1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
|
// Copyright (C) MongoDB, Inc. 2017-present.
//
// Licensed under the Apache License, Version 2.0 (the "License"); you may
// not use this file except in compliance with the License. You may obtain
// a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
package gridfs // import "go.mongodb.org/mongo-driver/mongo/gridfs"
import (
"bytes"
"context"
"errors"
"fmt"
"io"
"time"
"go.mongodb.org/mongo-driver/bson"
"go.mongodb.org/mongo-driver/bson/primitive"
"go.mongodb.org/mongo-driver/mongo"
"go.mongodb.org/mongo-driver/mongo/options"
"go.mongodb.org/mongo-driver/mongo/readconcern"
"go.mongodb.org/mongo-driver/mongo/readpref"
"go.mongodb.org/mongo-driver/mongo/writeconcern"
"go.mongodb.org/mongo-driver/x/bsonx"
"go.mongodb.org/mongo-driver/x/bsonx/bsoncore"
)
// TODO: add sessions options
// DefaultChunkSize is the default size of each file chunk.
const DefaultChunkSize int32 = 255 * 1024 // 255 KiB
// ErrFileNotFound occurs if a user asks to download a file with a file ID that isn't found in the files collection.
var ErrFileNotFound = errors.New("file with given parameters not found")
// ErrMissingChunkSize occurs when downloading a file if the files collection document is missing the "chunkSize" field.
var ErrMissingChunkSize = errors.New("files collection document does not contain a 'chunkSize' field")
// Bucket represents a GridFS bucket.
type Bucket struct {
db *mongo.Database
chunksColl *mongo.Collection // collection to store file chunks
filesColl *mongo.Collection // collection to store file metadata
name string
chunkSize int32
wc *writeconcern.WriteConcern
rc *readconcern.ReadConcern
rp *readpref.ReadPref
firstWriteDone bool
readBuf []byte
writeBuf []byte
readDeadline time.Time
writeDeadline time.Time
}
// Upload contains options to upload a file to a bucket.
type Upload struct {
chunkSize int32
metadata bsonx.Doc
}
// NewBucket creates a GridFS bucket.
func NewBucket(db *mongo.Database, opts ...*options.BucketOptions) (*Bucket, error) {
b := &Bucket{
name: "fs",
chunkSize: DefaultChunkSize,
db: db,
wc: db.WriteConcern(),
rc: db.ReadConcern(),
rp: db.ReadPreference(),
}
bo := options.MergeBucketOptions(opts...)
if bo.Name != nil {
b.name = *bo.Name
}
if bo.ChunkSizeBytes != nil {
b.chunkSize = *bo.ChunkSizeBytes
}
if bo.WriteConcern != nil {
b.wc = bo.WriteConcern
}
if bo.ReadConcern != nil {
b.rc = bo.ReadConcern
}
if bo.ReadPreference != nil {
b.rp = bo.ReadPreference
}
var collOpts = options.Collection().SetWriteConcern(b.wc).SetReadConcern(b.rc).SetReadPreference(b.rp)
b.chunksColl = db.Collection(b.name+".chunks", collOpts)
b.filesColl = db.Collection(b.name+".files", collOpts)
b.readBuf = make([]byte, b.chunkSize)
b.writeBuf = make([]byte, b.chunkSize)
return b, nil
}
// SetWriteDeadline sets the write deadline for this bucket.
func (b *Bucket) SetWriteDeadline(t time.Time) error {
b.writeDeadline = t
return nil
}
// SetReadDeadline sets the read deadline for this bucket
func (b *Bucket) SetReadDeadline(t time.Time) error {
b.readDeadline = t
return nil
}
// OpenUploadStream creates a file ID new upload stream for a file given the filename.
func (b *Bucket) OpenUploadStream(filename string, opts ...*options.UploadOptions) (*UploadStream, error) {
return b.OpenUploadStreamWithID(primitive.NewObjectID(), filename, opts...)
