-
Notifications
You must be signed in to change notification settings - Fork 0
/
model.go
469 lines (425 loc) · 12.3 KB
/
model.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
package spell
import (
"regexp"
"runtime"
"sort"
"strings"
"sync"
"unicode/utf8"
)
/**
Model represents misspell corrector structure
*/
type Model struct {
Terms []string
TermsDict map[string]int
Index map[string][][]int
IndexTail map[string]map[string]bool
Affects [][][]int // inputlen -> edit len -> term lens
KnownAffects []bool
Depth int
IndexSplitLen int
TermsCounts []float64
TotalTerms float64
distanceMeasurers []*DistanceMeasurer `binary:"-"`
currentMeasurerId int `binary:"-"`
measurersSemaphore chan bool `binary:"-"`
measurersMutex sync.Mutex `binary:"-"`
}
func InitModel() *Model {
model := Model{
Terms: []string{},
TermsDict: map[string]int{},
Index: map[string][][]int{},
IndexTail: map[string]map[string]bool{},
Affects: [][][]int{},
KnownAffects: make([]bool, 15),
Depth: 2,
IndexSplitLen: DefaultIndexSplitLen,
TermsCounts: make([]float64, 0),
}
model.InitMeasurers()
return &model
}
func (model *Model) InitMeasurers() {
procCount := runtime.GOMAXPROCS(-1)
model.distanceMeasurers = make([]*DistanceMeasurer, procCount)
for i := 0; i < procCount; i++ {
model.distanceMeasurers[i] = NewDistanceMeasurer()
}
model.currentMeasurerId = 0
model.measurersSemaphore = make(chan bool, procCount)
model.measurersMutex = sync.Mutex{}
}
/**
Has term been added to the model
*/
func (model *Model) HasTerm(term string) bool {
_, ok := model.TermsDict[strings.ToLower(term)]
return ok
}
/**
Parse text, split it to terms and add them to the model
*/
func (model *Model) TrainText(text []byte) {
var (
termsRegex = regexp.MustCompile(`(?m)[\p{L}-]+`)
rawTermsB = termsRegex.FindAll(text, -1)
terms = map[string]float64{}
)
for _, termB := range rawTermsB {
if utf8.RuneCount(termB) < DefaultMinTermLen {
continue
}
term := string(termB)
term = strings.ToLower(term)
terms[term] += 1
}
for term, count := range terms {
if count >= DefaultMinTermCount {
model.AddTerm(term, count)
}
}
}
/**
Directly add terms to the model
*/
func (model *Model) TrainTerms(terms []string) {
for _, term := range terms {
model.AddTerm(term, 1)
}
}
/**
Add one term to the model
*/
func (model *Model) AddTerm(term string, count float64) bool {
var (
termLen = utf8.RuneCountInString(term)
ok = false
termId = 0
termLo = strings.ToLower(term)
)
if termLen < DefaultMinTermLen {
return false
}
model.TotalTerms += count
if termId, ok = model.TermsDict[termLo]; ok {
model.TermsCounts[termId] += count
return true
}
var (
termsByLen [][]int
termsIndex []int
termI = termLen - 1
)
termId = len(model.Terms)
model.Terms = append(model.Terms, termLo)
model.TermsCounts = append(model.TermsCounts, count)
model.TermsDict[termLo] = termId
edits := GetMultiEdits(termLo, 0.0, float64(model.Depth))
for edit := range edits {
editHead, editTail := model.splitEdit(edit)
if termsByLen, ok = model.Index[editHead]; !ok {
termsByLen = make([][]int, termLen)
model.Index[editHead] = termsByLen
} else {
if termLen > len(termsByLen) {
model.Index[editHead] = make([][]int, termLen)
copy(model.Index[editHead], termsByLen)
termsByLen = model.Index[editHead]
}
}
termsIndex = termsByLen[termI]
if termsIndex == nil {
termsIndex = []int{termId}
termsByLen[termI] = termsIndex
} else if termsIndex[len(termsIndex) - 1] != termId {
termsIndex = append(termsIndex, termId)
}
termsByLen[termI] = termsIndex
// edit tail
if editTail != "" {
tailToHeads := model.IndexTail[editTail]
if tailToHeads == nil {
tailToHeads = make(map[string]bool)
model.IndexTail[editTail] = tailToHeads
}
tailToHeads[editHead] = true
}
}
// fill known affects
if termLen > len(model.KnownAffects) {
knownAffects := make([]bool, termLen)
copy(knownAffects, model.KnownAffects)
model.KnownAffects = knownAffects
}
if !model.KnownAffects[termI] {
