forked from globalsign/mgo
-
Notifications
You must be signed in to change notification settings - Fork 2
/
gridfs.go
888 lines (816 loc) · 24.4 KB
/
gridfs.go
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
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
// mgo - MongoDB driver for Go
//
// Copyright (c) 2010-2012 - Gustavo Niemeyer <[email protected]>
//
// 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 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.
package mgo
import (
"context"
"crypto/md5"
"encoding/hex"
"errors"
"hash"
"io"
"os"
"sync"
"time"
"github.com/izinga/mgo/bson"
mongoDriverBson "go.mongodb.org/mongo-driver/bson"
"go.mongodb.org/mongo-driver/mongo/gridfs"
)
// GridFS stores files in two collections:
//
// - chunks stores the binary chunks. For details, see the chunks Collection.
// - files stores the file’s metadata. For details, see the files Collection.
//
// GridFS places the collections in a common bucket by prefixing each with the bucket name.
// By default, GridFS uses two collections with a bucket named fs:
//
// - fs.files
// - fs.chunks
//
// You can choose a different bucket name, as well as create multiple buckets in a single database.
// The full collection name, which includes the bucket name, is subject to the namespace length limit.
//
// Relevant documentation:
//
// https://docs.mongodb.com/manual/core/gridfs/
// https://docs.mongodb.com/manual/core/gridfs/#gridfs-chunks-collection
// https://docs.mongodb.com/manual/core/gridfs/#gridfs-files-collection
type GridFS struct {
Files *Collection
Chunks *Collection
}
type gfsFileMode int
const (
gfsClosed gfsFileMode = 0
gfsReading gfsFileMode = 1
gfsWriting gfsFileMode = 2
)
// GridFile document in files collection
type GridFile struct {
m sync.Mutex
c sync.Cond
gfs *GridFS
mode gfsFileMode
err error
chunk int
offset int64
wpending int
wbuf []byte
wsum hash.Hash
rbuf []byte
rcache *gfsCachedChunk
doc gfsFile
uploadStream *gridfs.UploadStream
downloadStream *gridfs.DownloadStream
}
type gfsFile struct {
Id interface{} `bson:"_id"`
ChunkSize int `bson:"chunkSize"`
UploadDate time.Time `bson:"uploadDate"`
Length int64 `bson:",minsize"`
MD5 string
Filename string `bson:",omitempty"`
ContentType string `bson:"contentType,omitempty"`
Metadata *bson.Raw `bson:",omitempty"`
}
type gfsChunk struct {
Id interface{} `bson:"_id"`
FilesId interface{} `bson:"files_id"`
N int
Data []byte
}
type gfsCachedChunk struct {
wait sync.Mutex
n int
data []byte
err error
}
func newGridFS(db *Database, prefix string) *GridFS {
return &GridFS{db.C(prefix + ".files"), db.C(prefix + ".chunks")}
}
func (gfs *GridFS) newFile() *GridFile {
file := &GridFile{gfs: gfs}
file.c.L = &file.m
//runtime.SetFinalizer(file, finalizeFile)
return file
}
func finalizeFile(file *GridFile) {
file.Close()
}
// Create creates a new file with the provided name in the GridFS. If the file
// name already exists, a new version will be inserted with an up-to-date
// uploadDate that will cause it to be atomically visible to the Open and
// OpenId methods. If the file name is not important, an empty name may be
// provided and the file Id used instead.
//
// It's important to Close files whether they are being written to
// or read from, and to check the err result to ensure the operation
// completed successfully.
//
// A simple example inserting a new file:
//
// func check(err error) {
// if err != nil {
// panic(err.String())
// }
// }
// file, err := db.GridFS("fs").Create("myfile.txt")
// check(err)
// n, err := file.Write([]byte("Hello world!"))
