This repository has been archived by the owner on Sep 26, 2019. It is now read-only.
/
count.go
223 lines (180 loc) · 4.77 KB
/
count.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
// Copyright 2019 The go-pttai Authors
// This file is part of the go-pttai library.
//
// The go-pttai library is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-pttai library 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 Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-pttai library. If not, see <http://www.gnu.org/licenses/>.
package service
import (
"hash"
"math"
"sync"
"github.com/ailabstw/go-pttai/common"
"github.com/ailabstw/go-pttai/common/types"
"github.com/ailabstw/go-pttai/pttdb"
"github.com/spaolacci/murmur3"
)
type Count struct {
Bits types.BitVector `json:"B"` // bitvector to hold the occupied buckets
// IsFull types.Bool `json:"i"`
lock sync.RWMutex
hash hash.Hash64
p uint
m uint64
db *pttdb.LDBBatch
dbPrefixID *types.PttID
dbID *types.PttID
dbPrefix []byte
}
func NewCount(db *pttdb.LDBBatch, dbPrefixID *types.PttID, dbID *types.PttID, dbPrefix []byte, p uint, isNewBits bool) (*Count, error) {
c := &Count{}
m := uint64(1 << p)
c.m = m
c.p = p
c.SetHash()
c.SetDB(db, dbPrefixID, dbID, dbPrefix)
if isNewBits {
c.NewBits()
}
return c, nil
}
func (c *Count) NewBits() {
bits := types.NewBitVector(c.m)
c.Bits = bits
}
func (c *Count) SetBits(bits types.BitVector) {
c.Bits = bits
}
func (c *Count) SetHash() {
c.hash = murmur3.New64()
}
func (c *Count) SetDB(db *pttdb.LDBBatch, dbPrefixID *types.PttID, dbID *types.PttID, dbPrefix []byte) {
c.db = db
c.dbPrefixID = dbPrefixID
c.dbID = dbID
c.dbPrefix = dbPrefix
}
func (c *Count) Save() error {
c.lock.RLock()
defer c.lock.RUnlock()
key, err := c.MarshalKey()
if err != nil {
return err
}
marshaled, err := c.Bits.Marshal()
if err != nil {
return err
}
err = c.db.DB().Put(key, marshaled)
if err != nil {
return err
}
return nil
}
func (c *Count) MarshalKey() ([]byte, error) {
return common.Concat([][]byte{c.dbPrefix, c.dbPrefixID[:], c.dbID[:]})
}
func (c *Count) Delete() error {
key, err := c.MarshalKey()
if err != nil {
return err
}
c.db.DB().Delete(key)
return nil
}
func (c *Count) Load() error {
c.lock.Lock()
defer c.lock.Unlock()
key, err := c.MarshalKey()
if err != nil {
return err
}
val, err := c.db.DBGet(key)
if err != nil {
return err
}
bits, err := types.UnmarshalBitVector(val)
if err != nil {
return err
}
c.SetBits(bits)
return nil
}
/*
Return: isNew
*/
func (c *Count) AddWithIsNew(item []byte) bool {
c.lock.Lock()
defer c.lock.Unlock()
c.hash.Reset()
c.hash.Write(item)
hash := c.hash.Sum64()
bucket := hash >> (64 - c.p) // top p bits are the bucket
return c.Bits.SetWithIsNew(bucket)
}
/*
Return: isNew
*/
func (c *Count) Add(item []byte) {
c.lock.Lock()
defer c.lock.Unlock()
c.hash.Reset()
c.hash.Write(item)
hash := c.hash.Sum64()
bucket := hash >> (64 - c.p) // top p bits are the bucket
c.Bits.Set(bucket)
}
// Distinct returns the estimate of the number of distinct elements seen
// if the backing BitVector is full it returns m, the size of the BitVector
func (c *Count) Count() uint64 {
c.lock.RLock()
defer c.lock.RUnlock()
zeroCount := c.m - c.Bits.PopCount()
if zeroCount > 0 {
return uint64(float64(c.m) * math.Log(float64(c.m)/float64(zeroCount)))
}
return (1 << c.p)
}
// Union the estimate of two LinearCounting reducing the precision to the minimum of the two sets
// the function will return nil and an error if the hash functions mismatch
func (c *Count) Union(c2 *Count) (*Count, error) {
c.lock.RLock()
defer c.lock.RUnlock()
c2.lock.RLock()
defer c2.lock.RUnlock()
// for each bucket take the OR of the two LinearCounting
combinedC, err := NewCount(c.db, c.dbPrefixID, c.dbID, c.dbPrefix, c.p, true)
if err != nil {
return nil, err
}
for i := range combinedC.Bits {
combinedC.Bits[i] = c.Bits[i] | c2.Bits[i]
}
return combinedC, nil
}
// Intersect the estimate of two LinearCounting reducing the precision to the minimum of the two sets
// the function will return nil and an error if the hash functions mismatch
func (c *Count) Intersect(c2 *Count) (*Count, error) {
c.lock.RLock()
defer c.lock.RUnlock()
c2.lock.RLock()
defer c2.lock.RUnlock()
// for each bucket take the AND of the two LinearCounting
combinedC, err := NewCount(c.db, c.dbPrefixID, c.dbID, c.dbPrefix, c.p, true)
if err != nil {
return nil, err
}
for i := range combinedC.Bits {
combinedC.Bits[i] = c.Bits[i] & c2.Bits[i]
}
return combinedC, nil
}