-
Notifications
You must be signed in to change notification settings - Fork 16
/
pamreader.go
489 lines (449 loc) · 14.3 KB
/
pamreader.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
// Copyright 2018 GRAIL, Inc. All rights reserved.
// Use of this source code is governed by the Apache 2.0
// license that can be found in the LICENSE file.
package pam
import (
"context"
"fmt"
"sync"
"sync/atomic"
"unsafe"
"github.com/grailbio/base/errors"
"github.com/grailbio/base/file"
"github.com/grailbio/base/recordio/recordiozstd"
"github.com/grailbio/base/vcontext"
"github.com/grailbio/bio/biopb"
gbam "github.com/grailbio/bio/encoding/bam"
"github.com/grailbio/bio/encoding/pam/fieldio"
"github.com/grailbio/bio/encoding/pam/pamutil"
"github.com/grailbio/hts/sam"
"v.io/x/lib/vlog"
)
// ReadOpts defines configuration parameters for PAM readers.
type ReadOpts struct {
// DropFields causes the listed fields not to be filled in Read().
DropFields []gbam.FieldType
// Optional row shard range. Only records in this range will be returned
// by Scan() and Read().
//
// Examples:
// RecRange{{0,0},{InfinityRefID, InfinityPos}} : covers all possible sequences.
// RecRange{{0,0},{LimitValidRefID, InfinityPos}} : covers all mapped sequences.
// RecRange{{UnmappedRefID,0},{UnmappedRefID, InfinityPos}} : covers all unmapped sequences.
Range biopb.CoordRange
// When set, it causes the file package to keep retrying when the file is
// reported as not found. This flag is passed to file.Opts. See file.Opts for
// more details.
RetryWhenNotFound bool
}
// ShardReader is for reading one PAM rowshard. This class is generally hidden
// behind pam.Reader and is not used directly.
type ShardReader struct {
label string // for vlogging only.
// Requested range of records to read. It is a copy of ReadOpts.Range.
// It may span outside the range of this shard.
//
// INVARIANT: (requestedRange ∩ shardRange) != ∅
requestedRange biopb.CoordRange
// Fields to read. It is a complement of ReadOpts.DropFields.
needField [gbam.NumFields]bool
// Reader for each field. nil if !needField[f]
fieldReaders [gbam.NumFields]*fieldio.Reader
index biopb.PAMShardIndex
header *sam.Header // Parsed out of index.EmbeddedBamHeader
path string // PAM directory path.
shardRange biopb.CoordRange // row range parsed out of the filename.
nRecords int // # records read so far
err *errors.Once // Points to Reader.err
}
var (
dummyMu sync.Mutex
dummyQual unsafe.Pointer // stores *[]byte
dummySeq unsafe.Pointer // stores *[]sam.Doublet
)
// GetDummyQual returns a singleton qual string that contains dummy data. It is
// returned when the qual field is dropped in the options.
func GetDummyQual(length int) []byte {
buf := (*[]byte)(atomic.LoadPointer(&dummyQual))
if buf != nil && len(*buf) >= length {
return (*buf)[:length]
}
dummyMu.Lock()
newBuf := make([]byte, length)
for i := 0; i < length; i++ {
newBuf[i] = 0xff
}
atomic.StorePointer(&dummyQual, unsafe.Pointer(&newBuf))
dummyMu.Unlock()
return newBuf[:length]
}
// GetDummySeq returns a singleton seq string that contains dummy data. It is
// returned when the seq field is dropped in the options.
func GetDummySeq(length int) sam.Seq {
n := (length + 1) / 2
buf := (*[]sam.Doublet)(atomic.LoadPointer(&dummySeq))
if buf != nil && len(*buf) >= n {
return sam.Seq{Length: length, Seq: (*buf)[:n]}
}
dummyMu.Lock()
newBuf := make([]sam.Doublet, length)
for i := 0; i < length; i++ {
newBuf[i] = 0xf
}
atomic.StorePointer(&dummySeq, unsafe.Pointer(&newBuf))
dummyMu.Unlock()
return sam.Seq{Length: length, Seq: newBuf[:n]}
}
// Read one record from the buffer "rb". prevRec is used to delta-decode some
// fields.