// check(err)
// err = file.Close()
// check(err)
// fmt.Printf("%d bytes written\n", n)
//
// The io.Writer interface is implemented by *GridFile and may be used to
// help on the file creation. For example:
//
// file, err := db.GridFS("fs").Create("myfile.txt")
// check(err)
// messages, err := os.Open("/var/log/messages")
// check(err)
// defer messages.Close()
// err = io.Copy(file, messages)
// check(err)
// err = file.Close()
// check(err)
func (gfs *GridFS) Create(name string) (file *GridFile, err error) {
file = gfs.newFile()
file.mode = gfsWriting
file.wsum = md5.New()
file.doc = gfsFile{Id: bson.NewObjectId(), ChunkSize: 255 * 1024, Filename: name}
if UseMongoDriver {
db := file.gfs.Files.Database.Session.GetDriverDatabase()
bucket, _ := gridfs.NewBucket(db)
var uploadStream *gridfs.UploadStream
uploadStream, err = bucket.OpenUploadStream(file.doc.Filename)
if err != nil {
return file, err
}
file.uploadStream = uploadStream
}
return
}
// OpenId returns the file with the provided id, for reading.
// If the file isn't found, err will be set to mgo.ErrNotFound.
//
// It's important to Close files whether they are being written to
// or read from, and to check the err result to ensure the operation
// completed successfully.
//
// The following example will print the first 8192 bytes from the file:
//
// func check(err error) {
// if err != nil {
// panic(err.String())
// }
// }
// file, err := db.GridFS("fs").OpenId(objid)
// check(err)
// b := make([]byte, 8192)
// n, err := file.Read(b)
// check(err)
// fmt.Println(string(b))
// check(err)
// err = file.Close()
// check(err)
// fmt.Printf("%d bytes read\n", n)
//
// The io.Reader interface is implemented by *GridFile and may be used to
// deal with it. As an example, the following snippet will dump the whole
// file into the standard output:
//
// file, err := db.GridFS("fs").OpenId(objid)
// check(err)
// err = io.Copy(os.Stdout, file)
// check(err)
// err = file.Close()
// check(err)
func (gfs *GridFS) OpenId(id interface{}) (file *GridFile, err error) {
if UseMongoDriver {
var doc gfsFile
db := gfs.Files.Database.Session.GetDriverDatabase()
err = db.Collection(gfs.Files.Name).FindOne(context.Background(), bson.M{"_id": id}).Decode(&doc)
if err != nil {
return
}
file = gfs.newFile()
file.mode = gfsReading
file.doc = doc
return
} else {
var doc gfsFile
err = gfs.Files.Find(bson.M{"_id": id}).One(&doc)
if err != nil {
return
}
file = gfs.newFile()
file.mode = gfsReading
file.doc = doc
return
}
}
// Open returns the most recently uploaded file with the provided
// name, for reading. If the file isn't found, err will be set
// to mgo.ErrNotFound.
//
// It's important to Close files whether they are being written to
// or read from, and to check the err result to ensure the operation
// completed successfully.
//
// The following example will print the first 8192 bytes from the file:
//
// file, err := db.GridFS("fs").Open("myfile.txt")
// check(err)
// b := make([]byte, 8192)
// n, err := file.Read(b)
// check(err)
// fmt.Println(string(b))
// check(err)
// err = file.Close()
// check(err)
// fmt.Printf("%d bytes read\n", n)
//
// The io.Reader interface is implemented by *GridFile and may be used to
// deal with it. As an example, the following snippet will dump the whole
// file into the standard output:
//
// file, err := db.GridFS("fs").Open("myfile.txt")
// check(err)
// err = io.Copy(os.Stdout, file)
// check(err)
// err = file.Close()
// check(err)
func (gfs *GridFS) Open(name string) (file *GridFile, err error) {
var doc gfsFile
err = gfs.Files.Find(bson.M{"filename": name}).Sort("-uploadDate").One(&doc)
if err != nil {
return
}
file = gfs.newFile()
file.mode = gfsReading
file.doc = doc
return
}
// OpenNext opens the next file from iter for reading, sets *file to it,
// and returns true on the success case. If no more documents are available
// on iter or an error occurred, *file is set to nil and the result is false.
// Errors will be available via iter.Err().
//
// The iter parameter must be an iterator on the GridFS files collection.
// Using the GridFS.Find method is an easy way to obtain such an iterator,
// but any iterator on the collection will work.
//
// If the provided *file is non-nil, OpenNext will close it before attempting
// to iterate to the next element. This means that in a loop one only
// has to worry about closing files when breaking out of the loop early
// (break, return, or panic).
