File: cache.go

package info (click to toggle)
golang-github-biogo-hts 1.4.4%2Bdfsg1-1
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 1,740 kB
  • sloc: makefile: 3
file content (517 lines) | stat: -rw-r--r-- 11,123 bytes parent folder | download | duplicates (4)
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
// Copyright ©2015 The bíogo Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

// Package cache provides basic block cache types for the bgzf package.
package cache

import (
	"sync"

	"github.com/biogo/hts/bgzf"
)

var (
	_ Cache = (*LRU)(nil)
	_ Cache = (*FIFO)(nil)
	_ Cache = (*Random)(nil)
)

// Free attempts to drop as many blocks from c as needed allow
// n successful Put calls on c. It returns a boolean indicating
// whether n slots were made available.
func Free(n int, c Cache) bool {
	empty := c.Cap() - c.Len()
	if n <= empty {
		return true
	}
	c.Drop(n - empty)
	return c.Cap()-c.Len() >= n
}

// Cache is an extension of bgzf.Cache that allows inspection
// and manipulation of the cache.
type Cache interface {
	bgzf.Cache

	// Len returns the number of elements held by
	// the cache.
	Len() int

	// Cap returns the maximum number of elements
	// that can be held by the cache.
	Cap() int

	// Resize changes the capacity of the cache to n,
	// dropping excess blocks if n is less than the
	// number of cached blocks.
	Resize(n int)

	// Drop evicts n elements from the cache according
	// to the cache eviction policy.
	Drop(n int)
}

func insertAfter(pos, n *node) {
	n.prev = pos
	pos.next, n.next, pos.next.prev = n, pos.next, n
}

func remove(n *node, table map[int64]*node) {
	delete(table, n.b.Base())
	n.prev.next = n.next
	n.next.prev = n.prev
	n.next = nil
	n.prev = nil
}

// NewLRU returns an LRU cache with n slots. If n is less than 1
// a nil cache is returned.
func NewLRU(n int) Cache {
	if n < 1 {
		return nil
	}
	c := LRU{
		table: make(map[int64]*node, n),
		cap:   n,
	}
	c.root.next = &c.root
	c.root.prev = &c.root
	return &c
}

// LRU satisfies the Cache interface with least recently used eviction
// behavior where Unused Blocks are preferentially evicted.
type LRU struct {
	mu    sync.RWMutex
	root  node
	table map[int64]*node
	cap   int
}

type node struct {
	b bgzf.Block

	next, prev *node
}

// Len returns the number of elements held by the cache.
func (c *LRU) Len() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return len(c.table)
}

// Cap returns the maximum number of elements that can be held by the cache.
func (c *LRU) Cap() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return c.cap
}

// Resize changes the capacity of the cache to n, dropping excess blocks
// if n is less than the number of cached blocks.
func (c *LRU) Resize(n int) {
	c.mu.Lock()
	if n < len(c.table) {
		c.drop(len(c.table) - n)
	}
	c.cap = n
	c.mu.Unlock()
}

// Drop evicts n elements from the cache according to the cache eviction policy.
func (c *LRU) Drop(n int) {
	c.mu.Lock()
	c.drop(n)
	c.mu.Unlock()
}

func (c *LRU) drop(n int) {
	for ; n > 0 && c.Len() > 0; n-- {
		remove(c.root.prev, c.table)
	}
}

// Get returns the Block in the Cache with the specified base or a nil Block
// if it does not exist.
func (c *LRU) Get(base int64) bgzf.Block {
	c.mu.Lock()
	defer c.mu.Unlock()

	n, ok := c.table[base]
	if !ok {
		return nil
	}
	remove(n, c.table)
	return n.b
}

// Peek returns a boolean indicating whether a Block exists in the Cache for
// the given base offset and the expected offset for the subsequent Block in
// the BGZF stream.
func (c *LRU) Peek(base int64) (exist bool, next int64) {
	c.mu.RLock()
	defer c.mu.RUnlock()

	n, exist := c.table[base]
	if !exist {
		return false, -1
	}
	next = n.b.NextBase()
	return exist, next
}