trackingMultiEdits := GetTrackingMultiEdits(termLo, OperationAffectedChange{0, map[int]bool{}}, float64(model.Depth))
for edit, trackingEdit := range trackingMultiEdits {
editLen := len(edit)
for inputDiff := range trackingEdit.InputLens {
inputLen := termLen + inputDiff
if inputLen > len(model.Affects) {
affects := make([][][]int, inputLen)
copy(affects, model.Affects)
model.Affects = affects
}
inputAffects := model.Affects[inputLen-1]
if inputAffects == nil {
inputAffects = make([][]int, termLen)
model.Affects[inputLen-1] = inputAffects
} else if editLen > len(inputAffects) {
inputAffectsCp := make([][]int, editLen)
copy(inputAffectsCp, inputAffects)
inputAffects = inputAffectsCp
model.Affects[inputLen-1] = inputAffectsCp
}
termsLens := model.Affects[inputLen-1][editLen-1]
if termsLens == nil {
termsLens = make([]int, 0, 10)
model.Affects[inputLen-1][editLen-1] = termsLens
}
index := sort.SearchInts(termsLens, termI)
if index == len(termsLens) || termsLens[index] != termI {
termsLens := append([]int{termI}, termsLens...)
sort.Ints(termsLens)
model.Affects[inputLen-1][editLen-1] = termsLens
}
}
}
model.KnownAffects[termLen-1] = true
}
return true
}
/**
Calculate raw unsorted suggestions
*/
func (model *Model) GetRawSuggestions(input string, calcEditorialPrescription bool) map[string]Suggestion {
result := make(map[string]Suggestion)
input = strings.ToLower(input)
var (
termsByLen [][]int
termsIndex []int
ok = false
inputLen = utf8.RuneCountInString(input)
editLen int
inputAffects [][]int
editAffects []int
term string
)
// todo add min input len check
// exact match
if termIndex, ok := model.TermsDict[input]; ok {
result[input] = Suggestion{
Term: input,
Distance: 0,
Score: 0,
Count: model.TermsCounts[termIndex],
}
}
// Index doesn't have any term that can be potentially mathed to input
if inputLen > len(model.Affects) {
return result
}
inputAffects = model.Affects[inputLen-1]
if inputAffects == nil {
return result
}
edits := GetMultiEdits(input, 0.0, float64(model.Depth))
for edit := range edits {
editHead, editTail := model.splitEdit(edit)
if termsByLen, ok = model.Index[editHead]; !ok {
continue
}
if editTail != "" && (model.IndexTail[editTail] == nil || !model.IndexTail[editTail][editHead]) {
continue
}
editLen = utf8.RuneCountInString(edit)
if editLen > len(inputAffects) {
continue
}
editAffects = inputAffects[editLen-1]
if editAffects == nil {
continue
}
for _, termI := range editAffects {
if termI >= len(termsByLen) {
continue
}
termsIndex = termsByLen[termI]
if termsIndex == nil {
continue
}
for _, termIndex := range termsIndex {
term = model.Terms[termIndex]
model.measurersSemaphore <- true
model.measurersMutex.Lock()
model.currentMeasurerId++
if model.currentMeasurerId == len(model.distanceMeasurers) {
model.currentMeasurerId = 0
}
measurer := model.distanceMeasurers[model.currentMeasurerId]
model.measurersMutex.Unlock()
distance, editorialPrescription := measurer.Distance(term, input, calcEditorialPrescription)
<- model.measurersSemaphore
if distance > model.Depth {
continue
}
if _, ok = result[term]; !ok {
result[term] = Suggestion{
Term: term,
Distance: distance,
Prescription: editorialPrescription,
Score: 0,
Count: model.TermsCounts[termIndex],
}
}
}
}
}
return result
}
/**
Return suggestions sorted by given scorer
*/
func (model *Model) GetSuggestions(input string, scoreModel ScoreModel, calcEditorialPrescription bool) []Suggestion {
var rawSuggestions = model.GetRawSuggestions(input, calcEditorialPrescription)
suggestions := make([]Suggestion, 0, len(rawSuggestions))
for _, suggestion := range rawSuggestions {
suggestion.Score = scoreModel.Score(&suggestion)
suggestions = append(suggestions, suggestion)
}
sort.Slice(suggestions, func(i, j int) bool {
return suggestions[i].Score < suggestions[j].Score
})
return suggestions
}
