feat: support of ZIP list/get/head as S3 extension (#12267)

When enabled, it is possible to list/get files
inside a zip file without uncompressing it.

Signed-off-by: Anis Elleuch <anis@min.io>
This commit is contained in:
Anis Elleuch 2021-06-10 16:17:03 +01:00 committed by GitHub
parent c221633a8a
commit ba5fb2365c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
10 changed files with 742 additions and 39 deletions

View File

@ -236,12 +236,20 @@ func (api objectAPIHandlers) ListObjectsV2Handler(w http.ResponseWriter, r *http
return
}
listObjectsV2 := objectAPI.ListObjectsV2
var (
listObjectsV2Info ListObjectsV2Info
err error
)
if r.Header.Get(xMinIOExtract) == "true" && strings.Contains(prefix, archivePattern) {
// Inititate a list objects operation inside a zip file based in the input params
listObjectsV2Info, err = listObjectsV2InArchive(ctx, objectAPI, bucket, prefix, token, delimiter, maxKeys, fetchOwner, startAfter)
} else {
// Inititate a list objects operation based on the input params.
// On success would return back ListObjectsInfo object to be
// marshaled into S3 compatible XML header.
listObjectsV2Info, err := listObjectsV2(ctx, bucket, prefix, token, delimiter, maxKeys, fetchOwner, startAfter)
listObjectsV2Info, err = objectAPI.ListObjectsV2(ctx, bucket, prefix, token, delimiter, maxKeys, fetchOwner, startAfter)
}
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return

