
+
+package mtree
+
+import (
+	"os"
+	"syscall"
+)
+
+func statIsUID(stat os.FileInfo, uid int) bool {
+	statT := stat.Sys().(*syscall.Stat_t)
+	return statT.Uid == uint32(uid)
+}
+
+func statIsGID(stat os.FileInfo, gid int) bool {
+	statT := stat.Sys().(*syscall.Stat_t)
+	return statT.Gid == uint32(gid)
+}

stat_windows.go

@@ -0,0 +1,12 @@
+// +build windows
+
+package mtree
+
+import "os"
+
+func statIsUID(stat os.FileInfo, uid int) bool {
+	return false
+}
+func statIsGID(stat os.FileInfo, uid int) bool {
+	return false
+}

update.go

@@ -1,6 +1,7 @@
 package mtree
 
 import (
+	"container/heap"
 	"os"
 	"sort"
 
@@ -12,8 +13,9 @@ var DefaultUpdateKeywords = []Keyword{
 	"uid",
 	"gid",
 	"mode",
-	"time",
 	"xattr",
+	"link",
+	"time",
 }
 
 // Update attempts to set the attributes of root directory path, given the values of `keywords` in dh DirectoryHierarchy.
@@ -29,6 +31,11 @@ func Update(root string, dh *DirectoryHierarchy, keywords []Keyword, fs FsEval)
 	}
 	sort.Sort(byPos(creator.DH.Entries))
 
+	// This is for deferring the update of mtimes of directories, to unwind them
+	// in a most specific path first
+	h := &pathUpdateHeap{}
+	heap.Init(h)
+
 	results := []InodeDelta{}
 	for i, e := range creator.DH.Entries {
 		switch e.Type {
@@ -62,6 +69,19 @@ func Update(root string, dh *DirectoryHierarchy, keywords []Keyword, fs FsEval)
 					logrus.Debugf("no UpdateKeywordFunc for %s; skipping", kv.Keyword())
 					continue
 				}
+
+				// TODO check for the type=dir of the entry as well
+				if kv.Keyword().Prefix() == "time" && e.IsDir() {
+					heap.Push(h, pathUpdate{
+						Path: pathname,
+						E:    e,
+						KV:   kv,
+						Func: ukFunc,
+					})
+
+					continue
+				}
+
 				if _, err := ukFunc(pathname, kv); err != nil {
 					results = append(results, InodeDelta{
 						diff: ErrorDifference,
@@ -75,16 +95,60 @@ func Update(root string, dh *DirectoryHierarchy, keywords []Keyword, fs FsEval)
 							},
 						}})
 				}
+				// XXX really would be great to have a Check() or Compare() right here,
+				// to compare each entry as it is encountered, rather than just running
+				// Check() on this path after the whole update is finished.
 			}
 		}
 	}
 
+	for h.Len() > 0 {
+		pu := heap.Pop(h).(pathUpdate)
+		if _, err := pu.Func(pu.Path, pu.KV); err != nil {
+			results = append(results, InodeDelta{
+				diff: ErrorDifference,
+				path: pu.Path,
+				old:  pu.E,
+				keys: []KeyDelta{
+					{
+						diff: ErrorDifference,
+						name: pu.KV.Keyword(),
+						err:  err,
+					},
+				}})
+		}
+	}
 	return results, nil
 }
 
-// Result of an "update" returns the produced results
-type Result struct {
-	Path    string
-	Keyword Keyword
-	Got     string
+type pathUpdateHeap []pathUpdate
+
+func (h pathUpdateHeap) Len() int      { return len(h) }
+func (h pathUpdateHeap) Swap(i, j int) { h[i], h[j] = h[j], h[i] }
+
+// This may end up looking backwards, but for container/heap, Less evaluates
+// the negative priority. So when popping members of the array, it will be
+// sorted by least. For this use-case, we want the most-qualified-name popped
+// first (the longest path name), such that "." is the last entry popped.
+func (h pathUpdateHeap) Less(i, j int) bool {
+	return len(h[i].Path) > len(h[j].Path)
+}
+
+func (h *pathUpdateHeap) Push(x interface{}) {
+	*h = append(*h, x.(pathUpdate))
+}
+
+func (h *pathUpdateHeap) Pop() interface{} {
+	old := *h
+	n := len(old)
+	x := old[n-1]
+	*h = old[0 : n-1]
+	return x
+}
+
+type pathUpdate struct {
+	Path string
+	E    Entry
+	KV   KeyVal
+	Func UpdateKeywordFunc
 }

update_test.go

@@ -3,6 +3,7 @@
 package mtree
 
 import (
+	"container/heap"
 	"encoding/json"
 	"io/ioutil"
 	"os"
@@ -11,8 +12,14 @@ import (
 	"strconv"
 	"testing"
 	"time"
+
+	"github.com/Sirupsen/logrus"
 )
 
