using godep to manage dependencies

weed/Godeps/Godeps.json

@@ -0,0 +1,184 @@
+{
+	"ImportPath": "github.com/chrislusf/seaweedfs/weed",
+	"GoVersion": "go1.6",
+	"GodepVersion": "v60",
+	"Deps": [
+		{
+			"ImportPath": "bazil.org/fuse",
+			"Rev": "e1ba3783de6cb92c16f72802a4bb94fe639a8049"
+		},
+		{
+			"ImportPath": "bazil.org/fuse/fs",
+			"Rev": "e1ba3783de6cb92c16f72802a4bb94fe639a8049"
+		},
+		{
+			"ImportPath": "bazil.org/fuse/fuseutil",
+			"Rev": "e1ba3783de6cb92c16f72802a4bb94fe639a8049"
+		},
+		{
+			"ImportPath": "github.com/boltdb/bolt",
+			"Comment": "v1.1.0-22-g6e1ca38",
+			"Rev": "6e1ca38c6a73025366cd8705553b404746ee6e63"
+		},
+		{
+			"ImportPath": "github.com/chrislusf/raft",
+			"Rev": "90f631ee823c83f594f27257bab64911190856af"
+		},
+		{
+			"ImportPath": "github.com/chrislusf/raft/protobuf",
+			"Rev": "90f631ee823c83f594f27257bab64911190856af"
+		},
+		{
+			"ImportPath": "github.com/dgrijalva/jwt-go",
+			"Comment": "v2.4.0",
+			"Rev": "f164e17f59b82642a3895ba065c385db6c547344"
+		},
+		{
+			"ImportPath": "github.com/disintegration/imaging",
+			"Rev": "546cb3c5137b3f1232e123a26aa033aade6b3066"
+		},
+		{
+			"ImportPath": "github.com/gocql/gocql",
+			"Comment": "1st_gen_framing-383-g87cc185",
+			"Rev": "87cc1854b57c7a4d8f4ae1d0cc358ed6ecb0f8c3"
+		},
+		{
+			"ImportPath": "github.com/gocql/gocql/internal/lru",
+			"Comment": "1st_gen_framing-383-g87cc185",
+			"Rev": "87cc1854b57c7a4d8f4ae1d0cc358ed6ecb0f8c3"
+		},
+		{
+			"ImportPath": "github.com/gocql/gocql/internal/murmur",
+			"Comment": "1st_gen_framing-383-g87cc185",
+			"Rev": "87cc1854b57c7a4d8f4ae1d0cc358ed6ecb0f8c3"
+		},
+		{
+			"ImportPath": "github.com/gocql/gocql/internal/streams",
+			"Comment": "1st_gen_framing-383-g87cc185",
+			"Rev": "87cc1854b57c7a4d8f4ae1d0cc358ed6ecb0f8c3"
+		},
+		{
+			"ImportPath": "github.com/gogo/protobuf/proto",
+			"Comment": "v0.1-78-gd3235f0",
+			"Rev": "d3235f01ecae4901dd9f7ea6af57a352c0189deb"
+		},
+		{
+			"ImportPath": "github.com/golang/protobuf/proto",
+			"Rev": "8d92cf5fc15a4382f8964b08e1f42a75c0591aa3"
+		},
+		{
+			"ImportPath": "github.com/golang/snappy",
+			"Rev": "723cc1e459b8eea2dea4583200fd60757d40097a"
+		},
+		{
+			"ImportPath": "github.com/gorilla/context",
+			"Rev": "215affda49addc4c8ef7e2534915df2c8c35c6cd"
+		},
+		{
+			"ImportPath": "github.com/gorilla/mux",
+			"Rev": "e444e69cbd2e2e3e0749a2f3c717cec491552bbf"
+		},
+		{
+			"ImportPath": "github.com/hailocab/go-hostpool",
+			"Rev": "0637eae892be221164aff5fcbccc57171aea6406"
+		},
+		{
+			"ImportPath": "github.com/pierrec/lz4",
+			"Rev": "0b67ae4bb1ab03691079e38dddbc3909d68de64f"
+		},
+		{
+			"ImportPath": "github.com/pierrec/xxHash/xxHash32",
+			"Rev": "122b94d6aa20d9b33d3989da6d7cd6cf641d2277"
+		},
+		{
+			"ImportPath": "github.com/rwcarlsen/goexif/exif",
+			"Rev": "709fab3d192d7c62f86043caff1e7e3fb0f42bd8"
+		},
+		{
+			"ImportPath": "github.com/rwcarlsen/goexif/tiff",
+			"Rev": "709fab3d192d7c62f86043caff1e7e3fb0f42bd8"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/cache",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/comparer",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/errors",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/filter",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/iterator",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/journal",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/memdb",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/opt",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/storage",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/table",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "github.com/syndtr/goleveldb/leveldb/util",
+			"Rev": "1a9d62f03ea92815b46fcaab357cfd4df264b1a0"
+		},
+		{
+			"ImportPath": "golang.org/x/image/bmp",
+			"Rev": "baddd3465a05d84a6d8d3507547a91cb188c81ea"
+		},
+		{
+			"ImportPath": "golang.org/x/image/tiff",
+			"Rev": "baddd3465a05d84a6d8d3507547a91cb188c81ea"
+		},
+		{
+			"ImportPath": "golang.org/x/image/tiff/lzw",
+			"Rev": "baddd3465a05d84a6d8d3507547a91cb188c81ea"
+		},
+		{
+			"ImportPath": "golang.org/x/net/context",
+			"Rev": "c764672d0ee39ffd83cfcb375804d3181302b62b"
+		},
+		{
+			"ImportPath": "golang.org/x/sys/unix",
+			"Rev": "320cb01ddbbf0473674c2585f9b6e245721de355"
+		},
+		{
+			"ImportPath": "gopkg.in/bufio.v1",
+			"Comment": "v1",
+			"Rev": "567b2bfa514e796916c4747494d6ff5132a1dfce"
+		},
+		{
+			"ImportPath": "gopkg.in/inf.v0",
+			"Rev": "3887ee99ecf07df5b447e9b00d9c0b2adaa9f3e4"
+		},
+		{
+			"ImportPath": "gopkg.in/redis.v2",
+			"Comment": "v2.3.2",
+			"Rev": "e6179049628164864e6e84e973cfb56335748dea"
+		}
+	]
+}

weed/Godeps/Readme

@@ -0,0 +1,5 @@
+This directory tree is generated automatically by godep.
+
+Please do not edit.
+
+See https://github.com/tools/godep for more information.

weed/vendor/bazil.org/fuse/.gitattributes

@@ -0,0 +1,2 @@
+*.go filter=gofmt
+*.cgo filter=gofmt

weed/vendor/bazil.org/fuse/.gitignore

@@ -0,0 +1,11 @@
+*~
+.#*
+## the next line needs to start with a backslash to avoid looking like
+## a comment
+\#*#
+.*.swp
+
+*.test
+
+/clockfs
+/hellofs

weed/vendor/bazil.org/fuse/LICENSE

@@ -0,0 +1,93 @@
+Copyright (c) 2013-2015 Tommi Virtanen.
+Copyright (c) 2009, 2011, 2012 The Go Authors.
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   * Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+   * Redistributions in binary form must reproduce the above
+copyright notice, this list of conditions and the following disclaimer
+in the documentation and/or other materials provided with the
+distribution.
+   * Neither the name of Google Inc. nor the names of its
+contributors may be used to endorse or promote products derived from
+this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+
+
+The following included software components have additional copyright
+notices and license terms that may differ from the above.
+
+
+File fuse.go:
+
+// Adapted from Plan 9 from User Space's src/cmd/9pfuse/fuse.c,
+// which carries this notice:
+//
+// The files in this directory are subject to the following license.
+//
+// The author of this software is Russ Cox.
+//
+//         Copyright (c) 2006 Russ Cox
+//
+// Permission to use, copy, modify, and distribute this software for any
+// purpose without fee is hereby granted, provided that this entire notice
+// is included in all copies of any software which is or includes a copy
+// or modification of this software and in all copies of the supporting
+// documentation for such software.
+//
+// THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
+// WARRANTY.  IN PARTICULAR, THE AUTHOR MAKES NO REPRESENTATION OR WARRANTY
+// OF ANY KIND CONCERNING THE MERCHANTABILITY OF THIS SOFTWARE OR ITS
+// FITNESS FOR ANY PARTICULAR PURPOSE.
+
+
+File fuse_kernel.go:
+
+// Derived from FUSE's fuse_kernel.h
+/*
+   This file defines the kernel interface of FUSE
+   Copyright (C) 2001-2007  Miklos Szeredi <miklos@szeredi.hu>
+
+
+   This -- and only this -- header file may also be distributed under
+   the terms of the BSD Licence as follows:
+
+   Copyright (C) 2001-2007 Miklos Szeredi. All rights reserved.
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in the
+      documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND
+   ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+   IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+   ARE DISCLAIMED.  IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
+   FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+   DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+   OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+   LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+   OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+   SUCH DAMAGE.
+*/

weed/vendor/bazil.org/fuse/README.md

@@ -0,0 +1,23 @@
+bazil.org/fuse -- Filesystems in Go
+===================================
+
+`bazil.org/fuse` is a Go library for writing FUSE userspace
+filesystems.
+
+It is a from-scratch implementation of the kernel-userspace
+communication protocol, and does not use the C library from the
+project called FUSE. `bazil.org/fuse` embraces Go fully for safety and
+ease of programming.
+
+Here’s how to get going:
+
+    go get bazil.org/fuse
+
+Website: http://bazil.org/fuse/
+
+Github repository: https://github.com/bazil/fuse
+
+API docs: http://godoc.org/bazil.org/fuse
+
+Our thanks to Russ Cox for his fuse library, which this project is
+based on.

weed/vendor/bazil.org/fuse/buffer.go

@@ -0,0 +1,35 @@
+package fuse
+
+import "unsafe"
+
+// buffer provides a mechanism for constructing a message from
+// multiple segments.
+type buffer []byte
+
+// alloc allocates size bytes and returns a pointer to the new
+// segment.
+func (w *buffer) alloc(size uintptr) unsafe.Pointer {
+	s := int(size)
+	if len(*w)+s > cap(*w) {
+		old := *w
+		*w = make([]byte, len(*w), 2*cap(*w)+s)
+		copy(*w, old)
+	}
+	l := len(*w)
+	*w = (*w)[:l+s]
+	return unsafe.Pointer(&(*w)[l])
+}
+
+// reset clears out the contents of the buffer.
+func (w *buffer) reset() {
+	for i := range (*w)[:cap(*w)] {
+		(*w)[i] = 0
+	}
+	*w = (*w)[:0]
+}
+
+func newBuffer(extra uintptr) buffer {
+	const hdrSize = unsafe.Sizeof(outHeader{})
+	buf := make(buffer, hdrSize, hdrSize+extra)
+	return buf
+}

weed/vendor/bazil.org/fuse/debug.go

@@ -0,0 +1,21 @@
+package fuse
+
+import (
+	"runtime"
+)
+
+func stack() string {
+	buf := make([]byte, 1024)
+	return string(buf[:runtime.Stack(buf, false)])
+}
+
+func nop(msg interface{}) {}
+
+// Debug is called to output debug messages, including protocol
+// traces. The default behavior is to do nothing.
+//
+// The messages have human-friendly string representations and are
+// safe to marshal to JSON.
+//
+// Implementations must not retain msg.
+var Debug func(msg interface{}) = nop

weed/vendor/bazil.org/fuse/error_darwin.go

@@ -0,0 +1,17 @@
+package fuse
+
+import (
+	"syscall"
+)
+
+const (
+	ENOATTR = Errno(syscall.ENOATTR)
+)
+
+const (
+	errNoXattr = ENOATTR
+)
+
+func init() {
+	errnoNames[errNoXattr] = "ENOATTR"
+}

weed/vendor/bazil.org/fuse/error_freebsd.go

@@ -0,0 +1,15 @@
+package fuse
+
+import "syscall"
+
+const (
+	ENOATTR = Errno(syscall.ENOATTR)
+)
+
+const (
+	errNoXattr = ENOATTR
+)
+
+func init() {
+	errnoNames[errNoXattr] = "ENOATTR"
+}

weed/vendor/bazil.org/fuse/error_linux.go

@@ -0,0 +1,17 @@
+package fuse
+
+import (
+	"syscall"
+)
+
+const (
+	ENODATA = Errno(syscall.ENODATA)
+)
+
+const (
+	errNoXattr = ENODATA
+)
+
+func init() {
+	errnoNames[errNoXattr] = "ENODATA"
+}

weed/vendor/bazil.org/fuse/error_std.go

@@ -0,0 +1,31 @@
+package fuse
+
+// There is very little commonality in extended attribute errors
+// across platforms.
+//
+// getxattr return value for "extended attribute does not exist" is
+// ENOATTR on OS X, and ENODATA on Linux and apparently at least
+// NetBSD. There may be a #define ENOATTR on Linux too, but the value
+// is ENODATA in the actual syscalls. FreeBSD and OpenBSD have no
+// ENODATA, only ENOATTR. ENOATTR is not in any of the standards,
+// ENODATA exists but is only used for STREAMs.
+//
+// Each platform will define it a errNoXattr constant, and this file
+// will enforce that it implements the right interfaces and hide the
+// implementation.
+//
+// https://developer.apple.com/library/mac/documentation/Darwin/Reference/ManPages/man2/getxattr.2.html
+// http://mail-index.netbsd.org/tech-kern/2012/04/30/msg013090.html
+// http://mail-index.netbsd.org/tech-kern/2012/04/30/msg013097.html
+// http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/errno.h.html
+// http://www.freebsd.org/cgi/man.cgi?query=extattr_get_file&sektion=2
+// http://nixdoc.net/man-pages/openbsd/man2/extattr_get_file.2.html
+
+// ErrNoXattr is a platform-independent error value meaning the
+// extended attribute was not found. It can be used to respond to
+// GetxattrRequest and such.
+const ErrNoXattr = errNoXattr
+
+var _ error = ErrNoXattr
+var _ Errno = ErrNoXattr
+var _ ErrorNumber = ErrNoXattr

weed/vendor/bazil.org/fuse/fs/tree.go

@@ -0,0 +1,99 @@
+// FUSE directory tree, for servers that wish to use it with the service loop.
+
+package fs
+
+import (
+	"os"
+	pathpkg "path"
+	"strings"
+
+	"golang.org/x/net/context"
+)
+
+import (
+	"bazil.org/fuse"
+)
+
+// A Tree implements a basic read-only directory tree for FUSE.
+// The Nodes contained in it may still be writable.
+type Tree struct {
+	tree
+}
+
+func (t *Tree) Root() (Node, error) {
+	return &t.tree, nil
+}
+
+// Add adds the path to the tree, resolving to the given node.
+// If path or a prefix of path has already been added to the tree,
+// Add panics.
+//
+// Add is only safe to call before starting to serve requests.
+func (t *Tree) Add(path string, node Node) {
+	path = pathpkg.Clean("/" + path)[1:]
+	elems := strings.Split(path, "/")
+	dir := Node(&t.tree)
+	for i, elem := range elems {
+		dt, ok := dir.(*tree)
+		if !ok {
+			panic("fuse: Tree.Add for " + strings.Join(elems[:i], "/") + " and " + path)
+		}
+		n := dt.lookup(elem)
+		if n != nil {
+			if i+1 == len(elems) {
+				panic("fuse: Tree.Add for " + path + " conflicts with " + elem)
+			}
+			dir = n
+		} else {
+			if i+1 == len(elems) {
+				dt.add(elem, node)
+			} else {
+				dir = &tree{}
+				dt.add(elem, dir)
+			}
+		}
+	}
+}
+
+type treeDir struct {
+	name string
+	node Node
+}
+
+type tree struct {
+	dir []treeDir
+}
+
+func (t *tree) lookup(name string) Node {
+	for _, d := range t.dir {
+		if d.name == name {
+			return d.node
+		}
+	}
+	return nil
+}
+
+func (t *tree) add(name string, n Node) {
+	t.dir = append(t.dir, treeDir{name, n})
+}
+
+func (t *tree) Attr(ctx context.Context, a *fuse.Attr) error {
+	a.Mode = os.ModeDir | 0555
+	return nil
+}
+
+func (t *tree) Lookup(ctx context.Context, name string) (Node, error) {
+	n := t.lookup(name)
+	if n != nil {
+		return n, nil
+	}
+	return nil, fuse.ENOENT
+}
+
+func (t *tree) ReadDirAll(ctx context.Context) ([]fuse.Dirent, error) {
+	var out []fuse.Dirent
+	for _, d := range t.dir {
+		out = append(out, fuse.Dirent{Name: d.name})
+	}
+	return out, nil
+}

weed/vendor/bazil.org/fuse/fuse_kernel.go

@@ -0,0 +1,773 @@
+// See the file LICENSE for copyright and licensing information.
+
+// Derived from FUSE's fuse_kernel.h, which carries this notice:
+/*
+   This file defines the kernel interface of FUSE
+   Copyright (C) 2001-2007  Miklos Szeredi <miklos@szeredi.hu>
+
+
+   This -- and only this -- header file may also be distributed under
+   the terms of the BSD Licence as follows:
+
+   Copyright (C) 2001-2007 Miklos Szeredi. All rights reserved.
+
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in the
+      documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND
+   ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+   IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+   ARE DISCLAIMED.  IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
+   FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+   DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+   OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+   HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+   LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+   OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+   SUCH DAMAGE.
+*/
+
+package fuse
+
+import (
+	"fmt"
+	"syscall"
+	"unsafe"
+)
+
+// The FUSE version implemented by the package.
+const (
+	protoVersionMinMajor = 7
+	protoVersionMinMinor = 8
+	protoVersionMaxMajor = 7
+	protoVersionMaxMinor = 12
+)
+
+const (
+	rootID = 1
+)
+
+type kstatfs struct {
+	Blocks  uint64
+	Bfree   uint64
+	Bavail  uint64
+	Files   uint64
+	Ffree   uint64
+	Bsize   uint32
+	Namelen uint32
+	Frsize  uint32
+	Padding uint32
+	Spare   [6]uint32
+}
+
+type fileLock struct {
+	Start uint64
+	End   uint64
+	Type  uint32
+	Pid   uint32
+}
+
+// GetattrFlags are bit flags that can be seen in GetattrRequest.
+type GetattrFlags uint32
+
+const (
+	// Indicates the handle is valid.
+	GetattrFh GetattrFlags = 1 << 0
+)
+
+var getattrFlagsNames = []flagName{
+	{uint32(GetattrFh), "GetattrFh"},
+}
+
+func (fl GetattrFlags) String() string {
+	return flagString(uint32(fl), getattrFlagsNames)
+}
+
+// The SetattrValid are bit flags describing which fields in the SetattrRequest
+// are included in the change.
+type SetattrValid uint32
+
+const (
+	SetattrMode   SetattrValid = 1 << 0
+	SetattrUid    SetattrValid = 1 << 1
+	SetattrGid    SetattrValid = 1 << 2
+	SetattrSize   SetattrValid = 1 << 3
+	SetattrAtime  SetattrValid = 1 << 4
+	SetattrMtime  SetattrValid = 1 << 5
+	SetattrHandle SetattrValid = 1 << 6
+
+	// Linux only(?)
+	SetattrAtimeNow  SetattrValid = 1 << 7
+	SetattrMtimeNow  SetattrValid = 1 << 8
+	SetattrLockOwner SetattrValid = 1 << 9 // http://www.mail-archive.com/git-commits-head@vger.kernel.org/msg27852.html
+
+	// OS X only
+	SetattrCrtime   SetattrValid = 1 << 28
+	SetattrChgtime  SetattrValid = 1 << 29
+	SetattrBkuptime SetattrValid = 1 << 30
+	SetattrFlags    SetattrValid = 1 << 31
+)
+
+func (fl SetattrValid) Mode() bool      { return fl&SetattrMode != 0 }
+func (fl SetattrValid) Uid() bool       { return fl&SetattrUid != 0 }
+func (fl SetattrValid) Gid() bool       { return fl&SetattrGid != 0 }
+func (fl SetattrValid) Size() bool      { return fl&SetattrSize != 0 }
+func (fl SetattrValid) Atime() bool     { return fl&SetattrAtime != 0 }
+func (fl SetattrValid) Mtime() bool     { return fl&SetattrMtime != 0 }
+func (fl SetattrValid) Handle() bool    { return fl&SetattrHandle != 0 }
+func (fl SetattrValid) AtimeNow() bool  { return fl&SetattrAtimeNow != 0 }
+func (fl SetattrValid) MtimeNow() bool  { return fl&SetattrMtimeNow != 0 }
+func (fl SetattrValid) LockOwner() bool { return fl&SetattrLockOwner != 0 }
+func (fl SetattrValid) Crtime() bool    { return fl&SetattrCrtime != 0 }
+func (fl SetattrValid) Chgtime() bool   { return fl&SetattrChgtime != 0 }
+func (fl SetattrValid) Bkuptime() bool  { return fl&SetattrBkuptime != 0 }
+func (fl SetattrValid) Flags() bool     { return fl&SetattrFlags != 0 }
+
+func (fl SetattrValid) String() string {
+	return flagString(uint32(fl), setattrValidNames)
+}
+
+var setattrValidNames = []flagName{
+	{uint32(SetattrMode), "SetattrMode"},
+	{uint32(SetattrUid), "SetattrUid"},
+	{uint32(SetattrGid), "SetattrGid"},
+	{uint32(SetattrSize), "SetattrSize"},
+	{uint32(SetattrAtime), "SetattrAtime"},
+	{uint32(SetattrMtime), "SetattrMtime"},
+	{uint32(SetattrHandle), "SetattrHandle"},
+	{uint32(SetattrAtimeNow), "SetattrAtimeNow"},
+	{uint32(SetattrMtimeNow), "SetattrMtimeNow"},
+	{uint32(SetattrLockOwner), "SetattrLockOwner"},
+	{uint32(SetattrCrtime), "SetattrCrtime"},
+	{uint32(SetattrChgtime), "SetattrChgtime"},
+	{uint32(SetattrBkuptime), "SetattrBkuptime"},
+	{uint32(SetattrFlags), "SetattrFlags"},
+}
+
+// Flags that can be seen in OpenRequest.Flags.
+const (
+	// Access modes. These are not 1-bit flags, but alternatives where
+	// only one can be chosen. See the IsReadOnly etc convenience
+	// methods.
+	OpenReadOnly  OpenFlags = syscall.O_RDONLY
+	OpenWriteOnly OpenFlags = syscall.O_WRONLY
+	OpenReadWrite OpenFlags = syscall.O_RDWR
+
+	// File was opened in append-only mode, all writes will go to end
+	// of file. OS X does not provide this information.
+	OpenAppend    OpenFlags = syscall.O_APPEND
+	OpenCreate    OpenFlags = syscall.O_CREAT
+	OpenDirectory OpenFlags = syscall.O_DIRECTORY
+	OpenExclusive OpenFlags = syscall.O_EXCL
+	OpenNonblock  OpenFlags = syscall.O_NONBLOCK
+	OpenSync      OpenFlags = syscall.O_SYNC
+	OpenTruncate  OpenFlags = syscall.O_TRUNC
+)
+
+// OpenAccessModeMask is a bitmask that separates the access mode
+// from the other flags in OpenFlags.
+const OpenAccessModeMask OpenFlags = syscall.O_ACCMODE
+
+// OpenFlags are the O_FOO flags passed to open/create/etc calls. For
+// example, os.O_WRONLY | os.O_APPEND.
+type OpenFlags uint32
+
+func (fl OpenFlags) String() string {
+	// O_RDONLY, O_RWONLY, O_RDWR are not flags
+	s := accModeName(fl & OpenAccessModeMask)
+	flags := uint32(fl &^ OpenAccessModeMask)
+	if flags != 0 {
+		s = s + "+" + flagString(flags, openFlagNames)
+	}
+	return s
+}
+
+// Return true if OpenReadOnly is set.
+func (fl OpenFlags) IsReadOnly() bool {
+	return fl&OpenAccessModeMask == OpenReadOnly
+}
+
+// Return true if OpenWriteOnly is set.
+func (fl OpenFlags) IsWriteOnly() bool {
+	return fl&OpenAccessModeMask == OpenWriteOnly
+}
+
+// Return true if OpenReadWrite is set.
+func (fl OpenFlags) IsReadWrite() bool {
+	return fl&OpenAccessModeMask == OpenReadWrite
+}
+
+func accModeName(flags OpenFlags) string {
+	switch flags {
+	case OpenReadOnly:
+		return "OpenReadOnly"
+	case OpenWriteOnly:
+		return "OpenWriteOnly"
+	case OpenReadWrite:
+		return "OpenReadWrite"
+	default:
+		return ""
+	}
+}
+
+var openFlagNames = []flagName{
+	{uint32(OpenAppend), "OpenAppend"},
+	{uint32(OpenCreate), "OpenCreate"},
+	{uint32(OpenDirectory), "OpenDirectory"},
+	{uint32(OpenExclusive), "OpenExclusive"},
+	{uint32(OpenNonblock), "OpenNonblock"},
+	{uint32(OpenSync), "OpenSync"},
+	{uint32(OpenTruncate), "OpenTruncate"},
+}
+
+// The OpenResponseFlags are returned in the OpenResponse.
+type OpenResponseFlags uint32
+
+const (
+	OpenDirectIO    OpenResponseFlags = 1 << 0 // bypass page cache for this open file
+	OpenKeepCache   OpenResponseFlags = 1 << 1 // don't invalidate the data cache on open
+	OpenNonSeekable OpenResponseFlags = 1 << 2 // mark the file as non-seekable (not supported on OS X)
+
+	OpenPurgeAttr OpenResponseFlags = 1 << 30 // OS X
+	OpenPurgeUBC  OpenResponseFlags = 1 << 31 // OS X
+)
+
+func (fl OpenResponseFlags) String() string {
+	return flagString(uint32(fl), openResponseFlagNames)
+}
+
+var openResponseFlagNames = []flagName{
+	{uint32(OpenDirectIO), "OpenDirectIO"},
+	{uint32(OpenKeepCache), "OpenKeepCache"},
+	{uint32(OpenNonSeekable), "OpenNonSeekable"},
+	{uint32(OpenPurgeAttr), "OpenPurgeAttr"},
+	{uint32(OpenPurgeUBC), "OpenPurgeUBC"},
+}
+
+// The InitFlags are used in the Init exchange.
+type InitFlags uint32
+
+const (
+	InitAsyncRead     InitFlags = 1 << 0
+	InitPosixLocks    InitFlags = 1 << 1
+	InitFileOps       InitFlags = 1 << 2
+	InitAtomicTrunc   InitFlags = 1 << 3
+	InitExportSupport InitFlags = 1 << 4
+	InitBigWrites     InitFlags = 1 << 5
+	// Do not mask file access modes with umask. Not supported on OS X.
+	InitDontMask        InitFlags = 1 << 6
+	InitSpliceWrite     InitFlags = 1 << 7
+	InitSpliceMove      InitFlags = 1 << 8
+	InitSpliceRead      InitFlags = 1 << 9
+	InitFlockLocks      InitFlags = 1 << 10
+	InitHasIoctlDir     InitFlags = 1 << 11
+	InitAutoInvalData   InitFlags = 1 << 12
+	InitDoReaddirplus   InitFlags = 1 << 13
+	InitReaddirplusAuto InitFlags = 1 << 14
+	InitAsyncDIO        InitFlags = 1 << 15
+	InitWritebackCache  InitFlags = 1 << 16
+	InitNoOpenSupport   InitFlags = 1 << 17
+
+	InitCaseSensitive InitFlags = 1 << 29 // OS X only
+	InitVolRename     InitFlags = 1 << 30 // OS X only
+	InitXtimes        InitFlags = 1 << 31 // OS X only
+)
+
+type flagName struct {
+	bit  uint32
+	name string
+}
+
+var initFlagNames = []flagName{
+	{uint32(InitAsyncRead), "InitAsyncRead"},
+	{uint32(InitPosixLocks), "InitPosixLocks"},
+	{uint32(InitFileOps), "InitFileOps"},
+	{uint32(InitAtomicTrunc), "InitAtomicTrunc"},
+	{uint32(InitExportSupport), "InitExportSupport"},
+	{uint32(InitBigWrites), "InitBigWrites"},
+	{uint32(InitDontMask), "InitDontMask"},
+	{uint32(InitSpliceWrite), "InitSpliceWrite"},
+	{uint32(InitSpliceMove), "InitSpliceMove"},
+	{uint32(InitSpliceRead), "InitSpliceRead"},
+	{uint32(InitFlockLocks), "InitFlockLocks"},
+	{uint32(InitHasIoctlDir), "InitHasIoctlDir"},
+	{uint32(InitAutoInvalData), "InitAutoInvalData"},
+	{uint32(InitDoReaddirplus), "InitDoReaddirplus"},
+	{uint32(InitReaddirplusAuto), "InitReaddirplusAuto"},
+	{uint32(InitAsyncDIO), "InitAsyncDIO"},
+	{uint32(InitWritebackCache), "InitWritebackCache"},
+	{uint32(InitNoOpenSupport), "InitNoOpenSupport"},
+
+	{uint32(InitCaseSensitive), "InitCaseSensitive"},
+	{uint32(InitVolRename), "InitVolRename"},
+	{uint32(InitXtimes), "InitXtimes"},
+}
+
+func (fl InitFlags) String() string {
+	return flagString(uint32(fl), initFlagNames)
+}
+
+func flagString(f uint32, names []flagName) string {
+	var s string
+
+	if f == 0 {
+		return "0"
+	}
+
+	for _, n := range names {
+		if f&n.bit != 0 {
+			s += "+" + n.name
+			f &^= n.bit
+		}
+	}
+	if f != 0 {
+		s += fmt.Sprintf("%+#x", f)
+	}
+	return s[1:]
+}
+
+// The ReleaseFlags are used in the Release exchange.
+type ReleaseFlags uint32
+
+const (
+	ReleaseFlush ReleaseFlags = 1 << 0
+)
+
+func (fl ReleaseFlags) String() string {
+	return flagString(uint32(fl), releaseFlagNames)
+}
+
+var releaseFlagNames = []flagName{
+	{uint32(ReleaseFlush), "ReleaseFlush"},
+}
+
+// Opcodes
+const (
+	opLookup      = 1
+	opForget      = 2 // no reply
+	opGetattr     = 3
+	opSetattr     = 4
+	opReadlink    = 5
+	opSymlink     = 6
+	opMknod       = 8
+	opMkdir       = 9
+	opUnlink      = 10
+	opRmdir       = 11
+	opRename      = 12
+	opLink        = 13
+	opOpen        = 14
+	opRead        = 15
+	opWrite       = 16
+	opStatfs      = 17
+	opRelease     = 18
+	opFsync       = 20
+	opSetxattr    = 21
+	opGetxattr    = 22
+	opListxattr   = 23
+	opRemovexattr = 24
+	opFlush       = 25
+	opInit        = 26
+	opOpendir     = 27
+	opReaddir     = 28
+	opReleasedir  = 29
+	opFsyncdir    = 30
+	opGetlk       = 31
+	opSetlk       = 32
+	opSetlkw      = 33
+	opAccess      = 34
+	opCreate      = 35
+	opInterrupt   = 36
+	opBmap        = 37
+	opDestroy     = 38
+	opIoctl       = 39 // Linux?
+	opPoll        = 40 // Linux?
+
+	// OS X
+	opSetvolname = 61
+	opGetxtimes  = 62
+	opExchange   = 63
+)
+
+type entryOut struct {
+	Nodeid         uint64 // Inode ID
+	Generation     uint64 // Inode generation
+	EntryValid     uint64 // Cache timeout for the name
+	AttrValid      uint64 // Cache timeout for the attributes
+	EntryValidNsec uint32
+	AttrValidNsec  uint32
+	Attr           attr
+}
+
+func entryOutSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 9}):
+		return unsafe.Offsetof(entryOut{}.Attr) + unsafe.Offsetof(entryOut{}.Attr.Blksize)
+	default:
+		return unsafe.Sizeof(entryOut{})
+	}
+}
+
+type forgetIn struct {
+	Nlookup uint64
+}
+
+type getattrIn struct {
+	GetattrFlags uint32
+	dummy        uint32
+	Fh           uint64
+}
+
+type attrOut struct {
+	AttrValid     uint64 // Cache timeout for the attributes
+	AttrValidNsec uint32
+	Dummy         uint32
+	Attr          attr
+}
+
+func attrOutSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 9}):
+		return unsafe.Offsetof(attrOut{}.Attr) + unsafe.Offsetof(attrOut{}.Attr.Blksize)
+	default:
+		return unsafe.Sizeof(attrOut{})
+	}
+}
+
+// OS X
+type getxtimesOut struct {
+	Bkuptime     uint64
+	Crtime       uint64
+	BkuptimeNsec uint32
+	CrtimeNsec   uint32
+}
+
+type mknodIn struct {
+	Mode    uint32
+	Rdev    uint32
+	Umask   uint32
+	padding uint32
+	// "filename\x00" follows.
+}
+
+func mknodInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 12}):
+		return unsafe.Offsetof(mknodIn{}.Umask)
+	default:
+		return unsafe.Sizeof(mknodIn{})
+	}
+}
+
+type mkdirIn struct {
+	Mode  uint32
+	Umask uint32
+	// filename follows
+}
+
+func mkdirInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 12}):
+		return unsafe.Offsetof(mkdirIn{}.Umask) + 4
+	default:
+		return unsafe.Sizeof(mkdirIn{})
+	}
+}
+
+type renameIn struct {
+	Newdir uint64
+	// "oldname\x00newname\x00" follows
+}
+
+// OS X
+type exchangeIn struct {
+	Olddir  uint64
+	Newdir  uint64
+	Options uint64
+}
+
+type linkIn struct {
+	Oldnodeid uint64
+}
+
+type setattrInCommon struct {
+	Valid     uint32
+	Padding   uint32
+	Fh        uint64
+	Size      uint64
+	LockOwner uint64 // unused on OS X?
+	Atime     uint64
+	Mtime     uint64
+	Unused2   uint64
+	AtimeNsec uint32
+	MtimeNsec uint32
+	Unused3   uint32
+	Mode      uint32
+	Unused4   uint32
+	Uid       uint32
+	Gid       uint32
+	Unused5   uint32
+}
+
+type openIn struct {
+	Flags  uint32
+	Unused uint32
+}
+
+type openOut struct {
+	Fh        uint64
+	OpenFlags uint32
+	Padding   uint32
+}
+
+type createIn struct {
+	Flags   uint32
+	Mode    uint32
+	Umask   uint32
+	padding uint32
+}
+
+func createInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 12}):
+		return unsafe.Offsetof(createIn{}.Umask)
+	default:
+		return unsafe.Sizeof(createIn{})
+	}
+}
+
+type releaseIn struct {
+	Fh           uint64
+	Flags        uint32
+	ReleaseFlags uint32
+	LockOwner    uint32
+}
+
+type flushIn struct {
+	Fh         uint64
+	FlushFlags uint32
+	Padding    uint32
+	LockOwner  uint64
+}
+
+type readIn struct {
+	Fh        uint64
+	Offset    uint64
+	Size      uint32
+	ReadFlags uint32
+	LockOwner uint64
+	Flags     uint32
+	padding   uint32
+}
+
+func readInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 9}):
+		return unsafe.Offsetof(readIn{}.ReadFlags) + 4
+	default:
+		return unsafe.Sizeof(readIn{})
+	}
+}
+
+// The ReadFlags are passed in ReadRequest.
+type ReadFlags uint32
+
+const (
+	// LockOwner field is valid.
+	ReadLockOwner ReadFlags = 1 << 1
+)
+
+var readFlagNames = []flagName{
+	{uint32(ReadLockOwner), "ReadLockOwner"},
+}
+
+func (fl ReadFlags) String() string {
+	return flagString(uint32(fl), readFlagNames)
+}
+
+type writeIn struct {
+	Fh         uint64
+	Offset     uint64
+	Size       uint32
+	WriteFlags uint32
+	LockOwner  uint64
+	Flags      uint32
+	padding    uint32
+}
+
+func writeInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 9}):
+		return unsafe.Offsetof(writeIn{}.LockOwner)
+	default:
+		return unsafe.Sizeof(writeIn{})
+	}
+}
+
+type writeOut struct {
+	Size    uint32
+	Padding uint32
+}
+
+// The WriteFlags are passed in WriteRequest.
+type WriteFlags uint32
+
+const (
+	WriteCache WriteFlags = 1 << 0
+	// LockOwner field is valid.
+	WriteLockOwner WriteFlags = 1 << 1
+)
+
+var writeFlagNames = []flagName{
+	{uint32(WriteCache), "WriteCache"},
+	{uint32(WriteLockOwner), "WriteLockOwner"},
+}
+
+func (fl WriteFlags) String() string {
+	return flagString(uint32(fl), writeFlagNames)
+}
+
+const compatStatfsSize = 48
+
+type statfsOut struct {
+	St kstatfs
+}
+
+type fsyncIn struct {
+	Fh         uint64
+	FsyncFlags uint32
+	Padding    uint32
+}
+
+type setxattrInCommon struct {
+	Size  uint32
+	Flags uint32
+}
+
+func (setxattrInCommon) position() uint32 {
+	return 0
+}
+
+type getxattrInCommon struct {
+	Size    uint32
+	Padding uint32
+}
+
+func (getxattrInCommon) position() uint32 {
+	return 0
+}
+
+type getxattrOut struct {
+	Size    uint32
+	Padding uint32
+}
+
+type lkIn struct {
+	Fh      uint64
+	Owner   uint64
+	Lk      fileLock
+	LkFlags uint32
+	padding uint32
+}
+
+func lkInSize(p Protocol) uintptr {
+	switch {
+	case p.LT(Protocol{7, 9}):
+		return unsafe.Offsetof(lkIn{}.LkFlags)
+	default:
+		return unsafe.Sizeof(lkIn{})
+	}
+}
+
+type lkOut struct {
+	Lk fileLock
+}
+
+type accessIn struct {
+	Mask    uint32
+	Padding uint32
+}
+
+type initIn struct {
+	Major        uint32
+	Minor        uint32
+	MaxReadahead uint32
+	Flags        uint32
+}
+
+const initInSize = int(unsafe.Sizeof(initIn{}))
+
+type initOut struct {
+	Major        uint32
+	Minor        uint32
+	MaxReadahead uint32
+	Flags        uint32
+	Unused       uint32
+	MaxWrite     uint32
+}
+
+type interruptIn struct {
+	Unique uint64
+}
+
+type bmapIn struct {
+	Block     uint64
+	BlockSize uint32
+	Padding   uint32
+}
+
+type bmapOut struct {
+	Block uint64
+}
+
+type inHeader struct {
+	Len     uint32
+	Opcode  uint32
+	Unique  uint64
+	Nodeid  uint64
+	Uid     uint32
+	Gid     uint32
+	Pid     uint32
+	Padding uint32
+}
+
+const inHeaderSize = int(unsafe.Sizeof(inHeader{}))
+
+type outHeader struct {
+	Len    uint32
+	Error  int32
+	Unique uint64
+}
+
+type dirent struct {
+	Ino     uint64
+	Off     uint64
+	Namelen uint32
+	Type    uint32
+	Name    [0]byte
+}
+
+const direntSize = 8 + 8 + 4 + 4
+
+const (
+	notifyCodePoll       int32 = 1
+	notifyCodeInvalInode int32 = 2
+	notifyCodeInvalEntry int32 = 3
+)
+
+type notifyInvalInodeOut struct {
+	Ino uint64
+	Off int64
+	Len int64
+}
+
+type notifyInvalEntryOut struct {
+	Parent  uint64
+	Namelen uint32
+	padding uint32
+}

weed/vendor/bazil.org/fuse/fuse_kernel_darwin.go

@@ -0,0 +1,88 @@
+package fuse
+
+import (
+	"time"
+)
+
+type attr struct {
+	Ino        uint64
+	Size       uint64
+	Blocks     uint64
+	Atime      uint64
+	Mtime      uint64
+	Ctime      uint64
+	Crtime_    uint64 // OS X only
+	AtimeNsec  uint32
+	MtimeNsec  uint32
+	CtimeNsec  uint32
+	CrtimeNsec uint32 // OS X only
+	Mode       uint32
+	Nlink      uint32
+	Uid        uint32
+	Gid        uint32
+	Rdev       uint32
+	Flags_     uint32 // OS X only; see chflags(2)
+	Blksize    uint32
+	padding    uint32
+}
+
+func (a *attr) SetCrtime(s uint64, ns uint32) {
+	a.Crtime_, a.CrtimeNsec = s, ns
+}
+
+func (a *attr) SetFlags(f uint32) {
+	a.Flags_ = f
+}
+
+type setattrIn struct {
+	setattrInCommon
+
+	// OS X only
+	Bkuptime_    uint64
+	Chgtime_     uint64
+	Crtime       uint64
+	BkuptimeNsec uint32
+	ChgtimeNsec  uint32
+	CrtimeNsec   uint32
+	Flags_       uint32 // see chflags(2)
+}
+
+func (in *setattrIn) BkupTime() time.Time {
+	return time.Unix(int64(in.Bkuptime_), int64(in.BkuptimeNsec))
+}
+
+func (in *setattrIn) Chgtime() time.Time {
+	return time.Unix(int64(in.Chgtime_), int64(in.ChgtimeNsec))
+}
+
+func (in *setattrIn) Flags() uint32 {
+	return in.Flags_
+}
+
+func openFlags(flags uint32) OpenFlags {
+	return OpenFlags(flags)
+}
+
+type getxattrIn struct {
+	getxattrInCommon
+
+	// OS X only
+	Position uint32
+	Padding  uint32
+}
+
+func (g *getxattrIn) position() uint32 {
+	return g.Position
+}
+
+type setxattrIn struct {
+	setxattrInCommon
+
+	// OS X only
+	Position uint32
+	Padding  uint32
+}
+
+func (s *setxattrIn) position() uint32 {
+	return s.Position
+}