View File

@ -309,20 +309,7 @@ func (api objectAPIHandlers) SelectObjectContentHandler(w http.ResponseWriter, r
})
}
// GetObjectHandler - GET Object
// ----------
// This implementation of the GET operation retrieves object. To use GET,
// you must have READ access to the object.
func (api objectAPIHandlers) GetObjectHandler(w http.ResponseWriter, r *http.Request) {
ctx := newContext(r, w, "GetObject")
defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
objectAPI := api.ObjectAPI()
if objectAPI == nil {
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL, guessIsBrowserReq(r))
return
}
func (api objectAPIHandlers) getObjectHandler(ctx context.Context, objectAPI ObjectLayer, bucket, object string, w http.ResponseWriter, r *http.Request) {
if crypto.S3.IsRequested(r.Header) || crypto.S3KMS.IsRequested(r.Header) { // If SSE-S3 or SSE-KMS present -> AWS fails with undefined error
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
@ -331,13 +318,6 @@ func (api objectAPIHandlers) GetObjectHandler(w http.ResponseWriter, r *http.Req
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
}
vars := mux.Vars(r)
bucket := vars["bucket"]
object, err := unescapePath(vars["object"])
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
// get gateway encryption options
opts, err := getOpts(ctx, r, bucket, object)
@ -555,19 +535,37 @@ func (api objectAPIHandlers) GetObjectHandler(w http.ResponseWriter, r *http.Req
})
}
// HeadObjectHandler - HEAD Object
// -----------
// The HEAD operation retrieves metadata from an object without returning the object itself.
func (api objectAPIHandlers) HeadObjectHandler(w http.ResponseWriter, r *http.Request) {
ctx := newContext(r, w, "HeadObject")
// GetObjectHandler - GET Object
// ----------
// This implementation of the GET operation retrieves object. To use GET,
// you must have READ access to the object.
func (api objectAPIHandlers) GetObjectHandler(w http.ResponseWriter, r *http.Request) {
ctx := newContext(r, w, "GetObject")
defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
objectAPI := api.ObjectAPI()
if objectAPI == nil {
writeErrorResponseHeadersOnly(w, errorCodes.ToAPIErr(ErrServerNotInitialized))
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrServerNotInitialized), r.URL, guessIsBrowserReq(r))
return
}
vars := mux.Vars(r)
bucket := vars["bucket"]
object, err := unescapePath(vars["object"])
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
if r.Header.Get(xMinIOExtract) == "true" && strings.Contains(object, archivePattern) {
api.getObjectInArchiveFileHandler(ctx, objectAPI, bucket, object, w, r)
} else {
api.getObjectHandler(ctx, objectAPI, bucket, object, w, r)
}
}
func (api objectAPIHandlers) headObjectHandler(ctx context.Context, objectAPI ObjectLayer, bucket, object string, w http.ResponseWriter, r *http.Request) {
if crypto.S3.IsRequested(r.Header) || crypto.S3KMS.IsRequested(r.Header) { // If SSE-S3 or SSE-KMS present -> AWS fails with undefined error
writeErrorResponseHeadersOnly(w, errorCodes.ToAPIErr(ErrBadRequest))
return
@ -576,13 +574,6 @@ func (api objectAPIHandlers) HeadObjectHandler(w http.ResponseWriter, r *http.Re
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
}
vars := mux.Vars(r)
bucket := vars["bucket"]
object, err := unescapePath(vars["object"])
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
getObjectInfo := objectAPI.GetObjectInfo
if api.CacheAPI() != nil {
@ -768,6 +759,34 @@ func (api objectAPIHandlers) HeadObjectHandler(w http.ResponseWriter, r *http.Re
})
}
// HeadObjectHandler - HEAD Object
// -----------
// The HEAD operation retrieves metadata from an object without returning the object itself.
func (api objectAPIHandlers) HeadObjectHandler(w http.ResponseWriter, r *http.Request) {
ctx := newContext(r, w, "HeadObject")
defer logger.AuditLog(ctx, w, r, mustGetClaimsFromToken(r))
objectAPI := api.ObjectAPI()
if objectAPI == nil {
writeErrorResponseHeadersOnly(w, errorCodes.ToAPIErr(ErrServerNotInitialized))
return
}
vars := mux.Vars(r)
bucket := vars["bucket"]
object, err := unescapePath(vars["object"])
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
if r.Header.Get(xMinIOExtract) == "true" && strings.Contains(object, archivePattern) {
api.headObjectInArchiveFileHandler(ctx, objectAPI, bucket, object, w, r)
} else {
api.headObjectHandler(ctx, objectAPI, bucket, object, w, r)
}
}
// Extract metadata relevant for an CopyObject operation based on conditional
// header values specified in X-Amz-Metadata-Directive.
func getCpObjMetadataFromHeader(ctx context.Context, r *http.Request, userMeta map[string]string) (map[string]string, error) {
@ -1704,6 +1723,14 @@ func (api objectAPIHandlers) PutObjectHandler(w http.ResponseWriter, r *http.Req
return
}
if r.Header.Get(xMinIOExtract) == "true" && strings.HasSuffix(object, archiveExt) {
opts := ObjectOptions{VersionID: objInfo.VersionID, MTime: objInfo.ModTime}
if _, err := updateObjectMetadataWithZipInfo(ctx, objectAPI, bucket, object, opts); err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
}
if kind, encrypted := crypto.IsEncrypted(objInfo.UserDefined); encrypted {
switch kind {
case crypto.S3:
@ -3125,6 +3152,14 @@ func (api objectAPIHandlers) CompleteMultipartUploadHandler(w http.ResponseWrite
}
}
if r.Header.Get(xMinIOExtract) == "true" && strings.HasSuffix(object, archiveExt) {
opts := ObjectOptions{VersionID: objInfo.VersionID, MTime: objInfo.ModTime}
if _, err := updateObjectMetadataWithZipInfo(ctx, objectAPI, bucket, object, opts); err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
}
setPutObjHeaders(w, objInfo, false)
if replicate, sync := mustReplicate(ctx, r, bucket, object, getMustReplicateOptions(objInfo, replication.ObjectReplicationType)); replicate {
scheduleReplication(ctx, objInfo.Clone(), objectAPI, sync, replication.ObjectReplicationType)