func (model *Model) splitEdit(edit string) (string, string) {
var (
editR = []rune(edit)
editRHead = editR
editHead = edit
editTail string
)
if len(editR) > model.IndexSplitLen {
editRHead = editR[:model.IndexSplitLen]
editHead = string(editRHead)
editTail = string(editR[model.IndexSplitLen:])
}
return editHead, editTail
}
func GetMultiEdits(term string, usedWeight float64, maxWeight float64) map[string]float64 {
edits := GetEdits(term, usedWeight, maxWeight)
if usedWeight < maxWeight {
traversalEdits := make(map[string]float64)
for k, v := range edits {
traversalEdits[k] = v
}
for term, weight := range traversalEdits {
subedits := GetMultiEdits(term, usedWeight+weight, maxWeight)
for subterm, v := range subedits {
edits[subterm] = v
}
}
}
return edits
}
func GetEdits(term string, usedWeight float64, maxWeight float64) map[string]float64 {
result := make(map[string]float64)
termR := []rune(term)
lenF := float64(len(termR))
for _, operationWeight := range OperationWeights {
if lenF-operationWeight.Weight < DefaultMinSpanningLen {
break
}
if usedWeight+operationWeight.Weight > maxWeight {
break
}
edit := string(termR[operationWeight.AffectedLen:])
if existingWeight, ok := result[edit]; !ok || existingWeight > operationWeight.Weight {
result[edit] = usedWeight + operationWeight.Weight
}
lenR := len(termR) - operationWeight.AffectedLen + 1
for i := 1; i < lenR; i++ {
edit := string(string(termR[0:i]) + string(termR[i+operationWeight.AffectedLen:]))
if existingWeight, ok := result[edit]; !ok || existingWeight > operationWeight.Weight {
result[edit] = usedWeight + operationWeight.Weight
}
}
}
return result;
}
func GetTrackingMultiEdits(term string, usedAffectedChange OperationAffectedChange, maxWeight float64) map[string]*OperationAffectedChange {
edits := GetTrackingEdits(term, usedAffectedChange, maxWeight)
if usedAffectedChange.Weight < maxWeight {
traversalEdits := make(map[string]*OperationAffectedChange)
for k, v := range edits {
traversalEdits[k] = v
}
for term, affectedChange := range traversalEdits {
u := OperationAffectedChange{
Weight: usedAffectedChange.Weight + affectedChange.Weight,
InputLens: map[int]bool{},
}
for l := range affectedChange.InputLens {
u.InputLens[l] = true
}
subedits := GetTrackingEdits(term, u, maxWeight)
for subterm, v := range subedits {
edits[subterm] = v
}
}
}
return edits
}
func GetTrackingEdits(term string, usedAffectedChange OperationAffectedChange, maxWeight float64) map[string]*OperationAffectedChange {
result := make(map[string]*OperationAffectedChange)
termR := []rune(term)
lenF := float64(len(termR))
for _, operationWeight := range CheckOperationWeight {
if lenF-operationWeight.Weight < DefaultMinSpanningLen {
break
}
if usedAffectedChange.Weight+operationWeight.Weight > maxWeight {
break
}
edit := string(termR[operationWeight.AffectedLen:])
if existingWeight, ok := result[edit]; !ok {
existingWeight = &OperationAffectedChange{
Weight: usedAffectedChange.Weight + operationWeight.Weight,
InputLens: map[int]bool{},
}
for _, l1 := range operationWeight.MisspellLens {
if len(usedAffectedChange.InputLens) > 0 {
for l2 := range usedAffectedChange.InputLens {
existingWeight.InputLens[l1+l2] = true
}
} else {
existingWeight.InputLens[l1] = true
}
}
result[edit] = existingWeight
}
lenR := len(termR) - operationWeight.AffectedLen + 1
for i := 1; i < lenR; i++ {
edit := string(termR[0:i]) + string(termR[i+operationWeight.AffectedLen:])
if existingWeight, ok := result[edit]; !ok {
existingWeight = &OperationAffectedChange{
Weight: usedAffectedChange.Weight + operationWeight.Weight,
InputLens: map[int]bool{},
}
for _, l1 := range operationWeight.MisspellLens {
if len(usedAffectedChange.InputLens) > 0 {
for l2 := range usedAffectedChange.InputLens {
existingWeight.InputLens[l1+l2] = true
}
} else {
existingWeight.InputLens[l1] = true
}
}
result[edit] = existingWeight
}
}
}
return result;
}