}
// OpenUploadStreamWithID creates a new upload stream for a file given the file ID and filename.
func (b *Bucket) OpenUploadStreamWithID(fileID interface{}, filename string, opts ...*options.UploadOptions) (*UploadStream, error) {
ctx, cancel := deadlineContext(b.writeDeadline)
if cancel != nil {
defer cancel()
}
if err := b.checkFirstWrite(ctx); err != nil {
return nil, err
}
upload, err := b.parseUploadOptions(opts...)
if err != nil {
return nil, err
}
return newUploadStream(upload, fileID, filename, b.chunksColl, b.filesColl), nil
}
// UploadFromStream creates a fileID and uploads a file given a source stream.
//
// If this upload requires a custom write deadline to be set on the bucket, it cannot be done concurrently with other
// write operations operations on this bucket that also require a custom deadline.
func (b *Bucket) UploadFromStream(filename string, source io.Reader, opts ...*options.UploadOptions) (primitive.ObjectID, error) {
fileID := primitive.NewObjectID()
err := b.UploadFromStreamWithID(fileID, filename, source, opts...)
return fileID, err
}
// UploadFromStreamWithID uploads a file given a source stream.
//
// If this upload requires a custom write deadline to be set on the bucket, it cannot be done concurrently with other
// write operations operations on this bucket that also require a custom deadline.
func (b *Bucket) UploadFromStreamWithID(fileID interface{}, filename string, source io.Reader, opts ...*options.UploadOptions) error {
us, err := b.OpenUploadStreamWithID(fileID, filename, opts...)
if err != nil {
return err
}
err = us.SetWriteDeadline(b.writeDeadline)
if err != nil {
_ = us.Close()
return err
}
for {
n, err := source.Read(b.readBuf)
if err != nil && err != io.EOF {
_ = us.Abort() // upload considered aborted if source stream returns an error
return err
}
if n > 0 {
_, err := us.Write(b.readBuf[:n])
if err != nil {
return err
}
}
if n == 0 || err == io.EOF {
break
}
}
return us.Close()
}
// OpenDownloadStream creates a stream from which the contents of the file can be read.
func (b *Bucket) OpenDownloadStream(fileID interface{}) (*DownloadStream, error) {
id, err := convertFileID(fileID)
if err != nil {
return nil, err
}
return b.openDownloadStream(bsonx.Doc{
{"_id", id},
})
}
// DownloadToStream downloads the file with the specified fileID and writes it to the provided io.Writer.
// Returns the number of bytes written to the steam and an error, or nil if there was no error.
//
// If this download requires a custom read deadline to be set on the bucket, it cannot be done concurrently with other
// read operations operations on this bucket that also require a custom deadline.
func (b *Bucket) DownloadToStream(fileID interface{}, stream io.Writer) (int64, error) {
ds, err := b.OpenDownloadStream(fileID)
if err != nil {
return 0, err
}
return b.downloadToStream(ds, stream)
}
// OpenDownloadStreamByName opens a download stream for the file with the given filename.
func (b *Bucket) OpenDownloadStreamByName(filename string, opts ...*options.NameOptions) (*DownloadStream, error) {
var numSkip int32 = -1
var sortOrder int32 = 1
nameOpts := options.MergeNameOptions(opts...)
if nameOpts.Revision != nil {
numSkip = *nameOpts.Revision
}
if numSkip < 0 {
sortOrder = -1
numSkip = (-1 * numSkip) - 1
}
findOpts := options.Find().SetSkip(int64(numSkip)).SetSort(bsonx.Doc{{"uploadDate", bsonx.Int32(sortOrder)}})
return b.openDownloadStream(bsonx.Doc{{"filename", bsonx.String(filename)}}, findOpts)
}
// DownloadToStreamByName downloads the file with the given name to the given io.Writer.
