| 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
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916 |
1
1
1
1
1
1
1
30
30
30
30
30
1
1
1
1
6
6
6
6
3
3
6
6
6
6
6
6
17
2
15
15
15
15
33
33
33
33
22
22
11
11
6
6
5
6
6
6
1
14
14
14
14
14
14
14
1
1
22
22
74
64
10
22
1
2
1
6
6
2
61
61
4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
351
351
16
15
16
15
15
15
510
90
420
420
285
285
420
15
15
16
10
10
10
10
10
1
9
9
9
9
9
2
7
7
7
2
2
11
11
11
11
11
200
10
190
47
47
42
47
47
143
11
3
3
3
3
3
3
3
1
1
3
3
3
12
6
3
3
6
6
6
22
6
16
94
94
14
80
10
10
10
10
6
2
2
2
2
2
2
4
4
6
6
6
6
6
23
5
18
108
108
15
15
1
1
1
14
14
6
6
6
25
6
6
6
6
6
3
3
3
3
12
12
12
11
11
1
12
10
10
10
10
10
2
2
2
2
4
4
2
2
2
2
2
4
1
| "use strict";
/* globals localStorage */
var Q = require("q");
var FieldDBObject = require("../FieldDBObject").FieldDBObject;
var mapReduceFactory = require("./../map_reduce/MapReduce").MapReduceFactory;
var Lexicon = require("../lexicon/Lexicon").Lexicon;
// Load n grams map reduce which is used in both couchdb and in the codebase
var morphemeNGramsMapString = require("../../map_reduce_data/views/morpheme_n_grams/map").morpheme_n_grams;
var MORPHEMES_N_GRAMS_MAP_REDUCE = mapReduceFactory({
filename: "morpheme_n_grams",
mapString: morphemeNGramsMapString
});
/**
* @class The Glosser is able to guess morpheme segmentation and/or glosses from an orthography/transcription.
* The default Glosser (this one) uses a lexicon which is generated off of existing data in the corpus to do this.
*
* The Glosser can display a visual representation of itself if D3 is provided to it as options
* or D3 is loaded in the context of the app.
*
* @name Glosser
* @extends FieldDBObject
* @constructs
*/
var Glosser = function Glosser(options) {
Eif (!this._fieldDBtype) {
this._fieldDBtype = "Glosser";
}
options = options || {};
this.debug("Constructing Glosser length: ", options);
FieldDBObject.apply(this, arguments);
};
Glosser.morphemeBoundaryRegEX = /[-=]/g;
Glosser.Lexicon = Lexicon;
/**
* [morpheme_n_grams_mapReduce description]
* @type {Function}
*/
Glosser.morpheme_n_grams_mapReduce = MORPHEMES_N_GRAMS_MAP_REDUCE;
/**
* Finds all combinations of an utterance line by looping through all
* possible parses of all words in the utterance line.
*
* This results in a list of alternateMorphemeLines which can be shown to the user
* as a type-ahead on the utterance line.
*
* These alternateMorphemeLines can be further filtered using other information
* (likelyhood of long distance relationships between morphemes,
* likelyhood of long distance relationships between glosses).
*
* http://stackoverflow.com/questions/15298912/javascript-generating-combinations-from-n-arrays-with-m-elements
*
* @param {Object} options An object containing metadata,
* options.columns is used to generate the combinations
* options.sort is optionally used to sort the resulting combinations
* @return {Object} [description]
*/
Glosser.calculateAllAlternativeCombinations = function(options) {
var debug = function() {
// console.log(arguments);
};
options.alternateMorphemeLines = options.alternateMorphemeLines || [];
debug("Beginning with a set of alternatives " + options.alternateMorphemeLines);
if (options.alternateMorphemeSegmentationsByWord) {
options.columns = [];
for (var word in options.alternateMorphemeSegmentationsByWord) {
Eif (options.alternateMorphemeSegmentationsByWord.hasOwnProperty(word) && options.alternateMorphemeSegmentationsByWord[word].fullParses) {
options.columns.push(options.alternateMorphemeSegmentationsByWord[word].fullParses);
}
}
}
options.columns = options.columns || [];
var utteranceWordCount = options.columns.length - 1;
debug("Finding all alternative parses for ", options.columns);
var recurse = function(parseInProgress, currentColumnIndex) {
if (!options.columns[currentColumnIndex]) {
return;
}
Glosser.removeRedundantCopies(options.columns[currentColumnIndex]);
var numberOfParsesForThisWord = options.columns[currentColumnIndex].length;
debug(" There are " + numberOfParsesForThisWord + " alternative parses for word " + (currentColumnIndex + 1));
for (var possibleParseIndex = 0; possibleParseIndex < numberOfParsesForThisWord; possibleParseIndex++) {
var copyOfParseInProgress = parseInProgress.slice(0); // clone column
var possibleParseOfThisWord = options.columns[currentColumnIndex][possibleParseIndex];
copyOfParseInProgress.push(possibleParseOfThisWord);
if (currentColumnIndex === utteranceWordCount) {
// We have a full parse of all words
debug(" found a complete alternate", copyOfParseInProgress);
options.alternateMorphemeLines.push(copyOfParseInProgress.join(" "));
} else {
debug("\n" + copyOfParseInProgress + ": set");
recurse(copyOfParseInProgress, currentColumnIndex + 1);
}
}
};
recurse([], 0);
if (typeof options.sort === "function") {
options.alternateMorphemeLines = options.alternateMorphemeLines.sort(options.sort);
}
Glosser.removeRedundantCopies(options.alternateMorphemeLines);
debug(options.alternateMorphemeLines);
return options;
};
/**
* If passed to a alternateMorphemeLines.map() results in a sort order where more
* segmentations are listed first (on top in a typeahead).