weed/vendor/bazil.org/fuse/fuse_kernel_freebsd.go

@@ -0,0 +1,62 @@
+package fuse
+
+import "time"
+
+type attr struct {
+	Ino       uint64
+	Size      uint64
+	Blocks    uint64
+	Atime     uint64
+	Mtime     uint64
+	Ctime     uint64
+	AtimeNsec uint32
+	MtimeNsec uint32
+	CtimeNsec uint32
+	Mode      uint32
+	Nlink     uint32
+	Uid       uint32
+	Gid       uint32
+	Rdev      uint32
+	Blksize   uint32
+	padding   uint32
+}
+
+func (a *attr) Crtime() time.Time {
+	return time.Time{}
+}
+
+func (a *attr) SetCrtime(s uint64, ns uint32) {
+	// ignored on freebsd
+}
+
+func (a *attr) SetFlags(f uint32) {
+	// ignored on freebsd
+}
+
+type setattrIn struct {
+	setattrInCommon
+}
+
+func (in *setattrIn) BkupTime() time.Time {
+	return time.Time{}
+}
+
+func (in *setattrIn) Chgtime() time.Time {
+	return time.Time{}
+}
+
+func (in *setattrIn) Flags() uint32 {
+	return 0
+}
+
+func openFlags(flags uint32) OpenFlags {
+	return OpenFlags(flags)
+}
+
+type getxattrIn struct {
+	getxattrInCommon
+}
+
+type setxattrIn struct {
+	setxattrInCommon
+}

weed/vendor/bazil.org/fuse/fuse_kernel_linux.go

@@ -0,0 +1,70 @@
+package fuse
+
+import "time"
+
+type attr struct {
+	Ino       uint64
+	Size      uint64
+	Blocks    uint64
+	Atime     uint64
+	Mtime     uint64
+	Ctime     uint64
+	AtimeNsec uint32
+	MtimeNsec uint32
+	CtimeNsec uint32
+	Mode      uint32
+	Nlink     uint32
+	Uid       uint32
+	Gid       uint32
+	Rdev      uint32
+	Blksize   uint32
+	padding   uint32
+}
+
+func (a *attr) Crtime() time.Time {
+	return time.Time{}
+}
+
+func (a *attr) SetCrtime(s uint64, ns uint32) {
+	// Ignored on Linux.
+}
+
+func (a *attr) SetFlags(f uint32) {
+	// Ignored on Linux.
+}
+
+type setattrIn struct {
+	setattrInCommon
+}
+
+func (in *setattrIn) BkupTime() time.Time {
+	return time.Time{}
+}
+
+func (in *setattrIn) Chgtime() time.Time {
+	return time.Time{}
+}
+
+func (in *setattrIn) Flags() uint32 {
+	return 0
+}
+
+func openFlags(flags uint32) OpenFlags {
+	// on amd64, the 32-bit O_LARGEFILE flag is always seen;
+	// on i386, the flag probably depends on the app
+	// requesting, but in any case should be utterly
+	// uninteresting to us here; our kernel protocol messages
+	// are not directly related to the client app's kernel
+	// API/ABI
+	flags &^= 0x8000
+
+	return OpenFlags(flags)
+}
+
+type getxattrIn struct {
+	getxattrInCommon
+}
+
+type setxattrIn struct {
+	setxattrInCommon
+}

weed/vendor/bazil.org/fuse/fuse_kernel_std.go

@@ -0,0 +1 @@
+package fuse

weed/vendor/bazil.org/fuse/fuseutil/fuseutil.go

@@ -0,0 +1,20 @@
+package fuseutil
+
+import (
+	"bazil.org/fuse"
+)
+
+// HandleRead handles a read request assuming that data is the entire file content.
+// It adjusts the amount returned in resp according to req.Offset and req.Size.
+func HandleRead(req *fuse.ReadRequest, resp *fuse.ReadResponse, data []byte) {
+	if req.Offset >= int64(len(data)) {
+		data = nil
+	} else {
+		data = data[req.Offset:]
+	}
+	if len(data) > req.Size {
+		data = data[:req.Size]
+	}
+	n := copy(resp.Data[:req.Size], data)
+	resp.Data = resp.Data[:n]
+}

weed/vendor/bazil.org/fuse/mount.go

@@ -0,0 +1,28 @@
+package fuse
+
+import (
+	"bufio"
+	"io"
+	"log"
+	"sync"
+)
+
+func neverIgnoreLine(line string) bool {
+	return false
+}
+
+func lineLogger(wg *sync.WaitGroup, prefix string, ignore func(line string) bool, r io.ReadCloser) {
+	defer wg.Done()
+
+	scanner := bufio.NewScanner(r)
+	for scanner.Scan() {
+		line := scanner.Text()
+		if ignore(line) {
+			continue
+		}
+		log.Printf("%s: %s", prefix, line)
+	}
+	if err := scanner.Err(); err != nil {
+		log.Printf("%s, error reading: %v", prefix, err)
+	}
+}

weed/vendor/bazil.org/fuse/mount_darwin.go

@@ -0,0 +1,206 @@
+package fuse
+
+import (
+	"errors"
+	"fmt"
+	"log"
+	"os"
+	"os/exec"
+	"path"
+	"strconv"
+	"strings"
+	"sync"
+	"syscall"
+)
+
+var (
+	errNoAvail         = errors.New("no available fuse devices")
+	errNotLoaded       = errors.New("osxfuse is not loaded")
+	errOSXFUSENotFound = errors.New("cannot locate OSXFUSE")
+)
+
+func loadOSXFUSE(bin string) error {
+	cmd := exec.Command(bin)
+	cmd.Dir = "/"
+	cmd.Stdout = os.Stdout
+	cmd.Stderr = os.Stderr
+	err := cmd.Run()
+	return err
+}
+
+func openOSXFUSEDev(devPrefix string) (*os.File, error) {
+	var f *os.File
+	var err error
+	for i := uint64(0); ; i++ {
+		path := devPrefix + strconv.FormatUint(i, 10)
+		f, err = os.OpenFile(path, os.O_RDWR, 0000)
+		if os.IsNotExist(err) {
+			if i == 0 {
+				// not even the first device was found -> fuse is not loaded
+				return nil, errNotLoaded
+			}
+
+			// we've run out of kernel-provided devices
+			return nil, errNoAvail
+		}
+
+		if err2, ok := err.(*os.PathError); ok && err2.Err == syscall.EBUSY {
+			// try the next one
+			continue
+		}
+
+		if err != nil {
+			return nil, err
+		}
+		return f, nil
+	}
+}
+
+func handleMountOSXFUSE(helperName string, errCh chan<- error) func(line string) (ignore bool) {
+	var noMountpointPrefix = helperName + `: `
+	const noMountpointSuffix = `: No such file or directory`
+	return func(line string) (ignore bool) {
+		if strings.HasPrefix(line, noMountpointPrefix) && strings.HasSuffix(line, noMountpointSuffix) {
+			// re-extract it from the error message in case some layer
+			// changed the path
+			mountpoint := line[len(noMountpointPrefix) : len(line)-len(noMountpointSuffix)]
+			err := &MountpointDoesNotExistError{
+				Path: mountpoint,
+			}
+			select {
+			case errCh <- err:
+				return true
+			default:
+				// not the first error; fall back to logging it
+				return false
+			}
+		}
+
+		return false
+	}
+}
+
+// isBoringMountOSXFUSEError returns whether the Wait error is
+// uninteresting; exit status 64 is.
+func isBoringMountOSXFUSEError(err error) bool {
+	if err, ok := err.(*exec.ExitError); ok && err.Exited() {
+		if status, ok := err.Sys().(syscall.WaitStatus); ok && status.ExitStatus() == 64 {
+			return true
+		}
+	}
+	return false
+}
+
+func callMount(bin string, daemonVar string, dir string, conf *mountConfig, f *os.File, ready chan<- struct{}, errp *error) error {
+	for k, v := range conf.options {
+		if strings.Contains(k, ",") || strings.Contains(v, ",") {
+			// Silly limitation but the mount helper does not
+			// understand any escaping. See TestMountOptionCommaError.
+			return fmt.Errorf("mount options cannot contain commas on darwin: %q=%q", k, v)
+		}
+	}
+	cmd := exec.Command(
+		bin,
+		"-o", conf.getOptions(),
+		// Tell osxfuse-kext how large our buffer is. It must split
+		// writes larger than this into multiple writes.
+		//
+		// OSXFUSE seems to ignore InitResponse.MaxWrite, and uses
+		// this instead.
+		"-o", "iosize="+strconv.FormatUint(maxWrite, 10),
+		// refers to fd passed in cmd.ExtraFiles
+		"3",
+		dir,
+	)
+	cmd.ExtraFiles = []*os.File{f}
+	cmd.Env = os.Environ()
+	cmd.Env = append(cmd.Env, "MOUNT_FUSEFS_CALL_BY_LIB=")
+
+	daemon := os.Args[0]
+	if daemonVar != "" {
+		cmd.Env = append(cmd.Env, daemonVar+"="+daemon)
+	}
+
+	stdout, err := cmd.StdoutPipe()
+	if err != nil {
+		return fmt.Errorf("setting up mount_osxfusefs stderr: %v", err)
+	}
+	stderr, err := cmd.StderrPipe()
+	if err != nil {
+		return fmt.Errorf("setting up mount_osxfusefs stderr: %v", err)
+	}
+
+	if err := cmd.Start(); err != nil {
+		return fmt.Errorf("mount_osxfusefs: %v", err)
+	}
+	helperErrCh := make(chan error, 1)
+	go func() {
+		var wg sync.WaitGroup
+		wg.Add(2)
+		go lineLogger(&wg, "mount helper output", neverIgnoreLine, stdout)
+		helperName := path.Base(bin)
+		go lineLogger(&wg, "mount helper error", handleMountOSXFUSE(helperName, helperErrCh), stderr)
+		wg.Wait()
+		if err := cmd.Wait(); err != nil {
+			// see if we have a better error to report
+			select {
+			case helperErr := <-helperErrCh:
+				// log the Wait error if it's not what we expected
+				if !isBoringMountOSXFUSEError(err) {
+					log.Printf("mount helper failed: %v", err)
+				}
+				// and now return what we grabbed from stderr as the real
+				// error
+				*errp = helperErr
+				close(ready)
+				return
+			default:
+				// nope, fall back to generic message
+			}
+
+			*errp = fmt.Errorf("mount_osxfusefs: %v", err)
+			close(ready)
+			return
+		}
+
+		*errp = nil
+		close(ready)
+	}()
+	return nil
+}
+
+func mount(dir string, conf *mountConfig, ready chan<- struct{}, errp *error) (*os.File, error) {
+	locations := conf.osxfuseLocations
+	if locations == nil {
+		locations = []OSXFUSEPaths{
+			OSXFUSELocationV3,
+			OSXFUSELocationV2,
+		}
+	}
+	for _, loc := range locations {
+		if _, err := os.Stat(loc.Mount); os.IsNotExist(err) {
+			// try the other locations
+			continue
+		}
+
+		f, err := openOSXFUSEDev(loc.DevicePrefix)
+		if err == errNotLoaded {
+			err = loadOSXFUSE(loc.Load)
+			if err != nil {
+				return nil, err
+			}
+			// try again
+			f, err = openOSXFUSEDev(loc.DevicePrefix)
+		}
+		if err != nil {
+			return nil, err
+		}
+		err = callMount(loc.Mount, loc.DaemonVar, dir, conf, f, ready, errp)
+		if err != nil {
+			f.Close()
+			return nil, err
+		}
+		return f, nil
+	}
+	return nil, errOSXFUSENotFound
+}

weed/vendor/bazil.org/fuse/mount_freebsd.go

@@ -0,0 +1,111 @@
+package fuse
+
+import (
+	"fmt"
+	"log"
+	"os"
+	"os/exec"
+	"strings"
+	"sync"
+	"syscall"
+)
+
+func handleMountFusefsStderr(errCh chan<- error) func(line string) (ignore bool) {
+	return func(line string) (ignore bool) {
+		const (
+			noMountpointPrefix = `mount_fusefs: `
+			noMountpointSuffix = `: No such file or directory`
+		)
+		if strings.HasPrefix(line, noMountpointPrefix) && strings.HasSuffix(line, noMountpointSuffix) {
+			// re-extract it from the error message in case some layer
+			// changed the path
+			mountpoint := line[len(noMountpointPrefix) : len(line)-len(noMountpointSuffix)]
+			err := &MountpointDoesNotExistError{
+				Path: mountpoint,
+			}
+			select {
+			case errCh <- err:
+				return true
+			default:
+				// not the first error; fall back to logging it
+				return false
+			}
+		}
+
+		return false
+	}
+}
+
+// isBoringMountFusefsError returns whether the Wait error is
+// uninteresting; exit status 1 is.
+func isBoringMountFusefsError(err error) bool {
+	if err, ok := err.(*exec.ExitError); ok && err.Exited() {
+		if status, ok := err.Sys().(syscall.WaitStatus); ok && status.ExitStatus() == 1 {
+			return true
+		}
+	}
+	return false
+}
+
+func mount(dir string, conf *mountConfig, ready chan<- struct{}, errp *error) (*os.File, error) {
+	for k, v := range conf.options {
+		if strings.Contains(k, ",") || strings.Contains(v, ",") {
+			// Silly limitation but the mount helper does not
+			// understand any escaping. See TestMountOptionCommaError.
+			return nil, fmt.Errorf("mount options cannot contain commas on FreeBSD: %q=%q", k, v)
+		}
+	}
+
+	f, err := os.OpenFile("/dev/fuse", os.O_RDWR, 0000)
+	if err != nil {
+		*errp = err
+		return nil, err
+	}
+
+	cmd := exec.Command(
+		"/sbin/mount_fusefs",
+		"--safe",
+		"-o", conf.getOptions(),
+		"3",
+		dir,
+	)
+	cmd.ExtraFiles = []*os.File{f}
+
+	stdout, err := cmd.StdoutPipe()
+	if err != nil {
+		return nil, fmt.Errorf("setting up mount_fusefs stderr: %v", err)
+	}
+	stderr, err := cmd.StderrPipe()
+	if err != nil {
+		return nil, fmt.Errorf("setting up mount_fusefs stderr: %v", err)
+	}
+
+	if err := cmd.Start(); err != nil {
+		return nil, fmt.Errorf("mount_fusefs: %v", err)
+	}
+	helperErrCh := make(chan error, 1)
+	var wg sync.WaitGroup
+	wg.Add(2)
+	go lineLogger(&wg, "mount helper output", neverIgnoreLine, stdout)
+	go lineLogger(&wg, "mount helper error", handleMountFusefsStderr(helperErrCh), stderr)
+	wg.Wait()
+	if err := cmd.Wait(); err != nil {
+		// see if we have a better error to report
+		select {
+		case helperErr := <-helperErrCh:
+			// log the Wait error if it's not what we expected
+			if !isBoringMountFusefsError(err) {
+				log.Printf("mount helper failed: %v", err)
+			}
+			// and now return what we grabbed from stderr as the real
+			// error
+			return nil, helperErr
+		default:
+			// nope, fall back to generic message
+		}
+		return nil, fmt.Errorf("mount_fusefs: %v", err)
+	}
+
+	close(ready)
+	return f, nil
+}

weed/vendor/bazil.org/fuse/mount_linux.go

@@ -0,0 +1,150 @@
+package fuse
+
+import (
+	"fmt"
+	"log"
+	"net"
+	"os"
+	"os/exec"
+	"strings"
+	"sync"
+	"syscall"
+)
+
+func handleFusermountStderr(errCh chan<- error) func(line string) (ignore bool) {
+	return func(line string) (ignore bool) {
+		if line == `fusermount: failed to open /etc/fuse.conf: Permission denied` {
+			// Silence this particular message, it occurs way too
+			// commonly and isn't very relevant to whether the mount
+			// succeeds or not.
+			return true
+		}
+
+		const (
+			noMountpointPrefix = `fusermount: failed to access mountpoint `
+			noMountpointSuffix = `: No such file or directory`
+		)
+		if strings.HasPrefix(line, noMountpointPrefix) && strings.HasSuffix(line, noMountpointSuffix) {
+			// re-extract it from the error message in case some layer
+			// changed the path
+			mountpoint := line[len(noMountpointPrefix) : len(line)-len(noMountpointSuffix)]
+			err := &MountpointDoesNotExistError{
+				Path: mountpoint,
+			}
+			select {
+			case errCh <- err:
+				return true
+			default:
+				// not the first error; fall back to logging it
+				return false
+			}
+		}
+
+		return false
+	}
+}
+
+// isBoringFusermountError returns whether the Wait error is
+// uninteresting; exit status 1 is.
+func isBoringFusermountError(err error) bool {
+	if err, ok := err.(*exec.ExitError); ok && err.Exited() {
+		if status, ok := err.Sys().(syscall.WaitStatus); ok && status.ExitStatus() == 1 {
+			return true
+		}
+	}
+	return false
+}
+
+func mount(dir string, conf *mountConfig, ready chan<- struct{}, errp *error) (fusefd *os.File, err error) {
+	// linux mount is never delayed
+	close(ready)
+
+	fds, err := syscall.Socketpair(syscall.AF_FILE, syscall.SOCK_STREAM, 0)
+	if err != nil {
+		return nil, fmt.Errorf("socketpair error: %v", err)
+	}
+
+	writeFile := os.NewFile(uintptr(fds[0]), "fusermount-child-writes")
+	defer writeFile.Close()
+
+	readFile := os.NewFile(uintptr(fds[1]), "fusermount-parent-reads")
+	defer readFile.Close()
+
+	cmd := exec.Command(
+		"fusermount",
+		"-o", conf.getOptions(),
+		"--",
+		dir,
+	)
+	cmd.Env = append(os.Environ(), "_FUSE_COMMFD=3")
+
+	cmd.ExtraFiles = []*os.File{writeFile}
+
+	var wg sync.WaitGroup
+	stdout, err := cmd.StdoutPipe()
+	if err != nil {
+		return nil, fmt.Errorf("setting up fusermount stderr: %v", err)
+	}
+	stderr, err := cmd.StderrPipe()
+	if err != nil {
+		return nil, fmt.Errorf("setting up fusermount stderr: %v", err)
+	}
+
+	if err := cmd.Start(); err != nil {
+		return nil, fmt.Errorf("fusermount: %v", err)
+	}
+	helperErrCh := make(chan error, 1)
+	wg.Add(2)
+	go lineLogger(&wg, "mount helper output", neverIgnoreLine, stdout)
+	go lineLogger(&wg, "mount helper error", handleFusermountStderr(helperErrCh), stderr)
+	wg.Wait()
+	if err := cmd.Wait(); err != nil {
+		// see if we have a better error to report
+		select {
+		case helperErr := <-helperErrCh:
+			// log the Wait error if it's not what we expected
+			if !isBoringFusermountError(err) {
+				log.Printf("mount helper failed: %v", err)
+			}
+			// and now return what we grabbed from stderr as the real
+			// error
+			return nil, helperErr
+		default:
+			// nope, fall back to generic message
+		}
+
+		return nil, fmt.Errorf("fusermount: %v", err)
+	}
+
+	c, err := net.FileConn(readFile)
+	if err != nil {
+		return nil, fmt.Errorf("FileConn from fusermount socket: %v", err)
+	}
+	defer c.Close()
+
+	uc, ok := c.(*net.UnixConn)
+	if !ok {
+		return nil, fmt.Errorf("unexpected FileConn type; expected UnixConn, got %T", c)
+	}
+
+	buf := make([]byte, 32) // expect 1 byte
+	oob := make([]byte, 32) // expect 24 bytes
+	_, oobn, _, _, err := uc.ReadMsgUnix(buf, oob)
+	scms, err := syscall.ParseSocketControlMessage(oob[:oobn])
+	if err != nil {
+		return nil, fmt.Errorf("ParseSocketControlMessage: %v", err)
+	}
+	if len(scms) != 1 {
+		return nil, fmt.Errorf("expected 1 SocketControlMessage; got scms = %#v", scms)
+	}
+	scm := scms[0]
+	gotFds, err := syscall.ParseUnixRights(&scm)
+	if err != nil {
+		return nil, fmt.Errorf("syscall.ParseUnixRights: %v", err)
+	}
+	if len(gotFds) != 1 {
+		return nil, fmt.Errorf("wanted 1 fd; got %#v", gotFds)
+	}
+	f := os.NewFile(uintptr(gotFds[0]), "/dev/fuse")
+	return f, nil
+}

weed/vendor/bazil.org/fuse/options.go

@@ -0,0 +1,223 @@
+package fuse
+
+import (
+	"errors"
+	"strings"
+)
+
+func dummyOption(conf *mountConfig) error {
+	return nil
+}
+
+// mountConfig holds the configuration for a mount operation.
+// Use it by passing MountOption values to Mount.
+type mountConfig struct {
+	options          map[string]string
+	maxReadahead     uint32
+	initFlags        InitFlags
+	osxfuseLocations []OSXFUSEPaths
+}
+
+func escapeComma(s string) string {
+	s = strings.Replace(s, `\`, `\\`, -1)
+	s = strings.Replace(s, `,`, `\,`, -1)
+	return s
+}
+
+// getOptions makes a string of options suitable for passing to FUSE
+// mount flag `-o`. Returns an empty string if no options were set.
+// Any platform specific adjustments should happen before the call.
+func (m *mountConfig) getOptions() string {
+	var opts []string
+	for k, v := range m.options {
+		k = escapeComma(k)
+		if v != "" {
+			k += "=" + escapeComma(v)
+		}
+		opts = append(opts, k)
+	}
+	return strings.Join(opts, ",")
+}
+
+type mountOption func(*mountConfig) error
+
+// MountOption is passed to Mount to change the behavior of the mount.
+type MountOption mountOption
+
+// FSName sets the file system name (also called source) that is
+// visible in the list of mounted file systems.
+//
+// FreeBSD ignores this option.
+func FSName(name string) MountOption {
+	return func(conf *mountConfig) error {
+		conf.options["fsname"] = name
+		return nil
+	}
+}
+
+// Subtype sets the subtype of the mount. The main type is always
+// `fuse`. The type in a list of mounted file systems will look like
+// `fuse.foo`.
+//
+// OS X ignores this option.
+// FreeBSD ignores this option.
+func Subtype(fstype string) MountOption {
+	return func(conf *mountConfig) error {
+		conf.options["subtype"] = fstype
+		return nil
+	}
+}
+
+// LocalVolume sets the volume to be local (instead of network),
+// changing the behavior of Finder, Spotlight, and such.
+//
+// OS X only. Others ignore this option.
+func LocalVolume() MountOption {
+	return localVolume
+}
+
+// VolumeName sets the volume name shown in Finder.
+//
+// OS X only. Others ignore this option.
+func VolumeName(name string) MountOption {
+	return volumeName(name)
+}
+
+var ErrCannotCombineAllowOtherAndAllowRoot = errors.New("cannot combine AllowOther and AllowRoot")
+
+// AllowOther allows other users to access the file system.
+//
+// Only one of AllowOther or AllowRoot can be used.
+func AllowOther() MountOption {
+	return func(conf *mountConfig) error {
+		if _, ok := conf.options["allow_root"]; ok {
+			return ErrCannotCombineAllowOtherAndAllowRoot
+		}
+		conf.options["allow_other"] = ""
+		return nil
+	}
+}
+
+// AllowRoot allows other users to access the file system.
+//
+// Only one of AllowOther or AllowRoot can be used.
+//
+// FreeBSD ignores this option.
+func AllowRoot() MountOption {
+	return func(conf *mountConfig) error {
+		if _, ok := conf.options["allow_other"]; ok {
+			return ErrCannotCombineAllowOtherAndAllowRoot
+		}
+		conf.options["allow_root"] = ""
+		return nil
+	}
+}
+
+// DefaultPermissions makes the kernel enforce access control based on
+// the file mode (as in chmod).
+//
+// Without this option, the Node itself decides what is and is not
+// allowed. This is normally ok because FUSE file systems cannot be
+// accessed by other users without AllowOther/AllowRoot.
+//
+// FreeBSD ignores this option.
+func DefaultPermissions() MountOption {
+	return func(conf *mountConfig) error {
+		conf.options["default_permissions"] = ""
+		return nil
+	}
+}
+
+// ReadOnly makes the mount read-only.
+func ReadOnly() MountOption {
+	return func(conf *mountConfig) error {
+		conf.options["ro"] = ""
+		return nil
+	}
+}
+
+// MaxReadahead sets the number of bytes that can be prefetched for
+// sequential reads. The kernel can enforce a maximum value lower than
+// this.
+//
+// This setting makes the kernel perform speculative reads that do not
+// originate from any client process. This usually tremendously
+// improves read performance.
+func MaxReadahead(n uint32) MountOption {
+	return func(conf *mountConfig) error {
+		conf.maxReadahead = n
+		return nil
+	}
+}
+
+// AsyncRead enables multiple outstanding read requests for the same
+// handle. Without this, there is at most one request in flight at a
+// time.
+func AsyncRead() MountOption {
+	return func(conf *mountConfig) error {
+		conf.initFlags |= InitAsyncRead
+		return nil
+	}
+}
+
+// WritebackCache enables the kernel to buffer writes before sending
+// them to the FUSE server. Without this, writethrough caching is
+// used.
+func WritebackCache() MountOption {
+	return func(conf *mountConfig) error {
+		conf.initFlags |= InitWritebackCache
+		return nil
+	}
+}
+
+// OSXFUSEPaths describes the paths used by an installed OSXFUSE
+// version. See OSXFUSELocationV3 for typical values.
+type OSXFUSEPaths struct {
+	// Prefix for the device file. At mount time, an incrementing
+	// number is suffixed until a free FUSE device is found.
+	DevicePrefix string
+	// Path of the load helper, used to load the kernel extension if
+	// no device files are found.
+	Load string
+	// Path of the mount helper, used for the actual mount operation.
+	Mount string
+	// Environment variable used to pass the path to the executable
+	// calling the mount helper.
+	DaemonVar string
+}
+
+// Default paths for OSXFUSE. See OSXFUSELocations.
+var (
+	OSXFUSELocationV3 = OSXFUSEPaths{
+		DevicePrefix: "/dev/osxfuse",
+		Load:         "/Library/Filesystems/osxfuse.fs/Contents/Resources/load_osxfuse",
+		Mount:        "/Library/Filesystems/osxfuse.fs/Contents/Resources/mount_osxfuse",
+		DaemonVar:    "MOUNT_OSXFUSE_DAEMON_PATH",
+	}
+	OSXFUSELocationV2 = OSXFUSEPaths{
+		DevicePrefix: "/dev/osxfuse",
+		Load:         "/Library/Filesystems/osxfusefs.fs/Support/load_osxfusefs",
+		Mount:        "/Library/Filesystems/osxfusefs.fs/Support/mount_osxfusefs",
+		DaemonVar:    "MOUNT_FUSEFS_DAEMON_PATH",
+	}
+)
+
+// OSXFUSELocations sets where to look for OSXFUSE files. The
+// arguments are all the possible locations. The previous locations
+// are replaced.
+//
+// Without this option, OSXFUSELocationV3 and OSXFUSELocationV2 are
+// used.
+//
+// OS X only. Others ignore this option.
+func OSXFUSELocations(paths ...OSXFUSEPaths) MountOption {
+	return func(conf *mountConfig) error {
+		if len(paths) == 0 {
+			return errors.New("must specify at least one location for OSXFUSELocations")
+		}
+		// replace previous values, but make a copy so there's no
+		// worries about caller mutating their slice
+		conf.osxfuseLocations = append(conf.osxfuseLocations[:0], paths...)
+		return nil
+	}
+}

weed/vendor/bazil.org/fuse/options_darwin.go

@@ -0,0 +1,13 @@
+package fuse
+
+func localVolume(conf *mountConfig) error {
+	conf.options["local"] = ""
+	return nil
+}
+
+func volumeName(name string) MountOption {
+	return func(conf *mountConfig) error {
+		conf.options["volname"] = name
+		return nil
+	}
+}

weed/vendor/bazil.org/fuse/options_freebsd.go

@@ -0,0 +1,9 @@
+package fuse
+
+func localVolume(conf *mountConfig) error {
+	return nil
+}
+
+func volumeName(name string) MountOption {
+	return dummyOption
+}

weed/vendor/bazil.org/fuse/options_linux.go

@@ -0,0 +1,9 @@
+package fuse
+
+func localVolume(conf *mountConfig) error {
+	return nil
+}
+
+func volumeName(name string) MountOption {
+	return dummyOption
+}

weed/vendor/bazil.org/fuse/protocol.go

@@ -0,0 +1,75 @@
+package fuse
+
+import (
+	"fmt"
+)
+
+// Protocol is a FUSE protocol version number.
+type Protocol struct {
+	Major uint32
+	Minor uint32
+}
+
+func (p Protocol) String() string {
+	return fmt.Sprintf("%d.%d", p.Major, p.Minor)
+}
+
+// LT returns whether a is less than b.
+func (a Protocol) LT(b Protocol) bool {
+	return a.Major < b.Major ||
+		(a.Major == b.Major && a.Minor < b.Minor)
+}
+
+// GE returns whether a is greater than or equal to b.
+func (a Protocol) GE(b Protocol) bool {
+	return a.Major > b.Major ||
+		(a.Major == b.Major && a.Minor >= b.Minor)
+}
+
+func (a Protocol) is79() bool {
+	return a.GE(Protocol{7, 9})
+}
+
+// HasAttrBlockSize returns whether Attr.BlockSize is respected by the
+// kernel.
+func (a Protocol) HasAttrBlockSize() bool {
+	return a.is79()
+}
+
+// HasReadWriteFlags returns whether ReadRequest/WriteRequest
+// fields Flags and FileFlags are valid.
+func (a Protocol) HasReadWriteFlags() bool {
+	return a.is79()
+}
+
+// HasGetattrFlags returns whether GetattrRequest field Flags is
+// valid.
+func (a Protocol) HasGetattrFlags() bool {
+	return a.is79()
+}
+
+func (a Protocol) is710() bool {
+	return a.GE(Protocol{7, 10})
+}
+
+// HasOpenNonSeekable returns whether OpenResponse field Flags flag
+// OpenNonSeekable is supported.
+func (a Protocol) HasOpenNonSeekable() bool {
+	return a.is710()
+}
+
+func (a Protocol) is712() bool {
+	return a.GE(Protocol{7, 12})
+}
+
+// HasUmask returns whether CreateRequest/MkdirRequest/MknodRequest
+// field Umask is valid.
+func (a Protocol) HasUmask() bool {
+	return a.is712()
+}
+
+// HasInvalidate returns whether InvalidateNode/InvalidateEntry are
+// supported.
+func (a Protocol) HasInvalidate() bool {
+	return a.is712()
+}

weed/vendor/bazil.org/fuse/unmount.go

@@ -0,0 +1,6 @@
+package fuse
+
+// Unmount tries to unmount the filesystem mounted at dir.
+func Unmount(dir string) error {
+	return unmount(dir)
+}

weed/vendor/bazil.org/fuse/unmount_linux.go

@@ -0,0 +1,21 @@
+package fuse
+
+import (
+	"bytes"
+	"errors"
+	"os/exec"
+)
+
+func unmount(dir string) error {
+	cmd := exec.Command("fusermount", "-u", dir)
+	output, err := cmd.CombinedOutput()
+	if err != nil {
+		if len(output) > 0 {
+			output = bytes.TrimRight(output, "\n")
+			msg := err.Error() + ": " + string(output)
+			err = errors.New(msg)
+		}
+		return err
+	}
+	return nil
+}

weed/vendor/bazil.org/fuse/unmount_std.go

@@ -0,0 +1,17 @@
+// +build !linux
+
+package fuse
+
+import (
+	"os"
+	"syscall"
+)
+
+func unmount(dir string) error {
+	err := syscall.Unmount(dir, 0)
+	if err != nil {
+		err = &os.PathError{Op: "unmount", Path: dir, Err: err}
+		return err
+	}
+	return nil
+}

weed/vendor/github.com/boltdb/bolt/.gitignore

@@ -0,0 +1,4 @@
+*.prof
+*.test
+*.swp
+/bin/

weed/vendor/github.com/boltdb/bolt/LICENSE

@@ -0,0 +1,20 @@
+The MIT License (MIT)
+
+Copyright (c) 2013 Ben Johnson
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of
+this software and associated documentation files (the "Software"), to deal in
+the Software without restriction, including without limitation the rights to
+use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
+the Software, and to permit persons to whom the Software is furnished to do so,
+subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
+FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
+COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
+IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

weed/vendor/github.com/boltdb/bolt/Makefile

@@ -0,0 +1,54 @@
+TEST=.
+BENCH=.
+COVERPROFILE=/tmp/c.out
+BRANCH=`git rev-parse --abbrev-ref HEAD`
+COMMIT=`git rev-parse --short HEAD`
+GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)"
+
+default: build
+
+bench:
+	go test -v -test.run=NOTHINCONTAINSTHIS -test.bench=$(BENCH)
+
+# http://cloc.sourceforge.net/
+cloc:
+	@cloc --not-match-f='Makefile|_test.go' .
+
+cover: fmt
+	go test -coverprofile=$(COVERPROFILE) -test.run=$(TEST) $(COVERFLAG) .
+	go tool cover -html=$(COVERPROFILE)
+	rm $(COVERPROFILE)
+
+cpuprofile: fmt
+	@go test -c
+	@./bolt.test -test.v -test.run=$(TEST) -test.cpuprofile cpu.prof
+
+# go get github.com/kisielk/errcheck
+errcheck:
+	@echo "=== errcheck ==="
+	@errcheck github.com/boltdb/bolt
+
+fmt:
+	@go fmt ./...
+
+get:
+	@go get -d ./...
+
+build: get
+	@mkdir -p bin
+	@go build -ldflags=$(GOLDFLAGS) -a -o bin/bolt ./cmd/bolt
+
+test: fmt
+	@go get github.com/stretchr/testify/assert
+	@echo "=== TESTS ==="
+	@go test -v -cover -test.run=$(TEST)
+	@echo ""
+	@echo ""
+	@echo "=== CLI ==="
+	@go test -v -test.run=$(TEST) ./cmd/bolt
+	@echo ""
+	@echo ""
+	@echo "=== RACE DETECTOR ==="
+	@go test -v -race -test.run="TestSimulate_(100op|1000op)"
+
+.PHONY: bench cloc cover cpuprofile fmt memprofile test