//
// For example:
//
// gfs := db.GridFS("fs")
// query := gfs.Find(nil).Sort("filename")
// iter := query.Iter()
// var f *mgo.GridFile
// for gfs.OpenNext(iter, &f) {
// fmt.Printf("Filename: %s\n", f.Name())
// }
// if iter.Close() != nil {
// panic(iter.Close())
// }
func (gfs *GridFS) OpenNext(iter *Iter, file **GridFile) bool {
if *file != nil {
// Ignoring the error here shouldn't be a big deal
// as we're reading the file and the loop iteration
// for this file is finished.
_ = (*file).Close()
}
var doc gfsFile
if !iter.Next(&doc) {
*file = nil
return false
}
f := gfs.newFile()
f.mode = gfsReading
f.doc = doc
*file = f
return true
}
// Find runs query on GridFS's files collection and returns
// the resulting Query.
//
// This logic:
//
// gfs := db.GridFS("fs")
// iter := gfs.Find(nil).Iter()
//
// Is equivalent to:
//
// files := db.C("fs" + ".files")
// iter := files.Find(nil).Iter()
func (gfs *GridFS) Find(query interface{}) *Query {
return gfs.Files.Find(query)
}
// RemoveId deletes the file with the provided id from the GridFS.
func (gfs *GridFS) RemoveId(id interface{}) error {
if UseMongoDriver {
db := gfs.Files.Database.Session.GetDriverDatabase()
bucket, _ := gridfs.NewBucket(db)
return bucket.Delete(id)
} else {
err := gfs.Files.Remove(bson.M{"_id": id})
if err != nil {
return err
}
_, err = gfs.Chunks.RemoveAll(bson.D{{Name: "files_id", Value: id}})
return err
}
}
type gfsDocId struct {
Id interface{} `bson:"_id"`
}
// Remove deletes all files with the provided name from the GridFS.
func (gfs *GridFS) Remove(name string) (err error) {
iter := gfs.Files.Find(bson.M{"filename": name}).Select(bson.M{"_id": 1}).Iter()
var doc gfsDocId
for iter.Next(&doc) {
if e := gfs.RemoveId(doc.Id); e != nil {
err = e
}
}
if err == nil {
err = iter.Close()
}
return err
}
func (file *GridFile) assertMode(mode gfsFileMode) {
switch file.mode {
case mode:
return
case gfsWriting:
panic("GridFile is open for writing")
case gfsReading:
panic("GridFile is open for reading")
case gfsClosed:
panic("GridFile is closed")
default:
panic("internal error: missing GridFile mode")
}
}
// SetChunkSize sets size of saved chunks. Once the file is written to, it
// will be split in blocks of that size and each block saved into an
// independent chunk document. The default chunk size is 255kb.
//
// It is a runtime error to call this function once the file has started
// being written to.
func (file *GridFile) SetChunkSize(bytes int) {
file.assertMode(gfsWriting)
debugf("GridFile %p: setting chunk size to %d", file, bytes)
file.m.Lock()
file.doc.ChunkSize = bytes
file.m.Unlock()
}
// Id returns the current file Id.
func (file *GridFile) Id() interface{} {
return file.doc.Id
}
// SetId changes the current file Id.
//
// It is a runtime error to call this function once the file has started
// being written to, or when the file is not open for writing.
func (file *GridFile) SetId(id interface{}) {
file.assertMode(gfsWriting)
file.m.Lock()
file.doc.Id = id
if UseMongoDriver {
file.uploadStream.FileID = id
}
file.m.Unlock()
}
// Name returns the optional file name. An empty string will be returned
// in case it is unset.
func (file *GridFile) Name() string {
return file.doc.Filename
}
// SetName changes the optional file name. An empty string may be used to
// unset it.
//
// It is a runtime error to call this function when the file is not open
// for writing.
func (file *GridFile) SetName(name string) {
file.assertMode(gfsWriting)
file.m.Lock()
file.doc.Filename = name
file.m.Unlock()
}
// ContentType returns the optional file content type. An empty string will be
// returned in case it is unset.
func (file *GridFile) ContentType() string {
return file.doc.ContentType
}
// SetContentType changes the optional file content type. An empty string may be
// used to unset it.