+func init() {
+	logrus.SetLevel(logrus.DebugLevel)
+}
+
 func TestUpdate(t *testing.T) {
 	content := []byte("I know half of you half as well as I ought to")
 	dir, err := ioutil.TempDir("", "test-check-keywords")
@@ -103,3 +110,27 @@ func TestUpdate(t *testing.T) {
 	}
 
 }
+
+func TestPathUpdateHeap(t *testing.T) {
+	h := &pathUpdateHeap{
+		pathUpdate{Path: "not/the/longest"},
+		pathUpdate{Path: "almost/the/longest"},
+		pathUpdate{Path: "."},
+		pathUpdate{Path: "short"},
+	}
+	heap.Init(h)
+	v := "this/is/one/is/def/the/longest"
+	heap.Push(h, pathUpdate{Path: v})
+
+	longest := len(v)
+	var p string
+	for h.Len() > 0 {
+		p = heap.Pop(h).(pathUpdate).Path
+		if len(p) > longest {
+			t.Errorf("expected next path to be shorter, but it was not %q is longer than %d", p, longest)
+		}
+	}
+	if p != "." {
+		t.Errorf("expected \".\" to be the last, but got %q", p)
+	}
+}

updatefuncs.go

@@ -8,6 +8,7 @@ import (
 	"time"
 
 	"github.com/Sirupsen/logrus"
+	"github.com/vbatts/go-mtree/pkg/govis"
 )
 
 // UpdateKeywordFunc is the signature for a function that will restore a file's
@@ -24,6 +25,7 @@ var UpdateKeywordFuncs = map[Keyword]UpdateKeywordFunc{
 	"uid":      uidUpdateKeywordFunc,
 	"gid":      gidUpdateKeywordFunc,
 	"xattr":    xattrUpdateKeywordFunc,
+	"link":     linkUpdateKeywordFunc,
 }
 
 func uidUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
@@ -31,6 +33,15 @@ func uidUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	if err != nil {
 		return nil, err
 	}
+
+	stat, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+	if statIsUID(stat, uid) {
+		return stat, nil
+	}
+
 	if err := os.Lchown(path, uid, -1); err != nil {
 		return nil, err
 	}
@@ -42,6 +53,15 @@ func gidUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	if err != nil {
 		return nil, err
 	}
+
+	stat, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+	if statIsGID(stat, gid) {
+		return stat, nil
+	}
+
 	if err := os.Lchown(path, -1, gid); err != nil {
 		return nil, err
 	}
@@ -49,10 +69,28 @@ func gidUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 }
 
 func modeUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
+	info, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+
+	// don't set mode on symlinks, as it passes through to the backing file
+	if info.Mode()&os.ModeSymlink != 0 {
+		return info, nil
+	}
 	vmode, err := strconv.ParseInt(kv.Value(), 8, 32)
 	if err != nil {
 		return nil, err
 	}
+
+	stat, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+	if stat.Mode() == os.FileMode(vmode) {
+		return stat, nil
+	}
+
 	logrus.Debugf("path: %q, kv.Value(): %q, vmode: %o", path, kv.Value(), vmode)
 	if err := os.Chmod(path, os.FileMode(vmode)); err != nil {
 		return nil, err
@@ -85,7 +123,19 @@ func tartimeUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	}
 