weed/vendor/github.com/boltdb/bolt/README.md

@@ -0,0 +1,729 @@
+Bolt [![Build Status](https://drone.io/github.com/boltdb/bolt/status.png)](https://drone.io/github.com/boltdb/bolt/latest) [![Coverage Status](https://coveralls.io/repos/boltdb/bolt/badge.png?branch=master)](https://coveralls.io/r/boltdb/bolt?branch=master) [![GoDoc](https://godoc.org/github.com/boltdb/bolt?status.png)](https://godoc.org/github.com/boltdb/bolt) ![Version](http://img.shields.io/badge/version-1.0-green.png)
+====
+
+Bolt is a pure Go key/value store inspired by [Howard Chu's][hyc_symas] 
+[LMDB project][lmdb]. The goal of the project is to provide a simple,
+fast, and reliable database for projects that don't require a full database
+server such as Postgres or MySQL.
+
+Since Bolt is meant to be used as such a low-level piece of functionality,
+simplicity is key. The API will be small and only focus on getting values
+and setting values. That's it.
+
+[hyc_symas]: https://twitter.com/hyc_symas
+[lmdb]: http://symas.com/mdb/
+
+
+## Project Status
+
+Bolt is stable and the API is fixed. Full unit test coverage and randomized
+black box testing are used to ensure database consistency and thread safety.
+Bolt is currently in high-load production environments serving databases as
+large as 1TB. Many companies such as Shopify and Heroku use Bolt-backed
+services every day.
+
+
+## Getting Started
+
+### Installing
+
+To start using Bolt, install Go and run `go get`:
+
+```sh
+$ go get github.com/boltdb/bolt/...
+```
+
+This will retrieve the library and install the `bolt` command line utility into
+your `$GOBIN` path.
+
+
+### Opening a database
+
+The top-level object in Bolt is a `DB`. It is represented as a single file on
+your disk and represents a consistent snapshot of your data.
+
+To open your database, simply use the `bolt.Open()` function:
+
+```go
+package main
+
+import (
+	"log"
+
+	"github.com/boltdb/bolt"
+)
+
+func main() {
+	// Open the my.db data file in your current directory.
+	// It will be created if it doesn't exist.
+	db, err := bolt.Open("my.db", 0600, nil)
+	if err != nil {
+		log.Fatal(err)
+	}
+	defer db.Close()
+
+	...
+}
+```
+
+Please note that Bolt obtains a file lock on the data file so multiple processes
+cannot open the same database at the same time. Opening an already open Bolt
+database will cause it to hang until the other process closes it. To prevent
+an indefinite wait you can pass a timeout option to the `Open()` function:
+
+```go
+db, err := bolt.Open("my.db", 0600, &bolt.Options{Timeout: 1 * time.Second})
+```
+
+
+### Transactions
+
+Bolt allows only one read-write transaction at a time but allows as many
+read-only transactions as you want at a time. Each transaction has a consistent
+view of the data as it existed when the transaction started.
+
+Individual transactions and all objects created from them (e.g. buckets, keys)
+are not thread safe. To work with data in multiple goroutines you must start
+a transaction for each one or use locking to ensure only one goroutine accesses
+a transaction at a time. Creating transaction from the `DB` is thread safe.
+
+Read-only transactions and read-write transactions should not depend on one
+another and generally shouldn't be opened simultaneously in the same goroutine.
+This can cause a deadlock as the read-write transaction needs to periodically
+re-map the data file but it cannot do so while a read-only transaction is open.
+
+
+#### Read-write transactions
+
+To start a read-write transaction, you can use the `DB.Update()` function:
+
+```go
+err := db.Update(func(tx *bolt.Tx) error {
+	...
+	return nil
+})
+```
+
+Inside the closure, you have a consistent view of the database. You commit the
+transaction by returning `nil` at the end. You can also rollback the transaction
+at any point by returning an error. All database operations are allowed inside
+a read-write transaction.
+
+Always check the return error as it will report any disk failures that can cause
+your transaction to not complete. If you return an error within your closure
+it will be passed through.
+
+
+#### Read-only transactions
+
+To start a read-only transaction, you can use the `DB.View()` function:
+
+```go
+err := db.View(func(tx *bolt.Tx) error {
+	...
+	return nil
+})
+```
+
+You also get a consistent view of the database within this closure, however,
+no mutating operations are allowed within a read-only transaction. You can only
+retrieve buckets, retrieve values, and copy the database within a read-only
+transaction.
+
+
+#### Batch read-write transactions
+
+Each `DB.Update()` waits for disk to commit the writes. This overhead
+can be minimized by combining multiple updates with the `DB.Batch()`
+function:
+
+```go
+err := db.Batch(func(tx *bolt.Tx) error {
+	...
+	return nil
+})
+```
+
+Concurrent Batch calls are opportunistically combined into larger
+transactions. Batch is only useful when there are multiple goroutines
+calling it.
+
+The trade-off is that `Batch` can call the given
+function multiple times, if parts of the transaction fail. The
+function must be idempotent and side effects must take effect only
+after a successful return from `DB.Batch()`.
+
+For example: don't display messages from inside the function, instead
+set variables in the enclosing scope:
+
+```go
+var id uint64
+err := db.Batch(func(tx *bolt.Tx) error {
+	// Find last key in bucket, decode as bigendian uint64, increment
+	// by one, encode back to []byte, and add new key.
+	...
+	id = newValue
+	return nil
+})
+if err != nil {
+	return ...
+}
+fmt.Println("Allocated ID %d", id)
+```
+
+
+#### Managing transactions manually
+
+The `DB.View()` and `DB.Update()` functions are wrappers around the `DB.Begin()`
+function. These helper functions will start the transaction, execute a function,
+and then safely close your transaction if an error is returned. This is the
+recommended way to use Bolt transactions.
+
+However, sometimes you may want to manually start and end your transactions.
+You can use the `Tx.Begin()` function directly but **please** be sure to close
+the transaction.
+
+```go
+// Start a writable transaction.
+tx, err := db.Begin(true)
+if err != nil {
+    return err
+}
+defer tx.Rollback()
+
+// Use the transaction...
+_, err := tx.CreateBucket([]byte("MyBucket"))
+if err != nil {
+    return err
+}
+
+// Commit the transaction and check for error.
+if err := tx.Commit(); err != nil {
+    return err
+}
+```
+
+The first argument to `DB.Begin()` is a boolean stating if the transaction
+should be writable.
+
+
+### Using buckets
+
+Buckets are collections of key/value pairs within the database. All keys in a
+bucket must be unique. You can create a bucket using the `DB.CreateBucket()`
+function:
+
+```go
+db.Update(func(tx *bolt.Tx) error {
+	b, err := tx.CreateBucket([]byte("MyBucket"))
+	if err != nil {
+		return fmt.Errorf("create bucket: %s", err)
+	}
+	return nil
+})
+```
+
+You can also create a bucket only if it doesn't exist by using the
+`Tx.CreateBucketIfNotExists()` function. It's a common pattern to call this
+function for all your top-level buckets after you open your database so you can
+guarantee that they exist for future transactions.
+
+To delete a bucket, simply call the `Tx.DeleteBucket()` function.
+
+
+### Using key/value pairs
+
+To save a key/value pair to a bucket, use the `Bucket.Put()` function:
+
+```go
+db.Update(func(tx *bolt.Tx) error {
+	b := tx.Bucket([]byte("MyBucket"))
+	err := b.Put([]byte("answer"), []byte("42"))
+	return err
+})
+```
+
+This will set the value of the `"answer"` key to `"42"` in the `MyBucket`
+bucket. To retrieve this value, we can use the `Bucket.Get()` function:
+
+```go
+db.View(func(tx *bolt.Tx) error {
+	b := tx.Bucket([]byte("MyBucket"))
+	v := b.Get([]byte("answer"))
+	fmt.Printf("The answer is: %s\n", v)
+	return nil
+})
+```
+
+The `Get()` function does not return an error because its operation is
+guaranteed to work (unless there is some kind of system failure). If the key
+exists then it will return its byte slice value. If it doesn't exist then it
+will return `nil`. It's important to note that you can have a zero-length value
+set to a key which is different than the key not existing.
+
+Use the `Bucket.Delete()` function to delete a key from the bucket.
+
+Please note that values returned from `Get()` are only valid while the
+transaction is open. If you need to use a value outside of the transaction
+then you must use `copy()` to copy it to another byte slice.
+
+
+### Autoincrementing integer for the bucket
+By using the NextSequence() function, you can let Bolt determine a sequence
+which can be used as the unique identifier for your key/value pairs. See the
+example below.
+
+```go
+// CreateUser saves u to the store. The new user ID is set on u once the data is persisted.
+func (s *Store) CreateUser(u *User) error {
+    return s.db.Update(func(tx *bolt.Tx) error {
+        // Retrieve the users bucket.
+        // This should be created when the DB is first opened.
+        b := tx.Bucket([]byte("users"))
+
+        // Generate ID for the user.
+        // This returns an error only if the Tx is closed or not writeable.
+        // That can't happen in an Update() call so I ignore the error check.
+        id, _ = b.NextSequence()
+        u.ID = int(id)
+
+        // Marshal user data into bytes.
+        buf, err := json.Marshal(u)
+        if err != nil {
+            return err
+        }
+
+        // Persist bytes to users bucket.
+        return b.Put(itob(u.ID), buf)
+    })
+}
+
+// itob returns an 8-byte big endian representation of v.
+func itob(v int) []byte {
+    b := make([]byte, 8)
+    binary.BigEndian.PutUint64(b, uint64(v))
+    return b
+}
+
+type User struct {
+    ID int
+    ...
+}
+
+```
+
+### Iterating over keys
+
+Bolt stores its keys in byte-sorted order within a bucket. This makes sequential
+iteration over these keys extremely fast. To iterate over keys we'll use a
+`Cursor`:
+
+```go
+db.View(func(tx *bolt.Tx) error {
+	b := tx.Bucket([]byte("MyBucket"))
+	c := b.Cursor()
+
+	for k, v := c.First(); k != nil; k, v = c.Next() {
+		fmt.Printf("key=%s, value=%s\n", k, v)
+	}
+
+	return nil
+})
+```
+
+The cursor allows you to move to a specific point in the list of keys and move
+forward or backward through the keys one at a time.
+
+The following functions are available on the cursor:
+
+```
+First()  Move to the first key.
+Last()   Move to the last key.
+Seek()   Move to a specific key.
+Next()   Move to the next key.
+Prev()   Move to the previous key.
+```
+
+Each of those functions has a return signature of `(key []byte, value []byte)`.
+When you have iterated to the end of the cursor then `Next()` will return a
+`nil` key.  You must seek to a position using `First()`, `Last()`, or `Seek()`
+before calling `Next()` or `Prev()`. If you do not seek to a position then
+these functions will return a `nil` key.
+
+During iteration, if the key is non-`nil` but the value is `nil`, that means
+the key refers to a bucket rather than a value.  Use `Bucket.Bucket()` to
+access the sub-bucket.
+
+
+#### Prefix scans
+
+To iterate over a key prefix, you can combine `Seek()` and `bytes.HasPrefix()`:
+
+```go
+db.View(func(tx *bolt.Tx) error {
+	c := tx.Bucket([]byte("MyBucket")).Cursor()
+
+	prefix := []byte("1234")
+	for k, v := c.Seek(prefix); bytes.HasPrefix(k, prefix); k, v = c.Next() {
+		fmt.Printf("key=%s, value=%s\n", k, v)
+	}
+
+	return nil
+})
+```
+
+#### Range scans
+
+Another common use case is scanning over a range such as a time range. If you
+use a sortable time encoding such as RFC3339 then you can query a specific
+date range like this:
+
+```go
+db.View(func(tx *bolt.Tx) error {
+	// Assume our events bucket has RFC3339 encoded time keys.
+	c := tx.Bucket([]byte("Events")).Cursor()
+
+	// Our time range spans the 90's decade.
+	min := []byte("1990-01-01T00:00:00Z")
+	max := []byte("2000-01-01T00:00:00Z")
+
+	// Iterate over the 90's.
+	for k, v := c.Seek(min); k != nil && bytes.Compare(k, max) <= 0; k, v = c.Next() {
+		fmt.Printf("%s: %s\n", k, v)
+	}
+
+	return nil
+})
+```
+
+
+#### ForEach()
+
+You can also use the function `ForEach()` if you know you'll be iterating over
+all the keys in a bucket:
+
+```go
+db.View(func(tx *bolt.Tx) error {
+	b := tx.Bucket([]byte("MyBucket"))
+	b.ForEach(func(k, v []byte) error {
+		fmt.Printf("key=%s, value=%s\n", k, v)
+		return nil
+	})
+	return nil
+})
+```
+
+
+### Nested buckets
+
+You can also store a bucket in a key to create nested buckets. The API is the
+same as the bucket management API on the `DB` object:
+
+```go
+func (*Bucket) CreateBucket(key []byte) (*Bucket, error)
+func (*Bucket) CreateBucketIfNotExists(key []byte) (*Bucket, error)
+func (*Bucket) DeleteBucket(key []byte) error
+```
+
+
+### Database backups
+
+Bolt is a single file so it's easy to backup. You can use the `Tx.WriteTo()`
+function to write a consistent view of the database to a writer. If you call
+this from a read-only transaction, it will perform a hot backup and not block
+your other database reads and writes.
+
+By default, it will use a regular file handle which will utilize the operating
+system's page cache. See the [`Tx`](https://godoc.org/github.com/boltdb/bolt#Tx)
+documentation for information about optimizing for larger-than-RAM datasets. 
+
+One common use case is to backup over HTTP so you can use tools like `cURL` to
+do database backups:
+
+```go
+func BackupHandleFunc(w http.ResponseWriter, req *http.Request) {
+	err := db.View(func(tx *bolt.Tx) error {
+		w.Header().Set("Content-Type", "application/octet-stream")
+		w.Header().Set("Content-Disposition", `attachment; filename="my.db"`)
+		w.Header().Set("Content-Length", strconv.Itoa(int(tx.Size())))
+		_, err := tx.WriteTo(w)
+		return err
+	})
+	if err != nil {
+		http.Error(w, err.Error(), http.StatusInternalServerError)
+	}
+}
+```
+
+Then you can backup using this command:
+
+```sh
+$ curl http://localhost/backup > my.db
+```
+
+Or you can open your browser to `http://localhost/backup` and it will download
+automatically.
+
+If you want to backup to another file you can use the `Tx.CopyFile()` helper
+function.
+
+
+### Statistics
+
+The database keeps a running count of many of the internal operations it
+performs so you can better understand what's going on. By grabbing a snapshot
+of these stats at two points in time we can see what operations were performed
+in that time range.
+
+For example, we could start a goroutine to log stats every 10 seconds:
+
+```go
+go func() {
+	// Grab the initial stats.
+	prev := db.Stats()
+
+	for {
+		// Wait for 10s.
+		time.Sleep(10 * time.Second)
+
+		// Grab the current stats and diff them.
+		stats := db.Stats()
+		diff := stats.Sub(&prev)
+
+		// Encode stats to JSON and print to STDERR.
+		json.NewEncoder(os.Stderr).Encode(diff)
+
+		// Save stats for the next loop.
+		prev = stats
+	}
+}()
+```
+
+It's also useful to pipe these stats to a service such as statsd for monitoring
+or to provide an HTTP endpoint that will perform a fixed-length sample.
+
+
+### Read-Only Mode
+
+Sometimes it is useful to create a shared, read-only Bolt database. To this,
+set the `Options.ReadOnly` flag when opening your database. Read-only mode
+uses a shared lock to allow multiple processes to read from the database but
+it will block any processes from opening the database in read-write mode.
+
+```go
+db, err := bolt.Open("my.db", 0666, &bolt.Options{ReadOnly: true})
+if err != nil {
+	log.Fatal(err)
+}
+```
+
+
+## Resources
+
+For more information on getting started with Bolt, check out the following articles:
+
+* [Intro to BoltDB: Painless Performant Persistence](http://npf.io/2014/07/intro-to-boltdb-painless-performant-persistence/) by [Nate Finch](https://github.com/natefinch).
+* [Bolt -- an embedded key/value database for Go](https://www.progville.com/go/bolt-embedded-db-golang/) by Progville
+
+
+## Comparison with other databases
+
+### Postgres, MySQL, & other relational databases
+
+Relational databases structure data into rows and are only accessible through
+the use of SQL. This approach provides flexibility in how you store and query
+your data but also incurs overhead in parsing and planning SQL statements. Bolt
+accesses all data by a byte slice key. This makes Bolt fast to read and write
+data by key but provides no built-in support for joining values together.
+
+Most relational databases (with the exception of SQLite) are standalone servers
+that run separately from your application. This gives your systems
+flexibility to connect multiple application servers to a single database
+server but also adds overhead in serializing and transporting data over the
+network. Bolt runs as a library included in your application so all data access
+has to go through your application's process. This brings data closer to your
+application but limits multi-process access to the data.
+
+
+### LevelDB, RocksDB
+
+LevelDB and its derivatives (RocksDB, HyperLevelDB) are similar to Bolt in that
+they are libraries bundled into the application, however, their underlying
+structure is a log-structured merge-tree (LSM tree). An LSM tree optimizes
+random writes by using a write ahead log and multi-tiered, sorted files called
+SSTables. Bolt uses a B+tree internally and only a single file. Both approaches
+have trade-offs.
+
+If you require a high random write throughput (>10,000 w/sec) or you need to use
+spinning disks then LevelDB could be a good choice. If your application is
+read-heavy or does a lot of range scans then Bolt could be a good choice.
+
+One other important consideration is that LevelDB does not have transactions.
+It supports batch writing of key/values pairs and it supports read snapshots
+but it will not give you the ability to do a compare-and-swap operation safely.
+Bolt supports fully serializable ACID transactions.
+
+
+### LMDB
+
+Bolt was originally a port of LMDB so it is architecturally similar. Both use
+a B+tree, have ACID semantics with fully serializable transactions, and support
+lock-free MVCC using a single writer and multiple readers.
+
+The two projects have somewhat diverged. LMDB heavily focuses on raw performance
+while Bolt has focused on simplicity and ease of use. For example, LMDB allows
+several unsafe actions such as direct writes for the sake of performance. Bolt
+opts to disallow actions which can leave the database in a corrupted state. The
+only exception to this in Bolt is `DB.NoSync`.
+
+There are also a few differences in API. LMDB requires a maximum mmap size when
+opening an `mdb_env` whereas Bolt will handle incremental mmap resizing
+automatically. LMDB overloads the getter and setter functions with multiple
+flags whereas Bolt splits these specialized cases into their own functions.
+
+
+## Caveats & Limitations
+
+It's important to pick the right tool for the job and Bolt is no exception.
+Here are a few things to note when evaluating and using Bolt:
+
+* Bolt is good for read intensive workloads. Sequential write performance is
+  also fast but random writes can be slow. You can add a write-ahead log or
+  [transaction coalescer](https://github.com/boltdb/coalescer) in front of Bolt
+  to mitigate this issue.
+
+* Bolt uses a B+tree internally so there can be a lot of random page access.
+  SSDs provide a significant performance boost over spinning disks.
+
+* Try to avoid long running read transactions. Bolt uses copy-on-write so
+  old pages cannot be reclaimed while an old transaction is using them.
+
+* Byte slices returned from Bolt are only valid during a transaction. Once the
+  transaction has been committed or rolled back then the memory they point to
+  can be reused by a new page or can be unmapped from virtual memory and you'll
+  see an `unexpected fault address` panic when accessing it.
+
+* Be careful when using `Bucket.FillPercent`. Setting a high fill percent for
+  buckets that have random inserts will cause your database to have very poor
+  page utilization.
+
+* Use larger buckets in general. Smaller buckets causes poor page utilization
+  once they become larger than the page size (typically 4KB).
+
+* Bulk loading a lot of random writes into a new bucket can be slow as the
+  page will not split until the transaction is committed. Randomly inserting
+  more than 100,000 key/value pairs into a single new bucket in a single
+  transaction is not advised.
+
+* Bolt uses a memory-mapped file so the underlying operating system handles the
+  caching of the data. Typically, the OS will cache as much of the file as it
+  can in memory and will release memory as needed to other processes. This means
+  that Bolt can show very high memory usage when working with large databases.
+  However, this is expected and the OS will release memory as needed. Bolt can
+  handle databases much larger than the available physical RAM, provided its
+  memory-map fits in the process virtual address space. It may be problematic
+  on 32-bits systems.
+
+* The data structures in the Bolt database are memory mapped so the data file
+  will be endian specific. This means that you cannot copy a Bolt file from a
+  little endian machine to a big endian machine and have it work. For most 
+  users this is not a concern since most modern CPUs are little endian.
+
+* Because of the way pages are laid out on disk, Bolt cannot truncate data files
+  and return free pages back to the disk. Instead, Bolt maintains a free list
+  of unused pages within its data file. These free pages can be reused by later
+  transactions. This works well for many use cases as databases generally tend
+  to grow. However, it's important to note that deleting large chunks of data
+  will not allow you to reclaim that space on disk.
+
+  For more information on page allocation, [see this comment][page-allocation].
+
+[page-allocation]: https://github.com/boltdb/bolt/issues/308#issuecomment-74811638
+
+
+## Reading the Source
+
+Bolt is a relatively small code base (<3KLOC) for an embedded, serializable,
+transactional key/value database so it can be a good starting point for people
+interested in how databases work.
+
+The best places to start are the main entry points into Bolt:
+
+- `Open()` - Initializes the reference to the database. It's responsible for
+  creating the database if it doesn't exist, obtaining an exclusive lock on the
+  file, reading the meta pages, & memory-mapping the file.
+
+- `DB.Begin()` - Starts a read-only or read-write transaction depending on the
+  value of the `writable` argument. This requires briefly obtaining the "meta"
+  lock to keep track of open transactions. Only one read-write transaction can
+  exist at a time so the "rwlock" is acquired during the life of a read-write
+  transaction.
+
+- `Bucket.Put()` - Writes a key/value pair into a bucket. After validating the
+  arguments, a cursor is used to traverse the B+tree to the page and position
+  where they key & value will be written. Once the position is found, the bucket
+  materializes the underlying page and the page's parent pages into memory as
+  "nodes". These nodes are where mutations occur during read-write transactions.
+  These changes get flushed to disk during commit.
+
+- `Bucket.Get()` - Retrieves a key/value pair from a bucket. This uses a cursor
+  to move to the page & position of a key/value pair. During a read-only
+  transaction, the key and value data is returned as a direct reference to the
+  underlying mmap file so there's no allocation overhead. For read-write
+  transactions, this data may reference the mmap file or one of the in-memory
+  node values.
+
+- `Cursor` - This object is simply for traversing the B+tree of on-disk pages
+  or in-memory nodes. It can seek to a specific key, move to the first or last
+  value, or it can move forward or backward. The cursor handles the movement up
+  and down the B+tree transparently to the end user.
+
+- `Tx.Commit()` - Converts the in-memory dirty nodes and the list of free pages
+  into pages to be written to disk. Writing to disk then occurs in two phases.
+  First, the dirty pages are written to disk and an `fsync()` occurs. Second, a
+  new meta page with an incremented transaction ID is written and another
+  `fsync()` occurs. This two phase write ensures that partially written data
+  pages are ignored in the event of a crash since the meta page pointing to them
+  is never written. Partially written meta pages are invalidated because they
+  are written with a checksum.
+
+If you have additional notes that could be helpful for others, please submit
+them via pull request.
+
+
+## Other Projects Using Bolt
+
+Below is a list of public, open source projects that use Bolt:
+
+* [Operation Go: A Routine Mission](http://gocode.io) - An online programming game for Golang using Bolt for user accounts and a leaderboard.
+* [Bazil](https://bazil.org/) - A file system that lets your data reside where it is most convenient for it to reside.
+* [DVID](https://github.com/janelia-flyem/dvid) - Added Bolt as optional storage engine and testing it against Basho-tuned leveldb.
+* [Skybox Analytics](https://github.com/skybox/skybox) - A standalone funnel analysis tool for web analytics.
+* [Scuttlebutt](https://github.com/benbjohnson/scuttlebutt) - Uses Bolt to store and process all Twitter mentions of GitHub projects.
+* [Wiki](https://github.com/peterhellberg/wiki) - A tiny wiki using Goji, BoltDB and Blackfriday.
+* [ChainStore](https://github.com/nulayer/chainstore) - Simple key-value interface to a variety of storage engines organized as a chain of operations.
+* [MetricBase](https://github.com/msiebuhr/MetricBase) - Single-binary version of Graphite.
+* [Gitchain](https://github.com/gitchain/gitchain) - Decentralized, peer-to-peer Git repositories aka "Git meets Bitcoin".
+* [event-shuttle](https://github.com/sclasen/event-shuttle) - A Unix system service to collect and reliably deliver messages to Kafka.
+* [ipxed](https://github.com/kelseyhightower/ipxed) - Web interface and api for ipxed.
+* [BoltStore](https://github.com/yosssi/boltstore) - Session store using Bolt.
+* [photosite/session](http://godoc.org/bitbucket.org/kardianos/photosite/session) - Sessions for a photo viewing site.
+* [LedisDB](https://github.com/siddontang/ledisdb) - A high performance NoSQL, using Bolt as optional storage.
+* [ipLocator](https://github.com/AndreasBriese/ipLocator) - A fast ip-geo-location-server using bolt with bloom filters.
+* [cayley](https://github.com/google/cayley) - Cayley is an open-source graph database using Bolt as optional backend.
+* [bleve](http://www.blevesearch.com/) - A pure Go search engine similar to ElasticSearch that uses Bolt as the default storage backend.
+* [tentacool](https://github.com/optiflows/tentacool) - REST api server to manage system stuff (IP, DNS, Gateway...) on a linux server.
+* [SkyDB](https://github.com/skydb/sky) - Behavioral analytics database.
+* [Seaweed File System](https://github.com/chrislusf/weed-fs) - Highly scalable distributed key~file system with O(1) disk read.
+* [InfluxDB](http://influxdb.com) - Scalable datastore for metrics, events, and real-time analytics.
+* [Freehold](http://tshannon.bitbucket.org/freehold/) - An open, secure, and lightweight platform for your files and data.
+* [Prometheus Annotation Server](https://github.com/oliver006/prom_annotation_server) - Annotation server for PromDash & Prometheus service monitoring system.
+* [Consul](https://github.com/hashicorp/consul) - Consul is service discovery and configuration made easy. Distributed, highly available, and datacenter-aware.
+* [Kala](https://github.com/ajvb/kala) - Kala is a modern job scheduler optimized to run on a single node. It is persistent, JSON over HTTP API, ISO 8601 duration notation, and dependent jobs.
+* [drive](https://github.com/odeke-em/drive) - drive is an unofficial Google Drive command line client for \*NIX operating systems.
+* [stow](https://github.com/djherbis/stow) -  a persistence manager for objects
+  backed by boltdb.
+* [buckets](https://github.com/joyrexus/buckets) - a bolt wrapper streamlining
+  simple tx and key scans.
+
+If you are using Bolt in a project please send a pull request to add it to the list.

weed/vendor/github.com/boltdb/bolt/batch.go

@@ -0,0 +1,138 @@
+package bolt
+
+import (
+	"errors"
+	"fmt"
+	"sync"
+	"time"
+)
+
+// Batch calls fn as part of a batch. It behaves similar to Update,
+// except:
+//
+// 1. concurrent Batch calls can be combined into a single Bolt
+// transaction.
+//
+// 2. the function passed to Batch may be called multiple times,
+// regardless of whether it returns error or not.
+//
+// This means that Batch function side effects must be idempotent and
+// take permanent effect only after a successful return is seen in
+// caller.
+//
+// The maximum batch size and delay can be adjusted with DB.MaxBatchSize
+// and DB.MaxBatchDelay, respectively.
+//
+// Batch is only useful when there are multiple goroutines calling it.
+func (db *DB) Batch(fn func(*Tx) error) error {
+	errCh := make(chan error, 1)
+
+	db.batchMu.Lock()
+	if (db.batch == nil) || (db.batch != nil && len(db.batch.calls) >= db.MaxBatchSize) {
+		// There is no existing batch, or the existing batch is full; start a new one.
+		db.batch = &batch{
+			db: db,
+		}
+		db.batch.timer = time.AfterFunc(db.MaxBatchDelay, db.batch.trigger)
+	}
+	db.batch.calls = append(db.batch.calls, call{fn: fn, err: errCh})
+	if len(db.batch.calls) >= db.MaxBatchSize {
+		// wake up batch, it's ready to run
+		go db.batch.trigger()
+	}
+	db.batchMu.Unlock()
+
+	err := <-errCh
+	if err == trySolo {
+		err = db.Update(fn)
+	}
+	return err
+}
+
+type call struct {
+	fn  func(*Tx) error
+	err chan<- error
+}
+
+type batch struct {
+	db    *DB
+	timer *time.Timer
+	start sync.Once
+	calls []call
+}
+
+// trigger runs the batch if it hasn't already been run.
+func (b *batch) trigger() {
+	b.start.Do(b.run)
+}
+
+// run performs the transactions in the batch and communicates results
+// back to DB.Batch.
+func (b *batch) run() {
+	b.db.batchMu.Lock()
+	b.timer.Stop()
+	// Make sure no new work is added to this batch, but don't break
+	// other batches.
+	if b.db.batch == b {
+		b.db.batch = nil
+	}
+	b.db.batchMu.Unlock()
+
+retry:
+	for len(b.calls) > 0 {
+		var failIdx = -1
+		err := b.db.Update(func(tx *Tx) error {
+			for i, c := range b.calls {
+				if err := safelyCall(c.fn, tx); err != nil {
+					failIdx = i
+					return err
+				}
+			}
+			return nil
+		})
+
+		if failIdx >= 0 {
+			// take the failing transaction out of the batch. it's
+			// safe to shorten b.calls here because db.batch no longer
+			// points to us, and we hold the mutex anyway.
+			c := b.calls[failIdx]
+			b.calls[failIdx], b.calls = b.calls[len(b.calls)-1], b.calls[:len(b.calls)-1]
+			// tell the submitter re-run it solo, continue with the rest of the batch
+			c.err <- trySolo
+			continue retry
+		}
+
+		// pass success, or bolt internal errors, to all callers
+		for _, c := range b.calls {
+			if c.err != nil {
+				c.err <- err
+			}
+		}
+		break retry
+	}
+}
+
+// trySolo is a special sentinel error value used for signaling that a
+// transaction function should be re-run. It should never be seen by
+// callers.
+var trySolo = errors.New("batch function returned an error and should be re-run solo")
+
+type panicked struct {
+	reason interface{}
+}
+
+func (p panicked) Error() string {
+	if err, ok := p.reason.(error); ok {
+		return err.Error()
+	}
+	return fmt.Sprintf("panic: %v", p.reason)
+}
+
+func safelyCall(fn func(*Tx) error, tx *Tx) (err error) {
+	defer func() {
+		if p := recover(); p != nil {
+			err = panicked{p}
+		}
+	}()
+	return fn(tx)
+}

weed/vendor/github.com/boltdb/bolt/bolt_386.go

@@ -0,0 +1,7 @@
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x7FFFFFFF // 2GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_amd64.go

@@ -0,0 +1,7 @@
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_arm.go

@@ -0,0 +1,7 @@
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x7FFFFFFF // 2GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_arm64.go

@@ -0,0 +1,9 @@
+// +build arm64
+
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_linux.go

@@ -0,0 +1,10 @@
+package bolt
+
+import (
+	"syscall"
+)
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return syscall.Fdatasync(int(db.file.Fd()))
+}

weed/vendor/github.com/boltdb/bolt/bolt_openbsd.go

@@ -0,0 +1,27 @@
+package bolt
+
+import (
+	"syscall"
+	"unsafe"
+)
+
+const (
+	msAsync      = 1 << iota // perform asynchronous writes
+	msSync                   // perform synchronous writes
+	msInvalidate             // invalidate cached data
+)
+
+func msync(db *DB) error {
+	_, _, errno := syscall.Syscall(syscall.SYS_MSYNC, uintptr(unsafe.Pointer(db.data)), uintptr(db.datasz), msInvalidate)
+	if errno != 0 {
+		return errno
+	}
+	return nil
+}
+
+func fdatasync(db *DB) error {
+	if db.data != nil {
+		return msync(db)
+	}
+	return db.file.Sync()
+}

weed/vendor/github.com/boltdb/bolt/bolt_ppc64le.go

@@ -0,0 +1,9 @@
+// +build ppc64le
+
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_s390x.go

@@ -0,0 +1,9 @@
+// +build s390x
+
+package bolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

weed/vendor/github.com/boltdb/bolt/bolt_unix.go

@@ -0,0 +1,100 @@
+// +build !windows,!plan9,!solaris
+
+package bolt
+
+import (
+	"fmt"
+	"os"
+	"syscall"
+	"time"
+	"unsafe"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(f *os.File, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	for {
+		// If we're beyond our timeout then return an error.
+		// This can only occur after we've attempted a flock once.
+		if t.IsZero() {
+			t = time.Now()
+		} else if timeout > 0 && time.Since(t) > timeout {
+			return ErrTimeout
+		}
+		flag := syscall.LOCK_SH
+		if exclusive {
+			flag = syscall.LOCK_EX
+		}
+
+		// Otherwise attempt to obtain an exclusive lock.
+		err := syscall.Flock(int(f.Fd()), flag|syscall.LOCK_NB)
+		if err == nil {
+			return nil
+		} else if err != syscall.EWOULDBLOCK {
+			return err
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(50 * time.Millisecond)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(f *os.File) error {
+	return syscall.Flock(int(f.Fd()), syscall.LOCK_UN)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Truncate and fsync to ensure file size metadata is flushed.
+	// https://github.com/boltdb/bolt/issues/284
+	if !db.NoGrowSync && !db.readOnly {
+		if err := db.file.Truncate(int64(sz)); err != nil {
+			return fmt.Errorf("file resize error: %s", err)
+		}
+		if err := db.file.Sync(); err != nil {
+			return fmt.Errorf("file sync error: %s", err)
+		}
+	}
+
+	// Map the data file to memory.
+	b, err := syscall.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	if err := madvise(b, syscall.MADV_RANDOM); err != nil {
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := syscall.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}
+
+// NOTE: This function is copied from stdlib because it is not available on darwin.
+func madvise(b []byte, advice int) (err error) {
+	_, _, e1 := syscall.Syscall(syscall.SYS_MADVISE, uintptr(unsafe.Pointer(&b[0])), uintptr(len(b)), uintptr(advice))
+	if e1 != 0 {
+		err = e1
+	}
+	return
+}

weed/vendor/github.com/boltdb/bolt/bolt_unix_solaris.go

@@ -0,0 +1,101 @@
+package bolt
+
+import (
+	"fmt"
+	"os"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(f *os.File, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	for {
+		// If we're beyond our timeout then return an error.
+		// This can only occur after we've attempted a flock once.
+		if t.IsZero() {
+			t = time.Now()
+		} else if timeout > 0 && time.Since(t) > timeout {
+			return ErrTimeout
+		}
+		var lock syscall.Flock_t
+		lock.Start = 0
+		lock.Len = 0
+		lock.Pid = 0
+		lock.Whence = 0
+		lock.Pid = 0
+		if exclusive {
+			lock.Type = syscall.F_WRLCK
+		} else {
+			lock.Type = syscall.F_RDLCK
+		}
+		err := syscall.FcntlFlock(f.Fd(), syscall.F_SETLK, &lock)
+		if err == nil {
+			return nil
+		} else if err != syscall.EAGAIN {
+			return err
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(50 * time.Millisecond)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(f *os.File) error {
+	var lock syscall.Flock_t
+	lock.Start = 0
+	lock.Len = 0
+	lock.Type = syscall.F_UNLCK
+	lock.Whence = 0
+	return syscall.FcntlFlock(uintptr(f.Fd()), syscall.F_SETLK, &lock)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Truncate and fsync to ensure file size metadata is flushed.
+	// https://github.com/boltdb/bolt/issues/284
+	if !db.NoGrowSync && !db.readOnly {
+		if err := db.file.Truncate(int64(sz)); err != nil {
+			return fmt.Errorf("file resize error: %s", err)
+		}
+		if err := db.file.Sync(); err != nil {
+			return fmt.Errorf("file sync error: %s", err)
+		}
+	}
+
+	// Map the data file to memory.
+	b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	if err := unix.Madvise(b, syscall.MADV_RANDOM); err != nil {
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := unix.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}

weed/vendor/github.com/boltdb/bolt/bolt_windows.go

@@ -0,0 +1,130 @@
+package bolt
+
+import (
+	"fmt"
+	"os"
+	"syscall"
+	"time"
+	"unsafe"
+)
+
+// LockFileEx code derived from golang build filemutex_windows.go @ v1.5.1
+var (
+	modkernel32      = syscall.NewLazyDLL("kernel32.dll")
+	procLockFileEx   = modkernel32.NewProc("LockFileEx")
+	procUnlockFileEx = modkernel32.NewProc("UnlockFileEx")
+)
+
+const (
+	// see https://msdn.microsoft.com/en-us/library/windows/desktop/aa365203(v=vs.85).aspx
+	flagLockExclusive       = 2
+	flagLockFailImmediately = 1
+
+	// see https://msdn.microsoft.com/en-us/library/windows/desktop/ms681382(v=vs.85).aspx
+	errLockViolation syscall.Errno = 0x21
+)
+
+func lockFileEx(h syscall.Handle, flags, reserved, locklow, lockhigh uint32, ol *syscall.Overlapped) (err error) {
+	r, _, err := procLockFileEx.Call(uintptr(h), uintptr(flags), uintptr(reserved), uintptr(locklow), uintptr(lockhigh), uintptr(unsafe.Pointer(ol)))
+	if r == 0 {
+		return err
+	}
+	return nil
+}
+
+func unlockFileEx(h syscall.Handle, reserved, locklow, lockhigh uint32, ol *syscall.Overlapped) (err error) {
+	r, _, err := procUnlockFileEx.Call(uintptr(h), uintptr(reserved), uintptr(locklow), uintptr(lockhigh), uintptr(unsafe.Pointer(ol)), 0)
+	if r == 0 {
+		return err
+	}
+	return nil
+}
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return db.file.Sync()
+}
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(f *os.File, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	for {
+		// If we're beyond our timeout then return an error.
+		// This can only occur after we've attempted a flock once.
+		if t.IsZero() {
+			t = time.Now()
+		} else if timeout > 0 && time.Since(t) > timeout {
+			return ErrTimeout
+		}
+
+		var flag uint32 = flagLockFailImmediately
+		if exclusive {
+			flag |= flagLockExclusive
+		}
+
+		err := lockFileEx(syscall.Handle(f.Fd()), flag, 0, 1, 0, &syscall.Overlapped{})
+		if err == nil {
+			return nil
+		} else if err != errLockViolation {
+			return err
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(50 * time.Millisecond)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(f *os.File) error {
+	return unlockFileEx(syscall.Handle(f.Fd()), 0, 1, 0, &syscall.Overlapped{})
+}
+
+// mmap memory maps a DB's data file.
+// Based on: https://github.com/edsrzf/mmap-go
+func mmap(db *DB, sz int) error {
+	if !db.readOnly {
+		// Truncate the database to the size of the mmap.
+		if err := db.file.Truncate(int64(sz)); err != nil {
+			return fmt.Errorf("truncate: %s", err)
+		}
+	}
+
+	// Open a file mapping handle.
+	sizelo := uint32(sz >> 32)
+	sizehi := uint32(sz) & 0xffffffff
+	h, errno := syscall.CreateFileMapping(syscall.Handle(db.file.Fd()), nil, syscall.PAGE_READONLY, sizelo, sizehi, nil)
+	if h == 0 {
+		return os.NewSyscallError("CreateFileMapping", errno)
+	}
+
+	// Create the memory map.
+	addr, errno := syscall.MapViewOfFile(h, syscall.FILE_MAP_READ, 0, 0, uintptr(sz))
+	if addr == 0 {
+		return os.NewSyscallError("MapViewOfFile", errno)
+	}
+
+	// Close mapping handle.
+	if err := syscall.CloseHandle(syscall.Handle(h)); err != nil {
+		return os.NewSyscallError("CloseHandle", err)
+	}
+
+	// Convert to a byte array.
+	db.data = ((*[maxMapSize]byte)(unsafe.Pointer(addr)))
+	db.datasz = sz
+
+	return nil
+}
+
+// munmap unmaps a pointer from a file.
+// Based on: https://github.com/edsrzf/mmap-go
+func munmap(db *DB) error {
+	if db.data == nil {
+		return nil
+	}
+
+	addr := (uintptr)(unsafe.Pointer(&db.data[0]))
+	if err := syscall.UnmapViewOfFile(addr); err != nil {
+		return os.NewSyscallError("UnmapViewOfFile", err)
+	}
+	return nil
+}

weed/vendor/github.com/boltdb/bolt/boltsync_unix.go

@@ -0,0 +1,8 @@
+// +build !windows,!plan9,!linux,!openbsd
+
+package bolt
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return db.file.Sync()
+}