//
// It is a runtime error to call this function when the file is not open
// for writing.
func (file *GridFile) SetContentType(ctype string) {
file.assertMode(gfsWriting)
file.m.Lock()
file.doc.ContentType = ctype
file.m.Unlock()
}
// GetMeta unmarshals the optional "metadata" field associated with the
// file into the result parameter. The meaning of keys under that field
// is user-defined. For example:
//
// result := struct{ INode int }{}
// err = file.GetMeta(&result)
// if err != nil {
// panic(err.String())
// }
// fmt.Printf("inode: %d\n", result.INode)
func (file *GridFile) GetMeta(result interface{}) (err error) {
file.m.Lock()
if file.doc.Metadata != nil {
err = bson.Unmarshal(file.doc.Metadata.Data, result)
}
file.m.Unlock()
return
}
// SetMeta changes the optional "metadata" field associated with the
// file. The meaning of keys under that field is user-defined.
// For example:
//
// file.SetMeta(bson.M{"inode": inode})
//
// It is a runtime error to call this function when the file is not open
// for writing.
func (file *GridFile) SetMeta(metadata interface{}) {
file.assertMode(gfsWriting)
data, err := bson.Marshal(metadata)
file.m.Lock()
if err != nil && file.err == nil {
file.err = err
} else {
file.doc.Metadata = &bson.Raw{Data: data}
}
file.m.Unlock()
}
// Size returns the file size in bytes.
func (file *GridFile) Size() (bytes int64) {
file.m.Lock()
bytes = file.doc.Length
file.m.Unlock()
return
}
// MD5 returns the file MD5 as a hex-encoded string.
func (file *GridFile) MD5() (md5 string) {
return file.doc.MD5
}
// UploadDate returns the file upload time.
func (file *GridFile) UploadDate() time.Time {
return file.doc.UploadDate
}
// SetUploadDate changes the file upload time.
//
// It is a runtime error to call this function when the file is not open
// for writing.
func (file *GridFile) SetUploadDate(t time.Time) {
file.assertMode(gfsWriting)
file.m.Lock()
file.doc.UploadDate = t
file.m.Unlock()
}
// Close flushes any pending changes in case the file is being written
// to, waits for any background operations to finish, and closes the file.
//
// It's important to Close files whether they are being written to
// or read from, and to check the err result to ensure the operation
// completed successfully.
func (file *GridFile) Close() (err error) {
file.m.Lock()
defer file.m.Unlock()
if UseMongoDriver {
if file.uploadStream != nil {
return file.uploadStream.Close()
}
if file.downloadStream != nil {
return file.downloadStream.Close()
}
return nil
} else {
if file.mode == gfsWriting {
if len(file.wbuf) > 0 && file.err == nil {
file.insertChunk(file.wbuf)
file.wbuf = file.wbuf[0:0]
}
file.completeWrite()
} else if file.mode == gfsReading && file.rcache != nil {
file.rcache.wait.Lock()
file.rcache = nil
}
file.mode = gfsClosed
debugf("GridFile %p: closed", file)
return file.err
}
}
func (file *GridFile) completeWrite() {
for file.wpending > 0 {
debugf("GridFile %p: waiting for %d pending chunks to complete file write", file, file.wpending)
file.c.Wait()
}
if file.err == nil {
hexsum := hex.EncodeToString(file.wsum.Sum(nil))
if file.doc.UploadDate.IsZero() {
file.doc.UploadDate = bson.Now()
}
file.doc.MD5 = hexsum
file.err = file.gfs.Files.Insert(file.doc)
}
if file.err != nil {
file.gfs.Chunks.RemoveAll(bson.D{{Name: "files_id", Value: file.doc.Id}})
}
if file.err == nil {
index := Index{
Key: []string{"files_id", "n"},
Unique: true,
}
file.err = file.gfs.Chunks.EnsureIndex(index)
}
}
// Abort cancels an in-progress write, preventing the file from being
// automically created and ensuring previously written chunks are
// removed when the file is closed.
