seaweedfs/weed/command/export.go

214 lines
5.9 KiB
Go
Raw Normal View History

package command
import (
"archive/tar"
"bytes"
2014-10-26 18:25:02 +00:00
"fmt"
"os"
"path"
2015-07-09 06:19:54 +00:00
"path/filepath"
"strconv"
"strings"
"text/template"
"time"
"github.com/chrislusf/seaweedfs/weed/glog"
"github.com/chrislusf/seaweedfs/weed/storage"
)
const (
defaultFnFormat = `{{.Mime}}/{{.Id}}:{{.Name}}`
2015-05-26 06:53:45 +00:00
timeFormat = "2006-01-02T15:04:05"
)
var (
export ExportOptions
)
type ExportOptions struct {
dir *string
collection *string
volumeId *int
}
var cmdExport = &Command{
2015-05-26 06:53:45 +00:00
UsageLine: "export -dir=/tmp -volumeId=234 -o=/dir/name.tar -fileNameFormat={{.Name}} -newer='" + timeFormat + "'",
2013-01-21 03:51:27 +00:00
Short: "list or export files from one volume data file",
Long: `List all files in a volume, or Export all files in a volume to a tar file if the output is specified.
2013-03-19 04:29:25 +00:00
The format of file name in the tar file can be customized. Default is {{.Mime}}/{{.Id}}:{{.Name}}. Also available is {{.Key}}.
`,
}
func init() {
cmdExport.Run = runExport // break init cycle
export.dir = cmdExport.Flag.String("dir", ".", "input data directory to store volume data files")
export.collection = cmdExport.Flag.String("collection", "", "the volume collection name")
export.volumeId = cmdExport.Flag.Int("volumeId", -1, "a volume id. The volume .dat and .idx files should already exist in the dir.")
}
var (
2015-06-02 07:33:13 +00:00
output = cmdExport.Flag.String("o", "", "output tar file name, must ends with .tar, or just a \"-\" for stdout")
2015-07-09 06:19:54 +00:00
format = cmdExport.Flag.String("fileNameFormat", defaultFnFormat, "filename formatted with {{.Mime}} {{.Id}} {{.Name}} {{.Ext}}")
2016-09-08 01:05:57 +00:00
newer = cmdExport.Flag.String("newer", "", "export only files newer than this time, default is all files. Must be specified in RFC3339 without timezone, e.g. 2006-01-02T15:04:05")
2015-06-02 07:23:41 +00:00
2015-06-02 07:33:13 +00:00
tarOutputFile *tar.Writer
tarHeader tar.Header
fileNameTemplate *template.Template
fileNameTemplateBuffer = bytes.NewBuffer(nil)
newerThan time.Time
newerThanUnix int64 = -1
localLocation, _ = time.LoadLocation("Local")
)
func runExport(cmd *Command, args []string) bool {
var err error
if *newer != "" {
if newerThan, err = time.ParseInLocation(timeFormat, *newer, localLocation); err != nil {
fmt.Println("cannot parse 'newer' argument: " + err.Error())
return false
}
newerThanUnix = newerThan.Unix()
}
2015-06-02 07:23:41 +00:00
if *export.volumeId == -1 {
return false
}
2015-06-02 07:33:13 +00:00
if *output != "" {
if *output != "-" && !strings.HasSuffix(*output, ".tar") {
fmt.Println("the output file", *output, "should be '-' or end with .tar")
return false
}
2015-06-02 07:33:13 +00:00
if fileNameTemplate, err = template.New("name").Parse(*format); err != nil {
fmt.Println("cannot parse format " + *format + ": " + err.Error())
return false
}
2015-06-02 07:33:13 +00:00
var outputFile *os.File
if *output == "-" {
outputFile = os.Stdout
} else {
2015-06-02 07:33:13 +00:00
if outputFile, err = os.Create(*output); err != nil {
glog.Fatalf("cannot open output tar %s: %s", *output, err)
}
}
2015-06-02 07:33:13 +00:00
defer outputFile.Close()
tarOutputFile = tar.NewWriter(outputFile)