506
cmd/s3-zip-handlers.go Normal file
View File

@ -0,0 +1,506 @@
// Copyright (c) 2015-2021 MinIO, Inc.
//
// This file is part of MinIO Object Storage stack
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
package cmd
import (
"bytes"
"context"
"errors"
"fmt"
"io"
stdioutil "io/ioutil"
"net/http"
"sort"
"strings"
"github.com/minio/minio/internal/crypto"
"github.com/minio/minio/internal/ioutil"
"github.com/minio/minio/internal/logger"
xnet "github.com/minio/minio/internal/net"
"github.com/minio/pkg/bucket/policy"
"github.com/minio/zipindex"
)
const (
archiveType = "zip"
archiveExt = "." + archiveType // ".zip"
archiveSeparator = "/"
archivePattern = archiveExt + archiveSeparator // ".zip/"
archiveTypeMetadataKey = ReservedMetadataPrefixLower + "archive-type" // "x-minio-internal-archive-type"
archiveInfoMetadataKey = ReservedMetadataPrefixLower + "archive-info" // "x-minio-internal-archive-info"
// Peek into a zip archive
xMinIOExtract = "x-minio-extract"
)
// splitZipExtensionPath splits the S3 path to the zip file and the path inside the zip:
// e.g /path/to/archive.zip/backup-2021/myimage.png => /path/to/archive.zip, backup/myimage.png
func splitZipExtensionPath(input string) (zipPath, object string, err error) {
idx := strings.Index(input, archivePattern)
if idx < 0 {
// Should never happen
return "", "", errors.New("unable to parse zip path")
}
return input[:idx+len(archivePattern)-1], input[idx+len(archivePattern):], nil
}
// getObjectInArchiveFileHandler - GET Object in the archive file
func (api objectAPIHandlers) getObjectInArchiveFileHandler(ctx context.Context, objectAPI ObjectLayer, bucket, object string, w http.ResponseWriter, r *http.Request) {
if crypto.S3.IsRequested(r.Header) || crypto.S3KMS.IsRequested(r.Header) { // If SSE-S3 or SSE-KMS present -> AWS fails with undefined error
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
}
if _, ok := crypto.IsRequested(r.Header); !objectAPI.IsEncryptionSupported() && ok {
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
}
zipPath, object, err := splitZipExtensionPath(object)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
// get gateway encryption options
opts, err := getOpts(ctx, r, bucket, zipPath)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
getObjectInfo := objectAPI.GetObjectInfo
if api.CacheAPI() != nil {
getObjectInfo = api.CacheAPI().GetObjectInfo
}
// Check for auth type to return S3 compatible error.
// type to return the correct error (NoSuchKey vs AccessDenied)
if s3Error := checkRequestAuthType(ctx, r, policy.GetObjectAction, bucket, zipPath); s3Error != ErrNone {
if getRequestAuthType(r) == authTypeAnonymous {
// As per "Permission" section in
// https://docs.aws.amazon.com/AmazonS3/latest/API/RESTObjectGET.html
// If the object you request does not exist,
// the error Amazon S3 returns depends on
// whether you also have the s3:ListBucket
// permission.
// * If you have the s3:ListBucket permission
// on the bucket, Amazon S3 will return an
// HTTP status code 404 ("no such key")
// error.
// * if you dont have the s3:ListBucket
// permission, Amazon S3 will return an HTTP
// status code 403 ("access denied") error.`
if globalPolicySys.IsAllowed(policy.Args{
Action: policy.ListBucketAction,
BucketName: bucket,
ConditionValues: getConditionValues(r, "", "", nil),
IsOwner: false,
}) {
_, err = getObjectInfo(ctx, bucket, zipPath, opts)
if toAPIError(ctx, err).Code == "NoSuchKey" {
s3Error = ErrNoSuchKey
}
}
}
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(s3Error), r.URL, guessIsBrowserReq(r))
return
}
// Validate pre-conditions if any.
opts.CheckPrecondFn = func(oi ObjectInfo) bool {
if objectAPI.IsEncryptionSupported() {
if _, err := DecryptObjectInfo(&oi, r); err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return true
}
}
return checkPreconditions(ctx, w, r, oi, opts)
}
zipObjInfo, err := getObjectInfo(ctx, bucket, zipPath, opts)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
var zipInfo []byte
if z, ok := zipObjInfo.UserDefined[archiveInfoMetadataKey]; ok {
zipInfo = []byte(z)
} else {
zipInfo, err = updateObjectMetadataWithZipInfo(ctx, objectAPI, bucket, zipPath, opts)
}
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
file, err := zipindex.FindSerialized(zipInfo, object)
if err != nil {
if err == io.EOF {
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrNoSuchKey), r.URL, guessIsBrowserReq(r))
} else {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
}
return
}
// New object info
fileObjInfo := ObjectInfo{
Bucket: bucket,
Name: object,
Size: int64(file.UncompressedSize64),