//
// It is a runtime error to call Abort when the file was not opened
// for writing.
func (file *GridFile) Abort() {
if file.mode != gfsWriting {
panic("file.Abort must be called on file opened for writing")
}
file.err = errors.New("write aborted")
}
// Write writes the provided data to the file and returns the
// number of bytes written and an error in case something
// wrong happened.
//
// The file will internally cache the data so that all but the last
// chunk sent to the database have the size defined by SetChunkSize.
// This also means that errors may be deferred until a future call
// to Write or Close.
//
// The parameters and behavior of this function turn the file
// into an io.Writer.
func (file *GridFile) Write(data []byte) (n int, err error) {
if UseMongoDriver {
debugf("GridFile %p: writing %d bytes", file, len(data))
return file.uploadStream.Write(data)
} else {
file.assertMode(gfsWriting)
file.m.Lock()
debugf("GridFile %p: writing %d bytes", file, len(data))
defer file.m.Unlock()
if file.err != nil {
return 0, file.err
}
n = len(data)
file.doc.Length += int64(n)
chunkSize := file.doc.ChunkSize
if len(file.wbuf)+len(data) < chunkSize {
file.wbuf = append(file.wbuf, data...)
return
}
// First, flush file.wbuf complementing with data.
if len(file.wbuf) > 0 {
missing := chunkSize - len(file.wbuf)
if missing > len(data) {
missing = len(data)
}
file.wbuf = append(file.wbuf, data[:missing]...)
data = data[missing:]
file.insertChunk(file.wbuf)
file.wbuf = file.wbuf[0:0]
}
// Then, flush all chunks from data without copying.
for len(data) > chunkSize {
size := chunkSize
if size > len(data) {
size = len(data)
}
file.insertChunk(data[:size])
data = data[size:]
}
// And append the rest for a future call.
file.wbuf = append(file.wbuf, data...)
return n, file.err
}
}
func (file *GridFile) insertChunk(data []byte) {
n := file.chunk
file.chunk++
debugf("GridFile %p: adding to checksum: %q", file, string(data))
file.wsum.Write(data)
for file.doc.ChunkSize*file.wpending >= 1024*1024 {
// Hold on.. we got a MB pending.
file.c.Wait()
if file.err != nil {
return
}
}
file.wpending++
debugf("GridFile %p: inserting chunk %d with %d bytes", file, n, len(data))
// We may not own the memory of data, so rather than
// simply copying it, we'll marshal the document ahead of time.
data, err := bson.Marshal(gfsChunk{bson.NewObjectId(), file.doc.Id, n, data})
if err != nil {
file.err = err
return
}
go func() {
err := file.gfs.Chunks.Insert(bson.Raw{Data: data})
file.m.Lock()
file.wpending--
if err != nil && file.err == nil {
file.err = err
}
file.c.Broadcast()
file.m.Unlock()
}()
}
// Seek sets the offset for the next Read or Write on file to
// offset, interpreted according to whence: 0 means relative to
// the origin of the file, 1 means relative to the current offset,
// and 2 means relative to the end. It returns the new offset and
// an error, if any.
func (file *GridFile) Seek(offset int64, whence int) (pos int64, err error) {
file.m.Lock()
debugf("GridFile %p: seeking for %s (whence=%d)", file, offset, whence)
defer file.m.Unlock()
switch whence {
case os.SEEK_SET:
case os.SEEK_CUR:
offset += file.offset
case os.SEEK_END:
offset += file.doc.Length
default:
panic("unsupported whence value")
}
if offset > file.doc.Length {
return file.offset, errors.New("seek past end of file")
}
if offset == file.doc.Length {
// If we're seeking to the end of the file,
// no need to read anything. This enables
// a client to find the size of the file using only the
// io.ReadSeeker interface with low overhead.
file.offset = offset
return file.offset, nil
}
chunk := int(offset / int64(file.doc.ChunkSize))
if chunk+1 == file.chunk && offset >= file.offset {
file.rbuf = file.rbuf[int(offset-file.offset):]
file.offset = offset
return file.offset, nil
}
file.offset = offset
file.chunk = chunk
file.rbuf = nil
file.rbuf, err = file.getChunk()
if err == nil {
file.rbuf = file.rbuf[int(file.offset-int64(chunk)*int64(file.doc.ChunkSize)):]
}
return file.offset, err
}
// Read reads into b the next available data from the file and
// returns the number of bytes written and an error in case
// something wrong happened. At the end of the file, n will
// be zero and err will be set to io.EOF.