weed/vendor/github.com/boltdb/bolt/bucket.go

@@ -0,0 +1,748 @@
+package bolt
+
+import (
+	"bytes"
+	"fmt"
+	"unsafe"
+)
+
+const (
+	// MaxKeySize is the maximum length of a key, in bytes.
+	MaxKeySize = 32768
+
+	// MaxValueSize is the maximum length of a value, in bytes.
+	MaxValueSize = (1 << 31) - 2
+)
+
+const (
+	maxUint = ^uint(0)
+	minUint = 0
+	maxInt  = int(^uint(0) >> 1)
+	minInt  = -maxInt - 1
+)
+
+const bucketHeaderSize = int(unsafe.Sizeof(bucket{}))
+
+const (
+	minFillPercent = 0.1
+	maxFillPercent = 1.0
+)
+
+// DefaultFillPercent is the percentage that split pages are filled.
+// This value can be changed by setting Bucket.FillPercent.
+const DefaultFillPercent = 0.5
+
+// Bucket represents a collection of key/value pairs inside the database.
+type Bucket struct {
+	*bucket
+	tx       *Tx                // the associated transaction
+	buckets  map[string]*Bucket // subbucket cache
+	page     *page              // inline page reference
+	rootNode *node              // materialized node for the root page.
+	nodes    map[pgid]*node     // node cache
+
+	// Sets the threshold for filling nodes when they split. By default,
+	// the bucket will fill to 50% but it can be useful to increase this
+	// amount if you know that your write workloads are mostly append-only.
+	//
+	// This is non-persisted across transactions so it must be set in every Tx.
+	FillPercent float64
+}
+
+// bucket represents the on-file representation of a bucket.
+// This is stored as the "value" of a bucket key. If the bucket is small enough,
+// then its root page can be stored inline in the "value", after the bucket
+// header. In the case of inline buckets, the "root" will be 0.
+type bucket struct {
+	root     pgid   // page id of the bucket's root-level page
+	sequence uint64 // monotonically incrementing, used by NextSequence()
+}
+
+// newBucket returns a new bucket associated with a transaction.
+func newBucket(tx *Tx) Bucket {
+	var b = Bucket{tx: tx, FillPercent: DefaultFillPercent}
+	if tx.writable {
+		b.buckets = make(map[string]*Bucket)
+		b.nodes = make(map[pgid]*node)
+	}
+	return b
+}
+
+// Tx returns the tx of the bucket.
+func (b *Bucket) Tx() *Tx {
+	return b.tx
+}
+
+// Root returns the root of the bucket.
+func (b *Bucket) Root() pgid {
+	return b.root
+}
+
+// Writable returns whether the bucket is writable.
+func (b *Bucket) Writable() bool {
+	return b.tx.writable
+}
+
+// Cursor creates a cursor associated with the bucket.
+// The cursor is only valid as long as the transaction is open.
+// Do not use a cursor after the transaction is closed.
+func (b *Bucket) Cursor() *Cursor {
+	// Update transaction statistics.
+	b.tx.stats.CursorCount++
+
+	// Allocate and return a cursor.
+	return &Cursor{
+		bucket: b,
+		stack:  make([]elemRef, 0),
+	}
+}
+
+// Bucket retrieves a nested bucket by name.
+// Returns nil if the bucket does not exist.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (b *Bucket) Bucket(name []byte) *Bucket {
+	if b.buckets != nil {
+		if child := b.buckets[string(name)]; child != nil {
+			return child
+		}
+	}
+
+	// Move cursor to key.
+	c := b.Cursor()
+	k, v, flags := c.seek(name)
+
+	// Return nil if the key doesn't exist or it is not a bucket.
+	if !bytes.Equal(name, k) || (flags&bucketLeafFlag) == 0 {
+		return nil
+	}
+
+	// Otherwise create a bucket and cache it.
+	var child = b.openBucket(v)
+	if b.buckets != nil {
+		b.buckets[string(name)] = child
+	}
+
+	return child
+}
+
+// Helper method that re-interprets a sub-bucket value
+// from a parent into a Bucket
+func (b *Bucket) openBucket(value []byte) *Bucket {
+	var child = newBucket(b.tx)
+
+	// If this is a writable transaction then we need to copy the bucket entry.
+	// Read-only transactions can point directly at the mmap entry.
+	if b.tx.writable {
+		child.bucket = &bucket{}
+		*child.bucket = *(*bucket)(unsafe.Pointer(&value[0]))
+	} else {
+		child.bucket = (*bucket)(unsafe.Pointer(&value[0]))
+	}
+
+	// Save a reference to the inline page if the bucket is inline.
+	if child.root == 0 {
+		child.page = (*page)(unsafe.Pointer(&value[bucketHeaderSize]))
+	}
+
+	return &child
+}
+
+// CreateBucket creates a new bucket at the given key and returns the new bucket.
+// Returns an error if the key already exists, if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (b *Bucket) CreateBucket(key []byte) (*Bucket, error) {
+	if b.tx.db == nil {
+		return nil, ErrTxClosed
+	} else if !b.tx.writable {
+		return nil, ErrTxNotWritable
+	} else if len(key) == 0 {
+		return nil, ErrBucketNameRequired
+	}
+
+	// Move cursor to correct position.
+	c := b.Cursor()
+	k, _, flags := c.seek(key)
+
+	// Return an error if there is an existing key.
+	if bytes.Equal(key, k) {
+		if (flags & bucketLeafFlag) != 0 {
+			return nil, ErrBucketExists
+		} else {
+			return nil, ErrIncompatibleValue
+		}
+	}
+
+	// Create empty, inline bucket.
+	var bucket = Bucket{
+		bucket:      &bucket{},
+		rootNode:    &node{isLeaf: true},
+		FillPercent: DefaultFillPercent,
+	}
+	var value = bucket.write()
+
+	// Insert into node.
+	key = cloneBytes(key)
+	c.node().put(key, key, value, 0, bucketLeafFlag)
+
+	// Since subbuckets are not allowed on inline buckets, we need to
+	// dereference the inline page, if it exists. This will cause the bucket
+	// to be treated as a regular, non-inline bucket for the rest of the tx.
+	b.page = nil
+
+	return b.Bucket(key), nil
+}
+
+// CreateBucketIfNotExists creates a new bucket if it doesn't already exist and returns a reference to it.
+// Returns an error if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (b *Bucket) CreateBucketIfNotExists(key []byte) (*Bucket, error) {
+	child, err := b.CreateBucket(key)
+	if err == ErrBucketExists {
+		return b.Bucket(key), nil
+	} else if err != nil {
+		return nil, err
+	}
+	return child, nil
+}
+
+// DeleteBucket deletes a bucket at the given key.
+// Returns an error if the bucket does not exists, or if the key represents a non-bucket value.
+func (b *Bucket) DeleteBucket(key []byte) error {
+	if b.tx.db == nil {
+		return ErrTxClosed
+	} else if !b.Writable() {
+		return ErrTxNotWritable
+	}
+
+	// Move cursor to correct position.
+	c := b.Cursor()
+	k, _, flags := c.seek(key)
+
+	// Return an error if bucket doesn't exist or is not a bucket.
+	if !bytes.Equal(key, k) {
+		return ErrBucketNotFound
+	} else if (flags & bucketLeafFlag) == 0 {
+		return ErrIncompatibleValue
+	}
+
+	// Recursively delete all child buckets.
+	child := b.Bucket(key)
+	err := child.ForEach(func(k, v []byte) error {
+		if v == nil {
+			if err := child.DeleteBucket(k); err != nil {
+				return fmt.Errorf("delete bucket: %s", err)
+			}
+		}
+		return nil
+	})
+	if err != nil {
+		return err
+	}
+
+	// Remove cached copy.
+	delete(b.buckets, string(key))
+
+	// Release all bucket pages to freelist.
+	child.nodes = nil
+	child.rootNode = nil
+	child.free()
+
+	// Delete the node if we have a matching key.
+	c.node().del(key)
+
+	return nil
+}
+
+// Get retrieves the value for a key in the bucket.
+// Returns a nil value if the key does not exist or if the key is a nested bucket.
+// The returned value is only valid for the life of the transaction.
+func (b *Bucket) Get(key []byte) []byte {
+	k, v, flags := b.Cursor().seek(key)
+
+	// Return nil if this is a bucket.
+	if (flags & bucketLeafFlag) != 0 {
+		return nil
+	}
+
+	// If our target node isn't the same key as what's passed in then return nil.
+	if !bytes.Equal(key, k) {
+		return nil
+	}
+	return v
+}
+
+// Put sets the value for a key in the bucket.
+// If the key exist then its previous value will be overwritten.
+// Supplied value must remain valid for the life of the transaction.
+// Returns an error if the bucket was created from a read-only transaction, if the key is blank, if the key is too large, or if the value is too large.
+func (b *Bucket) Put(key []byte, value []byte) error {
+	if b.tx.db == nil {
+		return ErrTxClosed
+	} else if !b.Writable() {
+		return ErrTxNotWritable
+	} else if len(key) == 0 {
+		return ErrKeyRequired
+	} else if len(key) > MaxKeySize {
+		return ErrKeyTooLarge
+	} else if int64(len(value)) > MaxValueSize {
+		return ErrValueTooLarge
+	}
+
+	// Move cursor to correct position.
+	c := b.Cursor()
+	k, _, flags := c.seek(key)
+
+	// Return an error if there is an existing key with a bucket value.
+	if bytes.Equal(key, k) && (flags&bucketLeafFlag) != 0 {
+		return ErrIncompatibleValue
+	}
+
+	// Insert into node.
+	key = cloneBytes(key)
+	c.node().put(key, key, value, 0, 0)
+
+	return nil
+}
+
+// Delete removes a key from the bucket.
+// If the key does not exist then nothing is done and a nil error is returned.
+// Returns an error if the bucket was created from a read-only transaction.
+func (b *Bucket) Delete(key []byte) error {
+	if b.tx.db == nil {
+		return ErrTxClosed
+	} else if !b.Writable() {
+		return ErrTxNotWritable
+	}
+
+	// Move cursor to correct position.
+	c := b.Cursor()
+	_, _, flags := c.seek(key)
+
+	// Return an error if there is already existing bucket value.
+	if (flags & bucketLeafFlag) != 0 {
+		return ErrIncompatibleValue
+	}
+
+	// Delete the node if we have a matching key.
+	c.node().del(key)
+
+	return nil
+}
+
+// NextSequence returns an autoincrementing integer for the bucket.
+func (b *Bucket) NextSequence() (uint64, error) {
+	if b.tx.db == nil {
+		return 0, ErrTxClosed
+	} else if !b.Writable() {
+		return 0, ErrTxNotWritable
+	}
+
+	// Materialize the root node if it hasn't been already so that the
+	// bucket will be saved during commit.
+	if b.rootNode == nil {
+		_ = b.node(b.root, nil)
+	}
+
+	// Increment and return the sequence.
+	b.bucket.sequence++
+	return b.bucket.sequence, nil
+}
+
+// ForEach executes a function for each key/value pair in a bucket.
+// If the provided function returns an error then the iteration is stopped and
+// the error is returned to the caller. The provided function must not modify
+// the bucket; this will result in undefined behavior.
+func (b *Bucket) ForEach(fn func(k, v []byte) error) error {
+	if b.tx.db == nil {
+		return ErrTxClosed
+	}
+	c := b.Cursor()
+	for k, v := c.First(); k != nil; k, v = c.Next() {
+		if err := fn(k, v); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+// Stat returns stats on a bucket.
+func (b *Bucket) Stats() BucketStats {
+	var s, subStats BucketStats
+	pageSize := b.tx.db.pageSize
+	s.BucketN += 1
+	if b.root == 0 {
+		s.InlineBucketN += 1
+	}
+	b.forEachPage(func(p *page, depth int) {
+		if (p.flags & leafPageFlag) != 0 {
+			s.KeyN += int(p.count)
+
+			// used totals the used bytes for the page
+			used := pageHeaderSize
+
+			if p.count != 0 {
+				// If page has any elements, add all element headers.
+				used += leafPageElementSize * int(p.count-1)
+
+				// Add all element key, value sizes.
+				// The computation takes advantage of the fact that the position
+				// of the last element's key/value equals to the total of the sizes
+				// of all previous elements' keys and values.
+				// It also includes the last element's header.
+				lastElement := p.leafPageElement(p.count - 1)
+				used += int(lastElement.pos + lastElement.ksize + lastElement.vsize)
+			}
+
+			if b.root == 0 {
+				// For inlined bucket just update the inline stats
+				s.InlineBucketInuse += used
+			} else {
+				// For non-inlined bucket update all the leaf stats
+				s.LeafPageN++
+				s.LeafInuse += used
+				s.LeafOverflowN += int(p.overflow)
+
+				// Collect stats from sub-buckets.
+				// Do that by iterating over all element headers
+				// looking for the ones with the bucketLeafFlag.
+				for i := uint16(0); i < p.count; i++ {
+					e := p.leafPageElement(i)
+					if (e.flags & bucketLeafFlag) != 0 {
+						// For any bucket element, open the element value
+						// and recursively call Stats on the contained bucket.
+						subStats.Add(b.openBucket(e.value()).Stats())
+					}
+				}
+			}
+		} else if (p.flags & branchPageFlag) != 0 {
+			s.BranchPageN++
+			lastElement := p.branchPageElement(p.count - 1)
+
+			// used totals the used bytes for the page
+			// Add header and all element headers.
+			used := pageHeaderSize + (branchPageElementSize * int(p.count-1))
+
+			// Add size of all keys and values.
+			// Again, use the fact that last element's position equals to
+			// the total of key, value sizes of all previous elements.
+			used += int(lastElement.pos + lastElement.ksize)
+			s.BranchInuse += used
+			s.BranchOverflowN += int(p.overflow)
+		}
+
+		// Keep track of maximum page depth.
+		if depth+1 > s.Depth {
+			s.Depth = (depth + 1)
+		}
+	})
+
+	// Alloc stats can be computed from page counts and pageSize.
+	s.BranchAlloc = (s.BranchPageN + s.BranchOverflowN) * pageSize
+	s.LeafAlloc = (s.LeafPageN + s.LeafOverflowN) * pageSize
+
+	// Add the max depth of sub-buckets to get total nested depth.
+	s.Depth += subStats.Depth
+	// Add the stats for all sub-buckets
+	s.Add(subStats)
+	return s
+}
+
+// forEachPage iterates over every page in a bucket, including inline pages.
+func (b *Bucket) forEachPage(fn func(*page, int)) {
+	// If we have an inline page then just use that.
+	if b.page != nil {
+		fn(b.page, 0)
+		return
+	}
+
+	// Otherwise traverse the page hierarchy.
+	b.tx.forEachPage(b.root, 0, fn)
+}
+
+// forEachPageNode iterates over every page (or node) in a bucket.
+// This also includes inline pages.
+func (b *Bucket) forEachPageNode(fn func(*page, *node, int)) {
+	// If we have an inline page or root node then just use that.
+	if b.page != nil {
+		fn(b.page, nil, 0)
+		return
+	}
+	b._forEachPageNode(b.root, 0, fn)
+}
+
+func (b *Bucket) _forEachPageNode(pgid pgid, depth int, fn func(*page, *node, int)) {
+	var p, n = b.pageNode(pgid)
+
+	// Execute function.
+	fn(p, n, depth)
+
+	// Recursively loop over children.
+	if p != nil {
+		if (p.flags & branchPageFlag) != 0 {
+			for i := 0; i < int(p.count); i++ {
+				elem := p.branchPageElement(uint16(i))
+				b._forEachPageNode(elem.pgid, depth+1, fn)
+			}
+		}
+	} else {
+		if !n.isLeaf {
+			for _, inode := range n.inodes {
+				b._forEachPageNode(inode.pgid, depth+1, fn)
+			}
+		}
+	}
+}
+
+// spill writes all the nodes for this bucket to dirty pages.
+func (b *Bucket) spill() error {
+	// Spill all child buckets first.
+	for name, child := range b.buckets {
+		// If the child bucket is small enough and it has no child buckets then
+		// write it inline into the parent bucket's page. Otherwise spill it
+		// like a normal bucket and make the parent value a pointer to the page.
+		var value []byte
+		if child.inlineable() {
+			child.free()
+			value = child.write()
+		} else {
+			if err := child.spill(); err != nil {
+				return err
+			}
+
+			// Update the child bucket header in this bucket.
+			value = make([]byte, unsafe.Sizeof(bucket{}))
+			var bucket = (*bucket)(unsafe.Pointer(&value[0]))
+			*bucket = *child.bucket
+		}
+
+		// Skip writing the bucket if there are no materialized nodes.
+		if child.rootNode == nil {
+			continue
+		}
+
+		// Update parent node.
+		var c = b.Cursor()
+		k, _, flags := c.seek([]byte(name))
+		if !bytes.Equal([]byte(name), k) {
+			panic(fmt.Sprintf("misplaced bucket header: %x -> %x", []byte(name), k))
+		}
+		if flags&bucketLeafFlag == 0 {
+			panic(fmt.Sprintf("unexpected bucket header flag: %x", flags))
+		}
+		c.node().put([]byte(name), []byte(name), value, 0, bucketLeafFlag)
+	}
+
+	// Ignore if there's not a materialized root node.
+	if b.rootNode == nil {
+		return nil
+	}
+
+	// Spill nodes.
+	if err := b.rootNode.spill(); err != nil {
+		return err
+	}
+	b.rootNode = b.rootNode.root()
+
+	// Update the root node for this bucket.
+	if b.rootNode.pgid >= b.tx.meta.pgid {
+		panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", b.rootNode.pgid, b.tx.meta.pgid))
+	}
+	b.root = b.rootNode.pgid
+
+	return nil
+}
+
+// inlineable returns true if a bucket is small enough to be written inline
+// and if it contains no subbuckets. Otherwise returns false.
+func (b *Bucket) inlineable() bool {
+	var n = b.rootNode
+
+	// Bucket must only contain a single leaf node.
+	if n == nil || !n.isLeaf {
+		return false
+	}
+
+	// Bucket is not inlineable if it contains subbuckets or if it goes beyond
+	// our threshold for inline bucket size.
+	var size = pageHeaderSize
+	for _, inode := range n.inodes {
+		size += leafPageElementSize + len(inode.key) + len(inode.value)
+
+		if inode.flags&bucketLeafFlag != 0 {
+			return false
+		} else if size > b.maxInlineBucketSize() {
+			return false
+		}
+	}
+
+	return true
+}
+
+// Returns the maximum total size of a bucket to make it a candidate for inlining.
+func (b *Bucket) maxInlineBucketSize() int {
+	return b.tx.db.pageSize / 4
+}
+
+// write allocates and writes a bucket to a byte slice.
+func (b *Bucket) write() []byte {
+	// Allocate the appropriate size.
+	var n = b.rootNode
+	var value = make([]byte, bucketHeaderSize+n.size())
+
+	// Write a bucket header.
+	var bucket = (*bucket)(unsafe.Pointer(&value[0]))
+	*bucket = *b.bucket
+
+	// Convert byte slice to a fake page and write the root node.
+	var p = (*page)(unsafe.Pointer(&value[bucketHeaderSize]))
+	n.write(p)
+
+	return value
+}
+
+// rebalance attempts to balance all nodes.
+func (b *Bucket) rebalance() {
+	for _, n := range b.nodes {
+		n.rebalance()
+	}
+	for _, child := range b.buckets {
+		child.rebalance()
+	}
+}
+
+// node creates a node from a page and associates it with a given parent.
+func (b *Bucket) node(pgid pgid, parent *node) *node {
+	_assert(b.nodes != nil, "nodes map expected")
+
+	// Retrieve node if it's already been created.
+	if n := b.nodes[pgid]; n != nil {
+		return n
+	}
+
+	// Otherwise create a node and cache it.
+	n := &node{bucket: b, parent: parent}
+	if parent == nil {
+		b.rootNode = n
+	} else {
+		parent.children = append(parent.children, n)
+	}
+
+	// Use the inline page if this is an inline bucket.
+	var p = b.page
+	if p == nil {
+		p = b.tx.page(pgid)
+	}
+
+	// Read the page into the node and cache it.
+	n.read(p)
+	b.nodes[pgid] = n
+
+	// Update statistics.
+	b.tx.stats.NodeCount++
+
+	return n
+}
+
+// free recursively frees all pages in the bucket.
+func (b *Bucket) free() {
+	if b.root == 0 {
+		return
+	}
+
+	var tx = b.tx
+	b.forEachPageNode(func(p *page, n *node, _ int) {
+		if p != nil {
+			tx.db.freelist.free(tx.meta.txid, p)
+		} else {
+			n.free()
+		}
+	})
+	b.root = 0
+}
+
+// dereference removes all references to the old mmap.
+func (b *Bucket) dereference() {
+	if b.rootNode != nil {
+		b.rootNode.root().dereference()
+	}
+
+	for _, child := range b.buckets {
+		child.dereference()
+	}
+}
+
+// pageNode returns the in-memory node, if it exists.
+// Otherwise returns the underlying page.
+func (b *Bucket) pageNode(id pgid) (*page, *node) {
+	// Inline buckets have a fake page embedded in their value so treat them
+	// differently. We'll return the rootNode (if available) or the fake page.
+	if b.root == 0 {
+		if id != 0 {
+			panic(fmt.Sprintf("inline bucket non-zero page access(2): %d != 0", id))
+		}
+		if b.rootNode != nil {
+			return nil, b.rootNode
+		}
+		return b.page, nil
+	}
+
+	// Check the node cache for non-inline buckets.
+	if b.nodes != nil {
+		if n := b.nodes[id]; n != nil {
+			return nil, n
+		}
+	}
+
+	// Finally lookup the page from the transaction if no node is materialized.
+	return b.tx.page(id), nil
+}
+
+// BucketStats records statistics about resources used by a bucket.
+type BucketStats struct {
+	// Page count statistics.
+	BranchPageN     int // number of logical branch pages
+	BranchOverflowN int // number of physical branch overflow pages
+	LeafPageN       int // number of logical leaf pages
+	LeafOverflowN   int // number of physical leaf overflow pages
+
+	// Tree statistics.
+	KeyN  int // number of keys/value pairs
+	Depth int // number of levels in B+tree
+
+	// Page size utilization.
+	BranchAlloc int // bytes allocated for physical branch pages
+	BranchInuse int // bytes actually used for branch data
+	LeafAlloc   int // bytes allocated for physical leaf pages
+	LeafInuse   int // bytes actually used for leaf data
+
+	// Bucket statistics
+	BucketN           int // total number of buckets including the top bucket
+	InlineBucketN     int // total number on inlined buckets
+	InlineBucketInuse int // bytes used for inlined buckets (also accounted for in LeafInuse)
+}
+
+func (s *BucketStats) Add(other BucketStats) {
+	s.BranchPageN += other.BranchPageN
+	s.BranchOverflowN += other.BranchOverflowN
+	s.LeafPageN += other.LeafPageN
+	s.LeafOverflowN += other.LeafOverflowN
+	s.KeyN += other.KeyN
+	if s.Depth < other.Depth {
+		s.Depth = other.Depth
+	}
+	s.BranchAlloc += other.BranchAlloc
+	s.BranchInuse += other.BranchInuse
+	s.LeafAlloc += other.LeafAlloc
+	s.LeafInuse += other.LeafInuse
+
+	s.BucketN += other.BucketN
+	s.InlineBucketN += other.InlineBucketN
+	s.InlineBucketInuse += other.InlineBucketInuse
+}
+
+// cloneBytes returns a copy of a given slice.
+func cloneBytes(v []byte) []byte {
+	var clone = make([]byte, len(v))
+	copy(clone, v)
+	return clone
+}

weed/vendor/github.com/boltdb/bolt/cursor.go

@@ -0,0 +1,400 @@
+package bolt
+
+import (
+	"bytes"
+	"fmt"
+	"sort"
+)
+
+// Cursor represents an iterator that can traverse over all key/value pairs in a bucket in sorted order.
+// Cursors see nested buckets with value == nil.
+// Cursors can be obtained from a transaction and are valid as long as the transaction is open.
+//
+// Keys and values returned from the cursor are only valid for the life of the transaction.
+//
+// Changing data while traversing with a cursor may cause it to be invalidated
+// and return unexpected keys and/or values. You must reposition your cursor
+// after mutating data.
+type Cursor struct {
+	bucket *Bucket
+	stack  []elemRef
+}
+
+// Bucket returns the bucket that this cursor was created from.
+func (c *Cursor) Bucket() *Bucket {
+	return c.bucket
+}
+
+// First moves the cursor to the first item in the bucket and returns its key and value.
+// If the bucket is empty then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) First() (key []byte, value []byte) {
+	_assert(c.bucket.tx.db != nil, "tx closed")
+	c.stack = c.stack[:0]
+	p, n := c.bucket.pageNode(c.bucket.root)
+	c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
+	c.first()
+
+	// If we land on an empty page then move to the next value.
+	// https://github.com/boltdb/bolt/issues/450
+	if c.stack[len(c.stack)-1].count() == 0 {
+		c.next()
+	}
+
+	k, v, flags := c.keyValue()
+	if (flags & uint32(bucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+
+}
+
+// Last moves the cursor to the last item in the bucket and returns its key and value.
+// If the bucket is empty then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Last() (key []byte, value []byte) {
+	_assert(c.bucket.tx.db != nil, "tx closed")
+	c.stack = c.stack[:0]
+	p, n := c.bucket.pageNode(c.bucket.root)
+	ref := elemRef{page: p, node: n}
+	ref.index = ref.count() - 1
+	c.stack = append(c.stack, ref)
+	c.last()
+	k, v, flags := c.keyValue()
+	if (flags & uint32(bucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Next moves the cursor to the next item in the bucket and returns its key and value.
+// If the cursor is at the end of the bucket then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Next() (key []byte, value []byte) {
+	_assert(c.bucket.tx.db != nil, "tx closed")
+	k, v, flags := c.next()
+	if (flags & uint32(bucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Prev moves the cursor to the previous item in the bucket and returns its key and value.
+// If the cursor is at the beginning of the bucket then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Prev() (key []byte, value []byte) {
+	_assert(c.bucket.tx.db != nil, "tx closed")
+
+	// Attempt to move back one element until we're successful.
+	// Move up the stack as we hit the beginning of each page in our stack.
+	for i := len(c.stack) - 1; i >= 0; i-- {
+		elem := &c.stack[i]
+		if elem.index > 0 {
+			elem.index--
+			break
+		}
+		c.stack = c.stack[:i]
+	}
+
+	// If we've hit the end then return nil.
+	if len(c.stack) == 0 {
+		return nil, nil
+	}
+
+	// Move down the stack to find the last element of the last leaf under this branch.
+	c.last()
+	k, v, flags := c.keyValue()
+	if (flags & uint32(bucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Seek moves the cursor to a given key and returns it.
+// If the key does not exist then the next key is used. If no keys
+// follow, a nil key is returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Seek(seek []byte) (key []byte, value []byte) {
+	k, v, flags := c.seek(seek)
+
+	// If we ended up after the last element of a page then move to the next one.
+	if ref := &c.stack[len(c.stack)-1]; ref.index >= ref.count() {
+		k, v, flags = c.next()
+	}
+
+	if k == nil {
+		return nil, nil
+	} else if (flags & uint32(bucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Delete removes the current key/value under the cursor from the bucket.
+// Delete fails if current key/value is a bucket or if the transaction is not writable.
+func (c *Cursor) Delete() error {
+	if c.bucket.tx.db == nil {
+		return ErrTxClosed
+	} else if !c.bucket.Writable() {
+		return ErrTxNotWritable
+	}
+
+	key, _, flags := c.keyValue()
+	// Return an error if current value is a bucket.
+	if (flags & bucketLeafFlag) != 0 {
+		return ErrIncompatibleValue
+	}
+	c.node().del(key)
+
+	return nil
+}
+
+// seek moves the cursor to a given key and returns it.
+// If the key does not exist then the next key is used.
+func (c *Cursor) seek(seek []byte) (key []byte, value []byte, flags uint32) {
+	_assert(c.bucket.tx.db != nil, "tx closed")
+
+	// Start from root page/node and traverse to correct page.
+	c.stack = c.stack[:0]
+	c.search(seek, c.bucket.root)
+	ref := &c.stack[len(c.stack)-1]
+
+	// If the cursor is pointing to the end of page/node then return nil.
+	if ref.index >= ref.count() {
+		return nil, nil, 0
+	}
+
+	// If this is a bucket then return a nil value.
+	return c.keyValue()
+}
+
+// first moves the cursor to the first leaf element under the last page in the stack.
+func (c *Cursor) first() {
+	for {
+		// Exit when we hit a leaf page.
+		var ref = &c.stack[len(c.stack)-1]
+		if ref.isLeaf() {
+			break
+		}
+
+		// Keep adding pages pointing to the first element to the stack.
+		var pgid pgid
+		if ref.node != nil {
+			pgid = ref.node.inodes[ref.index].pgid
+		} else {
+			pgid = ref.page.branchPageElement(uint16(ref.index)).pgid
+		}
+		p, n := c.bucket.pageNode(pgid)
+		c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
+	}
+}
+
+// last moves the cursor to the last leaf element under the last page in the stack.
+func (c *Cursor) last() {
+	for {
+		// Exit when we hit a leaf page.
+		ref := &c.stack[len(c.stack)-1]
+		if ref.isLeaf() {
+			break
+		}
+
+		// Keep adding pages pointing to the last element in the stack.
+		var pgid pgid
+		if ref.node != nil {
+			pgid = ref.node.inodes[ref.index].pgid
+		} else {
+			pgid = ref.page.branchPageElement(uint16(ref.index)).pgid
+		}
+		p, n := c.bucket.pageNode(pgid)
+
+		var nextRef = elemRef{page: p, node: n}
+		nextRef.index = nextRef.count() - 1
+		c.stack = append(c.stack, nextRef)
+	}
+}
+
+// next moves to the next leaf element and returns the key and value.
+// If the cursor is at the last leaf element then it stays there and returns nil.
+func (c *Cursor) next() (key []byte, value []byte, flags uint32) {
+	for {
+		// Attempt to move over one element until we're successful.
+		// Move up the stack as we hit the end of each page in our stack.
+		var i int
+		for i = len(c.stack) - 1; i >= 0; i-- {
+			elem := &c.stack[i]
+			if elem.index < elem.count()-1 {
+				elem.index++
+				break
+			}
+		}
+
+		// If we've hit the root page then stop and return. This will leave the
+		// cursor on the last element of the last page.
+		if i == -1 {
+			return nil, nil, 0
+		}
+
+		// Otherwise start from where we left off in the stack and find the
+		// first element of the first leaf page.
+		c.stack = c.stack[:i+1]
+		c.first()
+
+		// If this is an empty page then restart and move back up the stack.
+		// https://github.com/boltdb/bolt/issues/450
+		if c.stack[len(c.stack)-1].count() == 0 {
+			continue
+		}
+
+		return c.keyValue()
+	}
+}
+
+// search recursively performs a binary search against a given page/node until it finds a given key.
+func (c *Cursor) search(key []byte, pgid pgid) {
+	p, n := c.bucket.pageNode(pgid)
+	if p != nil && (p.flags&(branchPageFlag|leafPageFlag)) == 0 {
+		panic(fmt.Sprintf("invalid page type: %d: %x", p.id, p.flags))
+	}
+	e := elemRef{page: p, node: n}
+	c.stack = append(c.stack, e)
+
+	// If we're on a leaf page/node then find the specific node.
+	if e.isLeaf() {
+		c.nsearch(key)
+		return
+	}
+
+	if n != nil {
+		c.searchNode(key, n)
+		return
+	}
+	c.searchPage(key, p)
+}
+
+func (c *Cursor) searchNode(key []byte, n *node) {
+	var exact bool
+	index := sort.Search(len(n.inodes), func(i int) bool {
+		// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
+		// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
+		ret := bytes.Compare(n.inodes[i].key, key)
+		if ret == 0 {
+			exact = true
+		}
+		return ret != -1
+	})
+	if !exact && index > 0 {
+		index--
+	}
+	c.stack[len(c.stack)-1].index = index
+
+	// Recursively search to the next page.
+	c.search(key, n.inodes[index].pgid)
+}
+
+func (c *Cursor) searchPage(key []byte, p *page) {
+	// Binary search for the correct range.
+	inodes := p.branchPageElements()
+
+	var exact bool
+	index := sort.Search(int(p.count), func(i int) bool {
+		// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
+		// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
+		ret := bytes.Compare(inodes[i].key(), key)
+		if ret == 0 {
+			exact = true
+		}
+		return ret != -1
+	})
+	if !exact && index > 0 {
+		index--
+	}
+	c.stack[len(c.stack)-1].index = index
+
+	// Recursively search to the next page.
+	c.search(key, inodes[index].pgid)
+}
+
+// nsearch searches the leaf node on the top of the stack for a key.
+func (c *Cursor) nsearch(key []byte) {
+	e := &c.stack[len(c.stack)-1]
+	p, n := e.page, e.node
+
+	// If we have a node then search its inodes.
+	if n != nil {
+		index := sort.Search(len(n.inodes), func(i int) bool {
+			return bytes.Compare(n.inodes[i].key, key) != -1
+		})
+		e.index = index
+		return
+	}
+
+	// If we have a page then search its leaf elements.
+	inodes := p.leafPageElements()
+	index := sort.Search(int(p.count), func(i int) bool {
+		return bytes.Compare(inodes[i].key(), key) != -1
+	})
+	e.index = index
+}
+
+// keyValue returns the key and value of the current leaf element.
+func (c *Cursor) keyValue() ([]byte, []byte, uint32) {
+	ref := &c.stack[len(c.stack)-1]
+	if ref.count() == 0 || ref.index >= ref.count() {
+		return nil, nil, 0
+	}
+
+	// Retrieve value from node.
+	if ref.node != nil {
+		inode := &ref.node.inodes[ref.index]
+		return inode.key, inode.value, inode.flags
+	}
+
+	// Or retrieve value from page.
+	elem := ref.page.leafPageElement(uint16(ref.index))
+	return elem.key(), elem.value(), elem.flags
+}
+
+// node returns the node that the cursor is currently positioned on.
+func (c *Cursor) node() *node {
+	_assert(len(c.stack) > 0, "accessing a node with a zero-length cursor stack")
+
+	// If the top of the stack is a leaf node then just return it.
+	if ref := &c.stack[len(c.stack)-1]; ref.node != nil && ref.isLeaf() {
+		return ref.node
+	}
+
+	// Start from root and traverse down the hierarchy.
+	var n = c.stack[0].node
+	if n == nil {
+		n = c.bucket.node(c.stack[0].page.id, nil)
+	}
+	for _, ref := range c.stack[:len(c.stack)-1] {
+		_assert(!n.isLeaf, "expected branch node")
+		n = n.childAt(int(ref.index))
+	}
+	_assert(n.isLeaf, "expected leaf node")
+	return n
+}
+
+// elemRef represents a reference to an element on a given page/node.
+type elemRef struct {
+	page  *page
+	node  *node
+	index int
+}
+
+// isLeaf returns whether the ref is pointing at a leaf page/node.
+func (r *elemRef) isLeaf() bool {
+	if r.node != nil {
+		return r.node.isLeaf
+	}
+	return (r.page.flags & leafPageFlag) != 0
+}
+
+// count returns the number of inodes or page elements.
+func (r *elemRef) count() int {
+	if r.node != nil {
+		return len(r.node.inodes)
+	}
+	return int(r.page.count)
+}

weed/vendor/github.com/boltdb/bolt/db.go

@@ -0,0 +1,800 @@
+package bolt
+
+import (
+	"fmt"
+	"hash/fnv"
+	"os"
+	"runtime"
+	"runtime/debug"
+	"strings"
+	"sync"
+	"time"
+	"unsafe"
+)
+
+// The largest step that can be taken when remapping the mmap.
+const maxMmapStep = 1 << 30 // 1GB
+
+// The data file format version.
+const version = 2
+
+// Represents a marker value to indicate that a file is a Bolt DB.
+const magic uint32 = 0xED0CDAED
+
+// IgnoreNoSync specifies whether the NoSync field of a DB is ignored when
+// syncing changes to a file.  This is required as some operating systems,
+// such as OpenBSD, do not have a unified buffer cache (UBC) and writes
+// must be synchronzied using the msync(2) syscall.
+const IgnoreNoSync = runtime.GOOS == "openbsd"
+
+// Default values if not set in a DB instance.
+const (
+	DefaultMaxBatchSize  int = 1000
+	DefaultMaxBatchDelay     = 10 * time.Millisecond
+)
+
+// DB represents a collection of buckets persisted to a file on disk.
+// All data access is performed through transactions which can be obtained through the DB.
+// All the functions on DB will return a ErrDatabaseNotOpen if accessed before Open() is called.
+type DB struct {
+	// When enabled, the database will perform a Check() after every commit.
+	// A panic is issued if the database is in an inconsistent state. This
+	// flag has a large performance impact so it should only be used for
+	// debugging purposes.
+	StrictMode bool
+
+	// Setting the NoSync flag will cause the database to skip fsync()
+	// calls after each commit. This can be useful when bulk loading data
+	// into a database and you can restart the bulk load in the event of
+	// a system failure or database corruption. Do not set this flag for
+	// normal use.
+	//
+	// If the package global IgnoreNoSync constant is true, this value is
+	// ignored.  See the comment on that constant for more details.
+	//
+	// THIS IS UNSAFE. PLEASE USE WITH CAUTION.
+	NoSync bool
+
+	// When true, skips the truncate call when growing the database.
+	// Setting this to true is only safe on non-ext3/ext4 systems.
+	// Skipping truncation avoids preallocation of hard drive space and
+	// bypasses a truncate() and fsync() syscall on remapping.
+	//
+	// https://github.com/boltdb/bolt/issues/284
+	NoGrowSync bool
+
+	// If you want to read the entire database fast, you can set MmapFlag to
+	// syscall.MAP_POPULATE on Linux 2.6.23+ for sequential read-ahead.
+	MmapFlags int
+
+	// MaxBatchSize is the maximum size of a batch. Default value is
+	// copied from DefaultMaxBatchSize in Open.
+	//
+	// If <=0, disables batching.
+	//
+	// Do not change concurrently with calls to Batch.
+	MaxBatchSize int
+
+	// MaxBatchDelay is the maximum delay before a batch starts.
+	// Default value is copied from DefaultMaxBatchDelay in Open.
+	//
+	// If <=0, effectively disables batching.
+	//
+	// Do not change concurrently with calls to Batch.
+	MaxBatchDelay time.Duration
+
+	path     string
+	file     *os.File
+	dataref  []byte // mmap'ed readonly, write throws SEGV
+	data     *[maxMapSize]byte
+	datasz   int
+	meta0    *meta
+	meta1    *meta
+	pageSize int
+	opened   bool
+	rwtx     *Tx
+	txs      []*Tx
+	freelist *freelist
+	stats    Stats
+
+	batchMu sync.Mutex
+	batch   *batch
+
+	rwlock   sync.Mutex   // Allows only one writer at a time.
+	metalock sync.Mutex   // Protects meta page access.
+	mmaplock sync.RWMutex // Protects mmap access during remapping.
+	statlock sync.RWMutex // Protects stats access.
+
+	ops struct {
+		writeAt func(b []byte, off int64) (n int, err error)
+	}
+
+	// Read only mode.
+	// When true, Update() and Begin(true) return ErrDatabaseReadOnly immediately.
+	readOnly bool
+}
+
+// Path returns the path to currently open database file.
+func (db *DB) Path() string {
+	return db.path
+}
+
+// GoString returns the Go string representation of the database.
+func (db *DB) GoString() string {
+	return fmt.Sprintf("bolt.DB{path:%q}", db.path)
+}
+
+// String returns the string representation of the database.
+func (db *DB) String() string {
+	return fmt.Sprintf("DB<%q>", db.path)
+}
+
+// Open creates and opens a database at the given path.
+// If the file does not exist then it will be created automatically.
+// Passing in nil options will cause Bolt to open the database with the default options.
+func Open(path string, mode os.FileMode, options *Options) (*DB, error) {
+	var db = &DB{opened: true}
+
+	// Set default options if no options are provided.
+	if options == nil {
+		options = DefaultOptions
+	}
+	db.NoGrowSync = options.NoGrowSync
+	db.MmapFlags = options.MmapFlags
+
+	// Set default values for later DB operations.
+	db.MaxBatchSize = DefaultMaxBatchSize
+	db.MaxBatchDelay = DefaultMaxBatchDelay
+
+	flag := os.O_RDWR
+	if options.ReadOnly {
+		flag = os.O_RDONLY
+		db.readOnly = true
+	}
+
+	// Open data file and separate sync handler for metadata writes.
+	db.path = path
+	var err error
+	if db.file, err = os.OpenFile(db.path, flag|os.O_CREATE, mode); err != nil {
+		_ = db.close()
+		return nil, err
+	}
+
+	// Lock file so that other processes using Bolt in read-write mode cannot
+	// use the database  at the same time. This would cause corruption since
+	// the two processes would write meta pages and free pages separately.
+	// The database file is locked exclusively (only one process can grab the lock)
+	// if !options.ReadOnly.
+	// The database file is locked using the shared lock (more than one process may
+	// hold a lock at the same time) otherwise (options.ReadOnly is set).
+	if err := flock(db.file, !db.readOnly, options.Timeout); err != nil {
+		_ = db.close()
+		return nil, err
+	}
+
+	// Default values for test hooks
+	db.ops.writeAt = db.file.WriteAt
+
+	// Initialize the database if it doesn't exist.
+	if info, err := db.file.Stat(); err != nil {
+		return nil, fmt.Errorf("stat error: %s", err)
+	} else if info.Size() == 0 {
+		// Initialize new files with meta pages.
+		if err := db.init(); err != nil {
+			return nil, err
+		}
+	} else {
+		// Read the first meta page to determine the page size.
+		var buf [0x1000]byte
+		if _, err := db.file.ReadAt(buf[:], 0); err == nil {
+			m := db.pageInBuffer(buf[:], 0).meta()
+			if err := m.validate(); err != nil {
+				return nil, fmt.Errorf("meta0 error: %s", err)
+			}
+			db.pageSize = int(m.pageSize)
+		}
+	}
+
+	// Memory map the data file.
+	if err := db.mmap(0); err != nil {
+		_ = db.close()
+		return nil, err
+	}
+
+	// Read in the freelist.
+	db.freelist = newFreelist()
+	db.freelist.read(db.page(db.meta().freelist))
+
+	// Mark the database as opened and return.
+	return db, nil
+}
+
+// mmap opens the underlying memory-mapped file and initializes the meta references.
+// minsz is the minimum size that the new mmap can be.
+func (db *DB) mmap(minsz int) error {
+	db.mmaplock.Lock()
+	defer db.mmaplock.Unlock()
+
+	info, err := db.file.Stat()
+	if err != nil {
+		return fmt.Errorf("mmap stat error: %s", err)
+	} else if int(info.Size()) < db.pageSize*2 {
+		return fmt.Errorf("file size too small")
+	}
+
+	// Ensure the size is at least the minimum size.
+	var size = int(info.Size())
+	if size < minsz {
+		size = minsz
+	}
+	size, err = db.mmapSize(size)
+	if err != nil {
+		return err
+	}
+
+	// Dereference all mmap references before unmapping.
+	if db.rwtx != nil {
+		db.rwtx.root.dereference()
+	}
+
+	// Unmap existing data before continuing.
+	if err := db.munmap(); err != nil {
+		return err
+	}
+
+	// Memory-map the data file as a byte slice.
+	if err := mmap(db, size); err != nil {
+		return err
+	}
+
+	// Save references to the meta pages.
+	db.meta0 = db.page(0).meta()
+	db.meta1 = db.page(1).meta()
+
+	// Validate the meta pages.
+	if err := db.meta0.validate(); err != nil {
+		return fmt.Errorf("meta0 error: %s", err)
+	}
+	if err := db.meta1.validate(); err != nil {
+		return fmt.Errorf("meta1 error: %s", err)
+	}
+
+	return nil
+}
+
+// munmap unmaps the data file from memory.
+func (db *DB) munmap() error {
+	if err := munmap(db); err != nil {
+		return fmt.Errorf("unmap error: " + err.Error())
+	}
+	return nil
+}
+
+// mmapSize determines the appropriate size for the mmap given the current size
+// of the database. The minimum size is 1MB and doubles until it reaches 1GB.
+// Returns an error if the new mmap size is greater than the max allowed.
+func (db *DB) mmapSize(size int) (int, error) {
+	// Double the size from 32KB until 1GB.
+	for i := uint(15); i <= 30; i++ {
+		if size <= 1<<i {
+			return 1 << i, nil
+		}
+	}
+
+	// Verify the requested size is not above the maximum allowed.
+	if size > maxMapSize {
+		return 0, fmt.Errorf("mmap too large")
+	}
+
+	// If larger than 1GB then grow by 1GB at a time.
+	sz := int64(size)
+	if remainder := sz % int64(maxMmapStep); remainder > 0 {
+		sz += int64(maxMmapStep) - remainder
+	}
+
+	// Ensure that the mmap size is a multiple of the page size.
+	// This should always be true since we're incrementing in MBs.
+	pageSize := int64(db.pageSize)
+	if (sz % pageSize) != 0 {
+		sz = ((sz / pageSize) + 1) * pageSize
+	}
+
+	// If we've exceeded the max size then only grow up to the max size.
+	if sz > maxMapSize {
+		sz = maxMapSize
+	}
+
+	return int(sz), nil
+}
+
+// init creates a new database file and initializes its meta pages.
+func (db *DB) init() error {
+	// Set the page size to the OS page size.
+	db.pageSize = os.Getpagesize()
+
+	// Create two meta pages on a buffer.
+	buf := make([]byte, db.pageSize*4)
+	for i := 0; i < 2; i++ {
+		p := db.pageInBuffer(buf[:], pgid(i))
+		p.id = pgid(i)
+		p.flags = metaPageFlag
+
+		// Initialize the meta page.
+		m := p.meta()
+		m.magic = magic
+		m.version = version
+		m.pageSize = uint32(db.pageSize)
+		m.freelist = 2
+		m.root = bucket{root: 3}
+		m.pgid = 4
+		m.txid = txid(i)
+	}
+
+	// Write an empty freelist at page 3.
+	p := db.pageInBuffer(buf[:], pgid(2))
+	p.id = pgid(2)
+	p.flags = freelistPageFlag
+	p.count = 0
+
+	// Write an empty leaf page at page 4.
+	p = db.pageInBuffer(buf[:], pgid(3))
+	p.id = pgid(3)
+	p.flags = leafPageFlag
+	p.count = 0
+
+	// Write the buffer to our data file.
+	if _, err := db.ops.writeAt(buf, 0); err != nil {
+		return err
+	}
+	if err := fdatasync(db); err != nil {
+		return err
+	}
+
+	return nil
+}
+
+// Close releases all database resources.
+// All transactions must be closed before closing the database.
+func (db *DB) Close() error {
+	db.rwlock.Lock()
+	defer db.rwlock.Unlock()
+
+	db.metalock.Lock()
+	defer db.metalock.Unlock()
+
+	db.mmaplock.RLock()
+	defer db.mmaplock.RUnlock()
+
+	return db.close()
+}
+
+func (db *DB) close() error {
+	db.opened = false
+
+	db.freelist = nil
+	db.path = ""
+
+	// Clear ops.
+	db.ops.writeAt = nil
+
+	// Close the mmap.
+	if err := db.munmap(); err != nil {
+		return err
+	}
+
+	// Close file handles.
+	if db.file != nil {
+		// No need to unlock read-only file.
+		if !db.readOnly {
+			// Unlock the file.
+			_ = funlock(db.file)
+		}
+
+		// Close the file descriptor.
+		if err := db.file.Close(); err != nil {
+			return fmt.Errorf("db file close: %s", err)
+		}
+		db.file = nil
+	}
+
+	return nil
+}
+
+// Begin starts a new transaction.
+// Multiple read-only transactions can be used concurrently but only one
+// write transaction can be used at a time. Starting multiple write transactions
+// will cause the calls to block and be serialized until the current write
+// transaction finishes.
+//
+// Transactions should not be depedent on one another. Opening a read
+// transaction and a write transaction in the same goroutine can cause the
+// writer to deadlock because the database periodically needs to re-mmap itself
+// as it grows and it cannot do that while a read transaction is open.
+//
+// IMPORTANT: You must close read-only transactions after you are finished or
+// else the database will not reclaim old pages.
+func (db *DB) Begin(writable bool) (*Tx, error) {
+	if writable {
+		return db.beginRWTx()
+	}
+	return db.beginTx()
+}
+
+func (db *DB) beginTx() (*Tx, error) {
+	// Lock the meta pages while we initialize the transaction. We obtain
+	// the meta lock before the mmap lock because that's the order that the
+	// write transaction will obtain them.
+	db.metalock.Lock()
+
+	// Obtain a read-only lock on the mmap. When the mmap is remapped it will
+	// obtain a write lock so all transactions must finish before it can be
+	// remapped.
+	db.mmaplock.RLock()
+
+	// Exit if the database is not open yet.
+	if !db.opened {
+		db.mmaplock.RUnlock()
+		db.metalock.Unlock()
+		return nil, ErrDatabaseNotOpen
+	}
+
+	// Create a transaction associated with the database.
+	t := &Tx{}
+	t.init(db)
+
+	// Keep track of transaction until it closes.
+	db.txs = append(db.txs, t)
+	n := len(db.txs)
+
+	// Unlock the meta pages.
+	db.metalock.Unlock()
+
+	// Update the transaction stats.
+	db.statlock.Lock()
+	db.stats.TxN++
+	db.stats.OpenTxN = n
+	db.statlock.Unlock()
+
+	return t, nil
+}
+
+func (db *DB) beginRWTx() (*Tx, error) {
+	// If the database was opened with Options.ReadOnly, return an error.
+	if db.readOnly {
+		return nil, ErrDatabaseReadOnly
+	}
+
+	// Obtain writer lock. This is released by the transaction when it closes.
+	// This enforces only one writer transaction at a time.
+	db.rwlock.Lock()
+
+	// Once we have the writer lock then we can lock the meta pages so that
+	// we can set up the transaction.
+	db.metalock.Lock()
+	defer db.metalock.Unlock()
+
+	// Exit if the database is not open yet.
+	if !db.opened {
+		db.rwlock.Unlock()
+		return nil, ErrDatabaseNotOpen
+	}
+
+	// Create a transaction associated with the database.
+	t := &Tx{writable: true}
+	t.init(db)
+	db.rwtx = t
+
+	// Free any pages associated with closed read-only transactions.
+	var minid txid = 0xFFFFFFFFFFFFFFFF
+	for _, t := range db.txs {
+		if t.meta.txid < minid {
+			minid = t.meta.txid
+		}
+	}
+	if minid > 0 {
+		db.freelist.release(minid - 1)
+	}
+
+	return t, nil
+}
+
+// removeTx removes a transaction from the database.
+func (db *DB) removeTx(tx *Tx) {
+	// Release the read lock on the mmap.
+	db.mmaplock.RUnlock()
+
+	// Use the meta lock to restrict access to the DB object.
+	db.metalock.Lock()
+
+	// Remove the transaction.
+	for i, t := range db.txs {
+		if t == tx {
+			db.txs = append(db.txs[:i], db.txs[i+1:]...)
+			break
+		}
+	}
+	n := len(db.txs)
+
+	// Unlock the meta pages.
+	db.metalock.Unlock()
+
+	// Merge statistics.
+	db.statlock.Lock()
+	db.stats.OpenTxN = n
+	db.stats.TxStats.add(&tx.stats)
+	db.statlock.Unlock()
+}
+
+// Update executes a function within the context of a read-write managed transaction.
+// If no error is returned from the function then the transaction is committed.
+// If an error is returned then the entire transaction is rolled back.
+// Any error that is returned from the function or returned from the commit is
+// returned from the Update() method.
+//
+// Attempting to manually commit or rollback within the function will cause a panic.
+func (db *DB) Update(fn func(*Tx) error) error {
+	t, err := db.Begin(true)
+	if err != nil {
+		return err
+	}
+
+	// Make sure the transaction rolls back in the event of a panic.
+	defer func() {
+		if t.db != nil {
+			t.rollback()
+		}
+	}()
+
+	// Mark as a managed tx so that the inner function cannot manually commit.
+	t.managed = true
+
+	// If an error is returned from the function then rollback and return error.
+	err = fn(t)
+	t.managed = false
+	if err != nil {
+		_ = t.Rollback()
+		return err
+	}
+
+	return t.Commit()
+}
+
+// View executes a function within the context of a managed read-only transaction.
+// Any error that is returned from the function is returned from the View() method.
+//
+// Attempting to manually rollback within the function will cause a panic.
+func (db *DB) View(fn func(*Tx) error) error {
+	t, err := db.Begin(false)
+	if err != nil {
+		return err
+	}
+
+	// Make sure the transaction rolls back in the event of a panic.
+	defer func() {
+		if t.db != nil {
+			t.rollback()
+		}
+	}()
+
+	// Mark as a managed tx so that the inner function cannot manually rollback.
+	t.managed = true
+
+	// If an error is returned from the function then pass it through.
+	err = fn(t)
+	t.managed = false
+	if err != nil {
+		_ = t.Rollback()
+		return err
+	}
+
+	if err := t.Rollback(); err != nil {
+		return err
+	}
+
+	return nil
+}
+
+// Sync executes fdatasync() against the database file handle.
+//
+// This is not necessary under normal operation, however, if you use NoSync
+// then it allows you to force the database file to sync against the disk.
+func (db *DB) Sync() error { return fdatasync(db) }
+
+// Stats retrieves ongoing performance stats for the database.
+// This is only updated when a transaction closes.
+func (db *DB) Stats() Stats {
+	db.statlock.RLock()
+	defer db.statlock.RUnlock()
+	return db.stats
+}
+
+// This is for internal access to the raw data bytes from the C cursor, use
+// carefully, or not at all.
+func (db *DB) Info() *Info {
+	return &Info{uintptr(unsafe.Pointer(&db.data[0])), db.pageSize}
+}
+
+// page retrieves a page reference from the mmap based on the current page size.
+func (db *DB) page(id pgid) *page {
+	pos := id * pgid(db.pageSize)
+	return (*page)(unsafe.Pointer(&db.data[pos]))
+}
+
+// pageInBuffer retrieves a page reference from a given byte array based on the current page size.
+func (db *DB) pageInBuffer(b []byte, id pgid) *page {
+	return (*page)(unsafe.Pointer(&b[id*pgid(db.pageSize)]))
+}
+
+// meta retrieves the current meta page reference.
+func (db *DB) meta() *meta {
+	if db.meta0.txid > db.meta1.txid {
+		return db.meta0
+	}
+	return db.meta1
+}
+
+// allocate returns a contiguous block of memory starting at a given page.
+func (db *DB) allocate(count int) (*page, error) {
+	// Allocate a temporary buffer for the page.
+	buf := make([]byte, count*db.pageSize)
+	p := (*page)(unsafe.Pointer(&buf[0]))
+	p.overflow = uint32(count - 1)
+
+	// Use pages from the freelist if they are available.
+	if p.id = db.freelist.allocate(count); p.id != 0 {
+		return p, nil
+	}
+
+	// Resize mmap() if we're at the end.
+	p.id = db.rwtx.meta.pgid
+	var minsz = int((p.id+pgid(count))+1) * db.pageSize
+	if minsz >= db.datasz {
+		if err := db.mmap(minsz); err != nil {
+			return nil, fmt.Errorf("mmap allocate error: %s", err)
+		}
+	}
+
+	// Move the page id high water mark.
+	db.rwtx.meta.pgid += pgid(count)
+
+	return p, nil
+}
+
+func (db *DB) IsReadOnly() bool {
+	return db.readOnly
+}
+
+// Options represents the options that can be set when opening a database.
+type Options struct {
+	// Timeout is the amount of time to wait to obtain a file lock.
+	// When set to zero it will wait indefinitely. This option is only
+	// available on Darwin and Linux.
+	Timeout time.Duration
+
+	// Sets the DB.NoGrowSync flag before memory mapping the file.
+	NoGrowSync bool
+
+	// Open database in read-only mode. Uses flock(..., LOCK_SH |LOCK_NB) to
+	// grab a shared lock (UNIX).
+	ReadOnly bool
+
+	// Sets the DB.MmapFlags flag before memory mapping the file.
+	MmapFlags int
+}
+
+// DefaultOptions represent the options used if nil options are passed into Open().
+// No timeout is used which will cause Bolt to wait indefinitely for a lock.
+var DefaultOptions = &Options{
+	Timeout:    0,
+	NoGrowSync: false,
+}
+
+// Stats represents statistics about the database.
+type Stats struct {
+	// Freelist stats
+	FreePageN     int // total number of free pages on the freelist
+	PendingPageN  int // total number of pending pages on the freelist
+	FreeAlloc     int // total bytes allocated in free pages
+	FreelistInuse int // total bytes used by the freelist
+
+	// Transaction stats
+	TxN     int // total number of started read transactions
+	OpenTxN int // number of currently open read transactions
+
+	TxStats TxStats // global, ongoing stats.
+}
+
+// Sub calculates and returns the difference between two sets of database stats.
+// This is useful when obtaining stats at two different points and time and
+// you need the performance counters that occurred within that time span.
+func (s *Stats) Sub(other *Stats) Stats {
+	if other == nil {
+		return *s
+	}
+	var diff Stats
+	diff.FreePageN = s.FreePageN
+	diff.PendingPageN = s.PendingPageN
+	diff.FreeAlloc = s.FreeAlloc
+	diff.FreelistInuse = s.FreelistInuse
+	diff.TxN = other.TxN - s.TxN
+	diff.TxStats = s.TxStats.Sub(&other.TxStats)
+	return diff
+}
+
+func (s *Stats) add(other *Stats) {
+	s.TxStats.add(&other.TxStats)
+}
+
+type Info struct {
+	Data     uintptr
+	PageSize int
+}
+
+type meta struct {
+	magic    uint32
+	version  uint32
+	pageSize uint32
+	flags    uint32
+	root     bucket
+	freelist pgid
+	pgid     pgid
+	txid     txid
+	checksum uint64
+}
+
+// validate checks the marker bytes and version of the meta page to ensure it matches this binary.
+func (m *meta) validate() error {
+	if m.checksum != 0 && m.checksum != m.sum64() {
+		return ErrChecksum
+	} else if m.magic != magic {
+		return ErrInvalid
+	} else if m.version != version {
+		return ErrVersionMismatch
+	}
+	return nil
+}
+
+// copy copies one meta object to another.
+func (m *meta) copy(dest *meta) {
+	*dest = *m
+}
+
+// write writes the meta onto a page.
+func (m *meta) write(p *page) {
+	if m.root.root >= m.pgid {
+		panic(fmt.Sprintf("root bucket pgid (%d) above high water mark (%d)", m.root.root, m.pgid))
+	} else if m.freelist >= m.pgid {
+		panic(fmt.Sprintf("freelist pgid (%d) above high water mark (%d)", m.freelist, m.pgid))
+	}
+
+	// Page id is either going to be 0 or 1 which we can determine by the transaction ID.
+	p.id = pgid(m.txid % 2)
+	p.flags |= metaPageFlag
+
+	// Calculate the checksum.
+	m.checksum = m.sum64()
+
+	m.copy(p.meta())
+}
+
+// generates the checksum for the meta.
+func (m *meta) sum64() uint64 {
+	var h = fnv.New64a()
+	_, _ = h.Write((*[unsafe.Offsetof(meta{}.checksum)]byte)(unsafe.Pointer(m))[:])
+	return h.Sum64()
+}
+
+// _assert will panic with a given formatted message if the given condition is false.
+func _assert(condition bool, msg string, v ...interface{}) {
+	if !condition {
+		panic(fmt.Sprintf("assertion failed: "+msg, v...))
+	}
+}
+
+func warn(v ...interface{})              { fmt.Fprintln(os.Stderr, v...) }
+func warnf(msg string, v ...interface{}) { fmt.Fprintf(os.Stderr, msg+"\n", v...) }
+
+func printstack() {
+	stack := strings.Join(strings.Split(string(debug.Stack()), "\n")[2:], "\n")
+	fmt.Fprintln(os.Stderr, stack)
+}