ModTime: zipObjInfo.ModTime,
}
var rc io.ReadCloser
if file.UncompressedSize64 > 0 {
rs := &HTTPRangeSpec{Start: file.Offset, End: file.Offset + int64(file.UncompressedSize64) - 1}
gr, err := objectAPI.GetObjectNInfo(ctx, bucket, zipPath, rs, nil, readLock, opts)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
defer gr.Close()
rc, err = file.Open(gr)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
} else {
rc = stdioutil.NopCloser(bytes.NewReader([]byte{}))
}
defer rc.Close()
if err = setObjectHeaders(w, fileObjInfo, nil, opts); err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
setHeadGetRespHeaders(w, r.URL.Query())
httpWriter := ioutil.WriteOnClose(w)
// Write object content to response body
if _, err = io.Copy(httpWriter, rc); err != nil {
if !httpWriter.HasWritten() {
// write error response only if no data or headers has been written to client yet
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
if !xnet.IsNetworkOrHostDown(err, true) { // do not need to log disconnected clients
logger.LogIf(ctx, fmt.Errorf("Unable to write all the data to client %w", err))
}
return
}
if err = httpWriter.Close(); err != nil {
if !httpWriter.HasWritten() { // write error response only if no data or headers has been written to client yet
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
if !xnet.IsNetworkOrHostDown(err, true) { // do not need to log disconnected clients
logger.LogIf(ctx, fmt.Errorf("Unable to write all the data to client %w", err))
}
return
}
}
// listObjectsV2InArchive generates S3 listing result ListObjectsV2Info from zip file, all parameters are already validated by the caller.
func listObjectsV2InArchive(ctx context.Context, objectAPI ObjectLayer, bucket, prefix, token, delimiter string, maxKeys int, fetchOwner bool, startAfter string) (ListObjectsV2Info, error) {
zipPath, _, err := splitZipExtensionPath(prefix)
if err != nil {
// Return empty listing
return ListObjectsV2Info{}, nil
}
zipObjInfo, err := objectAPI.GetObjectInfo(ctx, bucket, zipPath, ObjectOptions{})
if err != nil {
// Return empty listing
return ListObjectsV2Info{}, nil
}
var zipInfo []byte
if z, ok := zipObjInfo.UserDefined[archiveInfoMetadataKey]; ok {
zipInfo = []byte(z)
} else {
// Always update the latest version
zipInfo, err = updateObjectMetadataWithZipInfo(ctx, objectAPI, bucket, zipPath, ObjectOptions{})
}
if err != nil {
return ListObjectsV2Info{}, err
}
files, err := zipindex.DeserializeFiles(zipInfo)
if err != nil {
return ListObjectsV2Info{}, err
}
sort.Slice(files, func(i, j int) bool {
return files[i].Name < files[j].Name
})
var (
count int
isTruncated bool
nextToken string
listObjectsInfo ListObjectsV2Info
)
// Always set this
listObjectsInfo.ContinuationToken = token
// Open and iterate through the files in the archive.
for _, file := range files {
objName := zipObjInfo.Name + archiveSeparator + file.Name
if objName <= startAfter || objName <= token {
continue
}
if strings.HasPrefix(objName, prefix) {
if count == maxKeys {
isTruncated = true
break
}
if delimiter != "" {
i := strings.Index(objName[len(prefix):], delimiter)
if i >= 0 {
commonPrefix := objName[:len(prefix)+i+1]
if len(listObjectsInfo.Prefixes) == 0 || commonPrefix != listObjectsInfo.Prefixes[len(listObjectsInfo.Prefixes)-1] {
listObjectsInfo.Prefixes = append(listObjectsInfo.Prefixes, commonPrefix)
count++
}
goto next
}
}
listObjectsInfo.Objects = append(listObjectsInfo.Objects, ObjectInfo{
Bucket: bucket,
Name: objName,
Size: int64(file.UncompressedSize64),
ModTime: zipObjInfo.ModTime,
})
count++
}
next:
nextToken = objName
}
if isTruncated {
listObjectsInfo.IsTruncated = true
listObjectsInfo.NextContinuationToken = nextToken
}
return listObjectsInfo, nil
}
// getFilesFromZIPObject reads a partial stream of a zip file to build the zipindex.Files index
func getFilesListFromZIPObject(ctx context.Context, objectAPI ObjectLayer, bucket, object string, opts ObjectOptions) (zipindex.Files, ObjectInfo, error) {
var size = 1 << 20
var objSize int64
for {
rs := &HTTPRangeSpec{IsSuffixLength: true, Start: int64(-size)}
gr, err := objectAPI.GetObjectNInfo(ctx, bucket, object, rs, nil, readLock, opts)
if err != nil {
return nil, ObjectInfo{}, err
}
b, err := stdioutil.ReadAll(gr)
if err != nil {
gr.Close()
return nil, ObjectInfo{}, err
}
gr.Close()
if size > len(b) {
size = len(b)
}
// Calculate the object real size if encrypted
if _, ok := crypto.IsEncrypted(gr.ObjInfo.UserDefined); ok {
objSize, err = gr.ObjInfo.DecryptedSize()
if err != nil {
return nil, ObjectInfo{}, err
}
} else {
objSize = gr.ObjInfo.Size
}
files, err := zipindex.ReadDir(b[len(b)-size:], objSize, nil)