func (r *ShardReader) readRecord() *sam.Record {
refs := r.header.Refs()
rec := sam.GetFromFreePool()
coord, ok := r.fieldReaders[gbam.FieldCoord].ReadCoordField()
if !ok {
return nil
}
if coord.RefId >= 0 {
rec.Ref = refs[coord.RefId]
rec.Pos = int(coord.Pos)
} else {
rec.Ref = nil
rec.Pos = -1
}
if r.needField[gbam.FieldFlags] {
flags, ok := r.fieldReaders[gbam.FieldFlags].ReadUint16Field()
if !ok {
return nil
}
rec.Flags = sam.Flags(flags)
}
if r.needField[gbam.FieldMapq] {
rec.MapQ, ok = r.fieldReaders[gbam.FieldMapq].ReadUint8Field()
if !ok {
return nil
}
}
if r.needField[gbam.FieldMateRefID] {
mateRefID, ok := r.fieldReaders[gbam.FieldMateRefID].ReadVarintDeltaField()
if !ok {
return nil
}
rec.MateRef = nil
if mateRefID >= 0 {
rec.MateRef = refs[mateRefID]
}
}
if r.needField[gbam.FieldMatePos] {
matePos, ok := r.fieldReaders[gbam.FieldMatePos].ReadVarintDeltaField()
if !ok {
return nil
}
rec.MatePos = int(matePos)
}
if r.needField[gbam.FieldTempLen] {
tempLen, ok := r.fieldReaders[gbam.FieldTempLen].ReadVarintField()
if !ok {
return nil
}
rec.TempLen = int(tempLen)
}
// Collect the length info for variable-length fields.
arenaBytes := 0
nCigarOps := 0
if r.needField[gbam.FieldCigar] {
if nCigarOps, ok = r.fieldReaders[gbam.FieldCigar].ReadCigarMetadata(); !ok {
return nil
}
arenaBytes += nCigarOps * gbam.CigarOpSize
}
nameMd := fieldio.StringDeltaMetadata{}
if r.needField[gbam.FieldName] {
if nameMd, ok = r.fieldReaders[gbam.FieldName].ReadStringDeltaMetadata(); !ok {
return nil
}
arenaBytes += nameMd.PrefixLen + nameMd.DeltaLen
}
rec.Seq.Length = 0
if r.needField[gbam.FieldSeq] {
if rec.Seq.Length, ok = r.fieldReaders[gbam.FieldSeq].ReadSeqMetadata(); !ok {
return nil
}
arenaBytes += fieldio.SeqBytes(rec.Seq.Length)
}
qualLen := 0
if r.needField[gbam.FieldQual] {
qualLen, ok = r.fieldReaders[gbam.FieldQual].ReadBytesMetadata()
if !ok {
return nil
}
arenaBytes += qualLen
}
auxMd := fieldio.AuxMetadata{}
if r.needField[gbam.FieldAux] {
if auxMd, ok = r.fieldReaders[gbam.FieldAux].ReadAuxMetadata(); !ok {
return nil
}
// Round up to the next CPU word boundary, since we will store
// pointers in the arena.
const pointerSize = int(unsafe.Sizeof(uintptr(0)))
arenaBytes += len(auxMd.Tags)*fieldio.SizeofSliceHeader + pointerSize
for _, tag := range auxMd.Tags {
arenaBytes += len(tag.Name) + tag.Len
}
}
sam.ResizeScratch(&rec.Scratch, arenaBytes)
arena := fieldio.NewUnsafeArena(rec.Scratch)
if r.needField[gbam.FieldCigar] {
rec.Cigar = r.fieldReaders[gbam.FieldCigar].ReadCigarField(nCigarOps, &arena)
}
if r.needField[gbam.FieldName] {
rec.Name = r.fieldReaders[gbam.FieldName].ReadStringDeltaField(nameMd, &arena)
}
// Qual and Seq must be of the same length, as per BAM file format spec. So
// when one of them is not read, fill it with dummy data.
switch {
case r.needField[gbam.FieldSeq] && r.needField[gbam.FieldQual]:
// Common case
rec.Seq = r.fieldReaders[gbam.FieldSeq].ReadSeqField(rec.Seq.Length, &arena)
rec.Qual = r.fieldReaders[gbam.FieldQual].ReadBytesField(qualLen, &arena)
case r.needField[gbam.FieldSeq] && !r.needField[gbam.FieldQual]:
// Fill qual with garbage data w/ the same length as seq
rec.Seq = r.fieldReaders[gbam.FieldSeq].ReadSeqField(rec.Seq.Length, &arena)
rec.Qual = GetDummyQual(rec.Seq.Length)
case !r.needField[gbam.FieldSeq] && r.needField[gbam.FieldQual]:
// Fill seq with garbage data w/ the same length as qual
rec.Qual = r.fieldReaders[gbam.FieldQual].ReadBytesField(qualLen, &arena)
rec.Seq = GetDummySeq(len(rec.Qual))
}
if r.needField[gbam.FieldAux] {
rec.AuxFields = r.fieldReaders[gbam.FieldAux].ReadAuxField(auxMd, &arena)
}
r.nRecords++
if coord.LT(r.requestedRange.Start) {
// This can't happen; seek() should have moved the read pointer >=
// requestedRange.Start.