//
// If this download requires a custom read deadline to be set on the bucket, it cannot be done concurrently with other
// read operations operations on this bucket that also require a custom deadline.
func (b *Bucket) DownloadToStreamByName(filename string, stream io.Writer, opts ...*options.NameOptions) (int64, error) {
ds, err := b.OpenDownloadStreamByName(filename, opts...)
if err != nil {
return 0, err
}
return b.downloadToStream(ds, stream)
}
// Delete deletes all chunks and metadata associated with the file with the given file ID.
//
// If this operation requires a custom write deadline to be set on the bucket, it cannot be done concurrently with other
// write operations operations on this bucket that also require a custom deadline.
func (b *Bucket) Delete(fileID interface{}) error {
// delete document in files collection and then chunks to minimize race conditions
ctx, cancel := deadlineContext(b.writeDeadline)
if cancel != nil {
defer cancel()
}
id, err := convertFileID(fileID)
if err != nil {
return err
}
res, err := b.filesColl.DeleteOne(ctx, bsonx.Doc{{"_id", id}})
if err == nil && res.DeletedCount == 0 {
err = ErrFileNotFound
}
if err != nil {
_ = b.deleteChunks(ctx, fileID) // can attempt to delete chunks even if no docs in files collection matched
return err
}
return b.deleteChunks(ctx, fileID)
}
// Find returns the files collection documents that match the given filter.
//
// If this download requires a custom read deadline to be set on the bucket, it cannot be done concurrently with other
// read operations operations on this bucket that also require a custom deadline.
func (b *Bucket) Find(filter interface{}, opts ...*options.GridFSFindOptions) (*mongo.Cursor, error) {
ctx, cancel := deadlineContext(b.readDeadline)
if cancel != nil {
defer cancel()
}
gfsOpts := options.MergeGridFSFindOptions(opts...)
find := options.Find()
if gfsOpts.AllowDiskUse != nil {
find.SetAllowDiskUse(*gfsOpts.AllowDiskUse)
}
if gfsOpts.BatchSize != nil {
find.SetBatchSize(*gfsOpts.BatchSize)
}
if gfsOpts.Limit != nil {
find.SetLimit(int64(*gfsOpts.Limit))
}
if gfsOpts.MaxTime != nil {
find.SetMaxTime(*gfsOpts.MaxTime)
}
if gfsOpts.NoCursorTimeout != nil {
find.SetNoCursorTimeout(*gfsOpts.NoCursorTimeout)
}
if gfsOpts.Skip != nil {
find.SetSkip(int64(*gfsOpts.Skip))
}
if gfsOpts.Sort != nil {
find.SetSort(gfsOpts.Sort)
}
return b.filesColl.Find(ctx, filter, find)
}
// Rename renames the stored file with the specified file ID.
//
// If this operation requires a custom write deadline to be set on the bucket, it cannot be done concurrently with other
// write operations operations on this bucket that also require a custom deadline
func (b *Bucket) Rename(fileID interface{}, newFilename string) error {
ctx, cancel := deadlineContext(b.writeDeadline)
if cancel != nil {
defer cancel()
}
id, err := convertFileID(fileID)
if err != nil {
return err
}
res, err := b.filesColl.UpdateOne(ctx,
bsonx.Doc{{"_id", id}},
bsonx.Doc{{"$set", bsonx.Document(bsonx.Doc{{"filename", bsonx.String(newFilename)}})}},
)
if err != nil {
return err
}
if res.MatchedCount == 0 {
return ErrFileNotFound
}
return nil
}
// Drop drops the files and chunks collections associated with this bucket.
//
// If this operation requires a custom write deadline to be set on the bucket, it cannot be done concurrently with other
// write operations operations on this bucket that also require a custom deadline
func (b *Bucket) Drop() error {
ctx, cancel := deadlineContext(b.writeDeadline)
if cancel != nil {
defer cancel()
}
err := b.filesColl.Drop(ctx)
if err != nil {
return err
}
return b.chunksColl.Drop(ctx)
}
// GetFilesCollection returns a handle to the collection that stores the file documents for this bucket.
func (b *Bucket) GetFilesCollection() *mongo.Collection {
return b.filesColl
}
// GetChunksCollection returns a handle to the collection that stores the file chunks for this bucket.