if err == nil {
return files, gr.ObjInfo, nil
}
var terr zipindex.ErrNeedMoreData
if errors.As(err, &terr) {
size = int(terr.FromEnd)
if size <= 0 || size > 100<<20 {
return nil, ObjectInfo{}, errors.New("zip directory too large")
}
} else {
return nil, ObjectInfo{}, err
}
}
}
// headObjectInArchiveFileHandler - HEAD Object in an archive file
func (api objectAPIHandlers) headObjectInArchiveFileHandler(ctx context.Context, objectAPI ObjectLayer, bucket, object string, w http.ResponseWriter, r *http.Request) {
if crypto.S3.IsRequested(r.Header) || crypto.S3KMS.IsRequested(r.Header) { // If SSE-S3 or SSE-KMS present -> AWS fails with undefined error
writeErrorResponseHeadersOnly(w, errorCodes.ToAPIErr(ErrBadRequest))
return
}
if _, ok := crypto.IsRequested(r.Header); !objectAPI.IsEncryptionSupported() && ok {
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrBadRequest), r.URL, guessIsBrowserReq(r))
return
}
zipPath, object, err := splitZipExtensionPath(object)
if err != nil {
writeErrorResponseHeadersOnly(w, toAPIError(ctx, err))
return
}
getObjectInfo := objectAPI.GetObjectInfo
if api.CacheAPI() != nil {
getObjectInfo = api.CacheAPI().GetObjectInfo
}
opts, err := getOpts(ctx, r, bucket, zipPath)
if err != nil {
writeErrorResponseHeadersOnly(w, toAPIError(ctx, err))
return
}
if s3Error := checkRequestAuthType(ctx, r, policy.GetObjectAction, bucket, zipPath); s3Error != ErrNone {
if getRequestAuthType(r) == authTypeAnonymous {
// As per "Permission" section in
// https://docs.aws.amazon.com/AmazonS3/latest/API/RESTObjectHEAD.html
// If the object you request does not exist,
// the error Amazon S3 returns depends on
// whether you also have the s3:ListBucket
// permission.
// * If you have the s3:ListBucket permission
// on the bucket, Amazon S3 will return an
// HTTP status code 404 ("no such key")
// error.
// * if you dont have the s3:ListBucket
// permission, Amazon S3 will return an HTTP
// status code 403 ("access denied") error.`
if globalPolicySys.IsAllowed(policy.Args{
Action: policy.ListBucketAction,
BucketName: bucket,
ConditionValues: getConditionValues(r, "", "", nil),
IsOwner: false,
}) {
_, err = getObjectInfo(ctx, bucket, zipPath, opts)
if toAPIError(ctx, err).Code == "NoSuchKey" {
s3Error = ErrNoSuchKey
}
}
}
writeErrorResponseHeadersOnly(w, errorCodes.ToAPIErr(s3Error))
return
}
var rs *HTTPRangeSpec
// Validate pre-conditions if any.
opts.CheckPrecondFn = func(oi ObjectInfo) bool {
return checkPreconditions(ctx, w, r, oi, opts)
}
zipObjInfo, err := getObjectInfo(ctx, bucket, zipPath, opts)
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
var zipInfo []byte
if z, ok := zipObjInfo.UserDefined[archiveInfoMetadataKey]; ok {
zipInfo = []byte(z)
} else {
zipInfo, err = updateObjectMetadataWithZipInfo(ctx, objectAPI, bucket, zipPath, opts)
}
if err != nil {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
return
}
file, err := zipindex.FindSerialized(zipInfo, object)
if err != nil {
if err == io.EOF {
writeErrorResponse(ctx, w, errorCodes.ToAPIErr(ErrNoSuchKey), r.URL, guessIsBrowserReq(r))
} else {
writeErrorResponse(ctx, w, toAPIError(ctx, err), r.URL, guessIsBrowserReq(r))
}
return
}
objInfo := ObjectInfo{
Bucket: bucket,
Name: file.Name,
Size: int64(file.UncompressedSize64),
ModTime: zipObjInfo.ModTime,
}
// Set standard object headers.
if err = setObjectHeaders(w, objInfo, nil, opts); err != nil {
writeErrorResponseHeadersOnly(w, toAPIError(ctx, err))
return
}
// Set any additional requested response headers.
setHeadGetRespHeaders(w, r.URL.Query())
// Successful response.
if rs != nil {
w.WriteHeader(http.StatusPartialContent)
} else {
w.WriteHeader(http.StatusOK)
}
}
// Update the passed zip object metadata with the zip contents info, file name, modtime, size, etc..
func updateObjectMetadataWithZipInfo(ctx context.Context, objectAPI ObjectLayer, bucket, object string, opts ObjectOptions) ([]byte, error) {
files, srcInfo, err := getFilesListFromZIPObject(ctx, objectAPI, bucket, object, opts)
if err != nil {
return nil, err
}
files.OptimizeSize()
zipInfo, err := files.Serialize()
if err != nil {
return nil, err
}
srcInfo.UserDefined[archiveTypeMetadataKey] = archiveType
srcInfo.UserDefined[archiveInfoMetadataKey] = string(zipInfo)
srcInfo.metadataOnly = true
// Always update the same version id & modtime
// Passing opts twice as source & destination options will update the metadata
// of the same object version to avoid creating a new version.
_, err = objectAPI.CopyObject(ctx, bucket, object, bucket, object, srcInfo, opts, opts)
if err != nil {
return nil, err
}
return zipInfo, nil
}