weed/vendor/github.com/boltdb/bolt/doc.go

@@ -0,0 +1,44 @@
+/*
+Package bolt implements a low-level key/value store in pure Go. It supports
+fully serializable transactions, ACID semantics, and lock-free MVCC with
+multiple readers and a single writer. Bolt can be used for projects that
+want a simple data store without the need to add large dependencies such as
+Postgres or MySQL.
+
+Bolt is a single-level, zero-copy, B+tree data store. This means that Bolt is
+optimized for fast read access and does not require recovery in the event of a
+system crash. Transactions which have not finished committing will simply be
+rolled back in the event of a crash.
+
+The design of Bolt is based on Howard Chu's LMDB database project.
+
+Bolt currently works on Windows, Mac OS X, and Linux.
+
+
+Basics
+
+There are only a few types in Bolt: DB, Bucket, Tx, and Cursor. The DB is
+a collection of buckets and is represented by a single file on disk. A bucket is
+a collection of unique keys that are associated with values.
+
+Transactions provide either read-only or read-write access to the database.
+Read-only transactions can retrieve key/value pairs and can use Cursors to
+iterate over the dataset sequentially. Read-write transactions can create and
+delete buckets and can insert and remove keys. Only one read-write transaction
+is allowed at a time.
+
+
+Caveats
+
+The database uses a read-only, memory-mapped data file to ensure that
+applications cannot corrupt the database, however, this means that keys and
+values returned from Bolt cannot be changed. Writing to a read-only byte slice
+will cause Go to panic.
+
+Keys and values retrieved from the database are only valid for the life of
+the transaction. When used outside the transaction, these byte slices can
+point to different data or can point to invalid memory which will cause a panic.
+
+
+*/
+package bolt

weed/vendor/github.com/boltdb/bolt/errors.go

@@ -0,0 +1,70 @@
+package bolt
+
+import "errors"
+
+// These errors can be returned when opening or calling methods on a DB.
+var (
+	// ErrDatabaseNotOpen is returned when a DB instance is accessed before it
+	// is opened or after it is closed.
+	ErrDatabaseNotOpen = errors.New("database not open")
+
+	// ErrDatabaseOpen is returned when opening a database that is
+	// already open.
+	ErrDatabaseOpen = errors.New("database already open")
+
+	// ErrInvalid is returned when a data file is not a Bolt-formatted database.
+	ErrInvalid = errors.New("invalid database")
+
+	// ErrVersionMismatch is returned when the data file was created with a
+	// different version of Bolt.
+	ErrVersionMismatch = errors.New("version mismatch")
+
+	// ErrChecksum is returned when either meta page checksum does not match.
+	ErrChecksum = errors.New("checksum error")
+
+	// ErrTimeout is returned when a database cannot obtain an exclusive lock
+	// on the data file after the timeout passed to Open().
+	ErrTimeout = errors.New("timeout")
+)
+
+// These errors can occur when beginning or committing a Tx.
+var (
+	// ErrTxNotWritable is returned when performing a write operation on a
+	// read-only transaction.
+	ErrTxNotWritable = errors.New("tx not writable")
+
+	// ErrTxClosed is returned when committing or rolling back a transaction
+	// that has already been committed or rolled back.
+	ErrTxClosed = errors.New("tx closed")
+
+	// ErrDatabaseReadOnly is returned when a mutating transaction is started on a
+	// read-only database.
+	ErrDatabaseReadOnly = errors.New("database is in read-only mode")
+)
+
+// These errors can occur when putting or deleting a value or a bucket.
+var (
+	// ErrBucketNotFound is returned when trying to access a bucket that has
+	// not been created yet.
+	ErrBucketNotFound = errors.New("bucket not found")
+
+	// ErrBucketExists is returned when creating a bucket that already exists.
+	ErrBucketExists = errors.New("bucket already exists")
+
+	// ErrBucketNameRequired is returned when creating a bucket with a blank name.
+	ErrBucketNameRequired = errors.New("bucket name required")
+
+	// ErrKeyRequired is returned when inserting a zero-length key.
+	ErrKeyRequired = errors.New("key required")
+
+	// ErrKeyTooLarge is returned when inserting a key that is larger than MaxKeySize.
+	ErrKeyTooLarge = errors.New("key too large")
+
+	// ErrValueTooLarge is returned when inserting a value that is larger than MaxValueSize.
+	ErrValueTooLarge = errors.New("value too large")
+
+	// ErrIncompatibleValue is returned when trying create or delete a bucket
+	// on an existing non-bucket key or when trying to create or delete a
+	// non-bucket key on an existing bucket key.
+	ErrIncompatibleValue = errors.New("incompatible value")
+)

weed/vendor/github.com/boltdb/bolt/freelist.go

@@ -0,0 +1,242 @@
+package bolt
+
+import (
+	"fmt"
+	"sort"
+	"unsafe"
+)
+
+// freelist represents a list of all pages that are available for allocation.
+// It also tracks pages that have been freed but are still in use by open transactions.
+type freelist struct {
+	ids     []pgid          // all free and available free page ids.
+	pending map[txid][]pgid // mapping of soon-to-be free page ids by tx.
+	cache   map[pgid]bool   // fast lookup of all free and pending page ids.
+}
+
+// newFreelist returns an empty, initialized freelist.
+func newFreelist() *freelist {
+	return &freelist{
+		pending: make(map[txid][]pgid),
+		cache:   make(map[pgid]bool),
+	}
+}
+
+// size returns the size of the page after serialization.
+func (f *freelist) size() int {
+	return pageHeaderSize + (int(unsafe.Sizeof(pgid(0))) * f.count())
+}
+
+// count returns count of pages on the freelist
+func (f *freelist) count() int {
+	return f.free_count() + f.pending_count()
+}
+
+// free_count returns count of free pages
+func (f *freelist) free_count() int {
+	return len(f.ids)
+}
+
+// pending_count returns count of pending pages
+func (f *freelist) pending_count() int {
+	var count int
+	for _, list := range f.pending {
+		count += len(list)
+	}
+	return count
+}
+
+// all returns a list of all free ids and all pending ids in one sorted list.
+func (f *freelist) all() []pgid {
+	m := make(pgids, 0)
+
+	for _, list := range f.pending {
+		m = append(m, list...)
+	}
+
+	sort.Sort(m)
+	return pgids(f.ids).merge(m)
+}
+
+// allocate returns the starting page id of a contiguous list of pages of a given size.
+// If a contiguous block cannot be found then 0 is returned.
+func (f *freelist) allocate(n int) pgid {
+	if len(f.ids) == 0 {
+		return 0
+	}
+
+	var initial, previd pgid
+	for i, id := range f.ids {
+		if id <= 1 {
+			panic(fmt.Sprintf("invalid page allocation: %d", id))
+		}
+
+		// Reset initial page if this is not contiguous.
+		if previd == 0 || id-previd != 1 {
+			initial = id
+		}
+
+		// If we found a contiguous block then remove it and return it.
+		if (id-initial)+1 == pgid(n) {
+			// If we're allocating off the beginning then take the fast path
+			// and just adjust the existing slice. This will use extra memory
+			// temporarily but the append() in free() will realloc the slice
+			// as is necessary.
+			if (i + 1) == n {
+				f.ids = f.ids[i+1:]
+			} else {
+				copy(f.ids[i-n+1:], f.ids[i+1:])
+				f.ids = f.ids[:len(f.ids)-n]
+			}
+
+			// Remove from the free cache.
+			for i := pgid(0); i < pgid(n); i++ {
+				delete(f.cache, initial+i)
+			}
+
+			return initial
+		}
+
+		previd = id
+	}
+	return 0
+}
+
+// free releases a page and its overflow for a given transaction id.
+// If the page is already free then a panic will occur.
+func (f *freelist) free(txid txid, p *page) {
+	if p.id <= 1 {
+		panic(fmt.Sprintf("cannot free page 0 or 1: %d", p.id))
+	}
+
+	// Free page and all its overflow pages.
+	var ids = f.pending[txid]
+	for id := p.id; id <= p.id+pgid(p.overflow); id++ {
+		// Verify that page is not already free.
+		if f.cache[id] {
+			panic(fmt.Sprintf("page %d already freed", id))
+		}
+
+		// Add to the freelist and cache.
+		ids = append(ids, id)
+		f.cache[id] = true
+	}
+	f.pending[txid] = ids
+}
+
+// release moves all page ids for a transaction id (or older) to the freelist.
+func (f *freelist) release(txid txid) {
+	m := make(pgids, 0)
+	for tid, ids := range f.pending {
+		if tid <= txid {
+			// Move transaction's pending pages to the available freelist.
+			// Don't remove from the cache since the page is still free.
+			m = append(m, ids...)
+			delete(f.pending, tid)
+		}
+	}
+	sort.Sort(m)
+	f.ids = pgids(f.ids).merge(m)
+}
+
+// rollback removes the pages from a given pending tx.
+func (f *freelist) rollback(txid txid) {
+	// Remove page ids from cache.
+	for _, id := range f.pending[txid] {
+		delete(f.cache, id)
+	}
+
+	// Remove pages from pending list.
+	delete(f.pending, txid)
+}
+
+// freed returns whether a given page is in the free list.
+func (f *freelist) freed(pgid pgid) bool {
+	return f.cache[pgid]
+}
+
+// read initializes the freelist from a freelist page.
+func (f *freelist) read(p *page) {
+	// If the page.count is at the max uint16 value (64k) then it's considered
+	// an overflow and the size of the freelist is stored as the first element.
+	idx, count := 0, int(p.count)
+	if count == 0xFFFF {
+		idx = 1
+		count = int(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0])
+	}
+
+	// Copy the list of page ids from the freelist.
+	ids := ((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[idx:count]
+	f.ids = make([]pgid, len(ids))
+	copy(f.ids, ids)
+
+	// Make sure they're sorted.
+	sort.Sort(pgids(f.ids))
+
+	// Rebuild the page cache.
+	f.reindex()
+}
+
+// write writes the page ids onto a freelist page. All free and pending ids are
+// saved to disk since in the event of a program crash, all pending ids will
+// become free.
+func (f *freelist) write(p *page) error {
+	// Combine the old free pgids and pgids waiting on an open transaction.
+	ids := f.all()
+
+	// Update the header flag.
+	p.flags |= freelistPageFlag
+
+	// The page.count can only hold up to 64k elements so if we overflow that
+	// number then we handle it by putting the size in the first element.
+	if len(ids) < 0xFFFF {
+		p.count = uint16(len(ids))
+		copy(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[:], ids)
+	} else {
+		p.count = 0xFFFF
+		((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[0] = pgid(len(ids))
+		copy(((*[maxAllocSize]pgid)(unsafe.Pointer(&p.ptr)))[1:], ids)
+	}
+
+	return nil
+}
+
+// reload reads the freelist from a page and filters out pending items.
+func (f *freelist) reload(p *page) {
+	f.read(p)
+
+	// Build a cache of only pending pages.
+	pcache := make(map[pgid]bool)
+	for _, pendingIDs := range f.pending {
+		for _, pendingID := range pendingIDs {
+			pcache[pendingID] = true
+		}
+	}
+
+	// Check each page in the freelist and build a new available freelist
+	// with any pages not in the pending lists.
+	var a []pgid
+	for _, id := range f.ids {
+		if !pcache[id] {
+			a = append(a, id)
+		}
+	}
+	f.ids = a
+
+	// Once the available list is rebuilt then rebuild the free cache so that
+	// it includes the available and pending free pages.
+	f.reindex()
+}
+
+// reindex rebuilds the free cache based on available and pending free lists.
+func (f *freelist) reindex() {
+	f.cache = make(map[pgid]bool)
+	for _, id := range f.ids {
+		f.cache[id] = true
+	}
+	for _, pendingIDs := range f.pending {
+		for _, pendingID := range pendingIDs {
+			f.cache[pendingID] = true
+		}
+	}
+}

weed/vendor/github.com/boltdb/bolt/node.go

@@ -0,0 +1,636 @@
+package bolt
+
+import (
+	"bytes"
+	"fmt"
+	"sort"
+	"unsafe"
+)
+
+// node represents an in-memory, deserialized page.
+type node struct {
+	bucket     *Bucket
+	isLeaf     bool
+	unbalanced bool
+	spilled    bool
+	key        []byte
+	pgid       pgid
+	parent     *node
+	children   nodes
+	inodes     inodes
+}
+
+// root returns the top-level node this node is attached to.
+func (n *node) root() *node {
+	if n.parent == nil {
+		return n
+	}
+	return n.parent.root()
+}
+
+// minKeys returns the minimum number of inodes this node should have.
+func (n *node) minKeys() int {
+	if n.isLeaf {
+		return 1
+	}
+	return 2
+}
+
+// size returns the size of the node after serialization.
+func (n *node) size() int {
+	sz, elsz := pageHeaderSize, n.pageElementSize()
+	for i := 0; i < len(n.inodes); i++ {
+		item := &n.inodes[i]
+		sz += elsz + len(item.key) + len(item.value)
+	}
+	return sz
+}
+
+// sizeLessThan returns true if the node is less than a given size.
+// This is an optimization to avoid calculating a large node when we only need
+// to know if it fits inside a certain page size.
+func (n *node) sizeLessThan(v int) bool {
+	sz, elsz := pageHeaderSize, n.pageElementSize()
+	for i := 0; i < len(n.inodes); i++ {
+		item := &n.inodes[i]
+		sz += elsz + len(item.key) + len(item.value)
+		if sz >= v {
+			return false
+		}
+	}
+	return true
+}
+
+// pageElementSize returns the size of each page element based on the type of node.
+func (n *node) pageElementSize() int {
+	if n.isLeaf {
+		return leafPageElementSize
+	}
+	return branchPageElementSize
+}
+
+// childAt returns the child node at a given index.
+func (n *node) childAt(index int) *node {
+	if n.isLeaf {
+		panic(fmt.Sprintf("invalid childAt(%d) on a leaf node", index))
+	}
+	return n.bucket.node(n.inodes[index].pgid, n)
+}
+
+// childIndex returns the index of a given child node.
+func (n *node) childIndex(child *node) int {
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, child.key) != -1 })
+	return index
+}
+
+// numChildren returns the number of children.
+func (n *node) numChildren() int {
+	return len(n.inodes)
+}
+
+// nextSibling returns the next node with the same parent.
+func (n *node) nextSibling() *node {
+	if n.parent == nil {
+		return nil
+	}
+	index := n.parent.childIndex(n)
+	if index >= n.parent.numChildren()-1 {
+		return nil
+	}
+	return n.parent.childAt(index + 1)
+}
+
+// prevSibling returns the previous node with the same parent.
+func (n *node) prevSibling() *node {
+	if n.parent == nil {
+		return nil
+	}
+	index := n.parent.childIndex(n)
+	if index == 0 {
+		return nil
+	}
+	return n.parent.childAt(index - 1)
+}
+
+// put inserts a key/value.
+func (n *node) put(oldKey, newKey, value []byte, pgid pgid, flags uint32) {
+	if pgid >= n.bucket.tx.meta.pgid {
+		panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", pgid, n.bucket.tx.meta.pgid))
+	} else if len(oldKey) <= 0 {
+		panic("put: zero-length old key")
+	} else if len(newKey) <= 0 {
+		panic("put: zero-length new key")
+	}
+
+	// Find insertion index.
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, oldKey) != -1 })
+
+	// Add capacity and shift nodes if we don't have an exact match and need to insert.
+	exact := (len(n.inodes) > 0 && index < len(n.inodes) && bytes.Equal(n.inodes[index].key, oldKey))
+	if !exact {
+		n.inodes = append(n.inodes, inode{})
+		copy(n.inodes[index+1:], n.inodes[index:])
+	}
+
+	inode := &n.inodes[index]
+	inode.flags = flags
+	inode.key = newKey
+	inode.value = value
+	inode.pgid = pgid
+	_assert(len(inode.key) > 0, "put: zero-length inode key")
+}
+
+// del removes a key from the node.
+func (n *node) del(key []byte) {
+	// Find index of key.
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].key, key) != -1 })
+
+	// Exit if the key isn't found.
+	if index >= len(n.inodes) || !bytes.Equal(n.inodes[index].key, key) {
+		return
+	}
+
+	// Delete inode from the node.
+	n.inodes = append(n.inodes[:index], n.inodes[index+1:]...)
+
+	// Mark the node as needing rebalancing.
+	n.unbalanced = true
+}
+
+// read initializes the node from a page.
+func (n *node) read(p *page) {
+	n.pgid = p.id
+	n.isLeaf = ((p.flags & leafPageFlag) != 0)
+	n.inodes = make(inodes, int(p.count))
+
+	for i := 0; i < int(p.count); i++ {
+		inode := &n.inodes[i]
+		if n.isLeaf {
+			elem := p.leafPageElement(uint16(i))
+			inode.flags = elem.flags
+			inode.key = elem.key()
+			inode.value = elem.value()
+		} else {
+			elem := p.branchPageElement(uint16(i))
+			inode.pgid = elem.pgid
+			inode.key = elem.key()
+		}
+		_assert(len(inode.key) > 0, "read: zero-length inode key")
+	}
+
+	// Save first key so we can find the node in the parent when we spill.
+	if len(n.inodes) > 0 {
+		n.key = n.inodes[0].key
+		_assert(len(n.key) > 0, "read: zero-length node key")
+	} else {
+		n.key = nil
+	}
+}
+
+// write writes the items onto one or more pages.
+func (n *node) write(p *page) {
+	// Initialize page.
+	if n.isLeaf {
+		p.flags |= leafPageFlag
+	} else {
+		p.flags |= branchPageFlag
+	}
+
+	if len(n.inodes) >= 0xFFFF {
+		panic(fmt.Sprintf("inode overflow: %d (pgid=%d)", len(n.inodes), p.id))
+	}
+	p.count = uint16(len(n.inodes))
+
+	// Loop over each item and write it to the page.
+	b := (*[maxAllocSize]byte)(unsafe.Pointer(&p.ptr))[n.pageElementSize()*len(n.inodes):]
+	for i, item := range n.inodes {
+		_assert(len(item.key) > 0, "write: zero-length inode key")
+
+		// Write the page element.
+		if n.isLeaf {
+			elem := p.leafPageElement(uint16(i))
+			elem.pos = uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem)))
+			elem.flags = item.flags
+			elem.ksize = uint32(len(item.key))
+			elem.vsize = uint32(len(item.value))
+		} else {
+			elem := p.branchPageElement(uint16(i))
+			elem.pos = uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem)))
+			elem.ksize = uint32(len(item.key))
+			elem.pgid = item.pgid
+			_assert(elem.pgid != p.id, "write: circular dependency occurred")
+		}
+
+		// If the length of key+value is larger than the max allocation size
+		// then we need to reallocate the byte array pointer.
+		//
+		// See: https://github.com/boltdb/bolt/pull/335
+		klen, vlen := len(item.key), len(item.value)
+		if len(b) < klen+vlen {
+			b = (*[maxAllocSize]byte)(unsafe.Pointer(&b[0]))[:]
+		}
+
+		// Write data for the element to the end of the page.
+		copy(b[0:], item.key)
+		b = b[klen:]
+		copy(b[0:], item.value)
+		b = b[vlen:]
+	}
+
+	// DEBUG ONLY: n.dump()
+}
+
+// split breaks up a node into multiple smaller nodes, if appropriate.
+// This should only be called from the spill() function.
+func (n *node) split(pageSize int) []*node {
+	var nodes []*node
+
+	node := n
+	for {
+		// Split node into two.
+		a, b := node.splitTwo(pageSize)
+		nodes = append(nodes, a)
+
+		// If we can't split then exit the loop.
+		if b == nil {
+			break
+		}
+
+		// Set node to b so it gets split on the next iteration.
+		node = b
+	}
+
+	return nodes
+}
+
+// splitTwo breaks up a node into two smaller nodes, if appropriate.
+// This should only be called from the split() function.
+func (n *node) splitTwo(pageSize int) (*node, *node) {
+	// Ignore the split if the page doesn't have at least enough nodes for
+	// two pages or if the nodes can fit in a single page.
+	if len(n.inodes) <= (minKeysPerPage*2) || n.sizeLessThan(pageSize) {
+		return n, nil
+	}
+
+	// Determine the threshold before starting a new node.
+	var fillPercent = n.bucket.FillPercent
+	if fillPercent < minFillPercent {
+		fillPercent = minFillPercent
+	} else if fillPercent > maxFillPercent {
+		fillPercent = maxFillPercent
+	}
+	threshold := int(float64(pageSize) * fillPercent)
+
+	// Determine split position and sizes of the two pages.
+	splitIndex, _ := n.splitIndex(threshold)
+
+	// Split node into two separate nodes.
+	// If there's no parent then we'll need to create one.
+	if n.parent == nil {
+		n.parent = &node{bucket: n.bucket, children: []*node{n}}
+	}
+
+	// Create a new node and add it to the parent.
+	next := &node{bucket: n.bucket, isLeaf: n.isLeaf, parent: n.parent}
+	n.parent.children = append(n.parent.children, next)
+
+	// Split inodes across two nodes.
+	next.inodes = n.inodes[splitIndex:]
+	n.inodes = n.inodes[:splitIndex]
+
+	// Update the statistics.
+	n.bucket.tx.stats.Split++
+
+	return n, next
+}
+
+// splitIndex finds the position where a page will fill a given threshold.
+// It returns the index as well as the size of the first page.
+// This is only be called from split().
+func (n *node) splitIndex(threshold int) (index, sz int) {
+	sz = pageHeaderSize
+
+	// Loop until we only have the minimum number of keys required for the second page.
+	for i := 0; i < len(n.inodes)-minKeysPerPage; i++ {
+		index = i
+		inode := n.inodes[i]
+		elsize := n.pageElementSize() + len(inode.key) + len(inode.value)
+
+		// If we have at least the minimum number of keys and adding another
+		// node would put us over the threshold then exit and return.
+		if i >= minKeysPerPage && sz+elsize > threshold {
+			break
+		}
+
+		// Add the element size to the total size.
+		sz += elsize
+	}
+
+	return
+}
+
+// spill writes the nodes to dirty pages and splits nodes as it goes.
+// Returns an error if dirty pages cannot be allocated.
+func (n *node) spill() error {
+	var tx = n.bucket.tx
+	if n.spilled {
+		return nil
+	}
+
+	// Spill child nodes first. Child nodes can materialize sibling nodes in
+	// the case of split-merge so we cannot use a range loop. We have to check
+	// the children size on every loop iteration.
+	sort.Sort(n.children)
+	for i := 0; i < len(n.children); i++ {
+		if err := n.children[i].spill(); err != nil {
+			return err
+		}
+	}
+
+	// We no longer need the child list because it's only used for spill tracking.
+	n.children = nil
+
+	// Split nodes into appropriate sizes. The first node will always be n.
+	var nodes = n.split(tx.db.pageSize)
+	for _, node := range nodes {
+		// Add node's page to the freelist if it's not new.
+		if node.pgid > 0 {
+			tx.db.freelist.free(tx.meta.txid, tx.page(node.pgid))
+			node.pgid = 0
+		}
+
+		// Allocate contiguous space for the node.
+		p, err := tx.allocate((node.size() / tx.db.pageSize) + 1)
+		if err != nil {
+			return err
+		}
+
+		// Write the node.
+		if p.id >= tx.meta.pgid {
+			panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", p.id, tx.meta.pgid))
+		}
+		node.pgid = p.id
+		node.write(p)
+		node.spilled = true
+
+		// Insert into parent inodes.
+		if node.parent != nil {
+			var key = node.key
+			if key == nil {
+				key = node.inodes[0].key
+			}
+
+			node.parent.put(key, node.inodes[0].key, nil, node.pgid, 0)
+			node.key = node.inodes[0].key
+			_assert(len(node.key) > 0, "spill: zero-length node key")
+		}
+
+		// Update the statistics.
+		tx.stats.Spill++
+	}
+
+	// If the root node split and created a new root then we need to spill that
+	// as well. We'll clear out the children to make sure it doesn't try to respill.
+	if n.parent != nil && n.parent.pgid == 0 {
+		n.children = nil
+		return n.parent.spill()
+	}
+
+	return nil
+}
+
+// rebalance attempts to combine the node with sibling nodes if the node fill
+// size is below a threshold or if there are not enough keys.
+func (n *node) rebalance() {
+	if !n.unbalanced {
+		return
+	}
+	n.unbalanced = false
+
+	// Update statistics.
+	n.bucket.tx.stats.Rebalance++
+
+	// Ignore if node is above threshold (25%) and has enough keys.
+	var threshold = n.bucket.tx.db.pageSize / 4
+	if n.size() > threshold && len(n.inodes) > n.minKeys() {
+		return
+	}
+
+	// Root node has special handling.
+	if n.parent == nil {
+		// If root node is a branch and only has one node then collapse it.
+		if !n.isLeaf && len(n.inodes) == 1 {
+			// Move root's child up.
+			child := n.bucket.node(n.inodes[0].pgid, n)
+			n.isLeaf = child.isLeaf
+			n.inodes = child.inodes[:]
+			n.children = child.children
+
+			// Reparent all child nodes being moved.
+			for _, inode := range n.inodes {
+				if child, ok := n.bucket.nodes[inode.pgid]; ok {
+					child.parent = n
+				}
+			}
+
+			// Remove old child.
+			child.parent = nil
+			delete(n.bucket.nodes, child.pgid)
+			child.free()
+		}
+
+		return
+	}
+
+	// If node has no keys then just remove it.
+	if n.numChildren() == 0 {
+		n.parent.del(n.key)
+		n.parent.removeChild(n)
+		delete(n.bucket.nodes, n.pgid)
+		n.free()
+		n.parent.rebalance()
+		return
+	}
+
+	_assert(n.parent.numChildren() > 1, "parent must have at least 2 children")
+
+	// Destination node is right sibling if idx == 0, otherwise left sibling.
+	var target *node
+	var useNextSibling = (n.parent.childIndex(n) == 0)
+	if useNextSibling {
+		target = n.nextSibling()
+	} else {
+		target = n.prevSibling()
+	}
+
+	// If target node has extra nodes then just move one over.
+	if target.numChildren() > target.minKeys() {
+		if useNextSibling {
+			// Reparent and move node.
+			if child, ok := n.bucket.nodes[target.inodes[0].pgid]; ok {
+				child.parent.removeChild(child)
+				child.parent = n
+				child.parent.children = append(child.parent.children, child)
+			}
+			n.inodes = append(n.inodes, target.inodes[0])
+			target.inodes = target.inodes[1:]
+
+			// Update target key on parent.
+			target.parent.put(target.key, target.inodes[0].key, nil, target.pgid, 0)
+			target.key = target.inodes[0].key
+			_assert(len(target.key) > 0, "rebalance(1): zero-length node key")
+		} else {
+			// Reparent and move node.
+			if child, ok := n.bucket.nodes[target.inodes[len(target.inodes)-1].pgid]; ok {
+				child.parent.removeChild(child)
+				child.parent = n
+				child.parent.children = append(child.parent.children, child)
+			}
+			n.inodes = append(n.inodes, inode{})
+			copy(n.inodes[1:], n.inodes)
+			n.inodes[0] = target.inodes[len(target.inodes)-1]
+			target.inodes = target.inodes[:len(target.inodes)-1]
+		}
+
+		// Update parent key for node.
+		n.parent.put(n.key, n.inodes[0].key, nil, n.pgid, 0)
+		n.key = n.inodes[0].key
+		_assert(len(n.key) > 0, "rebalance(2): zero-length node key")
+
+		return
+	}
+
+	// If both this node and the target node are too small then merge them.
+	if useNextSibling {
+		// Reparent all child nodes being moved.
+		for _, inode := range target.inodes {
+			if child, ok := n.bucket.nodes[inode.pgid]; ok {
+				child.parent.removeChild(child)
+				child.parent = n
+				child.parent.children = append(child.parent.children, child)
+			}
+		}
+
+		// Copy over inodes from target and remove target.
+		n.inodes = append(n.inodes, target.inodes...)
+		n.parent.del(target.key)
+		n.parent.removeChild(target)
+		delete(n.bucket.nodes, target.pgid)
+		target.free()
+	} else {
+		// Reparent all child nodes being moved.
+		for _, inode := range n.inodes {
+			if child, ok := n.bucket.nodes[inode.pgid]; ok {
+				child.parent.removeChild(child)
+				child.parent = target
+				child.parent.children = append(child.parent.children, child)
+			}
+		}
+
+		// Copy over inodes to target and remove node.
+		target.inodes = append(target.inodes, n.inodes...)
+		n.parent.del(n.key)
+		n.parent.removeChild(n)
+		delete(n.bucket.nodes, n.pgid)
+		n.free()
+	}
+
+	// Either this node or the target node was deleted from the parent so rebalance it.
+	n.parent.rebalance()
+}
+
+// removes a node from the list of in-memory children.
+// This does not affect the inodes.
+func (n *node) removeChild(target *node) {
+	for i, child := range n.children {
+		if child == target {
+			n.children = append(n.children[:i], n.children[i+1:]...)
+			return
+		}
+	}
+}
+
+// dereference causes the node to copy all its inode key/value references to heap memory.
+// This is required when the mmap is reallocated so inodes are not pointing to stale data.
+func (n *node) dereference() {
+	if n.key != nil {
+		key := make([]byte, len(n.key))
+		copy(key, n.key)
+		n.key = key
+		_assert(n.pgid == 0 || len(n.key) > 0, "dereference: zero-length node key on existing node")
+	}
+
+	for i := range n.inodes {
+		inode := &n.inodes[i]
+
+		key := make([]byte, len(inode.key))
+		copy(key, inode.key)
+		inode.key = key
+		_assert(len(inode.key) > 0, "dereference: zero-length inode key")
+
+		value := make([]byte, len(inode.value))
+		copy(value, inode.value)
+		inode.value = value
+	}
+
+	// Recursively dereference children.
+	for _, child := range n.children {
+		child.dereference()
+	}
+
+	// Update statistics.
+	n.bucket.tx.stats.NodeDeref++
+}
+
+// free adds the node's underlying page to the freelist.
+func (n *node) free() {
+	if n.pgid != 0 {
+		n.bucket.tx.db.freelist.free(n.bucket.tx.meta.txid, n.bucket.tx.page(n.pgid))
+		n.pgid = 0
+	}
+}
+
+// dump writes the contents of the node to STDERR for debugging purposes.
+/*
+func (n *node) dump() {
+	// Write node header.
+	var typ = "branch"
+	if n.isLeaf {
+		typ = "leaf"
+	}
+	warnf("[NODE %d {type=%s count=%d}]", n.pgid, typ, len(n.inodes))
+
+	// Write out abbreviated version of each item.
+	for _, item := range n.inodes {
+		if n.isLeaf {
+			if item.flags&bucketLeafFlag != 0 {
+				bucket := (*bucket)(unsafe.Pointer(&item.value[0]))
+				warnf("+L %08x -> (bucket root=%d)", trunc(item.key, 4), bucket.root)
+			} else {
+				warnf("+L %08x -> %08x", trunc(item.key, 4), trunc(item.value, 4))
+			}
+		} else {
+			warnf("+B %08x -> pgid=%d", trunc(item.key, 4), item.pgid)
+		}
+	}
+	warn("")
+}
+*/
+
+type nodes []*node
+
+func (s nodes) Len() int           { return len(s) }
+func (s nodes) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
+func (s nodes) Less(i, j int) bool { return bytes.Compare(s[i].inodes[0].key, s[j].inodes[0].key) == -1 }
+
+// inode represents an internal node inside of a node.
+// It can be used to point to elements in a page or point
+// to an element which hasn't been added to a page yet.
+type inode struct {
+	flags uint32
+	pgid  pgid
+	key   []byte
+	value []byte
+}
+
+type inodes []inode