vlog.Panicf("Record too small: %v, requested %+v", r, r.requestedRange)
}
if coord.GE(r.requestedRange.Limit) {
return nil
}
return rec
}
func validateReadOpts(o *ReadOpts) error {
for _, fi := range o.DropFields {
if int(fi) < 0 || int(fi) >= gbam.NumFields {
return fmt.Errorf("invalid DropField %v in %+v", fi, *o)
}
if fi == gbam.FieldCoord {
// Coord field is needed to support range reads.
return fmt.Errorf("dropping Coord field is not supported in %+v", *o)
}
}
return pamutil.ValidateCoordRange(&o.Range)
}
type fieldSeeker struct {
r *fieldio.Reader
skip func(*fieldio.Reader)
}
func (f *fieldSeeker) Seek(requestedRange biopb.CoordRange) (biopb.Coord, bool) {
return f.r.Seek(requestedRange)
}
func (f *fieldSeeker) Skip() {
f.skip(f.r)
}
// Set up the reader so that next call to readRecord() will read the first
// record at or after requestedRange.Start.
func (r *ShardReader) seek(requestedRange biopb.CoordRange) {
var readers []fieldio.ColumnSeeker
fields := []struct {
field gbam.FieldType
skip func(*fieldio.Reader)
}{
{gbam.FieldFlags, (*fieldio.Reader).SkipUint16Field},
{gbam.FieldMapq, (*fieldio.Reader).SkipUint8Field},
{gbam.FieldMateRefID, func(fr *fieldio.Reader) { fr.ReadVarintDeltaField() }},
{gbam.FieldMatePos, func(fr *fieldio.Reader) { fr.ReadVarintDeltaField() }},
{gbam.FieldTempLen, func(fr *fieldio.Reader) { fr.ReadVarintField() }},
{gbam.FieldCigar, (*fieldio.Reader).SkipCigarField},
{gbam.FieldName, (*fieldio.Reader).SkipStringDeltaField},
{gbam.FieldSeq, (*fieldio.Reader).SkipSeqField},
{gbam.FieldQual, (*fieldio.Reader).SkipBytesField},
{gbam.FieldAux, (*fieldio.Reader).SkipAuxField},
}
for _, d := range fields {
if fr := r.fieldReaders[d.field]; fr != nil {
readers = append(readers, &fieldSeeker{fr, d.skip})
}
}
r.err.Set(fieldio.SeekReaders(requestedRange, r.fieldReaders[gbam.FieldCoord], readers))
}
// NewShardReader creates a reader for a rowshard. requestedRange and
// dropFields are fields from ReadOpts. pamIndex is the index file information,
// gleaned from its pathname. muPtr and errPtr are for reporting errors to the
// parent.