// Put inserts a Block into the Cache, returning the Block that was evicted or
// nil if no eviction was necessary and the Block was retained. Unused Blocks
// are not retained but are returned if the Cache is full.
func (c *LRU) Put(b bgzf.Block) (evicted bgzf.Block, retained bool) {
	c.mu.Lock()
	defer c.mu.Unlock()

	var d bgzf.Block
	if _, ok := c.table[b.Base()]; ok {
		return b, false
	}
	used := b.Used()
	if len(c.table) == c.cap {
		if !used {
			return b, false
		}
		d = c.root.prev.b
		remove(c.root.prev, c.table)
	}
	n := &node{b: b}
	c.table[b.Base()] = n
	if used {
		insertAfter(&c.root, n)
	} else {
		insertAfter(c.root.prev, n)
	}
	return d, true
}

// NewFIFO returns a FIFO cache with n slots. If n is less than 1
// a nil cache is returned.
func NewFIFO(n int) Cache {
	if n < 1 {
		return nil
	}
	c := FIFO{
		table: make(map[int64]*node, n),
		cap:   n,
	}
	c.root.next = &c.root
	c.root.prev = &c.root
	return &c
}

// FIFO satisfies the Cache interface with first in first out eviction
// behavior where Unused Blocks are preferentially evicted.
type FIFO struct {
	mu    sync.RWMutex
	root  node
	table map[int64]*node
	cap   int
}

// Len returns the number of elements held by the cache.
func (c *FIFO) Len() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return len(c.table)
}

// Cap returns the maximum number of elements that can be held by the cache.
func (c *FIFO) Cap() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return c.cap
}

// Resize changes the capacity of the cache to n, dropping excess blocks
// if n is less than the number of cached blocks.
func (c *FIFO) Resize(n int) {
	c.mu.Lock()
	if n < len(c.table) {
		c.drop(len(c.table) - n)
	}
	c.cap = n
	c.mu.Unlock()
}

// Drop evicts n elements from the cache according to the cache eviction policy.
func (c *FIFO) Drop(n int) {
	c.mu.Lock()
	c.drop(n)
	c.mu.Unlock()
}

func (c *FIFO) drop(n int) {
	for ; n > 0 && c.Len() > 0; n-- {
		remove(c.root.prev, c.table)
	}
}

// Get returns the Block in the Cache with the specified base or a nil Block
// if it does not exist.
func (c *FIFO) Get(base int64) bgzf.Block {
	c.mu.Lock()
	defer c.mu.Unlock()

	n, ok := c.table[base]
	if !ok {
		return nil
	}
	if !n.b.Used() {
		remove(n, c.table)
	}
	return n.b
}

// Peek returns a boolean indicating whether a Block exists in the Cache for
// the given base offset and the expected offset for the subsequent Block in
// the BGZF stream.
func (c *FIFO) Peek(base int64) (exist bool, next int64) {
	c.mu.RLock()
	defer c.mu.RUnlock()

	n, exist := c.table[base]
	if !exist {
		return false, -1
	}
	next = n.b.NextBase()
	return exist, next
}

// Put inserts a Block into the Cache, returning the Block that was evicted or
// nil if no eviction was necessary and the Block was retained. Unused Blocks
// are not retained but are returned if the Cache is full.
func (c *FIFO) Put(b bgzf.Block) (evicted bgzf.Block, retained bool) {
	c.mu.Lock()
	defer c.mu.Unlock()

	var d bgzf.Block
	if _, ok := c.table[b.Base()]; ok {
		return b, false
	}
	used := b.Used()
	if len(c.table) == c.cap {
		if !used {
			return b, false
		}
		d = c.root.prev.b
		remove(c.root.prev, c.table)
	}
	n := &node{b: b}
	c.table[b.Base()] = n
	if used {
		insertAfter(&c.root, n)
	} else {
		insertAfter(c.root.prev, n)
	}
	return d, true
}

// NewRandom returns a random eviction cache with n slots. If n is less than 1
// a nil cache is returned.
func NewRandom(n int) Cache {
	if n < 1 {
		return nil
	}
	return &Random{
		table: make(map[int64]bgzf.Block, n),
		cap:   n,
	}
}

// Random satisfies the Cache interface with random eviction behavior
// where Unused Blocks are preferentially evicted.
type Random struct {
	mu    sync.RWMutex
	table map[int64]bgzf.Block
	cap   int
}

// Len returns the number of elements held by the cache.
func (c *Random) Len() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return len(c.table)
}