weed/vendor/github.com/boltdb/bolt/page.go

@@ -0,0 +1,172 @@
+package bolt
+
+import (
+	"fmt"
+	"os"
+	"sort"
+	"unsafe"
+)
+
+const pageHeaderSize = int(unsafe.Offsetof(((*page)(nil)).ptr))
+
+const minKeysPerPage = 2
+
+const branchPageElementSize = int(unsafe.Sizeof(branchPageElement{}))
+const leafPageElementSize = int(unsafe.Sizeof(leafPageElement{}))
+
+const (
+	branchPageFlag   = 0x01
+	leafPageFlag     = 0x02
+	metaPageFlag     = 0x04
+	freelistPageFlag = 0x10
+)
+
+const (
+	bucketLeafFlag = 0x01
+)
+
+type pgid uint64
+
+type page struct {
+	id       pgid
+	flags    uint16
+	count    uint16
+	overflow uint32
+	ptr      uintptr
+}
+
+// typ returns a human readable page type string used for debugging.
+func (p *page) typ() string {
+	if (p.flags & branchPageFlag) != 0 {
+		return "branch"
+	} else if (p.flags & leafPageFlag) != 0 {
+		return "leaf"
+	} else if (p.flags & metaPageFlag) != 0 {
+		return "meta"
+	} else if (p.flags & freelistPageFlag) != 0 {
+		return "freelist"
+	}
+	return fmt.Sprintf("unknown<%02x>", p.flags)
+}
+
+// meta returns a pointer to the metadata section of the page.
+func (p *page) meta() *meta {
+	return (*meta)(unsafe.Pointer(&p.ptr))
+}
+
+// leafPageElement retrieves the leaf node by index
+func (p *page) leafPageElement(index uint16) *leafPageElement {
+	n := &((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[index]
+	return n
+}
+
+// leafPageElements retrieves a list of leaf nodes.
+func (p *page) leafPageElements() []leafPageElement {
+	return ((*[0x7FFFFFF]leafPageElement)(unsafe.Pointer(&p.ptr)))[:]
+}
+
+// branchPageElement retrieves the branch node by index
+func (p *page) branchPageElement(index uint16) *branchPageElement {
+	return &((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[index]
+}
+
+// branchPageElements retrieves a list of branch nodes.
+func (p *page) branchPageElements() []branchPageElement {
+	return ((*[0x7FFFFFF]branchPageElement)(unsafe.Pointer(&p.ptr)))[:]
+}
+
+// dump writes n bytes of the page to STDERR as hex output.
+func (p *page) hexdump(n int) {
+	buf := (*[maxAllocSize]byte)(unsafe.Pointer(p))[:n]
+	fmt.Fprintf(os.Stderr, "%x\n", buf)
+}
+
+type pages []*page
+
+func (s pages) Len() int           { return len(s) }
+func (s pages) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
+func (s pages) Less(i, j int) bool { return s[i].id < s[j].id }
+
+// branchPageElement represents a node on a branch page.
+type branchPageElement struct {
+	pos   uint32
+	ksize uint32
+	pgid  pgid
+}
+
+// key returns a byte slice of the node key.
+func (n *branchPageElement) key() []byte {
+	buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
+	return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos]))[:n.ksize]
+}
+
+// leafPageElement represents a node on a leaf page.
+type leafPageElement struct {
+	flags uint32
+	pos   uint32
+	ksize uint32
+	vsize uint32
+}
+
+// key returns a byte slice of the node key.
+func (n *leafPageElement) key() []byte {
+	buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
+	return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos]))[:n.ksize]
+}
+
+// value returns a byte slice of the node value.
+func (n *leafPageElement) value() []byte {
+	buf := (*[maxAllocSize]byte)(unsafe.Pointer(n))
+	return (*[maxAllocSize]byte)(unsafe.Pointer(&buf[n.pos+n.ksize]))[:n.vsize]
+}
+
+// PageInfo represents human readable information about a page.
+type PageInfo struct {
+	ID            int
+	Type          string
+	Count         int
+	OverflowCount int
+}
+
+type pgids []pgid
+
+func (s pgids) Len() int           { return len(s) }
+func (s pgids) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
+func (s pgids) Less(i, j int) bool { return s[i] < s[j] }
+
+// merge returns the sorted union of a and b.
+func (a pgids) merge(b pgids) pgids {
+	// Return the opposite slice if one is nil.
+	if len(a) == 0 {
+		return b
+	} else if len(b) == 0 {
+		return a
+	}
+
+	// Create a list to hold all elements from both lists.
+	merged := make(pgids, 0, len(a)+len(b))
+
+	// Assign lead to the slice with a lower starting value, follow to the higher value.
+	lead, follow := a, b
+	if b[0] < a[0] {
+		lead, follow = b, a
+	}
+
+	// Continue while there are elements in the lead.
+	for len(lead) > 0 {
+		// Merge largest prefix of lead that is ahead of follow[0].
+		n := sort.Search(len(lead), func(i int) bool { return lead[i] > follow[0] })
+		merged = append(merged, lead[:n]...)
+		if n >= len(lead) {
+			break
+		}
+
+		// Swap lead and follow.
+		lead, follow = follow, lead[n:]
+	}
+
+	// Append what's left in follow.
+	merged = append(merged, follow...)
+
+	return merged
+}

weed/vendor/github.com/boltdb/bolt/tx.go

@@ -0,0 +1,624 @@
+package bolt
+
+import (
+	"fmt"
+	"io"
+	"os"
+	"sort"
+	"time"
+	"unsafe"
+)
+
+// txid represents the internal transaction identifier.
+type txid uint64
+
+// Tx represents a read-only or read/write transaction on the database.
+// Read-only transactions can be used for retrieving values for keys and creating cursors.
+// Read/write transactions can create and remove buckets and create and remove keys.
+//
+// IMPORTANT: You must commit or rollback transactions when you are done with
+// them. Pages can not be reclaimed by the writer until no more transactions
+// are using them. A long running read transaction can cause the database to
+// quickly grow.
+type Tx struct {
+	writable       bool
+	managed        bool
+	db             *DB
+	meta           *meta
+	root           Bucket
+	pages          map[pgid]*page
+	stats          TxStats
+	commitHandlers []func()
+
+	// WriteFlag specifies the flag for write-related methods like WriteTo().
+	// Tx opens the database file with the specified flag to copy the data.
+	//
+	// By default, the flag is unset, which works well for mostly in-memory
+	// workloads. For databases that are much larger than available RAM,
+	// set the flag to syscall.O_DIRECT to avoid trashing the page cache.
+	WriteFlag int
+}
+
+// init initializes the transaction.
+func (tx *Tx) init(db *DB) {
+	tx.db = db
+	tx.pages = nil
+
+	// Copy the meta page since it can be changed by the writer.
+	tx.meta = &meta{}
+	db.meta().copy(tx.meta)
+
+	// Copy over the root bucket.
+	tx.root = newBucket(tx)
+	tx.root.bucket = &bucket{}
+	*tx.root.bucket = tx.meta.root
+
+	// Increment the transaction id and add a page cache for writable transactions.
+	if tx.writable {
+		tx.pages = make(map[pgid]*page)
+		tx.meta.txid += txid(1)
+	}
+}
+
+// ID returns the transaction id.
+func (tx *Tx) ID() int {
+	return int(tx.meta.txid)
+}
+
+// DB returns a reference to the database that created the transaction.
+func (tx *Tx) DB() *DB {
+	return tx.db
+}
+
+// Size returns current database size in bytes as seen by this transaction.
+func (tx *Tx) Size() int64 {
+	return int64(tx.meta.pgid) * int64(tx.db.pageSize)
+}
+
+// Writable returns whether the transaction can perform write operations.
+func (tx *Tx) Writable() bool {
+	return tx.writable
+}
+
+// Cursor creates a cursor associated with the root bucket.
+// All items in the cursor will return a nil value because all root bucket keys point to buckets.
+// The cursor is only valid as long as the transaction is open.
+// Do not use a cursor after the transaction is closed.
+func (tx *Tx) Cursor() *Cursor {
+	return tx.root.Cursor()
+}
+
+// Stats retrieves a copy of the current transaction statistics.
+func (tx *Tx) Stats() TxStats {
+	return tx.stats
+}
+
+// Bucket retrieves a bucket by name.
+// Returns nil if the bucket does not exist.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) Bucket(name []byte) *Bucket {
+	return tx.root.Bucket(name)
+}
+
+// CreateBucket creates a new bucket.
+// Returns an error if the bucket already exists, if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) CreateBucket(name []byte) (*Bucket, error) {
+	return tx.root.CreateBucket(name)
+}
+
+// CreateBucketIfNotExists creates a new bucket if it doesn't already exist.
+// Returns an error if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) CreateBucketIfNotExists(name []byte) (*Bucket, error) {
+	return tx.root.CreateBucketIfNotExists(name)
+}
+
+// DeleteBucket deletes a bucket.
+// Returns an error if the bucket cannot be found or if the key represents a non-bucket value.
+func (tx *Tx) DeleteBucket(name []byte) error {
+	return tx.root.DeleteBucket(name)
+}
+
+// ForEach executes a function for each bucket in the root.
+// If the provided function returns an error then the iteration is stopped and
+// the error is returned to the caller.
+func (tx *Tx) ForEach(fn func(name []byte, b *Bucket) error) error {
+	return tx.root.ForEach(func(k, v []byte) error {
+		if err := fn(k, tx.root.Bucket(k)); err != nil {
+			return err
+		}
+		return nil
+	})
+}
+
+// OnCommit adds a handler function to be executed after the transaction successfully commits.
+func (tx *Tx) OnCommit(fn func()) {
+	tx.commitHandlers = append(tx.commitHandlers, fn)
+}
+
+// Commit writes all changes to disk and updates the meta page.
+// Returns an error if a disk write error occurs, or if Commit is
+// called on a read-only transaction.
+func (tx *Tx) Commit() error {
+	_assert(!tx.managed, "managed tx commit not allowed")
+	if tx.db == nil {
+		return ErrTxClosed
+	} else if !tx.writable {
+		return ErrTxNotWritable
+	}
+
+	// TODO(benbjohnson): Use vectorized I/O to write out dirty pages.
+
+	// Rebalance nodes which have had deletions.
+	var startTime = time.Now()
+	tx.root.rebalance()
+	if tx.stats.Rebalance > 0 {
+		tx.stats.RebalanceTime += time.Since(startTime)
+	}
+
+	// spill data onto dirty pages.
+	startTime = time.Now()
+	if err := tx.root.spill(); err != nil {
+		tx.rollback()
+		return err
+	}
+	tx.stats.SpillTime += time.Since(startTime)
+
+	// Free the old root bucket.
+	tx.meta.root.root = tx.root.root
+
+	// Free the freelist and allocate new pages for it. This will overestimate
+	// the size of the freelist but not underestimate the size (which would be bad).
+	tx.db.freelist.free(tx.meta.txid, tx.db.page(tx.meta.freelist))
+	p, err := tx.allocate((tx.db.freelist.size() / tx.db.pageSize) + 1)
+	if err != nil {
+		tx.rollback()
+		return err
+	}
+	if err := tx.db.freelist.write(p); err != nil {
+		tx.rollback()
+		return err
+	}
+	tx.meta.freelist = p.id
+
+	// Write dirty pages to disk.
+	startTime = time.Now()
+	if err := tx.write(); err != nil {
+		tx.rollback()
+		return err
+	}
+
+	// If strict mode is enabled then perform a consistency check.
+	// Only the first consistency error is reported in the panic.
+	if tx.db.StrictMode {
+		if err, ok := <-tx.Check(); ok {
+			panic("check fail: " + err.Error())
+		}
+	}
+
+	// Write meta to disk.
+	if err := tx.writeMeta(); err != nil {
+		tx.rollback()
+		return err
+	}
+	tx.stats.WriteTime += time.Since(startTime)
+
+	// Finalize the transaction.
+	tx.close()
+
+	// Execute commit handlers now that the locks have been removed.
+	for _, fn := range tx.commitHandlers {
+		fn()
+	}
+
+	return nil
+}
+
+// Rollback closes the transaction and ignores all previous updates. Read-only
+// transactions must be rolled back and not committed.
+func (tx *Tx) Rollback() error {
+	_assert(!tx.managed, "managed tx rollback not allowed")
+	if tx.db == nil {
+		return ErrTxClosed
+	}
+	tx.rollback()
+	return nil
+}
+
+func (tx *Tx) rollback() {
+	if tx.db == nil {
+		return
+	}
+	if tx.writable {
+		tx.db.freelist.rollback(tx.meta.txid)
+		tx.db.freelist.reload(tx.db.page(tx.db.meta().freelist))
+	}
+	tx.close()
+}
+
+func (tx *Tx) close() {
+	if tx.db == nil {
+		return
+	}
+	if tx.writable {
+		// Grab freelist stats.
+		var freelistFreeN = tx.db.freelist.free_count()
+		var freelistPendingN = tx.db.freelist.pending_count()
+		var freelistAlloc = tx.db.freelist.size()
+
+		// Remove transaction ref & writer lock.
+		tx.db.rwtx = nil
+		tx.db.rwlock.Unlock()
+
+		// Merge statistics.
+		tx.db.statlock.Lock()
+		tx.db.stats.FreePageN = freelistFreeN
+		tx.db.stats.PendingPageN = freelistPendingN
+		tx.db.stats.FreeAlloc = (freelistFreeN + freelistPendingN) * tx.db.pageSize
+		tx.db.stats.FreelistInuse = freelistAlloc
+		tx.db.stats.TxStats.add(&tx.stats)
+		tx.db.statlock.Unlock()
+	} else {
+		tx.db.removeTx(tx)
+	}
+
+	// Clear all references.
+	tx.db = nil
+	tx.meta = nil
+	tx.root = Bucket{tx: tx}
+	tx.pages = nil
+}
+
+// Copy writes the entire database to a writer.
+// This function exists for backwards compatibility. Use WriteTo() instead.
+func (tx *Tx) Copy(w io.Writer) error {
+	_, err := tx.WriteTo(w)
+	return err
+}
+
+// WriteTo writes the entire database to a writer.
+// If err == nil then exactly tx.Size() bytes will be written into the writer.
+func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) {
+	// Attempt to open reader with WriteFlag
+	f, err := os.OpenFile(tx.db.path, os.O_RDONLY|tx.WriteFlag, 0)
+	if err != nil {
+		return 0, err
+	}
+	defer f.Close()
+
+	// Copy the meta pages.
+	tx.db.metalock.Lock()
+	n, err = io.CopyN(w, f, int64(tx.db.pageSize*2))
+	tx.db.metalock.Unlock()
+	if err != nil {
+		return n, fmt.Errorf("meta copy: %s", err)
+	}
+
+	// Copy data pages.
+	wn, err := io.CopyN(w, f, tx.Size()-int64(tx.db.pageSize*2))
+	n += wn
+	if err != nil {
+		return n, err
+	}
+
+	return n, f.Close()
+}
+
+// CopyFile copies the entire database to file at the given path.
+// A reader transaction is maintained during the copy so it is safe to continue
+// using the database while a copy is in progress.
+func (tx *Tx) CopyFile(path string, mode os.FileMode) error {
+	f, err := os.OpenFile(path, os.O_RDWR|os.O_CREATE|os.O_TRUNC, mode)
+	if err != nil {
+		return err
+	}
+
+	err = tx.Copy(f)
+	if err != nil {
+		_ = f.Close()
+		return err
+	}
+	return f.Close()
+}
+
+// Check performs several consistency checks on the database for this transaction.
+// An error is returned if any inconsistency is found.
+//
+// It can be safely run concurrently on a writable transaction. However, this
+// incurs a high cost for large databases and databases with a lot of subbuckets
+// because of caching. This overhead can be removed if running on a read-only
+// transaction, however, it is not safe to execute other writer transactions at
+// the same time.
+func (tx *Tx) Check() <-chan error {
+	ch := make(chan error)
+	go tx.check(ch)
+	return ch
+}
+
+func (tx *Tx) check(ch chan error) {
+	// Check if any pages are double freed.
+	freed := make(map[pgid]bool)
+	for _, id := range tx.db.freelist.all() {
+		if freed[id] {
+			ch <- fmt.Errorf("page %d: already freed", id)
+		}
+		freed[id] = true
+	}
+
+	// Track every reachable page.
+	reachable := make(map[pgid]*page)
+	reachable[0] = tx.page(0) // meta0
+	reachable[1] = tx.page(1) // meta1
+	for i := uint32(0); i <= tx.page(tx.meta.freelist).overflow; i++ {
+		reachable[tx.meta.freelist+pgid(i)] = tx.page(tx.meta.freelist)
+	}
+
+	// Recursively check buckets.
+	tx.checkBucket(&tx.root, reachable, freed, ch)
+
+	// Ensure all pages below high water mark are either reachable or freed.
+	for i := pgid(0); i < tx.meta.pgid; i++ {
+		_, isReachable := reachable[i]
+		if !isReachable && !freed[i] {
+			ch <- fmt.Errorf("page %d: unreachable unfreed", int(i))
+		}
+	}
+
+	// Close the channel to signal completion.
+	close(ch)
+}
+
+func (tx *Tx) checkBucket(b *Bucket, reachable map[pgid]*page, freed map[pgid]bool, ch chan error) {
+	// Ignore inline buckets.
+	if b.root == 0 {
+		return
+	}
+
+	// Check every page used by this bucket.
+	b.tx.forEachPage(b.root, 0, func(p *page, _ int) {
+		if p.id > tx.meta.pgid {
+			ch <- fmt.Errorf("page %d: out of bounds: %d", int(p.id), int(b.tx.meta.pgid))
+		}
+
+		// Ensure each page is only referenced once.
+		for i := pgid(0); i <= pgid(p.overflow); i++ {
+			var id = p.id + i
+			if _, ok := reachable[id]; ok {
+				ch <- fmt.Errorf("page %d: multiple references", int(id))
+			}
+			reachable[id] = p
+		}
+
+		// We should only encounter un-freed leaf and branch pages.
+		if freed[p.id] {
+			ch <- fmt.Errorf("page %d: reachable freed", int(p.id))
+		} else if (p.flags&branchPageFlag) == 0 && (p.flags&leafPageFlag) == 0 {
+			ch <- fmt.Errorf("page %d: invalid type: %s", int(p.id), p.typ())
+		}
+	})
+
+	// Check each bucket within this bucket.
+	_ = b.ForEach(func(k, v []byte) error {
+		if child := b.Bucket(k); child != nil {
+			tx.checkBucket(child, reachable, freed, ch)
+		}
+		return nil
+	})
+}
+
+// allocate returns a contiguous block of memory starting at a given page.
+func (tx *Tx) allocate(count int) (*page, error) {
+	p, err := tx.db.allocate(count)
+	if err != nil {
+		return nil, err
+	}
+
+	// Save to our page cache.
+	tx.pages[p.id] = p
+
+	// Update statistics.
+	tx.stats.PageCount++
+	tx.stats.PageAlloc += count * tx.db.pageSize
+
+	return p, nil
+}
+
+// write writes any dirty pages to disk.
+func (tx *Tx) write() error {
+	// Sort pages by id.
+	pages := make(pages, 0, len(tx.pages))
+	for _, p := range tx.pages {
+		pages = append(pages, p)
+	}
+	sort.Sort(pages)
+
+	// Write pages to disk in order.
+	for _, p := range pages {
+		size := (int(p.overflow) + 1) * tx.db.pageSize
+		offset := int64(p.id) * int64(tx.db.pageSize)
+
+		// Write out page in "max allocation" sized chunks.
+		ptr := (*[maxAllocSize]byte)(unsafe.Pointer(p))
+		for {
+			// Limit our write to our max allocation size.
+			sz := size
+			if sz > maxAllocSize-1 {
+				sz = maxAllocSize - 1
+			}
+
+			// Write chunk to disk.
+			buf := ptr[:sz]
+			if _, err := tx.db.ops.writeAt(buf, offset); err != nil {
+				return err
+			}
+
+			// Update statistics.
+			tx.stats.Write++
+
+			// Exit inner for loop if we've written all the chunks.
+			size -= sz
+			if size == 0 {
+				break
+			}
+
+			// Otherwise move offset forward and move pointer to next chunk.
+			offset += int64(sz)
+			ptr = (*[maxAllocSize]byte)(unsafe.Pointer(&ptr[sz]))
+		}
+	}
+
+	// Ignore file sync if flag is set on DB.
+	if !tx.db.NoSync || IgnoreNoSync {
+		if err := fdatasync(tx.db); err != nil {
+			return err
+		}
+	}
+
+	// Clear out page cache.
+	tx.pages = make(map[pgid]*page)
+
+	return nil
+}
+
+// writeMeta writes the meta to the disk.
+func (tx *Tx) writeMeta() error {
+	// Create a temporary buffer for the meta page.
+	buf := make([]byte, tx.db.pageSize)
+	p := tx.db.pageInBuffer(buf, 0)
+	tx.meta.write(p)
+
+	// Write the meta page to file.
+	if _, err := tx.db.ops.writeAt(buf, int64(p.id)*int64(tx.db.pageSize)); err != nil {
+		return err
+	}
+	if !tx.db.NoSync || IgnoreNoSync {
+		if err := fdatasync(tx.db); err != nil {
+			return err
+		}
+	}
+
+	// Update statistics.
+	tx.stats.Write++
+
+	return nil
+}
+
+// page returns a reference to the page with a given id.
+// If page has been written to then a temporary bufferred page is returned.
+func (tx *Tx) page(id pgid) *page {
+	// Check the dirty pages first.
+	if tx.pages != nil {
+		if p, ok := tx.pages[id]; ok {
+			return p
+		}
+	}
+
+	// Otherwise return directly from the mmap.
+	return tx.db.page(id)
+}
+
+// forEachPage iterates over every page within a given page and executes a function.
+func (tx *Tx) forEachPage(pgid pgid, depth int, fn func(*page, int)) {
+	p := tx.page(pgid)
+
+	// Execute function.
+	fn(p, depth)
+
+	// Recursively loop over children.
+	if (p.flags & branchPageFlag) != 0 {
+		for i := 0; i < int(p.count); i++ {
+			elem := p.branchPageElement(uint16(i))
+			tx.forEachPage(elem.pgid, depth+1, fn)
+		}
+	}
+}
+
+// Page returns page information for a given page number.
+// This is only safe for concurrent use when used by a writable transaction.
+func (tx *Tx) Page(id int) (*PageInfo, error) {
+	if tx.db == nil {
+		return nil, ErrTxClosed
+	} else if pgid(id) >= tx.meta.pgid {
+		return nil, nil
+	}
+
+	// Build the page info.
+	p := tx.db.page(pgid(id))
+	info := &PageInfo{
+		ID:            id,
+		Count:         int(p.count),
+		OverflowCount: int(p.overflow),
+	}
+
+	// Determine the type (or if it's free).
+	if tx.db.freelist.freed(pgid(id)) {
+		info.Type = "free"
+	} else {
+		info.Type = p.typ()
+	}
+
+	return info, nil
+}
+
+// TxStats represents statistics about the actions performed by the transaction.
+type TxStats struct {
+	// Page statistics.
+	PageCount int // number of page allocations
+	PageAlloc int // total bytes allocated
+
+	// Cursor statistics.
+	CursorCount int // number of cursors created
+
+	// Node statistics
+	NodeCount int // number of node allocations
+	NodeDeref int // number of node dereferences
+
+	// Rebalance statistics.
+	Rebalance     int           // number of node rebalances
+	RebalanceTime time.Duration // total time spent rebalancing
+
+	// Split/Spill statistics.
+	Split     int           // number of nodes split
+	Spill     int           // number of nodes spilled
+	SpillTime time.Duration // total time spent spilling
+
+	// Write statistics.
+	Write     int           // number of writes performed
+	WriteTime time.Duration // total time spent writing to disk
+}
+
+func (s *TxStats) add(other *TxStats) {
+	s.PageCount += other.PageCount
+	s.PageAlloc += other.PageAlloc
+	s.CursorCount += other.CursorCount
+	s.NodeCount += other.NodeCount
+	s.NodeDeref += other.NodeDeref
+	s.Rebalance += other.Rebalance
+	s.RebalanceTime += other.RebalanceTime
+	s.Split += other.Split
+	s.Spill += other.Spill
+	s.SpillTime += other.SpillTime
+	s.Write += other.Write
+	s.WriteTime += other.WriteTime
+}
+
+// Sub calculates and returns the difference between two sets of transaction stats.
+// This is useful when obtaining stats at two different points and time and
+// you need the performance counters that occurred within that time span.
+func (s *TxStats) Sub(other *TxStats) TxStats {
+	var diff TxStats
+	diff.PageCount = s.PageCount - other.PageCount
+	diff.PageAlloc = s.PageAlloc - other.PageAlloc
+	diff.CursorCount = s.CursorCount - other.CursorCount
+	diff.NodeCount = s.NodeCount - other.NodeCount
+	diff.NodeDeref = s.NodeDeref - other.NodeDeref
+	diff.Rebalance = s.Rebalance - other.Rebalance
+	diff.RebalanceTime = s.RebalanceTime - other.RebalanceTime
+	diff.Split = s.Split - other.Split
+	diff.Spill = s.Spill - other.Spill
+	diff.SpillTime = s.SpillTime - other.SpillTime
+	diff.Write = s.Write - other.Write
+	diff.WriteTime = s.WriteTime - other.WriteTime
+	return diff
+}

weed/vendor/github.com/chrislusf/raft/.gitignore

@@ -0,0 +1,25 @@
+# Compiled Object files, Static and Dynamic libs (Shared Objects)
+*.o
+*.a
+*.so
+
+# Folders
+_obj
+_test
+
+# Architecture specific extensions/prefixes
+*.[568vq]
+[568vq].out
+
+*.cgo1.go
+*.cgo2.c
+_cgo_defun.c
+_cgo_gotypes.go
+_cgo_export.*
+
+_testmain.go
+
+*.exe
+
+coverage.html
+coverprofile.out

weed/vendor/github.com/chrislusf/raft/LICENSE

@@ -0,0 +1,20 @@
+Copyright 2013 go-raft contributors
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

weed/vendor/github.com/chrislusf/raft/Makefile

@@ -0,0 +1,17 @@
+COVERPROFILE=cover.out
+
+default: test
+
+cover:
+	go test -coverprofile=$(COVERPROFILE) .
+	go tool cover -html=$(COVERPROFILE)
+	rm $(COVERPROFILE)
+
+dependencies:
+	go get -d .
+
+test:
+	go test -i ./...
+	go test -v ./...
+
+.PHONY: coverage dependencies test

weed/vendor/github.com/chrislusf/raft/README.md

@@ -0,0 +1,146 @@
+go-raft [![Build Status](https://drone.io/github.com/goraft/raft/status.png)](https://drone.io/github.com/goraft/raft/latest) [![Coverage Status](https://coveralls.io/repos/goraft/raft/badge.png?branch=master)](https://coveralls.io/r/goraft/raft?branch=master)
+=======
+
+## Overview
+
+![unmaintained](http://img.shields.io/badge/status-unmaintained-red.png)
+
+**NOTE**: This project is unmaintained. If you are using goraft in a project
+and want to carry the project forward please file an issue with your ideas and
+intentions. The original project authors have created new raft implementations
+now used in etcd and InfluxDB.
+
+This is a Go implementation of the Raft distributed consensus protocol.
+Raft is a protocol by which a cluster of nodes can maintain a replicated state machine.
+The state machine is kept in sync through the use of a replicated log.
+
+For more details on Raft, you can read [In Search of an Understandable Consensus Algorithm][raft-paper] by Diego Ongaro and John Ousterhout.
+
+## Project Status
+
+This library is feature complete but should be considered experimental until it has seen more usage.
+If you have any questions on implementing go-raft in your project please file an issue.
+There is an [active community][community] of developers who can help.
+go-raft is under the MIT license.
+
+[community]: https://github.com/goraft/raft/contributors
+
+### Features
+
+- Leader election
+- Log replication
+- Configuration changes
+- Log compaction
+- Unit tests
+- Fast Protobuf Log Encoding
+- HTTP transport
+
+### Projects
+
+These projects are built on go-raft:
+
+- [coreos/etcd 0.4.* and older](https://github.com/coreos/etcd) - A highly-available key value store for shared configuration and service discovery. Note: etcd from v2.0.0 onward has since started using a [new raft implementation](godoc.org/github.com/coreos/etcd/raft).
+- [goraft/raftd](https://github.com/goraft/raftd) - A reference implementation for using the go-raft library for distributed consensus.
+- [skynetservices/skydns](https://github.com/skynetservices/skydns) - DNS for skynet or any other service discovery.
+- [influxdb/influxdb](https://github.com/influxdb/influxdb) - An open-source, distributed, time series, events, and metrics database.
+- [Weed File System](https://weed-fs.googlecode.com) - A scalable distributed key-to-file system with O(1) disk access for each read.
+- [rqlite](https://github.com/otoolep/rqlite) - A replicated SQLite database, distributing the database replicas across multiple nodes.
+
+If you have a project that you're using go-raft in, please add it to this README so others can see implementation examples.
+
+## Contact and Resources
+
+- [raft-dev][raft-dev] is a mailing list for discussion about best practices
+  and implementation of Raft. Not goraft specific but helpful if you have
+  questions.
+- [Slides from Ben's talk][bens-talk] which includes easy to understand
+  diagrams of leader election and replication
+- The [Raft Consensus homepage][raft-home] has links to additional raft
+  implementations, slides to talks on Raft and general information
+
+[raft-home]:  http://raftconsensus.github.io/
+[raft-dev]: https://groups.google.com/forum/#!forum/raft-dev
+[bens-talk]: https://speakerdeck.com/benbjohnson/raft-the-understandable-distributed-consensus-protocol
+
+## The Raft Protocol
+
+This section provides a summary of the Raft protocol from a high level.
+For a more detailed explanation on the failover process and election terms please see the full paper describing the protocol: [In Search of an Understandable Consensus Algorithm][raft-paper].
+
+### Overview
+
+Maintaining state in a single process on a single server is easy.
+Your process is a single point of authority so there are no conflicts when reading and writing state.
+Even multi-threaded processes can rely on locks or coroutines to serialize access to the data.
+
+However, in a distributed system there is no single point of authority.
+Servers can crash or the network between two machines can become unavailable or any number of other problems can occur.
+
+A distributed consensus protocol is used for maintaining a consistent state across multiple servers in a cluster.
+Many distributed systems are built upon the Paxos protocol but Paxos can be difficult to understand and there are many gaps between Paxos and real world implementation.
+
+An alternative is the [Raft distributed consensus protocol][raft-paper] by Diego Ongaro and John Ousterhout.
+Raft is a protocol built with understandability as a primary tenet and it centers around two things:
+
+1. Leader Election
+2. Replicated Log
+
+With these two constructs, you can build a system that can maintain state across multiple servers -- even in the event of multiple failures.
+
+### Leader Election
+
+The Raft protocol effectively works as a master-slave system whereby state changes are written to a single server in the cluster and are distributed out to the rest of the servers in the cluster.
+This simplifies the protocol since there is only one data authority and conflicts will not have to be resolved.
+
+Raft ensures that there is only one leader at a time.
+It does this by performing elections among the nodes in the cluster and requiring that a node must receive a majority of the votes in order to become leader.
+For example, if you have 3 nodes in your cluster then a single node would need 2 votes in order to become the leader.
+For a 5 node cluster, a server would need 3 votes to become leader.
+
+### Replicated Log
+
+To maintain state, a log of commands is maintained.
+Each command makes a change to the state of the server and the command is deterministic.
+By ensuring that this log is replicated identically between all the nodes in the cluster we can replicate the state at any point in time in the log by running each command sequentially.
+
+Replicating the log under normal conditions is done by sending an `AppendEntries` RPC from the leader to each of the other servers in the cluster (called Peers).
+Each peer will append the entries from the leader through a 2-phase commit process which ensure that a majority of servers in the cluster have entries written to log.
+
+
+## Raft in Practice
+
+### Optimal Cluster Size
+
+The primary consideration when choosing the node count in your Raft cluster is the number of nodes that can simultaneously fail.
+Because Raft requires a majority of nodes to be available to make progress, the number of node failures the cluster can tolerate is `(n / 2) - 1`.
+
+This means that a 3-node cluster can tolerate 1 node failure.
+If 2 nodes fail then the cluster cannot commit entries or elect a new leader so progress stops.
+A 5-node cluster can tolerate 2 node failures. A 9-node cluster can tolerate 4 node failures.
+It is unlikely that 4 nodes will simultaneously fail so clusters larger than 9 nodes are not common.
+
+Another consideration is performance.
+The leader must replicate log entries for each follower node so CPU and networking resources can quickly be bottlenecked under stress in a large cluster.
+
+
+### Scaling Raft
+
+Once you grow beyond the maximum size of your cluster there are a few options for scaling Raft:
+
+1. *Core nodes with dumb replication.*
+   This option requires you to maintain a small cluster (e.g. 5 nodes) that is involved in the Raft process and then replicate only committed log entries to the remaining nodes in the cluster.
+   This works well if you have reads in your system that can be stale.
+
+2. *Sharding.*
+   This option requires that you segment your data into different clusters.
+   This option works well if you need very strong consistency and therefore need to read and write heavily from the leader.
+
+If you have a very large cluster that you need to replicate to using Option 1 then you may want to look at performing hierarchical replication so that nodes can better share the load.
+
+
+## History
+
+Ben Johnson started this library for use in his behavioral analytics database called [Sky](https://github.com/skydb/sky).
+He put it under the MIT license in the hopes that it would be useful for other projects too.
+
+[raft-paper]: https://ramcloud.stanford.edu/raft.pdf

weed/vendor/github.com/chrislusf/raft/append_entries.go

@@ -0,0 +1,146 @@
+package raft
+
+import (
+	"io"
+	"io/ioutil"
+
+	"github.com/chrislusf/raft/protobuf"
+	"github.com/gogo/protobuf/proto"
+)
+
+// The request sent to a server to append entries to the log.
+type AppendEntriesRequest struct {
+	Term         uint64
+	PrevLogIndex uint64
+	PrevLogTerm  uint64
+	CommitIndex  uint64
+	LeaderName   string
+	Entries      []*protobuf.LogEntry
+}
+
+// The response returned from a server appending entries to the log.
+type AppendEntriesResponse struct {
+	pb     *protobuf.AppendEntriesResponse
+	peer   string
+	append bool
+}
+
+// Creates a new AppendEntries request.
+func newAppendEntriesRequest(term uint64, prevLogIndex uint64, prevLogTerm uint64,
+	commitIndex uint64, leaderName string, entries []*LogEntry) *AppendEntriesRequest {
+	pbEntries := make([]*protobuf.LogEntry, len(entries))
+
+	for i := range entries {
+		pbEntries[i] = entries[i].pb
+	}
+
+	return &AppendEntriesRequest{
+		Term:         term,
+		PrevLogIndex: prevLogIndex,
+		PrevLogTerm:  prevLogTerm,
+		CommitIndex:  commitIndex,
+		LeaderName:   leaderName,
+		Entries:      pbEntries,
+	}
+}
+
+// Encodes the AppendEntriesRequest to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (req *AppendEntriesRequest) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.AppendEntriesRequest{
+		Term:         proto.Uint64(req.Term),
+		PrevLogIndex: proto.Uint64(req.PrevLogIndex),
+		PrevLogTerm:  proto.Uint64(req.PrevLogTerm),
+		CommitIndex:  proto.Uint64(req.CommitIndex),
+		LeaderName:   proto.String(req.LeaderName),
+		Entries:      req.Entries,
+	}
+
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the AppendEntriesRequest from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (req *AppendEntriesRequest) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return -1, err
+	}
+
+	pb := new(protobuf.AppendEntriesRequest)
+	if err := proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	req.Term = pb.GetTerm()
+	req.PrevLogIndex = pb.GetPrevLogIndex()
+	req.PrevLogTerm = pb.GetPrevLogTerm()
+	req.CommitIndex = pb.GetCommitIndex()
+	req.LeaderName = pb.GetLeaderName()
+	req.Entries = pb.GetEntries()
+
+	return len(data), nil
+}
+
+// Creates a new AppendEntries response.
+func newAppendEntriesResponse(term uint64, success bool, index uint64, commitIndex uint64) *AppendEntriesResponse {
+	pb := &protobuf.AppendEntriesResponse{
+		Term:        proto.Uint64(term),
+		Index:       proto.Uint64(index),
+		Success:     proto.Bool(success),
+		CommitIndex: proto.Uint64(commitIndex),
+	}
+
+	return &AppendEntriesResponse{
+		pb: pb,
+	}
+}
+
+func (aer *AppendEntriesResponse) Index() uint64 {
+	return aer.pb.GetIndex()
+}
+
+func (aer *AppendEntriesResponse) CommitIndex() uint64 {
+	return aer.pb.GetCommitIndex()
+}
+
+func (aer *AppendEntriesResponse) Term() uint64 {
+	return aer.pb.GetTerm()
+}
+
+func (aer *AppendEntriesResponse) Success() bool {
+	return aer.pb.GetSuccess()
+}
+
+// Encodes the AppendEntriesResponse to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (resp *AppendEntriesResponse) Encode(w io.Writer) (int, error) {
+	b, err := proto.Marshal(resp.pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(b)
+}
+
+// Decodes the AppendEntriesResponse from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (resp *AppendEntriesResponse) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+	if err != nil {
+		return -1, err
+	}
+
+	resp.pb = new(protobuf.AppendEntriesResponse)
+	if err := proto.Unmarshal(data, resp.pb); err != nil {
+		return -1, err
+	}
+
+	return len(data), nil
+}

weed/vendor/github.com/chrislusf/raft/command.go

@@ -0,0 +1,76 @@
+package raft
+
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"io"
+	"reflect"
+)
+
+var commandTypes map[string]Command
+
+func init() {
+	commandTypes = map[string]Command{}
+}
+
+// Command represents an action to be taken on the replicated state machine.
+type Command interface {
+	CommandName() string
+}
+
+// CommandApply represents the interface to apply a command to the server.
+type CommandApply interface {
+	Apply(Context) (interface{}, error)
+}
+
+// deprecatedCommandApply represents the old interface to apply a command to the server.
+type deprecatedCommandApply interface {
+	Apply(Server) (interface{}, error)
+}
+
+type CommandEncoder interface {
+	Encode(w io.Writer) error
+	Decode(r io.Reader) error
+}
+
+// Creates a new instance of a command by name.
+func newCommand(name string, data []byte) (Command, error) {
+	// Find the registered command.
+	command := commandTypes[name]
+	if command == nil {
+		return nil, fmt.Errorf("raft.Command: Unregistered command type: %s", name)
+	}
+
+	// Make a copy of the command.
+	v := reflect.New(reflect.Indirect(reflect.ValueOf(command)).Type()).Interface()
+	copy, ok := v.(Command)
+	if !ok {
+		panic(fmt.Sprintf("raft: Unable to copy command: %s (%v)", command.CommandName(), reflect.ValueOf(v).Kind().String()))
+	}
+
+	// If data for the command was passed in the decode it.
+	if data != nil {
+		if encoder, ok := copy.(CommandEncoder); ok {
+			if err := encoder.Decode(bytes.NewReader(data)); err != nil {
+				return nil, err
+			}
+		} else {
+			if err := json.NewDecoder(bytes.NewReader(data)).Decode(copy); err != nil {
+				return nil, err
+			}
+		}
+	}
+
+	return copy, nil
+}
+
+// Registers a command by storing a reference to an instance of it.
+func RegisterCommand(command Command) {
+	if command == nil {
+		panic(fmt.Sprintf("raft: Cannot register nil"))
+	} else if commandTypes[command.CommandName()] != nil {
+		panic(fmt.Sprintf("raft: Duplicate registration: %s", command.CommandName()))
+	}
+	commandTypes[command.CommandName()] = command
+}

weed/vendor/github.com/chrislusf/raft/commands.go

@@ -0,0 +1,78 @@
+package raft
+
+import (
+	"io"
+)
+
+// Join command interface
+type JoinCommand interface {
+	Command
+	NodeName() string
+}
+
+// Join command
+type DefaultJoinCommand struct {
+	Name             string `json:"name"`
+	ConnectionString string `json:"connectionString"`
+}
+
+// Leave command interface
+type LeaveCommand interface {
+	Command
+	NodeName() string
+}
+
+// Leave command
+type DefaultLeaveCommand struct {
+	Name string `json:"name"`
+}
+
+// NOP command
+type NOPCommand struct {
+}
+
+// The name of the Join command in the log
+func (c *DefaultJoinCommand) CommandName() string {
+	return "raft:join"
+}
+
+func (c *DefaultJoinCommand) Apply(server Server) (interface{}, error) {
+	err := server.AddPeer(c.Name, c.ConnectionString)
+
+	return []byte("join"), err
+}
+
+func (c *DefaultJoinCommand) NodeName() string {
+	return c.Name
+}
+
+// The name of the Leave command in the log
+func (c *DefaultLeaveCommand) CommandName() string {
+	return "raft:leave"
+}
+
+func (c *DefaultLeaveCommand) Apply(server Server) (interface{}, error) {
+	err := server.RemovePeer(c.Name)
+
+	return []byte("leave"), err
+}
+func (c *DefaultLeaveCommand) NodeName() string {
+	return c.Name
+}
+
+// The name of the NOP command in the log
+func (c NOPCommand) CommandName() string {
+	return "raft:nop"
+}
+
+func (c NOPCommand) Apply(server Server) (interface{}, error) {
+	return nil, nil
+}
+
+func (c NOPCommand) Encode(w io.Writer) error {
+	return nil
+}
+
+func (c NOPCommand) Decode(r io.Reader) error {
+	return nil
+}

weed/vendor/github.com/chrislusf/raft/config.go

@@ -0,0 +1,7 @@
+package raft
+
+type Config struct {
+	CommitIndex uint64 `json:"commitIndex"`
+	// TODO decide what we need to store in peer struct
+	Peers []*Peer `json:"peers"`
+}

weed/vendor/github.com/chrislusf/raft/context.go

@@ -0,0 +1,39 @@
+package raft
+
+// Context represents the current state of the server. It is passed into
+// a command when the command is being applied since the server methods
+// are locked.
+type Context interface {
+	Server() Server
+	CurrentTerm() uint64
+	CurrentIndex() uint64
+	CommitIndex() uint64
+}
+
+// context is the concrete implementation of Context.
+type context struct {
+	server       Server
+	currentIndex uint64
+	currentTerm  uint64
+	commitIndex  uint64
+}
+
+// Server returns a reference to the server.
+func (c *context) Server() Server {
+	return c.server
+}
+
+// CurrentTerm returns current term the server is in.
+func (c *context) CurrentTerm() uint64 {
+	return c.currentTerm
+}
+
+// CurrentIndex returns current index the server is at.
+func (c *context) CurrentIndex() uint64 {
+	return c.currentIndex
+}
+
+// CommitIndex returns last commit index the server is at.
+func (c *context) CommitIndex() uint64 {
+	return c.commitIndex
+}