func (b *Bucket) GetChunksCollection() *mongo.Collection {
return b.chunksColl
}
func (b *Bucket) openDownloadStream(filter interface{}, opts ...*options.FindOptions) (*DownloadStream, error) {
ctx, cancel := deadlineContext(b.readDeadline)
if cancel != nil {
defer cancel()
}
cursor, err := b.findFile(ctx, filter, opts...)
if err != nil {
return nil, err
}
// Unmarshal the data into a File instance, which can be passed to newDownloadStream. The _id value has to be
// parsed out separately because "_id" will not match the File.ID field and we want to avoid exposing BSON tags
// in the File type. After parsing it, use RawValue.Unmarshal to ensure File.ID is set to the appropriate value.
var foundFile File
if err = cursor.Decode(&foundFile); err != nil {
return nil, fmt.Errorf("error decoding files collection document: %v", err)
}
if foundFile.Length == 0 {
return newDownloadStream(nil, foundFile.ChunkSize, &foundFile), nil
}
// For a file with non-zero length, chunkSize must exist so we know what size to expect when downloading chunks.
if _, err := cursor.Current.LookupErr("chunkSize"); err != nil {
return nil, ErrMissingChunkSize
}
chunksCursor, err := b.findChunks(ctx, foundFile.ID)
if err != nil {
return nil, err
}
// The chunk size can be overridden for individual files, so the expected chunk size should be the "chunkSize"
// field from the files collection document, not the bucket's chunk size.
return newDownloadStream(chunksCursor, foundFile.ChunkSize, &foundFile), nil
}
func deadlineContext(deadline time.Time) (context.Context, context.CancelFunc) {
if deadline.Equal(time.Time{}) {
return context.Background(), nil
}
return context.WithDeadline(context.Background(), deadline)
}
func (b *Bucket) downloadToStream(ds *DownloadStream, stream io.Writer) (int64, error) {
err := ds.SetReadDeadline(b.readDeadline)
if err != nil {
_ = ds.Close()
return 0, err
}
copied, err := io.Copy(stream, ds)
if err != nil {
_ = ds.Close()
return 0, err
}
return copied, ds.Close()
}
func (b *Bucket) deleteChunks(ctx context.Context, fileID interface{}) error {
id, err := convertFileID(fileID)
if err != nil {
return err
}
_, err = b.chunksColl.DeleteMany(ctx, bsonx.Doc{{"files_id", id}})
return err
}
func (b *Bucket) findFile(ctx context.Context, filter interface{}, opts ...*options.FindOptions) (*mongo.Cursor, error) {
cursor, err := b.filesColl.Find(ctx, filter, opts...)
if err != nil {
return nil, err
}
if !cursor.Next(ctx) {
_ = cursor.Close(ctx)
return nil, ErrFileNotFound
}
return cursor, nil
}
func (b *Bucket) findChunks(ctx context.Context, fileID interface{}) (*mongo.Cursor, error) {
id, err := convertFileID(fileID)
if err != nil {
return nil, err
}
chunksCursor, err := b.chunksColl.Find(ctx,
bsonx.Doc{{"files_id", id}},
options.Find().SetSort(bsonx.Doc{{"n", bsonx.Int32(1)}})) // sort by chunk index
if err != nil {
return nil, err
}
return chunksCursor, nil
}
// returns true if the 2 index documents are equal
func numericalIndexDocsEqual(expected, actual bsoncore.Document) (bool, error) {
if bytes.Equal(expected, actual) {
return true, nil
}
actualElems, err := actual.Elements()
if err != nil {
return false, err
}
expectedElems, err := expected.Elements()
if err != nil {
return false, err
}
if len(actualElems) != len(expectedElems) {
return false, nil
}
for idx, expectedElem := range expectedElems {
actualElem := actualElems[idx]