View File

@ -0,0 +1,40 @@
# Perform S3 operations in a ZIP content[![Slack](https://slack.min.io/slack?type=svg)](https://slack.min.io) [![Docker Pulls](https://img.shields.io/docker/pulls/minio/minio.svg?maxAge=604800)](https://hub.docker.com/r/minio/minio/)
### Overview
MinIO implements an S3 extension to list, stat and download files inside a ZIP file stored in any bucket. A perfect use case scenario is when you have a lot of small files archived in multiple ZIP files. Uploading them is faster than uploading small files individually. Besides, your S3 applications will be able to access to the data with little performance overhead.
The main limitation is that to update or delete content of a file inside a ZIP file the entire ZIP file must be replaced.
### How to enable S3 ZIP behavior ?
Ensure to set the following header `x-minio-extract` to `true` in your S3 requests.
### How to access to files inside a ZIP archive
Accessing to contents inside an archive can be done using regular S3 API with a modified request path. You just need to append the path of the content inside the archive to the path of the archive itself.
e.g.:
To download `2021/taxes.csv` archived in `financial.zip` and stored under a bucket named `company-data`, you can issue a GET request using the following path 'company-data/financial.zip/2021/taxes.csv`
### Contents properties
All properties except the file size are tied to the zip file. This means that modification date, headers, tags, etc. can only be set for the zip file as a whole. In similar fashion, replication will replicate the zip file as a whole and not individual files.
### Code Examples
[Using minio-go library](https://github.com/minio/minio/blob/master/docs/extensions/s3zip/examples/minio-go/main.go)
[Using AWS JS SDK v2](https://github.com/minio/minio/blob/master/docs/extensions/s3zip/examples/aws-js/main.js)
[Using boto3](https://github.com/minio/minio/blob/master/docs/extensions/s3zip/examples/boto3/main.py)
### Requirements and limits
- ListObjectsV2 can only list the most recent ZIP archive version of your object, applicable only for versioned buckets.
- ListObjectsV2 API calls must be used to list zip file content.
- Names inside ZIP files are kept unmodified, but some may lead to invalid paths. See [Object key naming guidelines](https://docs.aws.amazon.com/AmazonS3/latest/userguide/object-keys.html) on safe names.
- This API behavior is limited for following **read** operations on files inside a zip archive:
- `HeadObject`
- `GetObject`
- `ListObjectsV2`
- A maximum of 100,000 files inside a single ZIP archive is recommended for best performance and memory usage trade-off.
- If the ZIP file directory isn't located within the last 100MB the file will not be parsed.