weed/vendor/github.com/chrislusf/raft/debug.go

@@ -0,0 +1,116 @@
+package raft
+
+import (
+	"log"
+	"os"
+)
+
+//------------------------------------------------------------------------------
+//
+// Variables
+//
+//------------------------------------------------------------------------------
+
+const (
+	Debug = 1
+	Trace = 2
+)
+
+var logLevel int = 0
+var logger *log.Logger
+
+func init() {
+	logger = log.New(os.Stdout, "[raft]", log.Lmicroseconds)
+}
+
+//------------------------------------------------------------------------------
+//
+// Functions
+//
+//------------------------------------------------------------------------------
+
+func LogLevel() int {
+	return logLevel
+}
+
+func SetLogLevel(level int) {
+	logLevel = level
+}
+
+//--------------------------------------
+// Warnings
+//--------------------------------------
+
+// Prints to the standard logger. Arguments are handled in the manner of
+// fmt.Print.
+func warn(v ...interface{}) {
+	logger.Print(v...)
+}
+
+// Prints to the standard logger. Arguments are handled in the manner of
+// fmt.Printf.
+func warnf(format string, v ...interface{}) {
+	logger.Printf(format, v...)
+}
+
+// Prints to the standard logger. Arguments are handled in the manner of
+// fmt.Println.
+func warnln(v ...interface{}) {
+	logger.Println(v...)
+}
+
+//--------------------------------------
+// Basic debugging
+//--------------------------------------
+
+// Prints to the standard logger if debug mode is enabled. Arguments
+// are handled in the manner of fmt.Print.
+func debug(v ...interface{}) {
+	if logLevel >= Debug {
+		logger.Print(v...)
+	}
+}
+
+// Prints to the standard logger if debug mode is enabled. Arguments
+// are handled in the manner of fmt.Printf.
+func debugf(format string, v ...interface{}) {
+	if logLevel >= Debug {
+		logger.Printf(format, v...)
+	}
+}
+
+// Prints to the standard logger if debug mode is enabled. Arguments
+// are handled in the manner of fmt.Println.
+func debugln(v ...interface{}) {
+	if logLevel >= Debug {
+		logger.Println(v...)
+	}
+}
+
+//--------------------------------------
+// Trace-level debugging
+//--------------------------------------
+
+// Prints to the standard logger if trace debugging is enabled. Arguments
+// are handled in the manner of fmt.Print.
+func trace(v ...interface{}) {
+	if logLevel >= Trace {
+		logger.Print(v...)
+	}
+}
+
+// Prints to the standard logger if trace debugging is enabled. Arguments
+// are handled in the manner of fmt.Printf.
+func tracef(format string, v ...interface{}) {
+	if logLevel >= Trace {
+		logger.Printf(format, v...)
+	}
+}
+
+// Prints to the standard logger if trace debugging is enabled. Arguments
+// are handled in the manner of debugln.
+func traceln(v ...interface{}) {
+	if logLevel >= Trace {
+		logger.Println(v...)
+	}
+}

weed/vendor/github.com/chrislusf/raft/event.go

@@ -0,0 +1,61 @@
+package raft
+
+const (
+	StateChangeEventType  = "stateChange"
+	LeaderChangeEventType = "leaderChange"
+	TermChangeEventType   = "termChange"
+	CommitEventType       = "commit"
+	AddPeerEventType      = "addPeer"
+	RemovePeerEventType   = "removePeer"
+
+	HeartbeatIntervalEventType        = "heartbeatInterval"
+	ElectionTimeoutThresholdEventType = "electionTimeoutThreshold"
+
+	HeartbeatEventType = "heartbeat"
+)
+
+// Event represents an action that occurred within the Raft library.
+// Listeners can subscribe to event types by using the Server.AddEventListener() function.
+type Event interface {
+	Type() string
+	Source() interface{}
+	Value() interface{}
+	PrevValue() interface{}
+}
+
+// event is the concrete implementation of the Event interface.
+type event struct {
+	typ       string
+	source    interface{}
+	value     interface{}
+	prevValue interface{}
+}
+
+// newEvent creates a new event.
+func newEvent(typ string, value interface{}, prevValue interface{}) *event {
+	return &event{
+		typ:       typ,
+		value:     value,
+		prevValue: prevValue,
+	}
+}
+
+// Type returns the type of event that occurred.
+func (e *event) Type() string {
+	return e.typ
+}
+
+// Source returns the object that dispatched the event.
+func (e *event) Source() interface{} {
+	return e.source
+}
+
+// Value returns the current value associated with the event, if applicable.
+func (e *event) Value() interface{} {
+	return e.value
+}
+
+// PrevValue returns the previous value associated with the event, if applicable.
+func (e *event) PrevValue() interface{} {
+	return e.prevValue
+}

weed/vendor/github.com/chrislusf/raft/event_dispatcher.go

@@ -0,0 +1,68 @@
+package raft
+
+import (
+	"reflect"
+	"sync"
+)
+
+// eventDispatcher is responsible for managing listeners for named events
+// and dispatching event notifications to those listeners.
+type eventDispatcher struct {
+	sync.RWMutex
+	source    interface{}
+	listeners map[string]eventListeners
+}
+
+// EventListener is a function that can receive event notifications.
+type EventListener func(Event)
+
+// EventListeners represents a collection of individual listeners.
+type eventListeners []EventListener
+
+// newEventDispatcher creates a new eventDispatcher instance.
+func newEventDispatcher(source interface{}) *eventDispatcher {
+	return &eventDispatcher{
+		source:    source,
+		listeners: make(map[string]eventListeners),
+	}
+}
+
+// AddEventListener adds a listener function for a given event type.
+func (d *eventDispatcher) AddEventListener(typ string, listener EventListener) {
+	d.Lock()
+	defer d.Unlock()
+	d.listeners[typ] = append(d.listeners[typ], listener)
+}
+
+// RemoveEventListener removes a listener function for a given event type.
+func (d *eventDispatcher) RemoveEventListener(typ string, listener EventListener) {
+	d.Lock()
+	defer d.Unlock()
+
+	// Grab a reference to the function pointer once.
+	ptr := reflect.ValueOf(listener).Pointer()
+
+	// Find listener by pointer and remove it.
+	listeners := d.listeners[typ]
+	for i, l := range listeners {
+		if reflect.ValueOf(l).Pointer() == ptr {
+			d.listeners[typ] = append(listeners[:i], listeners[i+1:]...)
+		}
+	}
+}
+
+// DispatchEvent dispatches an event.
+func (d *eventDispatcher) DispatchEvent(e Event) {
+	d.RLock()
+	defer d.RUnlock()
+
+	// Automatically set the event source.
+	if e, ok := e.(*event); ok {
+		e.source = d.source
+	}
+
+	// Dispatch the event to all listeners.
+	for _, l := range d.listeners[e.Type()] {
+		l(e)
+	}
+}

weed/vendor/github.com/chrislusf/raft/http_transporter.go

@@ -0,0 +1,325 @@
+package raft
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"path"
+	"time"
+)
+
+// Parts from this transporter were heavily influenced by Peter Bougon's
+// raft implementation: https://github.com/peterbourgon/raft
+
+//------------------------------------------------------------------------------
+//
+// Typedefs
+//
+//------------------------------------------------------------------------------
+
+// An HTTPTransporter is a default transport layer used to communicate between
+// multiple servers.
+type HTTPTransporter struct {
+	DisableKeepAlives    bool
+	prefix               string
+	appendEntriesPath    string
+	requestVotePath      string
+	snapshotPath         string
+	snapshotRecoveryPath string
+	httpClient           http.Client
+	Transport            *http.Transport
+}
+
+type HTTPMuxer interface {
+	HandleFunc(string, func(http.ResponseWriter, *http.Request))
+}
+
+//------------------------------------------------------------------------------
+//
+// Constructor
+//
+//------------------------------------------------------------------------------
+
+// Creates a new HTTP transporter with the given path prefix.
+func NewHTTPTransporter(prefix string, timeout time.Duration) *HTTPTransporter {
+	t := &HTTPTransporter{
+		DisableKeepAlives:    false,
+		prefix:               prefix,
+		appendEntriesPath:    joinPath(prefix, "/appendEntries"),
+		requestVotePath:      joinPath(prefix, "/requestVote"),
+		snapshotPath:         joinPath(prefix, "/snapshot"),
+		snapshotRecoveryPath: joinPath(prefix, "/snapshotRecovery"),
+		Transport:            &http.Transport{DisableKeepAlives: false},
+	}
+	t.httpClient.Transport = t.Transport
+	t.Transport.ResponseHeaderTimeout = timeout
+	return t
+}
+
+//------------------------------------------------------------------------------
+//
+// Accessors
+//
+//------------------------------------------------------------------------------
+
+// Retrieves the path prefix used by the transporter.
+func (t *HTTPTransporter) Prefix() string {
+	return t.prefix
+}
+
+// Retrieves the AppendEntries path.
+func (t *HTTPTransporter) AppendEntriesPath() string {
+	return t.appendEntriesPath
+}
+
+// Retrieves the RequestVote path.
+func (t *HTTPTransporter) RequestVotePath() string {
+	return t.requestVotePath
+}
+
+// Retrieves the Snapshot path.
+func (t *HTTPTransporter) SnapshotPath() string {
+	return t.snapshotPath
+}
+
+// Retrieves the SnapshotRecovery path.
+func (t *HTTPTransporter) SnapshotRecoveryPath() string {
+	return t.snapshotRecoveryPath
+}
+
+//------------------------------------------------------------------------------
+//
+// Methods
+//
+//------------------------------------------------------------------------------
+
+//--------------------------------------
+// Installation
+//--------------------------------------
+
+// Applies Raft routes to an HTTP router for a given server.
+func (t *HTTPTransporter) Install(server Server, mux HTTPMuxer) {
+	mux.HandleFunc(t.AppendEntriesPath(), t.appendEntriesHandler(server))
+	mux.HandleFunc(t.RequestVotePath(), t.requestVoteHandler(server))
+	mux.HandleFunc(t.SnapshotPath(), t.snapshotHandler(server))
+	mux.HandleFunc(t.SnapshotRecoveryPath(), t.snapshotRecoveryHandler(server))
+}
+
+//--------------------------------------
+// Outgoing
+//--------------------------------------
+
+// Sends an AppendEntries RPC to a peer.
+func (t *HTTPTransporter) SendAppendEntriesRequest(server Server, peer *Peer, req *AppendEntriesRequest) *AppendEntriesResponse {
+	var b bytes.Buffer
+	if _, err := req.Encode(&b); err != nil {
+		traceln("transporter.ae.encoding.error:", err)
+		return nil
+	}
+
+	url := joinPath(peer.ConnectionString, t.AppendEntriesPath())
+	traceln(server.Name(), "POST", url)
+
+	httpResp, err := t.httpClient.Post(url, "application/protobuf", &b)
+	if httpResp == nil || err != nil {
+		traceln("transporter.ae.response.error:", err)
+		return nil
+	}
+	defer httpResp.Body.Close()
+
+	resp := &AppendEntriesResponse{}
+	if _, err = resp.Decode(httpResp.Body); err != nil && err != io.EOF {
+		traceln("transporter.ae.decoding.error:", err)
+		return nil
+	}
+
+	return resp
+}
+
+// Sends a RequestVote RPC to a peer.
+func (t *HTTPTransporter) SendVoteRequest(server Server, peer *Peer, req *RequestVoteRequest) *RequestVoteResponse {
+	var b bytes.Buffer
+	if _, err := req.Encode(&b); err != nil {
+		traceln("transporter.rv.encoding.error:", err)
+		return nil
+	}
+
+	url := fmt.Sprintf("%s%s", peer.ConnectionString, t.RequestVotePath())
+	traceln(server.Name(), "POST", url)
+
+	httpResp, err := t.httpClient.Post(url, "application/protobuf", &b)
+	if httpResp == nil || err != nil {
+		traceln("transporter.rv.response.error:", err)
+		return nil
+	}
+	defer httpResp.Body.Close()
+
+	resp := &RequestVoteResponse{}
+	if _, err = resp.Decode(httpResp.Body); err != nil && err != io.EOF {
+		traceln("transporter.rv.decoding.error:", err)
+		return nil
+	}
+
+	return resp
+}
+
+func joinPath(connectionString, thePath string) string {
+	u, err := url.Parse(connectionString)
+	if err != nil {
+		panic(err)
+	}
+	u.Path = path.Join(u.Path, thePath)
+	return u.String()
+}
+
+// Sends a SnapshotRequest RPC to a peer.
+func (t *HTTPTransporter) SendSnapshotRequest(server Server, peer *Peer, req *SnapshotRequest) *SnapshotResponse {
+	var b bytes.Buffer
+	if _, err := req.Encode(&b); err != nil {
+		traceln("transporter.rv.encoding.error:", err)
+		return nil
+	}
+
+	url := joinPath(peer.ConnectionString, t.snapshotPath)
+	traceln(server.Name(), "POST", url)
+
+	httpResp, err := t.httpClient.Post(url, "application/protobuf", &b)
+	if httpResp == nil || err != nil {
+		traceln("transporter.rv.response.error:", err)
+		return nil
+	}
+	defer httpResp.Body.Close()
+
+	resp := &SnapshotResponse{}
+	if _, err = resp.Decode(httpResp.Body); err != nil && err != io.EOF {
+		traceln("transporter.rv.decoding.error:", err)
+		return nil
+	}
+
+	return resp
+}
+
+// Sends a SnapshotRequest RPC to a peer.
+func (t *HTTPTransporter) SendSnapshotRecoveryRequest(server Server, peer *Peer, req *SnapshotRecoveryRequest) *SnapshotRecoveryResponse {
+	var b bytes.Buffer
+	if _, err := req.Encode(&b); err != nil {
+		traceln("transporter.rv.encoding.error:", err)
+		return nil
+	}
+
+	url := joinPath(peer.ConnectionString, t.snapshotRecoveryPath)
+	traceln(server.Name(), "POST", url)
+
+	httpResp, err := t.httpClient.Post(url, "application/protobuf", &b)
+	if httpResp == nil || err != nil {
+		traceln("transporter.rv.response.error:", err)
+		return nil
+	}
+	defer httpResp.Body.Close()
+
+	resp := &SnapshotRecoveryResponse{}
+	if _, err = resp.Decode(httpResp.Body); err != nil && err != io.EOF {
+		traceln("transporter.rv.decoding.error:", err)
+		return nil
+	}
+
+	return resp
+}
+
+//--------------------------------------
+// Incoming
+//--------------------------------------
+
+// Handles incoming AppendEntries requests.
+func (t *HTTPTransporter) appendEntriesHandler(server Server) http.HandlerFunc {
+	return func(w http.ResponseWriter, r *http.Request) {
+		traceln(server.Name(), "RECV /appendEntries")
+
+		req := &AppendEntriesRequest{}
+		if _, err := req.Decode(r.Body); err != nil {
+			http.Error(w, "", http.StatusBadRequest)
+			return
+		}
+
+		resp := server.AppendEntries(req)
+		if resp == nil {
+			http.Error(w, "Failed creating response.", http.StatusInternalServerError)
+			return
+		}
+		if _, err := resp.Encode(w); err != nil {
+			http.Error(w, "", http.StatusInternalServerError)
+			return
+		}
+	}
+}
+
+// Handles incoming RequestVote requests.
+func (t *HTTPTransporter) requestVoteHandler(server Server) http.HandlerFunc {
+	return func(w http.ResponseWriter, r *http.Request) {
+		traceln(server.Name(), "RECV /requestVote")
+
+		req := &RequestVoteRequest{}
+		if _, err := req.Decode(r.Body); err != nil {
+			http.Error(w, "", http.StatusBadRequest)
+			return
+		}
+
+		resp := server.RequestVote(req)
+		if resp == nil {
+			http.Error(w, "Failed creating response.", http.StatusInternalServerError)
+			return
+		}
+		if _, err := resp.Encode(w); err != nil {
+			http.Error(w, "", http.StatusInternalServerError)
+			return
+		}
+	}
+}
+
+// Handles incoming Snapshot requests.
+func (t *HTTPTransporter) snapshotHandler(server Server) http.HandlerFunc {
+	return func(w http.ResponseWriter, r *http.Request) {
+		traceln(server.Name(), "RECV /snapshot")
+
+		req := &SnapshotRequest{}
+		if _, err := req.Decode(r.Body); err != nil {
+			http.Error(w, "", http.StatusBadRequest)
+			return
+		}
+
+		resp := server.RequestSnapshot(req)
+		if resp == nil {
+			http.Error(w, "Failed creating response.", http.StatusInternalServerError)
+			return
+		}
+		if _, err := resp.Encode(w); err != nil {
+			http.Error(w, "", http.StatusInternalServerError)
+			return
+		}
+	}
+}
+
+// Handles incoming SnapshotRecovery requests.
+func (t *HTTPTransporter) snapshotRecoveryHandler(server Server) http.HandlerFunc {
+	return func(w http.ResponseWriter, r *http.Request) {
+		traceln(server.Name(), "RECV /snapshotRecovery")
+
+		req := &SnapshotRecoveryRequest{}
+		if _, err := req.Decode(r.Body); err != nil {
+			http.Error(w, "", http.StatusBadRequest)
+			return
+		}
+
+		resp := server.SnapshotRecoveryRequest(req)
+		if resp == nil {
+			http.Error(w, "Failed creating response.", http.StatusInternalServerError)
+			return
+		}
+		if _, err := resp.Encode(w); err != nil {
+			http.Error(w, "", http.StatusInternalServerError)
+			return
+		}
+	}
+}

weed/vendor/github.com/chrislusf/raft/log.go

@@ -0,0 +1,632 @@
+package raft
+
+import (
+	"bufio"
+	"errors"
+	"fmt"
+	"io"
+	"os"
+	"sync"
+
+	"github.com/chrislusf/raft/protobuf"
+)
+
+//------------------------------------------------------------------------------
+//
+// Typedefs
+//
+//------------------------------------------------------------------------------
+
+// A log is a collection of log entries that are persisted to durable storage.
+type Log struct {
+	ApplyFunc   func(*LogEntry, Command) (interface{}, error)
+	file        *os.File
+	path        string
+	entries     []*LogEntry
+	commitIndex uint64
+	mutex       sync.RWMutex
+	startIndex  uint64 // the index before the first entry in the Log entries
+	startTerm   uint64
+	initialized bool
+}
+
+// The results of the applying a log entry.
+type logResult struct {
+	returnValue interface{}
+	err         error
+}
+
+//------------------------------------------------------------------------------
+//
+// Constructor
+//
+//------------------------------------------------------------------------------
+
+// Creates a new log.
+func newLog() *Log {
+	return &Log{
+		entries: make([]*LogEntry, 0),
+	}
+}
+
+//------------------------------------------------------------------------------
+//
+// Accessors
+//
+//------------------------------------------------------------------------------
+
+//--------------------------------------
+// Log Indices
+//--------------------------------------
+
+// The last committed index in the log.
+func (l *Log) CommitIndex() uint64 {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+	return l.commitIndex
+}
+
+// The current index in the log.
+func (l *Log) currentIndex() uint64 {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+	return l.internalCurrentIndex()
+}
+
+// The current index in the log without locking
+func (l *Log) internalCurrentIndex() uint64 {
+	if len(l.entries) == 0 {
+		return l.startIndex
+	}
+	return l.entries[len(l.entries)-1].Index()
+}
+
+// The next index in the log.
+func (l *Log) nextIndex() uint64 {
+	return l.currentIndex() + 1
+}
+
+// Determines if the log contains zero entries.
+func (l *Log) isEmpty() bool {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+	return (len(l.entries) == 0) && (l.startIndex == 0)
+}
+
+// The name of the last command in the log.
+func (l *Log) lastCommandName() string {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+	if len(l.entries) > 0 {
+		if entry := l.entries[len(l.entries)-1]; entry != nil {
+			return entry.CommandName()
+		}
+	}
+	return ""
+}
+
+//--------------------------------------
+// Log Terms
+//--------------------------------------
+
+// The current term in the log.
+func (l *Log) currentTerm() uint64 {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+
+	if len(l.entries) == 0 {
+		return l.startTerm
+	}
+	return l.entries[len(l.entries)-1].Term()
+}
+
+//------------------------------------------------------------------------------
+//
+// Methods
+//
+//------------------------------------------------------------------------------
+
+//--------------------------------------
+// State
+//--------------------------------------
+
+// Opens the log file and reads existing entries. The log can remain open and
+// continue to append entries to the end of the log.
+func (l *Log) open(path string) error {
+	// Read all the entries from the log if one exists.
+	var readBytes int64
+
+	var err error
+	debugln("log.open.open ", path)
+	// open log file
+	l.file, err = os.OpenFile(path, os.O_RDWR, 0600)
+	l.path = path
+
+	if err != nil {
+		// if the log file does not exist before
+		// we create the log file and set commitIndex to 0
+		if os.IsNotExist(err) {
+			l.file, err = os.OpenFile(path, os.O_WRONLY|os.O_CREATE, 0600)
+			debugln("log.open.create ", path)
+			if err == nil {
+				l.initialized = true
+			}
+			return err
+		}
+		return err
+	}
+	debugln("log.open.exist ", path)
+
+	// Read the file and decode entries.
+	for {
+		// Instantiate log entry and decode into it.
+		entry, _ := newLogEntry(l, nil, 0, 0, nil)
+		entry.Position, _ = l.file.Seek(0, os.SEEK_CUR)
+
+		n, err := entry.Decode(l.file)
+		if err != nil {
+			if err == io.EOF {
+				debugln("open.log.append: finish ")
+			} else {
+				if err = os.Truncate(path, readBytes); err != nil {
+					return fmt.Errorf("raft.Log: Unable to recover: %v", err)
+				}
+			}
+			break
+		}
+		if entry.Index() > l.startIndex {
+			// Append entry.
+			l.entries = append(l.entries, entry)
+			if entry.Index() <= l.commitIndex {
+				command, err := newCommand(entry.CommandName(), entry.Command())
+				if err != nil {
+					continue
+				}
+				l.ApplyFunc(entry, command)
+			}
+			debugln("open.log.append log index ", entry.Index())
+		}
+
+		readBytes += int64(n)
+	}
+	debugln("open.log.recovery number of log ", len(l.entries))
+	l.initialized = true
+	return nil
+}
+
+// Closes the log file.
+func (l *Log) close() {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+
+	if l.file != nil {
+		l.file.Close()
+		l.file = nil
+	}
+	l.entries = make([]*LogEntry, 0)
+}
+
+// sync to disk
+func (l *Log) sync() error {
+	return l.file.Sync()
+}
+
+//--------------------------------------
+// Entries
+//--------------------------------------
+
+// Creates a log entry associated with this log.
+func (l *Log) createEntry(term uint64, command Command, e *ev) (*LogEntry, error) {
+	return newLogEntry(l, e, l.nextIndex(), term, command)
+}
+
+// Retrieves an entry from the log. If the entry has been eliminated because
+// of a snapshot then nil is returned.
+func (l *Log) getEntry(index uint64) *LogEntry {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+
+	if index <= l.startIndex || index > (l.startIndex+uint64(len(l.entries))) {
+		return nil
+	}
+	return l.entries[index-l.startIndex-1]
+}
+
+// Checks if the log contains a given index/term combination.
+func (l *Log) containsEntry(index uint64, term uint64) bool {
+	entry := l.getEntry(index)
+	return (entry != nil && entry.Term() == term)
+}
+
+// Retrieves a list of entries after a given index as well as the term of the
+// index provided. A nil list of entries is returned if the index no longer
+// exists because a snapshot was made.
+func (l *Log) getEntriesAfter(index uint64, maxLogEntriesPerRequest uint64) ([]*LogEntry, uint64) {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+
+	// Return nil if index is before the start of the log.
+	if index < l.startIndex {
+		traceln("log.entriesAfter.before: ", index, " ", l.startIndex)
+		return nil, 0
+	}
+
+	// Return an error if the index doesn't exist.
+	if index > (uint64(len(l.entries)) + l.startIndex) {
+		panic(fmt.Sprintf("raft: Index is beyond end of log: %v %v", len(l.entries), index))
+	}
+
+	// If we're going from the beginning of the log then return the whole log.
+	if index == l.startIndex {
+		traceln("log.entriesAfter.beginning: ", index, " ", l.startIndex)
+		return l.entries, l.startTerm
+	}
+
+	traceln("log.entriesAfter.partial: ", index, " ", l.entries[len(l.entries)-1].Index)
+
+	entries := l.entries[index-l.startIndex:]
+	length := len(entries)
+
+	traceln("log.entriesAfter: startIndex:", l.startIndex, " length", len(l.entries))
+
+	if uint64(length) < maxLogEntriesPerRequest {
+		// Determine the term at the given entry and return a subslice.
+		return entries, l.entries[index-1-l.startIndex].Term()
+	} else {
+		return entries[:maxLogEntriesPerRequest], l.entries[index-1-l.startIndex].Term()
+	}
+}
+
+//--------------------------------------
+// Commit
+//--------------------------------------
+
+// Retrieves the last index and term that has been committed to the log.
+func (l *Log) commitInfo() (index uint64, term uint64) {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+	// If we don't have any committed entries then just return zeros.
+	if l.commitIndex == 0 {
+		return 0, 0
+	}
+
+	// No new commit log after snapshot
+	if l.commitIndex == l.startIndex {
+		return l.startIndex, l.startTerm
+	}
+
+	// Return the last index & term from the last committed entry.
+	debugln("commitInfo.get.[", l.commitIndex, "/", l.startIndex, "]")
+	entry := l.entries[l.commitIndex-1-l.startIndex]
+	return entry.Index(), entry.Term()
+}
+
+// Retrieves the last index and term that has been appended to the log.
+func (l *Log) lastInfo() (index uint64, term uint64) {
+	l.mutex.RLock()
+	defer l.mutex.RUnlock()
+
+	// If we don't have any entries then just return zeros.
+	if len(l.entries) == 0 {
+		return l.startIndex, l.startTerm
+	}
+
+	// Return the last index & term
+	entry := l.entries[len(l.entries)-1]
+	return entry.Index(), entry.Term()
+}
+
+// Updates the commit index
+func (l *Log) updateCommitIndex(index uint64) {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+	if index > l.commitIndex {
+		l.commitIndex = index
+	}
+	debugln("update.commit.index ", index)
+}
+
+// Updates the commit index and writes entries after that index to the stable storage.
+func (l *Log) setCommitIndex(index uint64) error {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+
+	// this is not error any more after limited the number of sending entries
+	// commit up to what we already have
+	if index > l.startIndex+uint64(len(l.entries)) {
+		debugln("raft.Log: Commit index", index, "set back to ", len(l.entries))
+		index = l.startIndex + uint64(len(l.entries))
+	}
+
+	// Do not allow previous indices to be committed again.
+
+	// This could happens, since the guarantee is that the new leader has up-to-dated
+	// log entries rather than has most up-to-dated committed index
+
+	// For example, Leader 1 send log 80 to follower 2 and follower 3
+	// follower 2 and follow 3 all got the new entries and reply
+	// leader 1 committed entry 80 and send reply to follower 2 and follower3
+	// follower 2 receive the new committed index and update committed index to 80
+	// leader 1 fail to send the committed index to follower 3
+	// follower 3 promote to leader (server 1 and server 2 will vote, since leader 3
+	// has up-to-dated the entries)
+	// when new leader 3 send heartbeat with committed index = 0 to follower 2,
+	// follower 2 should reply success and let leader 3 update the committed index to 80
+
+	if index < l.commitIndex {
+		return nil
+	}
+
+	// Find all entries whose index is between the previous index and the current index.
+	for i := l.commitIndex + 1; i <= index; i++ {
+		entryIndex := i - 1 - l.startIndex
+		entry := l.entries[entryIndex]
+
+		// Update commit index.
+		l.commitIndex = entry.Index()
+
+		// Decode the command.
+		command, err := newCommand(entry.CommandName(), entry.Command())
+		if err != nil {
+			return err
+		}
+
+		// Apply the changes to the state machine and store the error code.
+		returnValue, err := l.ApplyFunc(entry, command)
+
+		debugf("setCommitIndex.set.result index: %v, entries index: %v", i, entryIndex)
+		if entry.event != nil {
+			entry.event.returnValue = returnValue
+			entry.event.c <- err
+		}
+
+		_, isJoinCommand := command.(JoinCommand)
+
+		// we can only commit up to the most recent join command
+		// if there is a join in this batch of commands.
+		// after this commit, we need to recalculate the majority.
+		if isJoinCommand {
+			return nil
+		}
+	}
+	return nil
+}
+
+// Set the commitIndex at the head of the log file to the current
+// commit Index. This should be called after obtained a log lock
+func (l *Log) flushCommitIndex() {
+	l.file.Seek(0, os.SEEK_SET)
+	fmt.Fprintf(l.file, "%8x\n", l.commitIndex)
+	l.file.Seek(0, os.SEEK_END)
+}
+
+//--------------------------------------
+// Truncation
+//--------------------------------------
+
+// Truncates the log to the given index and term. This only works if the log
+// at the index has not been committed.
+func (l *Log) truncate(index uint64, term uint64) error {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+	debugln("log.truncate: ", index)
+
+	// Do not allow committed entries to be truncated.
+	if index < l.commitIndex {
+		debugln("log.truncate.before")
+		return fmt.Errorf("raft.Log: Index is already committed (%v): (IDX=%v, TERM=%v)", l.commitIndex, index, term)
+	}
+
+	// Do not truncate past end of entries.
+	if index > l.startIndex+uint64(len(l.entries)) {
+		debugln("log.truncate.after")
+		return fmt.Errorf("raft.Log: Entry index does not exist (MAX=%v): (IDX=%v, TERM=%v)", len(l.entries), index, term)
+	}
+
+	// If we're truncating everything then just clear the entries.
+	if index == l.startIndex {
+		debugln("log.truncate.clear")
+		l.file.Truncate(0)
+		l.file.Seek(0, os.SEEK_SET)
+
+		// notify clients if this node is the previous leader
+		for _, entry := range l.entries {
+			if entry.event != nil {
+				entry.event.c <- errors.New("command failed to be committed due to node failure")
+			}
+		}
+
+		l.entries = []*LogEntry{}
+	} else {
+		// Do not truncate if the entry at index does not have the matching term.
+		entry := l.entries[index-l.startIndex-1]
+		if len(l.entries) > 0 && entry.Term() != term {
+			debugln("log.truncate.termMismatch")
+			return fmt.Errorf("raft.Log: Entry at index does not have matching term (%v): (IDX=%v, TERM=%v)", entry.Term(), index, term)
+		}
+
+		// Otherwise truncate up to the desired entry.
+		if index < l.startIndex+uint64(len(l.entries)) {
+			debugln("log.truncate.finish")
+			position := l.entries[index-l.startIndex].Position
+			l.file.Truncate(position)
+			l.file.Seek(position, os.SEEK_SET)
+
+			// notify clients if this node is the previous leader
+			for i := index - l.startIndex; i < uint64(len(l.entries)); i++ {
+				entry := l.entries[i]
+				if entry.event != nil {
+					entry.event.c <- errors.New("command failed to be committed due to node failure")
+				}
+			}
+
+			l.entries = l.entries[0 : index-l.startIndex]
+		}
+	}
+
+	return nil
+}
+
+//--------------------------------------
+// Append
+//--------------------------------------
+
+// Appends a series of entries to the log.
+func (l *Log) appendEntries(entries []*protobuf.LogEntry) error {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+
+	startPosition, _ := l.file.Seek(0, os.SEEK_CUR)
+
+	w := bufio.NewWriter(l.file)
+
+	var size int64
+	var err error
+	// Append each entry but exit if we hit an error.
+	for i := range entries {
+		logEntry := &LogEntry{
+			log:      l,
+			Position: startPosition,
+			pb:       entries[i],
+		}
+
+		if size, err = l.writeEntry(logEntry, w); err != nil {
+			return err
+		}
+
+		startPosition += size
+	}
+	w.Flush()
+	err = l.sync()
+
+	if err != nil {
+		panic(err)
+	}
+
+	return nil
+}
+
+// Writes a single log entry to the end of the log.
+func (l *Log) appendEntry(entry *LogEntry) error {
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+
+	if l.file == nil {
+		return errors.New("raft.Log: Log is not open")
+	}
+
+	// Make sure the term and index are greater than the previous.
+	if len(l.entries) > 0 {
+		lastEntry := l.entries[len(l.entries)-1]
+		if entry.Term() < lastEntry.Term() {
+			return fmt.Errorf("raft.Log: Cannot append entry with earlier term (%x:%x <= %x:%x)", entry.Term(), entry.Index(), lastEntry.Term(), lastEntry.Index())
+		} else if entry.Term() == lastEntry.Term() && entry.Index() <= lastEntry.Index() {
+			return fmt.Errorf("raft.Log: Cannot append entry with earlier index in the same term (%x:%x <= %x:%x)", entry.Term(), entry.Index(), lastEntry.Term(), lastEntry.Index())
+		}
+	}
+
+	position, _ := l.file.Seek(0, os.SEEK_CUR)
+
+	entry.Position = position
+
+	// Write to storage.
+	if _, err := entry.Encode(l.file); err != nil {
+		return err
+	}
+
+	// Append to entries list if stored on disk.
+	l.entries = append(l.entries, entry)
+
+	return nil
+}
+
+// appendEntry with Buffered io
+func (l *Log) writeEntry(entry *LogEntry, w io.Writer) (int64, error) {
+	if l.file == nil {
+		return -1, errors.New("raft.Log: Log is not open")
+	}
+
+	// Make sure the term and index are greater than the previous.
+	if len(l.entries) > 0 {
+		lastEntry := l.entries[len(l.entries)-1]
+		if entry.Term() < lastEntry.Term() {
+			return -1, fmt.Errorf("raft.Log: Cannot append entry with earlier term (%x:%x <= %x:%x)", entry.Term(), entry.Index(), lastEntry.Term(), lastEntry.Index())
+		} else if entry.Term() == lastEntry.Term() && entry.Index() <= lastEntry.Index() {
+			return -1, fmt.Errorf("raft.Log: Cannot append entry with earlier index in the same term (%x:%x <= %x:%x)", entry.Term(), entry.Index(), lastEntry.Term(), lastEntry.Index())
+		}
+	}
+
+	// Write to storage.
+	size, err := entry.Encode(w)
+	if err != nil {
+		return -1, err
+	}
+
+	// Append to entries list if stored on disk.
+	l.entries = append(l.entries, entry)
+
+	return int64(size), nil
+}
+
+//--------------------------------------
+// Log compaction
+//--------------------------------------
+
+// compact the log before index (including index)
+func (l *Log) compact(index uint64, term uint64) error {
+	var entries []*LogEntry
+
+	l.mutex.Lock()
+	defer l.mutex.Unlock()
+
+	if index == 0 {
+		return nil
+	}
+	// nothing to compaction
+	// the index may be greater than the current index if
+	// we just recovery from on snapshot
+	if index >= l.internalCurrentIndex() {
+		entries = make([]*LogEntry, 0)
+	} else {
+		// get all log entries after index
+		entries = l.entries[index-l.startIndex:]
+	}
+
+	// create a new log file and add all the entries
+	new_file_path := l.path + ".new"
+	file, err := os.OpenFile(new_file_path, os.O_APPEND|os.O_CREATE|os.O_WRONLY, 0600)
+	if err != nil {
+		return err
+	}
+	for _, entry := range entries {
+		position, _ := l.file.Seek(0, os.SEEK_CUR)
+		entry.Position = position
+
+		if _, err = entry.Encode(file); err != nil {
+			file.Close()
+			os.Remove(new_file_path)
+			return err
+		}
+	}
+	file.Sync()
+
+	old_file := l.file
+
+	// rename the new log file
+	err = os.Rename(new_file_path, l.path)
+	if err != nil {
+		file.Close()
+		os.Remove(new_file_path)
+		return err
+	}
+	l.file = file
+
+	// close the old log file
+	old_file.Close()
+
+	// compaction the in memory log
+	l.entries = entries
+	l.startIndex = index
+	l.startTerm = term
+	return nil
+}

weed/vendor/github.com/chrislusf/raft/log_entry.go

@@ -0,0 +1,107 @@
+package raft
+
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"io"
+
+	"github.com/chrislusf/raft/protobuf"
+	"github.com/gogo/protobuf/proto"
+)
+
+// A log entry stores a single item in the log.
+type LogEntry struct {
+	pb       *protobuf.LogEntry
+	Position int64 // position in the log file
+	log      *Log
+	event    *ev
+}
+
+// Creates a new log entry associated with a log.
+func newLogEntry(log *Log, event *ev, index uint64, term uint64, command Command) (*LogEntry, error) {
+	var buf bytes.Buffer
+	var commandName string
+	if command != nil {
+		commandName = command.CommandName()
+		if encoder, ok := command.(CommandEncoder); ok {
+			if err := encoder.Encode(&buf); err != nil {
+				return nil, err
+			}
+		} else {
+			if err := json.NewEncoder(&buf).Encode(command); err != nil {
+				return nil, err
+			}
+		}
+	}
+
+	pb := &protobuf.LogEntry{
+		Index:       proto.Uint64(index),
+		Term:        proto.Uint64(term),
+		CommandName: proto.String(commandName),
+		Command:     buf.Bytes(),
+	}
+
+	e := &LogEntry{
+		pb:    pb,
+		log:   log,
+		event: event,
+	}
+
+	return e, nil
+}
+
+func (e *LogEntry) Index() uint64 {
+	return e.pb.GetIndex()
+}
+
+func (e *LogEntry) Term() uint64 {
+	return e.pb.GetTerm()
+}
+
+func (e *LogEntry) CommandName() string {
+	return e.pb.GetCommandName()
+}
+
+func (e *LogEntry) Command() []byte {
+	return e.pb.GetCommand()
+}
+
+// Encodes the log entry to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (e *LogEntry) Encode(w io.Writer) (int, error) {
+	b, err := proto.Marshal(e.pb)
+	if err != nil {
+		return -1, err
+	}
+
+	if _, err = fmt.Fprintf(w, "%8x\n", len(b)); err != nil {
+		return -1, err
+	}
+
+	return w.Write(b)
+}
+
+// Decodes the log entry from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (e *LogEntry) Decode(r io.Reader) (int, error) {
+
+	var length int
+	_, err := fmt.Fscanf(r, "%8x\n", &length)
+	if err != nil {
+		return -1, err
+	}
+
+	data := make([]byte, length)
+	_, err = io.ReadFull(r, data)
+
+	if err != nil {
+		return -1, err
+	}
+
+	if err = proto.Unmarshal(data, e.pb); err != nil {
+		return -1, err
+	}
+
+	return length + 8 + 1, nil
+}