// Cap returns the maximum number of elements that can be held by the cache.
func (c *Random) Cap() int {
	c.mu.RLock()
	defer c.mu.RUnlock()

	return c.cap
}

// Resize changes the capacity of the cache to n, dropping excess blocks
// if n is less than the number of cached blocks.
func (c *Random) Resize(n int) {
	c.mu.Lock()
	if n < len(c.table) {
		c.drop(len(c.table) - n)
	}
	c.cap = n
	c.mu.Unlock()
}

// Drop evicts n elements from the cache according to the cache eviction policy.
func (c *Random) Drop(n int) {
	c.mu.Lock()
	c.drop(n)
	c.mu.Unlock()
}

func (c *Random) drop(n int) {
	if n < 1 {
		return
	}
	for k, b := range c.table {
		if b.Used() {
			continue
		}
		delete(c.table, k)
		if n--; n == 0 {
			return
		}
	}
	for k := range c.table {
		delete(c.table, k)
		if n--; n == 0 {
			break
		}
	}
}

// Get returns the Block in the Cache with the specified base or a nil Block
// if it does not exist.
func (c *Random) Get(base int64) bgzf.Block {
	c.mu.Lock()
	defer c.mu.Unlock()

	b, ok := c.table[base]
	if !ok {
		return nil
	}
	delete(c.table, base)
	return b
}

// Peek returns a boolean indicating whether a Block exists in the Cache for
// the given base offset and the expected offset for the subsequent Block in
// the BGZF stream.
func (c *Random) Peek(base int64) (exist bool, next int64) {
	c.mu.RLock()
	defer c.mu.RUnlock()

	n, exist := c.table[base]
	if !exist {
		return false, -1
	}
	next = n.NextBase()
	return exist, next
}

// Put inserts a Block into the Cache, returning the Block that was evicted or
// nil if no eviction was necessary and the Block was retained. Unused Blocks
// are not retained but are returned if the Cache is full.
func (c *Random) Put(b bgzf.Block) (evicted bgzf.Block, retained bool) {
	c.mu.Lock()
	defer c.mu.Unlock()

	var d bgzf.Block
	if _, ok := c.table[b.Base()]; ok {
		return b, false
	}
	if len(c.table) == c.cap {
		if !b.Used() {
			return b, false
		}
		for k, v := range c.table {
			if v.Used() {
				continue
			}
			delete(c.table, k)
			d = v
			goto done
		}
		for k, v := range c.table {
			delete(c.table, k)
			d = v
			break
		}
	done:
	}
	c.table[b.Base()] = b
	return d, true
}

// StatsRecorder allows a bgzf.Cache to capture cache statistics.
type StatsRecorder struct {
	bgzf.Cache

	mu    sync.RWMutex
	stats Stats
}

// Stats represents statistics of a bgzf.Cache.
type Stats struct {
	Gets      int // number of Get operations
	Misses    int // number of cache misses
	Puts      int // number of Put operations
	Retains   int // number of times a Put has resulted in Block retention
	Evictions int // number of times a Put has resulted in a Block eviction
}

// Stats returns the current statistics for the cache.
func (s *StatsRecorder) Stats() Stats {
	s.mu.RLock()
	defer s.mu.RUnlock()
	return s.stats
}

// Reset zeros the statistics kept by the StatsRecorder.
func (s *StatsRecorder) Reset() {
	s.mu.Lock()
	s.stats = Stats{}
	s.mu.Unlock()
}

// Get returns the Block in the underlying Cache with the specified base or a nil
// Block if it does not exist. It updates the gets and misses statistics.
func (s *StatsRecorder) Get(base int64) bgzf.Block {
	s.mu.Lock()
	s.stats.Gets++
	blk := s.Cache.Get(base)
	if blk == nil {
		s.stats.Misses++
	}
	s.mu.Unlock()
	return blk
}

// Put inserts a Block into the underlying Cache, returning the Block and eviction
// status according to the underlying cache behavior. It updates the puts, retains and
// evictions statistics.
func (s *StatsRecorder) Put(b bgzf.Block) (evicted bgzf.Block, retained bool) {
	s.mu.Lock()
	s.stats.Puts++
	blk, retained := s.Cache.Put(b)
	if retained {
		s.stats.Retains++
		if blk != nil {
			s.stats.Evictions++
		}
	}
	s.mu.Unlock()
	return blk, retained
}