View File

@ -0,0 +1,31 @@
var AWS = require('aws-sdk');
var s3 = new AWS.S3({
accessKeyId: 'YOUR-ACCESSKEYID' ,
secretAccessKey: 'YOUR-SECRETACCESSKEY' ,
endpoint: 'http://127.0.0.1:9000' ,
s3ForcePathStyle: true,
signatureVersion: 'v4'
});
// List all contents stored in the zip archive
s3.listObjectsV2({Bucket : 'your-bucket', Prefix: 'path/to/file.zip/'}).
on('build', function(req) { req.httpRequest.headers['X-Minio-Extract'] = 'true'; }).
send(function(err, data) {
if (err) {
console.log("Error", err);
} else {
console.log("Success", data);
}
});
// Download a file in the archive and store it in /tmp/data.csv
var file = require('fs').createWriteStream('/tmp/data.csv');
s3.getObject({Bucket: 'your-bucket', Key: 'path/to/file.zip/data.csv'}).
on('build', function(req) { req.httpRequest.headers['X-Minio-Extract'] = 'true'; }).
on('httpData', function(chunk) { file.write(chunk); }).
on('httpDone', function() { file.end(); }).
send();

View File

@ -0,0 +1,8 @@
{
"name": "s3-zip-example",
"version": "1.0.0",
"main": "main.js",
"dependencies": {
"aws-sdk": "^2.924.0"
}
}

View File

@ -0,0 +1,25 @@
#!/usr/bin/env/python
import boto3
from botocore.client import Config
s3 = boto3.client('s3',
endpoint_url='http://localhost:9000',
aws_access_key_id='YOUR-ACCESSKEYID',
aws_secret_access_key='YOUR-SECRETACCESSKEY',
config=Config(signature_version='s3v4'),
region_name='us-east-1')
def _add_header(request, **kwargs):
request.headers.add_header('x-minio-extract', 'true')
event_system = s3.meta.events
event_system.register_first('before-sign.s3.*', _add_header)
# List zip contents
response = s3.list_objects_v2(Bucket="your-bucket", Prefix="path/to/file.zip/")
print(response)
# Downlaod data.csv stored in the zip file
s3.download_file(Bucket='your-bucket', Key='path/to/file.zip/data.csv', Filename='/tmp/data.csv')

View File

@ -0,0 +1,45 @@
package main
import (
"context"
"io"
"log"
"net/http"
"os"
"github.com/minio/minio-go/v7"
"github.com/minio/minio-go/v7/pkg/credentials"
)
type s3ExtensionTransport struct {
tr http.RoundTripper
}
func (t *s3ExtensionTransport) RoundTrip(req *http.Request) (*http.Response, error) {
req.Header.Add("x-minio-extract", "true")
return t.tr.RoundTrip(req)
}
func main() {
tr, _ := minio.DefaultTransport(false)
s3Client, err := minio.New("minio-server-address:9000", &minio.Options{
Creds: credentials.NewStaticV4("access-key", "secret-key", ""),
Transport: &s3ExtensionTransport{tr},
})
if err != nil {
log.Fatalln(err)
}
// Download API.md from the archive
rd, err := s3Client.GetObject(context.Background(), "your-bucket", "path/to/file.zip/data.csv", minio.GetObjectOptions{})
if err != nil {
log.Fatalln(err)
}
_, err = io.Copy(os.Stdout, rd)
if err != nil {
log.Fatalln(err)
}
return
}

1
go.mod
View File

@ -53,6 +53,7 @@ require (
github.com/minio/sha256-simd v1.0.0
github.com/minio/simdjson-go v0.2.1
github.com/minio/sio v0.3.0
github.com/minio/zipindex v0.2.0
github.com/mitchellh/go-homedir v1.1.0
github.com/montanaflynn/stats v0.5.0
github.com/nats-io/nats-server/v2 v2.1.9