weed/vendor/github.com/chrislusf/raft/peer.go

@@ -0,0 +1,320 @@
+package raft
+
+import (
+	"sync"
+	"time"
+)
+
+//------------------------------------------------------------------------------
+//
+// Typedefs
+//
+//------------------------------------------------------------------------------
+
+// A peer is a reference to another server involved in the consensus protocol.
+type Peer struct {
+	server            *server
+	Name              string `json:"name"`
+	ConnectionString  string `json:"connectionString"`
+	prevLogIndex      uint64
+	stopChan          chan bool
+	heartbeatInterval time.Duration
+	lastActivity      time.Time
+	sync.RWMutex
+}
+
+//------------------------------------------------------------------------------
+//
+// Constructor
+//
+//------------------------------------------------------------------------------
+
+// Creates a new peer.
+func newPeer(server *server, name string, connectionString string, heartbeatInterval time.Duration) *Peer {
+	return &Peer{
+		server:            server,
+		Name:              name,
+		ConnectionString:  connectionString,
+		heartbeatInterval: heartbeatInterval,
+	}
+}
+
+//------------------------------------------------------------------------------
+//
+// Accessors
+//
+//------------------------------------------------------------------------------
+
+// Sets the heartbeat timeout.
+func (p *Peer) setHeartbeatInterval(duration time.Duration) {
+	p.heartbeatInterval = duration
+}
+
+//--------------------------------------
+// Prev log index
+//--------------------------------------
+
+// Retrieves the previous log index.
+func (p *Peer) getPrevLogIndex() uint64 {
+	p.RLock()
+	defer p.RUnlock()
+	return p.prevLogIndex
+}
+
+// Sets the previous log index.
+func (p *Peer) setPrevLogIndex(value uint64) {
+	p.Lock()
+	defer p.Unlock()
+	p.prevLogIndex = value
+}
+
+func (p *Peer) setLastActivity(now time.Time) {
+	p.Lock()
+	defer p.Unlock()
+	p.lastActivity = now
+}
+
+//------------------------------------------------------------------------------
+//
+// Methods
+//
+//------------------------------------------------------------------------------
+
+//--------------------------------------
+// Heartbeat
+//--------------------------------------
+
+// Starts the peer heartbeat.
+func (p *Peer) startHeartbeat() {
+	p.stopChan = make(chan bool)
+	c := make(chan bool)
+
+	p.setLastActivity(time.Now())
+
+	p.server.routineGroup.Add(1)
+	go func() {
+		defer p.server.routineGroup.Done()
+		p.heartbeat(c)
+	}()
+	<-c
+}
+
+// Stops the peer heartbeat.
+func (p *Peer) stopHeartbeat(flush bool) {
+	p.setLastActivity(time.Time{})
+
+	p.stopChan <- flush
+}
+
+// LastActivity returns the last time any response was received from the peer.
+func (p *Peer) LastActivity() time.Time {
+	p.RLock()
+	defer p.RUnlock()
+	return p.lastActivity
+}
+
+//--------------------------------------
+// Copying
+//--------------------------------------
+
+// Clones the state of the peer. The clone is not attached to a server and
+// the heartbeat timer will not exist.
+func (p *Peer) clone() *Peer {
+	p.Lock()
+	defer p.Unlock()
+	return &Peer{
+		Name:             p.Name,
+		ConnectionString: p.ConnectionString,
+		prevLogIndex:     p.prevLogIndex,
+		lastActivity:     p.lastActivity,
+	}
+}
+
+//--------------------------------------
+// Heartbeat
+//--------------------------------------
+
+// Listens to the heartbeat timeout and flushes an AppendEntries RPC.
+func (p *Peer) heartbeat(c chan bool) {
+	stopChan := p.stopChan
+
+	c <- true
+
+	ticker := time.Tick(p.heartbeatInterval)
+
+	debugln("peer.heartbeat: ", p.Name, p.heartbeatInterval)
+
+	for {
+		select {
+		case flush := <-stopChan:
+			if flush {
+				// before we can safely remove a node
+				// we must flush the remove command to the node first
+				p.flush()
+				debugln("peer.heartbeat.stop.with.flush: ", p.Name)
+				return
+			} else {
+				debugln("peer.heartbeat.stop: ", p.Name)
+				return
+			}
+
+		case <-ticker:
+			start := time.Now()
+			p.flush()
+			duration := time.Now().Sub(start)
+			p.server.DispatchEvent(newEvent(HeartbeatEventType, duration, nil))
+		}
+	}
+}
+
+func (p *Peer) flush() {
+	debugln("peer.heartbeat.flush: ", p.Name)
+	prevLogIndex := p.getPrevLogIndex()
+	term := p.server.currentTerm
+
+	entries, prevLogTerm := p.server.log.getEntriesAfter(prevLogIndex, p.server.maxLogEntriesPerRequest)
+
+	if entries != nil {
+		p.sendAppendEntriesRequest(newAppendEntriesRequest(term, prevLogIndex, prevLogTerm, p.server.log.CommitIndex(), p.server.name, entries))
+	} else {
+		p.sendSnapshotRequest(newSnapshotRequest(p.server.name, p.server.snapshot))
+	}
+}
+
+//--------------------------------------
+// Append Entries
+//--------------------------------------
+
+// Sends an AppendEntries request to the peer through the transport.
+func (p *Peer) sendAppendEntriesRequest(req *AppendEntriesRequest) {
+	tracef("peer.append.send: %s->%s [prevLog:%v length: %v]\n",
+		p.server.Name(), p.Name, req.PrevLogIndex, len(req.Entries))
+
+	resp := p.server.Transporter().SendAppendEntriesRequest(p.server, p, req)
+	if resp == nil {
+		p.server.DispatchEvent(newEvent(HeartbeatIntervalEventType, p, nil))
+		debugln("peer.append.timeout: ", p.server.Name(), "->", p.Name)
+		return
+	}
+	traceln("peer.append.resp: ", p.server.Name(), "<-", p.Name)
+
+	p.setLastActivity(time.Now())
+	// If successful then update the previous log index.
+	p.Lock()
+	if resp.Success() {
+		if len(req.Entries) > 0 {
+			p.prevLogIndex = req.Entries[len(req.Entries)-1].GetIndex()
+
+			// if peer append a log entry from the current term
+			// we set append to true
+			if req.Entries[len(req.Entries)-1].GetTerm() == p.server.currentTerm {
+				resp.append = true
+			}
+		}
+		traceln("peer.append.resp.success: ", p.Name, "; idx =", p.prevLogIndex)
+		// If it was unsuccessful then decrement the previous log index and
+		// we'll try again next time.
+	} else {
+		if resp.Term() > p.server.Term() {
+			// this happens when there is a new leader comes up that this *leader* has not
+			// known yet.
+			// this server can know until the new leader send a ae with higher term
+			// or this server finish processing this response.
+			debugln("peer.append.resp.not.update: new.leader.found")
+		} else if resp.Term() == req.Term && resp.CommitIndex() >= p.prevLogIndex {
+			// we may miss a response from peer
+			// so maybe the peer has committed the logs we just sent
+			// but we did not receive the successful reply and did not increase
+			// the prevLogIndex
+
+			// peer failed to truncate the log and sent a fail reply at this time
+			// we just need to update peer's prevLog index to commitIndex
+
+			p.prevLogIndex = resp.CommitIndex()
+			debugln("peer.append.resp.update: ", p.Name, "; idx =", p.prevLogIndex)
+
+		} else if p.prevLogIndex > 0 {
+			// Decrement the previous log index down until we find a match. Don't
+			// let it go below where the peer's commit index is though. That's a
+			// problem.
+			p.prevLogIndex--
+			// if it not enough, we directly decrease to the index of the
+			if p.prevLogIndex > resp.Index() {
+				p.prevLogIndex = resp.Index()
+			}
+
+			debugln("peer.append.resp.decrement: ", p.Name, "; idx =", p.prevLogIndex)
+		}
+	}
+	p.Unlock()
+
+	// Attach the peer to resp, thus server can know where it comes from
+	resp.peer = p.Name
+	// Send response to server for processing.
+	p.server.sendAsync(resp)
+}
+
+// Sends an Snapshot request to the peer through the transport.
+func (p *Peer) sendSnapshotRequest(req *SnapshotRequest) {
+	debugln("peer.snap.send: ", p.Name)
+
+	resp := p.server.Transporter().SendSnapshotRequest(p.server, p, req)
+	if resp == nil {
+		debugln("peer.snap.timeout: ", p.Name)
+		return
+	}
+
+	debugln("peer.snap.recv: ", p.Name)
+
+	// If successful, the peer should have been to snapshot state
+	// Send it the snapshot!
+	p.setLastActivity(time.Now())
+
+	if resp.Success {
+		p.sendSnapshotRecoveryRequest()
+	} else {
+		debugln("peer.snap.failed: ", p.Name)
+		return
+	}
+
+}
+
+// Sends an Snapshot Recovery request to the peer through the transport.
+func (p *Peer) sendSnapshotRecoveryRequest() {
+	req := newSnapshotRecoveryRequest(p.server.name, p.server.snapshot)
+	debugln("peer.snap.recovery.send: ", p.Name)
+	resp := p.server.Transporter().SendSnapshotRecoveryRequest(p.server, p, req)
+
+	if resp == nil {
+		debugln("peer.snap.recovery.timeout: ", p.Name)
+		return
+	}
+
+	p.setLastActivity(time.Now())
+	if resp.Success {
+		p.prevLogIndex = req.LastIndex
+	} else {
+		debugln("peer.snap.recovery.failed: ", p.Name)
+		return
+	}
+
+	p.server.sendAsync(resp)
+}
+
+//--------------------------------------
+// Vote Requests
+//--------------------------------------
+
+// send VoteRequest Request
+func (p *Peer) sendVoteRequest(req *RequestVoteRequest, c chan *RequestVoteResponse) {
+	debugln("peer.vote: ", p.server.Name(), "->", p.Name)
+	req.peer = p
+	if resp := p.server.Transporter().SendVoteRequest(p.server, p, req); resp != nil {
+		debugln("peer.vote.recv: ", p.server.Name(), "<-", p.Name)
+		p.setLastActivity(time.Now())
+		resp.peer = p
+		c <- resp
+	} else {
+		debugln("peer.vote.failed: ", p.server.Name(), "<-", p.Name)
+	}
+}

weed/vendor/github.com/chrislusf/raft/protobuf/append_entries_request.pb.go

@@ -0,0 +1,90 @@
+// Code generated by protoc-gen-go.
+// source: append_entries_request.proto
+// DO NOT EDIT!
+
+/*
+Package protobuf is a generated protocol buffer package.
+
+It is generated from these files:
+	append_entries_request.proto
+	append_entries_responses.proto
+	log_entry.proto
+	request_vote_request.proto
+	request_vote_responses.proto
+	snapshot_recovery_request.proto
+	snapshot_recovery_response.proto
+	snapshot_request.proto
+	snapshot_response.proto
+
+It has these top-level messages:
+	AppendEntriesRequest
+*/
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type AppendEntriesRequest struct {
+	Term             *uint64     `protobuf:"varint,1,req" json:"Term,omitempty"`
+	PrevLogIndex     *uint64     `protobuf:"varint,2,req" json:"PrevLogIndex,omitempty"`
+	PrevLogTerm      *uint64     `protobuf:"varint,3,req" json:"PrevLogTerm,omitempty"`
+	CommitIndex      *uint64     `protobuf:"varint,4,req" json:"CommitIndex,omitempty"`
+	LeaderName       *string     `protobuf:"bytes,5,req" json:"LeaderName,omitempty"`
+	Entries          []*LogEntry `protobuf:"bytes,6,rep" json:"Entries,omitempty"`
+	XXX_unrecognized []byte      `json:"-"`
+}
+
+func (m *AppendEntriesRequest) Reset()         { *m = AppendEntriesRequest{} }
+func (m *AppendEntriesRequest) String() string { return proto.CompactTextString(m) }
+func (*AppendEntriesRequest) ProtoMessage()    {}
+
+func (m *AppendEntriesRequest) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *AppendEntriesRequest) GetPrevLogIndex() uint64 {
+	if m != nil && m.PrevLogIndex != nil {
+		return *m.PrevLogIndex
+	}
+	return 0
+}
+
+func (m *AppendEntriesRequest) GetPrevLogTerm() uint64 {
+	if m != nil && m.PrevLogTerm != nil {
+		return *m.PrevLogTerm
+	}
+	return 0
+}
+
+func (m *AppendEntriesRequest) GetCommitIndex() uint64 {
+	if m != nil && m.CommitIndex != nil {
+		return *m.CommitIndex
+	}
+	return 0
+}
+
+func (m *AppendEntriesRequest) GetLeaderName() string {
+	if m != nil && m.LeaderName != nil {
+		return *m.LeaderName
+	}
+	return ""
+}
+
+func (m *AppendEntriesRequest) GetEntries() []*LogEntry {
+	if m != nil {
+		return m.Entries
+	}
+	return nil
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/append_entries_request.proto

@@ -0,0 +1,25 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+import "log_entry.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message AppendEntriesRequest {
+	required uint64 Term=1;
+	required uint64 PrevLogIndex=2;
+	required uint64 PrevLogTerm=3;
+	required uint64 CommitIndex=4;
+	required string LeaderName=5;
+	repeated LogEntry Entries=6;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/append_entries_responses.pb.go

@@ -0,0 +1,57 @@
+// Code generated by protoc-gen-go.
+// source: append_entries_responses.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type AppendEntriesResponse struct {
+	Term             *uint64 `protobuf:"varint,1,req" json:"Term,omitempty"`
+	Index            *uint64 `protobuf:"varint,2,req" json:"Index,omitempty"`
+	CommitIndex      *uint64 `protobuf:"varint,3,req" json:"CommitIndex,omitempty"`
+	Success          *bool   `protobuf:"varint,4,req" json:"Success,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *AppendEntriesResponse) Reset()         { *m = AppendEntriesResponse{} }
+func (m *AppendEntriesResponse) String() string { return proto.CompactTextString(m) }
+func (*AppendEntriesResponse) ProtoMessage()    {}
+
+func (m *AppendEntriesResponse) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *AppendEntriesResponse) GetIndex() uint64 {
+	if m != nil && m.Index != nil {
+		return *m.Index
+	}
+	return 0
+}
+
+func (m *AppendEntriesResponse) GetCommitIndex() uint64 {
+	if m != nil && m.CommitIndex != nil {
+		return *m.CommitIndex
+	}
+	return 0
+}
+
+func (m *AppendEntriesResponse) GetSuccess() bool {
+	if m != nil && m.Success != nil {
+		return *m.Success
+	}
+	return false
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/append_entries_responses.proto

@@ -0,0 +1,22 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message AppendEntriesResponse {
+	required uint64 Term=1;
+	required uint64 Index=2;
+	required uint64 CommitIndex=3;
+	required bool   Success=4;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/log_entry.pb.go

@@ -0,0 +1,57 @@
+// Code generated by protoc-gen-go.
+// source: log_entry.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type LogEntry struct {
+	Index            *uint64 `protobuf:"varint,1,req" json:"Index,omitempty"`
+	Term             *uint64 `protobuf:"varint,2,req" json:"Term,omitempty"`
+	CommandName      *string `protobuf:"bytes,3,req" json:"CommandName,omitempty"`
+	Command          []byte  `protobuf:"bytes,4,opt" json:"Command,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *LogEntry) Reset()         { *m = LogEntry{} }
+func (m *LogEntry) String() string { return proto.CompactTextString(m) }
+func (*LogEntry) ProtoMessage()    {}
+
+func (m *LogEntry) GetIndex() uint64 {
+	if m != nil && m.Index != nil {
+		return *m.Index
+	}
+	return 0
+}
+
+func (m *LogEntry) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *LogEntry) GetCommandName() string {
+	if m != nil && m.CommandName != nil {
+		return *m.CommandName
+	}
+	return ""
+}
+
+func (m *LogEntry) GetCommand() []byte {
+	if m != nil {
+		return m.Command
+	}
+	return nil
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/log_entry.proto

@@ -0,0 +1,22 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message LogEntry {
+	required uint64 Index=1;
+	required uint64 Term=2;
+	required string CommandName=3;
+	optional bytes Command=4; // for nop-command
+}

weed/vendor/github.com/chrislusf/raft/protobuf/request_vote_request.pb.go

@@ -0,0 +1,57 @@
+// Code generated by protoc-gen-go.
+// source: request_vote_request.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type RequestVoteRequest struct {
+	Term             *uint64 `protobuf:"varint,1,req" json:"Term,omitempty"`
+	LastLogIndex     *uint64 `protobuf:"varint,2,req" json:"LastLogIndex,omitempty"`
+	LastLogTerm      *uint64 `protobuf:"varint,3,req" json:"LastLogTerm,omitempty"`
+	CandidateName    *string `protobuf:"bytes,4,req" json:"CandidateName,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *RequestVoteRequest) Reset()         { *m = RequestVoteRequest{} }
+func (m *RequestVoteRequest) String() string { return proto.CompactTextString(m) }
+func (*RequestVoteRequest) ProtoMessage()    {}
+
+func (m *RequestVoteRequest) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *RequestVoteRequest) GetLastLogIndex() uint64 {
+	if m != nil && m.LastLogIndex != nil {
+		return *m.LastLogIndex
+	}
+	return 0
+}
+
+func (m *RequestVoteRequest) GetLastLogTerm() uint64 {
+	if m != nil && m.LastLogTerm != nil {
+		return *m.LastLogTerm
+	}
+	return 0
+}
+
+func (m *RequestVoteRequest) GetCandidateName() string {
+	if m != nil && m.CandidateName != nil {
+		return *m.CandidateName
+	}
+	return ""
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/request_vote_request.proto

@@ -0,0 +1,22 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message RequestVoteRequest {
+	required uint64 Term=1;
+	required uint64 LastLogIndex=2;
+	required uint64 LastLogTerm=3;
+	required string CandidateName=4;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/request_vote_responses.pb.go

@@ -0,0 +1,41 @@
+// Code generated by protoc-gen-go.
+// source: request_vote_responses.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type RequestVoteResponse struct {
+	Term             *uint64 `protobuf:"varint,1,req" json:"Term,omitempty"`
+	VoteGranted      *bool   `protobuf:"varint,2,req" json:"VoteGranted,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *RequestVoteResponse) Reset()         { *m = RequestVoteResponse{} }
+func (m *RequestVoteResponse) String() string { return proto.CompactTextString(m) }
+func (*RequestVoteResponse) ProtoMessage()    {}
+
+func (m *RequestVoteResponse) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *RequestVoteResponse) GetVoteGranted() bool {
+	if m != nil && m.VoteGranted != nil {
+		return *m.VoteGranted
+	}
+	return false
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/request_vote_responses.proto

@@ -0,0 +1,20 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message RequestVoteResponse {
+	required uint64 Term=1;
+	required bool VoteGranted=2;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_recovery_request.pb.go

@@ -0,0 +1,89 @@
+// Code generated by protoc-gen-go.
+// source: snapshot_recovery_request.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type SnapshotRecoveryRequest struct {
+	LeaderName       *string                         `protobuf:"bytes,1,req" json:"LeaderName,omitempty"`
+	LastIndex        *uint64                         `protobuf:"varint,2,req" json:"LastIndex,omitempty"`
+	LastTerm         *uint64                         `protobuf:"varint,3,req" json:"LastTerm,omitempty"`
+	Peers            []*SnapshotRecoveryRequest_Peer `protobuf:"bytes,4,rep" json:"Peers,omitempty"`
+	State            []byte                          `protobuf:"bytes,5,req" json:"State,omitempty"`
+	XXX_unrecognized []byte                          `json:"-"`
+}
+
+func (m *SnapshotRecoveryRequest) Reset()         { *m = SnapshotRecoveryRequest{} }
+func (m *SnapshotRecoveryRequest) String() string { return proto.CompactTextString(m) }
+func (*SnapshotRecoveryRequest) ProtoMessage()    {}
+
+func (m *SnapshotRecoveryRequest) GetLeaderName() string {
+	if m != nil && m.LeaderName != nil {
+		return *m.LeaderName
+	}
+	return ""
+}
+
+func (m *SnapshotRecoveryRequest) GetLastIndex() uint64 {
+	if m != nil && m.LastIndex != nil {
+		return *m.LastIndex
+	}
+	return 0
+}
+
+func (m *SnapshotRecoveryRequest) GetLastTerm() uint64 {
+	if m != nil && m.LastTerm != nil {
+		return *m.LastTerm
+	}
+	return 0
+}
+
+func (m *SnapshotRecoveryRequest) GetPeers() []*SnapshotRecoveryRequest_Peer {
+	if m != nil {
+		return m.Peers
+	}
+	return nil
+}
+
+func (m *SnapshotRecoveryRequest) GetState() []byte {
+	if m != nil {
+		return m.State
+	}
+	return nil
+}
+
+type SnapshotRecoveryRequest_Peer struct {
+	Name             *string `protobuf:"bytes,1,req" json:"Name,omitempty"`
+	ConnectionString *string `protobuf:"bytes,2,req" json:"ConnectionString,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *SnapshotRecoveryRequest_Peer) Reset()         { *m = SnapshotRecoveryRequest_Peer{} }
+func (m *SnapshotRecoveryRequest_Peer) String() string { return proto.CompactTextString(m) }
+func (*SnapshotRecoveryRequest_Peer) ProtoMessage()    {}
+
+func (m *SnapshotRecoveryRequest_Peer) GetName() string {
+	if m != nil && m.Name != nil {
+		return *m.Name
+	}
+	return ""
+}
+
+func (m *SnapshotRecoveryRequest_Peer) GetConnectionString() string {
+	if m != nil && m.ConnectionString != nil {
+		return *m.ConnectionString
+	}
+	return ""
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_recovery_request.proto

@@ -0,0 +1,29 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message SnapshotRecoveryRequest {
+	required string  LeaderName=1;
+	required uint64  LastIndex=2;
+	required uint64  LastTerm=3;
+
+	message Peer {
+		required string Name=1;
+		required string ConnectionString=2;
+	}
+	repeated Peer  Peers=4;
+
+	required bytes   State=5;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_recovery_response.pb.go

@@ -0,0 +1,49 @@
+// Code generated by protoc-gen-go.
+// source: snapshot_recovery_response.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type SnapshotRecoveryResponse struct {
+	Term             *uint64 `protobuf:"varint,1,req" json:"Term,omitempty"`
+	Success          *bool   `protobuf:"varint,2,req" json:"Success,omitempty"`
+	CommitIndex      *uint64 `protobuf:"varint,3,req" json:"CommitIndex,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *SnapshotRecoveryResponse) Reset()         { *m = SnapshotRecoveryResponse{} }
+func (m *SnapshotRecoveryResponse) String() string { return proto.CompactTextString(m) }
+func (*SnapshotRecoveryResponse) ProtoMessage()    {}
+
+func (m *SnapshotRecoveryResponse) GetTerm() uint64 {
+	if m != nil && m.Term != nil {
+		return *m.Term
+	}
+	return 0
+}
+
+func (m *SnapshotRecoveryResponse) GetSuccess() bool {
+	if m != nil && m.Success != nil {
+		return *m.Success
+	}
+	return false
+}
+
+func (m *SnapshotRecoveryResponse) GetCommitIndex() uint64 {
+	if m != nil && m.CommitIndex != nil {
+		return *m.CommitIndex
+	}
+	return 0
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_recovery_response.proto

@@ -0,0 +1,21 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message SnapshotRecoveryResponse {
+	required uint64 Term=1;
+	required bool Success=2;
+	required uint64 CommitIndex=3;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_request.pb.go

@@ -0,0 +1,49 @@
+// Code generated by protoc-gen-go.
+// source: snapshot_request.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type SnapshotRequest struct {
+	LeaderName       *string `protobuf:"bytes,1,req" json:"LeaderName,omitempty"`
+	LastIndex        *uint64 `protobuf:"varint,2,req" json:"LastIndex,omitempty"`
+	LastTerm         *uint64 `protobuf:"varint,3,req" json:"LastTerm,omitempty"`
+	XXX_unrecognized []byte  `json:"-"`
+}
+
+func (m *SnapshotRequest) Reset()         { *m = SnapshotRequest{} }
+func (m *SnapshotRequest) String() string { return proto.CompactTextString(m) }
+func (*SnapshotRequest) ProtoMessage()    {}
+
+func (m *SnapshotRequest) GetLeaderName() string {
+	if m != nil && m.LeaderName != nil {
+		return *m.LeaderName
+	}
+	return ""
+}
+
+func (m *SnapshotRequest) GetLastIndex() uint64 {
+	if m != nil && m.LastIndex != nil {
+		return *m.LastIndex
+	}
+	return 0
+}
+
+func (m *SnapshotRequest) GetLastTerm() uint64 {
+	if m != nil && m.LastTerm != nil {
+		return *m.LastTerm
+	}
+	return 0
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_request.proto

@@ -0,0 +1,21 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message SnapshotRequest {
+	required string LeaderName=1;
+	required uint64 LastIndex=2;
+	required uint64 LastTerm=3;
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_response.pb.go

@@ -0,0 +1,33 @@
+// Code generated by protoc-gen-go.
+// source: snapshot_response.proto
+// DO NOT EDIT!
+
+package protobuf
+
+import proto "github.com/gogo/protobuf/proto"
+import math "math"
+
+// discarding unused import gogoproto "github.com/gogo/protobuf/gogoproto/gogo.pb"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = math.Inf
+
+type SnapshotResponse struct {
+	Success          *bool  `protobuf:"varint,1,req" json:"Success,omitempty"`
+	XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *SnapshotResponse) Reset()         { *m = SnapshotResponse{} }
+func (m *SnapshotResponse) String() string { return proto.CompactTextString(m) }
+func (*SnapshotResponse) ProtoMessage()    {}
+
+func (m *SnapshotResponse) GetSuccess() bool {
+	if m != nil && m.Success != nil {
+		return *m.Success
+	}
+	return false
+}
+
+func init() {
+}

weed/vendor/github.com/chrislusf/raft/protobuf/snapshot_response.proto

@@ -0,0 +1,19 @@
+package protobuf;
+
+import "github.com/gogo/protobuf/gogoproto/gogo.proto";
+
+option (gogoproto.gostring_all) = true;
+option (gogoproto.equal_all) = true;
+option (gogoproto.verbose_equal_all) = true;
+option (gogoproto.goproto_stringer_all) = false;
+option (gogoproto.stringer_all) =  true;
+option (gogoproto.populate_all) = true;
+option (gogoproto.testgen_all) = true;
+option (gogoproto.benchgen_all) = true;
+option (gogoproto.marshaler_all) = true;
+option (gogoproto.sizer_all) = true;
+option (gogoproto.unmarshaler_all) = true;
+
+message SnapshotResponse {
+	required bool Success=1;
+}

weed/vendor/github.com/chrislusf/raft/request_vote.go

@@ -0,0 +1,122 @@
+package raft
+
+import (
+	"io"
+	"io/ioutil"
+
+	"github.com/chrislusf/raft/protobuf"
+	"github.com/gogo/protobuf/proto"
+)
+
+// The request sent to a server to vote for a candidate to become a leader.
+type RequestVoteRequest struct {
+	peer          *Peer
+	Term          uint64
+	LastLogIndex  uint64
+	LastLogTerm   uint64
+	CandidateName string
+}
+
+// The response returned from a server after a vote for a candidate to become a leader.
+type RequestVoteResponse struct {
+	peer        *Peer
+	Term        uint64
+	VoteGranted bool
+}
+
+// Creates a new RequestVote request.
+func newRequestVoteRequest(term uint64, candidateName string, lastLogIndex uint64, lastLogTerm uint64) *RequestVoteRequest {
+	return &RequestVoteRequest{
+		Term:          term,
+		LastLogIndex:  lastLogIndex,
+		LastLogTerm:   lastLogTerm,
+		CandidateName: candidateName,
+	}
+}
+
+// Encodes the RequestVoteRequest to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (req *RequestVoteRequest) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.RequestVoteRequest{
+		Term:          proto.Uint64(req.Term),
+		LastLogIndex:  proto.Uint64(req.LastLogIndex),
+		LastLogTerm:   proto.Uint64(req.LastLogTerm),
+		CandidateName: proto.String(req.CandidateName),
+	}
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the RequestVoteRequest from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (req *RequestVoteRequest) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return -1, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.RequestVoteRequest{}
+	if err = proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	req.Term = pb.GetTerm()
+	req.LastLogIndex = pb.GetLastLogIndex()
+	req.LastLogTerm = pb.GetLastLogTerm()
+	req.CandidateName = pb.GetCandidateName()
+
+	return totalBytes, nil
+}
+
+// Creates a new RequestVote response.
+func newRequestVoteResponse(term uint64, voteGranted bool) *RequestVoteResponse {
+	return &RequestVoteResponse{
+		Term:        term,
+		VoteGranted: voteGranted,
+	}
+}
+
+// Encodes the RequestVoteResponse to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (resp *RequestVoteResponse) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.RequestVoteResponse{
+		Term:        proto.Uint64(resp.Term),
+		VoteGranted: proto.Bool(resp.VoteGranted),
+	}
+
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the RequestVoteResponse from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (resp *RequestVoteResponse) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return 0, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.RequestVoteResponse{}
+	if err = proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	resp.Term = pb.GetTerm()
+	resp.VoteGranted = pb.GetVoteGranted()
+
+	return totalBytes, nil
+}

weed/vendor/github.com/chrislusf/raft/snapshot.go

@@ -0,0 +1,304 @@
+package raft
+
+import (
+	"encoding/json"
+	"fmt"
+	"hash/crc32"
+	"io"
+	"io/ioutil"
+	"os"
+
+	"github.com/chrislusf/raft/protobuf"
+	"github.com/gogo/protobuf/proto"
+)
+
+// Snapshot represents an in-memory representation of the current state of the system.
+type Snapshot struct {
+	LastIndex uint64 `json:"lastIndex"`
+	LastTerm  uint64 `json:"lastTerm"`
+
+	// Cluster configuration.
+	Peers []*Peer `json:"peers"`
+	State []byte  `json:"state"`
+	Path  string  `json:"path"`
+}
+
+// The request sent to a server to start from the snapshot.
+type SnapshotRecoveryRequest struct {
+	LeaderName string
+	LastIndex  uint64
+	LastTerm   uint64
+	Peers      []*Peer
+	State      []byte
+}
+
+// The response returned from a server appending entries to the log.
+type SnapshotRecoveryResponse struct {
+	Term        uint64
+	Success     bool
+	CommitIndex uint64
+}
+
+// The request sent to a server to start from the snapshot.
+type SnapshotRequest struct {
+	LeaderName string
+	LastIndex  uint64
+	LastTerm   uint64
+}
+
+// The response returned if the follower entered snapshot state
+type SnapshotResponse struct {
+	Success bool `json:"success"`
+}
+
+// save writes the snapshot to file.
+func (ss *Snapshot) save() error {
+	// Open the file for writing.
+	file, err := os.OpenFile(ss.Path, os.O_CREATE|os.O_WRONLY, 0600)
+	if err != nil {
+		return err
+	}
+	defer file.Close()
+
+	// Serialize to JSON.
+	b, err := json.Marshal(ss)
+	if err != nil {
+		return err
+	}
+
+	// Generate checksum and write it to disk.
+	checksum := crc32.ChecksumIEEE(b)
+	if _, err = fmt.Fprintf(file, "%08x\n", checksum); err != nil {
+		return err
+	}
+
+	// Write the snapshot to disk.
+	if _, err = file.Write(b); err != nil {
+		return err
+	}
+
+	// Ensure that the snapshot has been flushed to disk before continuing.
+	if err := file.Sync(); err != nil {
+		return err
+	}
+
+	return nil
+}
+
+// remove deletes the snapshot file.
+func (ss *Snapshot) remove() error {
+	if err := os.Remove(ss.Path); err != nil {
+		return err
+	}
+	return nil
+}
+
+// Creates a new Snapshot request.
+func newSnapshotRecoveryRequest(leaderName string, snapshot *Snapshot) *SnapshotRecoveryRequest {
+	return &SnapshotRecoveryRequest{
+		LeaderName: leaderName,
+		LastIndex:  snapshot.LastIndex,
+		LastTerm:   snapshot.LastTerm,
+		Peers:      snapshot.Peers,
+		State:      snapshot.State,
+	}
+}
+
+// Encodes the SnapshotRecoveryRequest to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (req *SnapshotRecoveryRequest) Encode(w io.Writer) (int, error) {
+
+	protoPeers := make([]*protobuf.SnapshotRecoveryRequest_Peer, len(req.Peers))
+
+	for i, peer := range req.Peers {
+		protoPeers[i] = &protobuf.SnapshotRecoveryRequest_Peer{
+			Name:             proto.String(peer.Name),
+			ConnectionString: proto.String(peer.ConnectionString),
+		}
+	}
+
+	pb := &protobuf.SnapshotRecoveryRequest{
+		LeaderName: proto.String(req.LeaderName),
+		LastIndex:  proto.Uint64(req.LastIndex),
+		LastTerm:   proto.Uint64(req.LastTerm),
+		Peers:      protoPeers,
+		State:      req.State,
+	}
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the SnapshotRecoveryRequest from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (req *SnapshotRecoveryRequest) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return 0, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.SnapshotRecoveryRequest{}
+	if err = proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	req.LeaderName = pb.GetLeaderName()
+	req.LastIndex = pb.GetLastIndex()
+	req.LastTerm = pb.GetLastTerm()
+	req.State = pb.GetState()
+
+	req.Peers = make([]*Peer, len(pb.Peers))
+
+	for i, peer := range pb.Peers {
+		req.Peers[i] = &Peer{
+			Name:             peer.GetName(),
+			ConnectionString: peer.GetConnectionString(),
+		}
+	}
+
+	return totalBytes, nil
+}
+
+// Creates a new Snapshot response.
+func newSnapshotRecoveryResponse(term uint64, success bool, commitIndex uint64) *SnapshotRecoveryResponse {
+	return &SnapshotRecoveryResponse{
+		Term:        term,
+		Success:     success,
+		CommitIndex: commitIndex,
+	}
+}
+
+// Encode writes the response to a writer.
+// Returns the number of bytes written and any error that occurs.
+func (req *SnapshotRecoveryResponse) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.SnapshotRecoveryResponse{
+		Term:        proto.Uint64(req.Term),
+		Success:     proto.Bool(req.Success),
+		CommitIndex: proto.Uint64(req.CommitIndex),
+	}
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the SnapshotRecoveryResponse from a buffer.
+func (req *SnapshotRecoveryResponse) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return 0, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.SnapshotRecoveryResponse{}
+	if err := proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	req.Term = pb.GetTerm()
+	req.Success = pb.GetSuccess()
+	req.CommitIndex = pb.GetCommitIndex()
+
+	return totalBytes, nil
+}
+
+// Creates a new Snapshot request.
+func newSnapshotRequest(leaderName string, snapshot *Snapshot) *SnapshotRequest {
+	return &SnapshotRequest{
+		LeaderName: leaderName,
+		LastIndex:  snapshot.LastIndex,
+		LastTerm:   snapshot.LastTerm,
+	}
+}
+
+// Encodes the SnapshotRequest to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (req *SnapshotRequest) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.SnapshotRequest{
+		LeaderName: proto.String(req.LeaderName),
+		LastIndex:  proto.Uint64(req.LastIndex),
+		LastTerm:   proto.Uint64(req.LastTerm),
+	}
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the SnapshotRequest from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (req *SnapshotRequest) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return 0, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.SnapshotRequest{}
+
+	if err := proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	req.LeaderName = pb.GetLeaderName()
+	req.LastIndex = pb.GetLastIndex()
+	req.LastTerm = pb.GetLastTerm()
+
+	return totalBytes, nil
+}
+
+// Creates a new Snapshot response.
+func newSnapshotResponse(success bool) *SnapshotResponse {
+	return &SnapshotResponse{
+		Success: success,
+	}
+}
+
+// Encodes the SnapshotResponse to a buffer. Returns the number of bytes
+// written and any error that may have occurred.
+func (resp *SnapshotResponse) Encode(w io.Writer) (int, error) {
+	pb := &protobuf.SnapshotResponse{
+		Success: proto.Bool(resp.Success),
+	}
+	p, err := proto.Marshal(pb)
+	if err != nil {
+		return -1, err
+	}
+
+	return w.Write(p)
+}
+
+// Decodes the SnapshotResponse from a buffer. Returns the number of bytes read and
+// any error that occurs.
+func (resp *SnapshotResponse) Decode(r io.Reader) (int, error) {
+	data, err := ioutil.ReadAll(r)
+
+	if err != nil {
+		return 0, err
+	}
+
+	totalBytes := len(data)
+
+	pb := &protobuf.SnapshotResponse{}
+	if err := proto.Unmarshal(data, pb); err != nil {
+		return -1, err
+	}
+
+	resp.Success = pb.GetSuccess()
+
+	return totalBytes, nil
+}

weed/vendor/github.com/chrislusf/raft/statemachine.go

@@ -0,0 +1,9 @@
+package raft
+
+// StateMachine is the interface for allowing the host application to save and
+// recovery the state machine. This makes it possible to make snapshots
+// and compact the log.
+type StateMachine interface {
+	Save() ([]byte, error)
+	Recovery([]byte) error
+}

weed/vendor/github.com/chrislusf/raft/test.go

@@ -0,0 +1,197 @@
+package raft
+
+import (
+	"fmt"
+	"io/ioutil"
+	"os"
+	"time"
+)
+
+const (
+	testHeartbeatInterval = 50 * time.Millisecond
+	testElectionTimeout   = 200 * time.Millisecond
+)
+
+const (
+	testListenerLoggerEnabled = false
+)
+
+func init() {
+	RegisterCommand(&testCommand1{})
+	RegisterCommand(&testCommand2{})
+}
+
+//------------------------------------------------------------------------------
+//
+// Helpers
+//
+//------------------------------------------------------------------------------
+
+//--------------------------------------
+// Logs
+//--------------------------------------
+
+func getLogPath() string {
+	f, _ := ioutil.TempFile("", "raft-log-")
+	f.Close()
+	os.Remove(f.Name())
+	return f.Name()
+}
+
+func setupLog(entries []*LogEntry) (*Log, string) {
+	f, _ := ioutil.TempFile("", "raft-log-")
+
+	for _, entry := range entries {
+		entry.Encode(f)
+	}
+	err := f.Close()
+
+	if err != nil {
+		panic(err)
+	}
+
+	log := newLog()
+	log.ApplyFunc = func(e *LogEntry, c Command) (interface{}, error) {
+		return nil, nil
+	}
+	if err := log.open(f.Name()); err != nil {
+		panic(err)
+	}
+	return log, f.Name()
+}
+
+//--------------------------------------
+// Servers
+//--------------------------------------
+
+func newTestServer(name string, transporter Transporter) Server {
+	p, _ := ioutil.TempDir("", "raft-server-")
+	if err := os.MkdirAll(p, 0644); err != nil {
+		panic(err.Error())
+	}
+	server, _ := NewServer(name, p, transporter, nil, nil, "")
+	if testListenerLoggerEnabled {
+		fn := func(e Event) {
+			server := e.Source().(Server)
+			warnf("[%s] %s %v -> %v\n", server.Name(), e.Type(), e.PrevValue(), e.Value())
+		}
+		server.AddEventListener(StateChangeEventType, fn)
+		server.AddEventListener(LeaderChangeEventType, fn)
+		server.AddEventListener(TermChangeEventType, fn)
+	}
+	return server
+}
+
+func newTestServerWithPath(name string, transporter Transporter, p string) Server {
+	server, _ := NewServer(name, p, transporter, nil, nil, "")
+	return server
+}
+
+func newTestServerWithLog(name string, transporter Transporter, entries []*LogEntry) Server {
+	server := newTestServer(name, transporter)
+	f, err := os.Create(server.LogPath())
+	if err != nil {
+		panic(err)
+	}
+
+	for _, entry := range entries {
+		entry.Encode(f)
+	}
+	f.Close()
+	return server
+}
+
+func newTestCluster(names []string, transporter Transporter, lookup map[string]Server) []Server {
+	servers := []Server{}
+	e0, _ := newLogEntry(newLog(), nil, 1, 1, &testCommand1{Val: "foo", I: 20})
+
+	for _, name := range names {
+		if lookup[name] != nil {
+			panic(fmt.Sprintf("raft: Duplicate server in test cluster! %v", name))
+		}
+		server := newTestServerWithLog("1", transporter, []*LogEntry{e0})
+		server.SetElectionTimeout(testElectionTimeout)
+		servers = append(servers, server)
+		lookup[name] = server
+	}
+	for _, server := range servers {
+		server.SetHeartbeatInterval(testHeartbeatInterval)
+		server.Start()
+		for _, peer := range servers {
+			server.AddPeer(peer.Name(), "")
+		}
+	}
+	return servers
+}
+
+//--------------------------------------
+// Transporter
+//--------------------------------------
+
+type testTransporter struct {
+	sendVoteRequestFunc          func(server Server, peer *Peer, req *RequestVoteRequest) *RequestVoteResponse
+	sendAppendEntriesRequestFunc func(server Server, peer *Peer, req *AppendEntriesRequest) *AppendEntriesResponse
+	sendSnapshotRequestFunc      func(server Server, peer *Peer, req *SnapshotRequest) *SnapshotResponse
+}
+
+func (t *testTransporter) SendVoteRequest(server Server, peer *Peer, req *RequestVoteRequest) *RequestVoteResponse {
+	return t.sendVoteRequestFunc(server, peer, req)
+}
+
+func (t *testTransporter) SendAppendEntriesRequest(server Server, peer *Peer, req *AppendEntriesRequest) *AppendEntriesResponse {
+	return t.sendAppendEntriesRequestFunc(server, peer, req)
+}
+
+func (t *testTransporter) SendSnapshotRequest(server Server, peer *Peer, req *SnapshotRequest) *SnapshotResponse {
+	return t.sendSnapshotRequestFunc(server, peer, req)
+}
+
+func (t *testTransporter) SendSnapshotRecoveryRequest(server Server, peer *Peer, req *SnapshotRecoveryRequest) *SnapshotRecoveryResponse {
+	return t.SendSnapshotRecoveryRequest(server, peer, req)
+}
+
+type testStateMachine struct {
+	saveFunc     func() ([]byte, error)
+	recoveryFunc func([]byte) error
+}
+
+func (sm *testStateMachine) Save() ([]byte, error) {
+	return sm.saveFunc()
+}
+
+func (sm *testStateMachine) Recovery(state []byte) error {
+	return sm.recoveryFunc(state)
+}
+
+//--------------------------------------
+// Command1
+//--------------------------------------
+
+type testCommand1 struct {
+	Val string `json:"val"`
+	I   int    `json:"i"`
+}
+
+func (c *testCommand1) CommandName() string {
+	return "cmd_1"
+}
+
+func (c *testCommand1) Apply(server Server) (interface{}, error) {
+	return nil, nil
+}
+
+//--------------------------------------
+// Command2
+//--------------------------------------
+
+type testCommand2 struct {
+	X int `json:"x"`
+}
+
+func (c *testCommand2) CommandName() string {
+	return "cmd_2"
+}
+
+func (c *testCommand2) Apply(server Server) (interface{}, error) {
+	return nil, nil
+}

weed/vendor/github.com/chrislusf/raft/transporter.go

@@ -0,0 +1,16 @@
+package raft
+
+//------------------------------------------------------------------------------
+//
+// Typedefs
+//
+//------------------------------------------------------------------------------
+
+// Transporter is the interface for allowing the host application to transport
+// requests to other nodes.
+type Transporter interface {
+	SendVoteRequest(server Server, peer *Peer, req *RequestVoteRequest) *RequestVoteResponse
+	SendAppendEntriesRequest(server Server, peer *Peer, req *AppendEntriesRequest) *AppendEntriesResponse
+	SendSnapshotRequest(server Server, peer *Peer, req *SnapshotRequest) *SnapshotResponse
+	SendSnapshotRecoveryRequest(server Server, peer *Peer, req *SnapshotRecoveryRequest) *SnapshotRecoveryResponse
+}

weed/vendor/github.com/chrislusf/raft/util.go

@@ -0,0 +1,62 @@
+package raft
+
+import (
+	"fmt"
+	"io"
+	"math/rand"
+	"os"
+	"time"
+)
+
+// uint64Slice implements sort interface
+type uint64Slice []uint64
+
+func (p uint64Slice) Len() int           { return len(p) }
+func (p uint64Slice) Less(i, j int) bool { return p[i] < p[j] }
+func (p uint64Slice) Swap(i, j int)      { p[i], p[j] = p[j], p[i] }
+
+// WriteFile writes data to a file named by filename.
+// If the file does not exist, WriteFile creates it with permissions perm;
+// otherwise WriteFile truncates it before writing.
+// This is copied from ioutil.WriteFile with the addition of a Sync call to
+// ensure the data reaches the disk.
+func writeFileSynced(filename string, data []byte, perm os.FileMode) error {
+	f, err := os.OpenFile(filename, os.O_WRONLY|os.O_CREATE|os.O_TRUNC, perm)
+	if err != nil {
+		return err
+	}
+	defer f.Close() // Idempotent
+
+	n, err := f.Write(data)
+	if err == nil && n < len(data) {
+		return io.ErrShortWrite
+	} else if err != nil {
+		return err
+	}
+
+	if err = f.Sync(); err != nil {
+		return err
+	}
+
+	return f.Close()
+}
+
+// Waits for a random time between two durations and sends the current time on
+// the returned channel.
+func afterBetween(min time.Duration, max time.Duration) <-chan time.Time {
+	rand := rand.New(rand.NewSource(time.Now().UnixNano()))
+	d, delta := min, (max - min)
+	if delta > 0 {
+		d += time.Duration(rand.Int63n(int64(delta)))
+	}
+	return time.After(d)
+}
+
+// TODO(xiangli): Remove assertions when we reach version 1.0
+
+// _assert will panic with a given formatted message if the given condition is false.
+func _assert(condition bool, msg string, v ...interface{}) {
+	if !condition {
+		panic(fmt.Sprintf("assertion failed: "+msg, v...))
+	}
+}