 	vtime := time.Unix(sec, 0)
-	if err := os.Chtimes(path, vtime, vtime); err != nil {
+
+	// if times are same then don't modify anything
+	// comparing Unix, since it does not include Nano seconds
+	if info.ModTime().Unix() == vtime.Unix() {
+		return info, nil
+	}
+
+	// symlinks are strange and most of the time passes through to the backing file
+	if info.Mode()&os.ModeSymlink != 0 {
+		if err := lchtimes(path, vtime, vtime); err != nil {
+			return nil, err
+		}
+	} else if err := os.Chtimes(path, vtime, vtime); err != nil {
 		return nil, err
 	}
 	return os.Lstat(path)
@@ -93,6 +143,11 @@ func tartimeUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 
 // this is nano second precision
 func timeUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
+	info, err := os.Lstat(path)
+	if err != nil {
+		return nil, err
+	}
+
 	v := strings.SplitN(kv.Value(), ".", 2)
 	if len(v) != 2 {
 		return nil, fmt.Errorf("expected a number like 1469104727.871937272")
@@ -101,11 +156,46 @@ func timeUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	if err != nil {
 		return nil, fmt.Errorf("expected nano seconds, but got %q", v[0]+v[1])
 	}
-	logrus.Debugf("arg: %q; nsec: %q", v[0]+v[1], nsec)
+	logrus.Debugf("arg: %q; nsec: %d", v[0]+v[1], nsec)
 
 	vtime := time.Unix(0, nsec)
-	if err := os.Chtimes(path, vtime, vtime); err != nil {
+
+	// if times are same then don't modify anything
+	if info.ModTime().Equal(vtime) {
+		return info, nil
+	}
+
+	// symlinks are strange and most of the time passes through to the backing file
+	if info.Mode()&os.ModeSymlink != 0 {
+		if err := lchtimes(path, vtime, vtime); err != nil {
+			return nil, err
+		}
+	} else if err := os.Chtimes(path, vtime, vtime); err != nil {
 		return nil, err
 	}
 	return os.Lstat(path)
 }
+
+func linkUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
+	linkname, err := govis.Unvis(kv.Value(), DefaultVisFlags)
+	if err != nil {
+		return nil, err
+	}
+	got, err := os.Readlink(path)
+	if err != nil {
+		return nil, err
+	}
+	if got == linkname {
+		return os.Lstat(path)
+	}
+
+	logrus.Debugf("linkUpdateKeywordFunc: removing %q to link to %q", path, linkname)
+	if err := os.Remove(path); err != nil {
+		return nil, err
+	}
+	if err := os.Symlink(linkname, path); err != nil {
+		return nil, err
+	}
+
+	return os.Lstat(path)
+}

updatefuncs_linux.go

@@ -5,6 +5,9 @@ package mtree
 import (
 	"encoding/base64"
 	"os"
+	"syscall"
+	"time"
+	"unsafe"
 
 	"github.com/vbatts/go-mtree/xattr"
 )
@@ -19,3 +22,37 @@ func xattrUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	}
 	return os.Lstat(path)
 }
+
+func lchtimes(name string, atime time.Time, mtime time.Time) error {
+	var utimes [2]syscall.Timespec
+	utimes[0] = syscall.NsecToTimespec(atime.UnixNano())
+	utimes[1] = syscall.NsecToTimespec(mtime.UnixNano())
+	if e := utimensat(atFdCwd, name, (*[2]syscall.Timespec)(unsafe.Pointer(&utimes[0])), atSymlinkNofollow); e != nil {
+		return &os.PathError{Op: "chtimes", Path: name, Err: e}
+	}
+	return nil
+
+}
+
+// from uapi/linux/fcntl.h
+// don't follow symlinks
+const atSymlinkNofollow = 0x100
+
+// special value for utimes as the FD for the current working directory
+const atFdCwd = -0x64
+
+func utimensat(dirfd int, path string, times *[2]syscall.Timespec, flags int) (err error) {
+	if len(times) != 2 {
+		return syscall.EINVAL
+	}
+	var _p0 *byte
+	_p0, err = syscall.BytePtrFromString(path)
+	if err != nil {
+		return
+	}
+	_, _, e1 := syscall.Syscall6(syscall.SYS_UTIMENSAT, uintptr(dirfd), uintptr(unsafe.Pointer(_p0)), uintptr(unsafe.Pointer(times)), uintptr(flags), 0, 0)
+	if e1 != 0 {
+		err = syscall.Errno(e1)
+	}
+	return
+}

updatefuncs_unsupported.go

@@ -2,8 +2,15 @@
 
 package mtree
 
-import "os"
+import (
+	"os"
+	"time"
+)
 
 func xattrUpdateKeywordFunc(path string, kv KeyVal) (os.FileInfo, error) {
 	return os.Lstat(path)
 }
+
+func lchtimes(name string, atime time.Time, mtime time.Time) error {
+	return nil
+}