//
// The parameters and behavior of this function turn the file
// into an io.Reader.
func (file *GridFile) Read(b []byte) (n int, err error) {
if UseMongoDriver {
db := file.gfs.Files.Database.Session.GetDriverDatabase()
bucket, _ := gridfs.NewBucket(db)
var downloadStream *gridfs.DownloadStream
downloadStream, err = bucket.OpenDownloadStream(file.Id())
if err != nil {
return 0, err
}
file.downloadStream = downloadStream
return downloadStream.Read(b)
} else {
file.assertMode(gfsReading)
file.m.Lock()
debugf("GridFile %p: reading at offset %d into buffer of length %d", file, file.offset, len(b))
defer file.m.Unlock()
if file.offset == file.doc.Length {
return 0, io.EOF
}
for err == nil {
i := copy(b, file.rbuf)
n += i
file.offset += int64(i)
file.rbuf = file.rbuf[i:]
if i == len(b) || file.offset == file.doc.Length {
break
}
b = b[i:]
file.rbuf, err = file.getChunk()
}
return n, err
}
}
func (file *GridFile) getChunk() (data []byte, err error) {
if UseMongoDriver {
cache := file.rcache
file.rcache = nil
if cache != nil && cache.n == file.chunk {
debugf("GridFile %p: Getting chunk %d from cache", file, file.chunk)
cache.wait.Lock()
data, err = cache.data, cache.err
} else {
debugf("GridFile %p: Fetching chunk %d", file, file.chunk)
var doc gfsChunk
err = file.gfs.Chunks.Find(mongoDriverBson.D{{Key: "files_id", Value: file.doc.Id}, {Key: "n", Value: file.chunk}}).One(&doc)
data = doc.Data
}
file.chunk++
if int64(file.chunk)*int64(file.doc.ChunkSize) < file.doc.Length {
// Read the next one in background.
cache = &gfsCachedChunk{n: file.chunk}
cache.wait.Lock()
debugf("GridFile %p: Scheduling chunk %d for background caching", file, file.chunk)
// Clone the session to avoid having it closed in between.
chunks := file.gfs.Chunks
session := chunks.Database.Session.Clone()
go func(id interface{}, n int) {
defer session.Close()
chunks = chunks.With(session)
var doc gfsChunk
cache.err = chunks.Find(mongoDriverBson.D{{Key: "files_id", Value: id}, {Key: "n", Value: n}}).One(&doc)
cache.data = doc.Data
cache.wait.Unlock()
}(file.doc.Id, file.chunk)
file.rcache = cache
}
debugf("Returning err: %#v", err)
return
} else {
cache := file.rcache
file.rcache = nil
if cache != nil && cache.n == file.chunk {
debugf("GridFile %p: Getting chunk %d from cache", file, file.chunk)
cache.wait.Lock()
data, err = cache.data, cache.err
} else {
debugf("GridFile %p: Fetching chunk %d", file, file.chunk)
var doc gfsChunk
err = file.gfs.Chunks.Find(bson.D{{Name: "files_id", Value: file.doc.Id}, {Name: "n", Value: file.chunk}}).One(&doc)
data = doc.Data
}
file.chunk++
if int64(file.chunk)*int64(file.doc.ChunkSize) < file.doc.Length {
// Read the next one in background.
cache = &gfsCachedChunk{n: file.chunk}
cache.wait.Lock()
debugf("GridFile %p: Scheduling chunk %d for background caching", file, file.chunk)
// Clone the session to avoid having it closed in between.
chunks := file.gfs.Chunks
session := chunks.Database.Session.Clone()
go func(id interface{}, n int) {
defer session.Close()
chunks = chunks.With(session)
var doc gfsChunk
cache.err = chunks.Find(bson.D{{Name: "files_id", Value: id}, {Name: "n", Value: n}}).One(&doc)
cache.data = doc.Data
cache.wait.Unlock()
}(file.doc.Id, file.chunk)
file.rcache = cache
}
debugf("Returning err: %#v", err)
return
}
}