defer tarOutputFile.Close()
t := time.Now()
tarHeader = tar.Header{Mode: 0644,
ModTime: t, Uid: os.Getuid(), Gid: os.Getgid(),
Typeflag: tar.TypeReg,
AccessTime: t, ChangeTime: t}
}
2015-06-02 07:23:41 +00:00
fileName := strconv.Itoa(*export.volumeId)
if *export.collection != "" {
fileName = *export.collection + "_" + fileName
}
2015-06-02 07:23:41 +00:00
vid := storage.VolumeId(*export.volumeId)
indexFile, err := os.OpenFile(path.Join(*export.dir, fileName+".idx"), os.O_RDONLY, 0644)
if err != nil {
glog.Fatalf("Create Volume Index [ERROR] %s\n", err)
}
defer indexFile.Close()
2017-05-27 05:51:25 +00:00
needleMap, err := storage.LoadBtreeNeedleMap(indexFile)
2013-02-10 21:41:25 +00:00
if err != nil {
glog.Fatalf("cannot load needle map from %s: %s", indexFile.Name(), err)
2013-02-10 21:41:25 +00:00
}
var version storage.Version
2015-06-02 07:23:41 +00:00
err = storage.ScanVolumeFile(*export.dir, *export.collection, vid,
storage.NeedleMapInMemory,
func(superBlock storage.SuperBlock) error {
version = superBlock.Version()
return nil
}, true, func(n *storage.Needle, offset int64) error {
2015-06-02 07:33:13 +00:00
nv, ok := needleMap.Get(n.Id)
glog.V(3).Infof("key %d offset %d size %d disk_size %d gzip %v ok %v nv %+v",
n.Id, offset, n.Size, n.DiskSize(), n.IsGzipped(), ok, nv)
if ok && nv.Size > 0 && int64(nv.Offset)*8 == offset {
if newerThanUnix >= 0 && n.HasLastModifiedDate() && n.LastModified < uint64(newerThanUnix) {
glog.V(3).Infof("Skipping this file, as it's old enough: LastModified %d vs %d",
n.LastModified, newerThanUnix)
return nil
}
return walker(vid, n, version)
}
if !ok {
glog.V(2).Infof("This seems deleted %d size %d", n.Id, n.Size)
} else {
glog.V(2).Infof("Skipping later-updated Id %d size %d", n.Id, n.Size)
}
return nil
})
if err != nil {
glog.Fatalf("Export Volume File [ERROR] %s\n", err)
}
return true
}
type nameParams struct {
Name string
Id uint64
Mime string
Key string
2015-07-09 06:19:54 +00:00
Ext string
}
func walker(vid storage.VolumeId, n *storage.Needle, version storage.Version) (err error) {
2014-03-24 04:57:10 +00:00
key := storage.NewFileIdFromNeedle(vid, n).String()
2015-06-02 07:33:13 +00:00
if tarOutputFile != nil {
fileNameTemplateBuffer.Reset()
if err = fileNameTemplate.Execute(fileNameTemplateBuffer,
2015-07-09 06:19:54 +00:00
nameParams{
Name: string(n.Name),
Id: n.Id,
Mime: string(n.Mime),
Key: key,
2015-07-09 06:19:54 +00:00
Ext: filepath.Ext(string(n.Name)),
},
); err != nil {
return err
}
2015-06-02 07:33:13 +00:00
fileName := fileNameTemplateBuffer.String()
if n.IsGzipped() && path.Ext(fileName) != ".gz" {
fileName = fileName + ".gz"
}
2015-06-02 07:33:13 +00:00
tarHeader.Name, tarHeader.Size = fileName, int64(len(n.Data))
if n.HasLastModifiedDate() {
tarHeader.ModTime = time.Unix(int64(n.LastModified), 0)
} else {
2015-05-26 06:53:45 +00:00
tarHeader.ModTime = time.Unix(0, 0)
}
tarHeader.ChangeTime = tarHeader.ModTime
2015-06-02 07:33:13 +00:00
if err = tarOutputFile.WriteHeader(&tarHeader); err != nil {
return err
}
2015-06-02 07:33:13 +00:00
_, err = tarOutputFile.Write(n.Data)
} else {
size := n.DataSize
if version == storage.Version1 {
size = n.Size
}
fmt.Printf("key=%s Name=%s Size=%d gzip=%t mime=%s\n",
key,
n.Name,
size,
2013-01-21 03:47:04 +00:00
n.IsGzipped(),
n.Mime,
2013-01-21 03:47:04 +00:00
)
}
return
}