if actualElem.Key() != expectedElem.Key() {
return false, nil
}
actualVal := actualElem.Value()
expectedVal := expectedElem.Value()
actualInt, actualOK := actualVal.AsInt64OK()
expectedInt, expectedOK := expectedVal.AsInt64OK()
//GridFS indexes always have numeric values
if !actualOK || !expectedOK {
return false, nil
}
if actualInt != expectedInt {
return false, nil
}
}
return true, nil
}
// Create an index if it doesn't already exist
func createNumericalIndexIfNotExists(ctx context.Context, iv mongo.IndexView, model mongo.IndexModel) error {
c, err := iv.List(ctx)
if err != nil {
return err
}
defer func() {
_ = c.Close(ctx)
}()
modelKeysBytes, err := bson.Marshal(model.Keys)
if err != nil {
return err
}
modelKeysDoc := bsoncore.Document(modelKeysBytes)
for c.Next(ctx) {
keyElem, err := c.Current.LookupErr("key")
if err != nil {
return err
}
keyElemDoc := keyElem.Document()
found, err := numericalIndexDocsEqual(modelKeysDoc, bsoncore.Document(keyElemDoc))
if err != nil {
return err
}
if found {
return nil
}
}
_, err = iv.CreateOne(ctx, model)
return err
}
// create indexes on the files and chunks collection if needed
func (b *Bucket) createIndexes(ctx context.Context) error {
// must use primary read pref mode to check if files coll empty
cloned, err := b.filesColl.Clone(options.Collection().SetReadPreference(readpref.Primary()))
if err != nil {
return err
}
docRes := cloned.FindOne(ctx, bsonx.Doc{}, options.FindOne().SetProjection(bsonx.Doc{{"_id", bsonx.Int32(1)}}))
_, err = docRes.DecodeBytes()
if err != mongo.ErrNoDocuments {
// nil, or error that occured during the FindOne operation
return err
}
filesIv := b.filesColl.Indexes()
chunksIv := b.chunksColl.Indexes()
filesModel := mongo.IndexModel{
Keys: bson.D{
{"filename", int32(1)},
{"uploadDate", int32(1)},
},
}
chunksModel := mongo.IndexModel{
Keys: bson.D{
{"files_id", int32(1)},
{"n", int32(1)},
},
Options: options.Index().SetUnique(true),
}
if err = createNumericalIndexIfNotExists(ctx, filesIv, filesModel); err != nil {
return err
}
if err = createNumericalIndexIfNotExists(ctx, chunksIv, chunksModel); err != nil {
return err
}
return nil
}
func (b *Bucket) checkFirstWrite(ctx context.Context) error {
if !b.firstWriteDone {
// before the first write operation, must determine if files collection is empty
// if so, create indexes if they do not already exist
if err := b.createIndexes(ctx); err != nil {
return err
}
b.firstWriteDone = true
}
return nil
}
func (b *Bucket) parseUploadOptions(opts ...*options.UploadOptions) (*Upload, error) {
upload := &Upload{
chunkSize: b.chunkSize, // upload chunk size defaults to bucket's value
}
uo := options.MergeUploadOptions(opts...)
if uo.ChunkSizeBytes != nil {
upload.chunkSize = *uo.ChunkSizeBytes
}
if uo.Registry == nil {
uo.Registry = bson.DefaultRegistry
}
if uo.Metadata != nil {
raw, err := bson.MarshalWithRegistry(uo.Registry, uo.Metadata)
if err != nil {
return nil, err
}
doc, err := bsonx.ReadDoc(raw)
if err != nil {
return nil, err
}
upload.metadata = doc
}
return upload, nil
}
type _convertFileID struct {
ID interface{} `bson:"_id"`
}
func convertFileID(fileID interface{}) (bsonx.Val, error) {
id := _convertFileID{
ID: fileID,
}
b, err := bson.Marshal(id)
if err != nil {
return bsonx.Val{}, err
}
val := bsoncore.Document(b).Lookup("_id")
var res bsonx.Val
err = res.UnmarshalBSONValue(val.Type, val.Data)
return res, err
}
|