4
go.sum
View File

@ -506,6 +506,8 @@ github.com/minio/simdjson-go v0.2.1/go.mod h1:JPUSkRykfSPS+AhO0YPA1h0l5vY7NqrF4z
github.com/minio/sio v0.2.1/go.mod h1:8b0yPp2avGThviy/+OCJBI6OMpvxoUuiLvE6F1lebhw=
github.com/minio/sio v0.3.0 h1:syEFBewzOMOYVzSTFpp1MqpSZk8rUNbz8VIIc+PNzus=
github.com/minio/sio v0.3.0/go.mod h1:8b0yPp2avGThviy/+OCJBI6OMpvxoUuiLvE6F1lebhw=
github.com/minio/zipindex v0.2.0 h1:iqgIhPkYnZ6fNd6q8trtJc8+mtKJNFTaxXlNw4n0We8=
github.com/minio/zipindex v0.2.0/go.mod h1:s+b/Qyw9JtSEnYfaM4ASOWNO2xGnXCfzQ+SWAzVkVZc=
github.com/mitchellh/cli v1.0.0/go.mod h1:hNIlj7HEI86fIcpObd7a0FcrxTWetlwJDGcceTlRvqc=
github.com/mitchellh/copystructure v1.0.0/go.mod h1:SNtv71yrdKgLRyLFxmLdkAbkKEFWgYaq1OVrnRcwhnw=
github.com/mitchellh/go-homedir v1.0.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrkLzIz1N1q0pr0=
@ -696,6 +698,7 @@ github.com/tidwall/pretty v1.1.0/go.mod h1:XNkn88O1ChpSDQmQeStsy+sBenx6DDtFZJxhV
github.com/tidwall/sjson v1.1.6 h1:8fDdlahON04OZBlTQCIatW8FstSFJz8oxidj5h0rmSQ=
github.com/tidwall/sjson v1.1.6/go.mod h1:KN3FZ7odvXIHPbJdhNorK/M9lWweVUbXsXXhrJ/kGOA=
github.com/tinylib/msgp v1.1.3/go.mod h1:+d+yLhGm8mzTaHzB+wgMYrodPfmZrzkirds8fDWklFE=
github.com/tinylib/msgp v1.1.5/go.mod h1:eQsjooMTnV42mHu917E26IogZ2930nFyBQdofk10Udg=
github.com/tinylib/msgp v1.1.6-0.20210521143832-0becd170c402 h1:x5VlSgDgIGXNegkO4gigpYmb/RFkKGgy12Kkrbif7XE=
github.com/tinylib/msgp v1.1.6-0.20210521143832-0becd170c402/go.mod h1:75BAfg2hauQhs3qedfdDZmWAPcFMAvJE5b9rGOMufyw=
github.com/tklauser/go-sysconf v0.3.4/go.mod h1:Cl2c8ZRWfHD5IrfHo9VN+FX9kCFjIOyVklgXycLB6ek=
@ -705,6 +708,7 @@ github.com/tklauser/numcpus v0.2.1/go.mod h1:9aU+wOc6WjUIZEwWMP62PL/41d65P+iks1g
github.com/tklauser/numcpus v0.2.2 h1:oyhllyrScuYI6g+h/zUvNXNp1wy7x8qQy3t/piefldA=
github.com/tklauser/numcpus v0.2.2/go.mod h1:x3qojaO3uyYt0i56EW/VUYs7uBvdl2fkfZFu0T9wgjM=
github.com/tmc/grpc-websocket-proxy v0.0.0-20170815181823-89b8d40f7ca8/go.mod h1:ncp9v5uamzpCO7NfCPTXjqaC+bZgJeR0sMTm6dMHP7U=
github.com/ttacon/chalk v0.0.0-20160626202418-22c06c80ed31/go.mod h1:onvgF043R+lC5RZ8IT9rBXDaEDnpnw/Cl+HFiw+v/7Q=
github.com/tv42/httpunix v0.0.0-20150427012821-b75d8614f926/go.mod h1:9ESjWnEqriFuLhtthL60Sar/7RFoluCcXsuvEwTV5KM=
github.com/urfave/cli v1.20.0/go.mod h1:70zkFmudgCuE/ngEzBv17Jvp/497gISqfk5gWijbERA=
github.com/urfave/cli v1.22.1/go.mod h1:Gos4lmkARVdJ6EkW0WaNv/tZAAMe9V7XWyB60NtXRu0=