//
// REQUIRES: requestedRange ∩ pamIndex.Range != ∅
func NewShardReader(
ctx context.Context,
opts ReadOpts,
pamIndex pamutil.FileInfo,
errp *errors.Once) *ShardReader {
r := &ShardReader{
label: fmt.Sprintf("%s:s%s:u%s", file.Base(pamIndex.Dir), pamutil.CoordRangePathString(pamIndex.Range), pamutil.CoordRangePathString(opts.Range)),
path: pamIndex.Dir,
shardRange: pamIndex.Range,
requestedRange: opts.Range,
err: errp,
}
vlog.VI(1).Infof("%v: NewShardReader", r.label)
for i := range r.needField {
r.needField[i] = true
}
for _, f := range opts.DropFields {
r.needField[f] = false
}
var err error
if r.index, err = pamutil.ReadShardIndex(vcontext.Background(), r.path, r.shardRange); err != nil {
vlog.Errorf("Failed to read shard index: %v", err)
r.err.Set(errors.E(err, fmt.Sprintf("newshardreader %s: read shard index", r.path)))
return r
}
r.header, err = gbam.UnmarshalHeader(r.index.EncodedBamHeader)
if err != nil {
r.err.Set(errors.E(err, fmt.Sprintf("newshardeader %s: decode sam.Header in index", r.path)))
return r
}
if !r.requestedRange.Intersects(r.shardRange) {
vlog.Panicf("%v: Range doesn't intersect", r.label)
}
for f := range r.needField {
if r.needField[f] {
path := pamutil.FieldDataPath(pamIndex.Dir, pamIndex.Range, gbam.FieldType(f).String())
label := fmt.Sprintf("%s:s%s:u%s(%v)",
file.Base(pamIndex.Dir),
pamutil.CoordRangePathString(pamIndex.Range),
pamutil.CoordRangePathString(r.requestedRange),
gbam.FieldType(f))
fileOpts := file.Opts{RetryWhenNotFound: opts.RetryWhenNotFound}
r.fieldReaders[f], err = fieldio.NewReader(ctx, path, label, f == int(gbam.FieldCoord), fileOpts, errp)
if err != nil {
r.err.Set(err)
return r
} else if r.fieldReaders[f] == nil {
r.err.Set(fmt.Errorf("missing file for %s: %s", label, path))
return r
}
}
}
r.seek(r.requestedRange)
return r
}
// Close must be called exactly once. After close, no method may be called.
func (r *ShardReader) Close(ctx context.Context) {
for f := range r.fieldReaders {
if fr := r.fieldReaders[f]; fr != nil {
fr.Close(ctx)
}
}
}
// Reader is the main PAM reader class. It can read across multiple rowshard
// files.
type Reader struct {
label string // for vlogging only.
ctx context.Context // passed into NewReader
opts ReadOpts
// Sorted list of *.index files whose rowrange intersect with
// opts.Range.
indexFiles []pamutil.FileInfo
// Current rowshard reader.
r *ShardReader
// Object returned by Record().
rec *sam.Record
numRead int
err errors.Once
}
// NewReader creates a new Reader.
func NewReader(opts ReadOpts, dir string) *Reader {
r := &Reader{
ctx: vcontext.Background(),
opts: opts,
}
r.err.Set(validateReadOpts(&r.opts))
if r.err.Err() != nil {
return r
}
r.label = fmt.Sprintf("%s:u%s", file.Base(dir), pamutil.CoordRangePathString(r.opts.Range))
var err error
if r.indexFiles, err = pamutil.FindIndexFilesInRange(r.ctx, dir, r.opts.Range); err != nil {
r.err.Set(err)
return r
}
if len(r.indexFiles) == 0 {
r.err.Set(fmt.Errorf("newreader %v: no pam file found for range %+v", dir, r.opts))
return r
}
vlog.VI(1).Infof("Found index files in range %+v: %+v", r.opts.Range, r.indexFiles)
r.r = NewShardReader(r.ctx, r.opts, r.indexFiles[0], &r.err)
r.indexFiles = r.indexFiles[1:]
return r
}
// Record returns the most recent record read by Scan.
//
// REQUIRES: Scan() returned true.
func (r *Reader) Record() *sam.Record {
return r.rec
}
// Scan reads the next record. It returns true if a record is found. It returns
// false on EOF or an error. Call Record to get the record, and Err or Close to
// obtain the error code.
//
// Example:
// r := pam.NewReader(...)
// for r.Scan() {
// rec := r.Record()
// ... use the rec ...
// }
// err := r.Close()
func (r *Reader) Scan() bool {
for {
if r.Err() != nil {
// TODO(saito) do lock-free access
return false
}
r.rec = r.r.readRecord()
if r.rec != nil {
r.numRead++
return true
}
if r.Err() != nil {
return false
}
// End of shard
if len(r.indexFiles) == 0 {
if r.opts.Range.Limit.RefId < 0 {
vlog.VI(1).Infof("%v: Finished reading %d records for range %v", r.label, r.numRead, r.opts.Range)
}
return false
}
r.r.Close(r.ctx)
r.r = NewShardReader(r.ctx, r.opts, r.indexFiles[0], &r.err)
r.indexFiles = r.indexFiles[1:]
}
}
// Err returns any error encountered so far.
//
// Note: Err never returns io.EOF. On EOF, Scan() returns false, and Err()
// returns nil.
func (r *Reader) Err() error {
return r.err.Err()
}
// Close all the files. Must be called exactly once.
func (r *Reader) Close() error {
if r.r != nil {
r.r.Close(r.ctx)
}
return r.err.Err()
}
func init() {
recordiozstd.Init()
}