*
* This is only useful if the user doesn't realize that the app can segment
* the utterance line, and/or if the glosser's segmentations are conservative
* and only offer real segmentations.
*
* @param {String} a A string (normallly a morphemes line)
* @param {String} b A string (normallly a morphemes line)
* @return {boolean} If a has more segmentations than b
*/
Glosser.sortAlternatesByMoreSegmentation = function(a, b) {
var segmentationsA = 0;
Eif (a && typeof a.split === "function") {
segmentationsA = a.split(Glosser.morphemeBoundaryRegEX).length;
}
var segmentationsB = 0;
Eif (b && typeof b.split === "function") {
segmentationsB = b.split(Glosser.morphemeBoundaryRegEX).length;
}
return segmentationsB - segmentationsA;
};
/**
* If passed to a alternateMorphemeLines.map() results in a sort order where fewer segmentations
* are listed first (on top in a typeahead).
*
* This is useful if the user prefers to see segmentations, but will most often
* choose the utteracne line as the morphemes line and then add the segmentations
* themself.
*
* @param {String} a A string (normallly a morphemes line)
* @param {String} b A string (normallly a morphemes line)
* @return {boolean} If a has more segmentations than b
*/
Glosser.sortAlternatesByLessSegmentation = function(a, b) {
var segmentationsA = 0;
if (a && typeof a.split === "function") {
segmentationsA = a.split(Glosser.morphemeBoundaryRegEX).length;
}
var segmentationsB = 0;
if (b && typeof b.split === "function") {
segmentationsB = b.split(Glosser.morphemeBoundaryRegEX).length;
}
return segmentationsA - segmentationsB;
};
/**
* Removes redundant copies from an array
*
* @param {Array} input An array to be cleaned
* @return {Array} The same array after removing redundant copies
*/
Glosser.removeRedundantCopies = function(input) {
var hash = {};
// var copy = [];
for (var i = input.length - 1; i >= 0; --i) {
if (!hash.hasOwnProperty(input[i])) {
hash[input[i]] = true;
// copy.unshift(input[i]);
} else {
input.splice(i, 1);
}
}
return input;
};
Glosser.removeWordBoundaryPlaceholders = function(input) {
return input.replace(/-?@-?/g, " ").trim();
};
Glosser.prototype = Object.create(FieldDBObject.prototype, /** @lends Glosser.prototype */ {
constructor: {
value: Glosser
},
// Internal models: used by the parse function
INTERNAL_MODELS: {
value: {
lexicon: Lexicon,
}
},
dbname: {
get: function() {
Iif (this._dbname) {
return this._dbname;
}
if (this.corpus && this.corpus.dbname) {
return this.corpus.dbname;
}
},
set: function(value) {
this._dbname = value;
}
},
lexicon: {
get: function() {
this.debug("getting lexicon");
return this._lexicon;
},
set: function(value) {
this.ensureSetViaAppropriateType("lexicon", value);
}
},
downloadPrecedenceRules: {
value: function(glosserURL) {
this.warn("DEPRECATED downloadPrecedenceRules use fetch instead");
return this.fetch({
url: glosserURL
});
}
},
fetch: {
value: function(options) {
Iif (this.fetching && this.whenFetched) {
return this.whenFetched;
}
var self = this;
var deferred = Q.defer();
this.whenFetched = deferred.promise;
options = options || {};
var glosserURL = "";
Iif (options && options.url) {
glosserURL = options.url;
}
Eif (!glosserURL || glosserURL === "default") {
var error;
Iif (!this.dbname && !options.dbname) {
error = {
userFriendlyErrors: ["Glosser's webservice can't be guessed, there is no current database identifier so the URL must be defined."]
};
}
Iif (!this.corpus) {
error = {
userFriendlyErrors: ["Glosser's webservice can't be guessed, there is no corpus loaded so the URL must be defined."]
};
}
Iif (error) {
deferred.reject(error);
return this.whenFetched;
}
Iif (this.corpus.prefs && this.corpus.prefs.glosserURL) {
glosserURL = this.corpus.prefs.glosserURL;
} else {
glosserURL = this.corpus.url + "/_design/data/_view/" + MORPHEMES_N_GRAMS_MAP_REDUCE.filename + "?group=true";
}
}
this.fetching = true;
Lexicon.CORS.makeCORSRequest({
type: "GET",
url: glosserURL
})
.then(function(reducedCouchDBresult) {
if (!reducedCouchDBresult || !reducedCouchDBresult.rows || reducedCouchDBresult.rows.length === undefined) {
self.fetching = false;
deferred.reject({
details: reducedCouchDBresult,
userFriendlyErrors: ["The result from the server " + glosserURL + " was not a standard response. Please report this."]
});
return;
}
// Reduce the reducedCouchDBresult such that reducedCouchDBresult which are found in multiple source
// words are only used/included once.
self.morphemeSegmentationKnowledgeBase = reducedCouchDBresult;
self.fetching = false;
// Save the reduced precedence self.morphemeSegmentationKnowledgeBase in localStorage
if (self.dbname) {
try {
localStorage.setItem(self.dbname + "morphemeSegmentationKnowledgeBase", JSON.stringify(self.morphemeSegmentationKnowledgeBase));
} catch (error) {
if (error.message === "localStorage is not defined") {
console.warn("cannot save precedence self.morphemeSegmentationKnowledgeBase for offline use");
} else {
//remove other corpora's self.morphemeSegmentationKnowledgeBase to make space for the most recent...
for (var x in localStorage) {
if (x.indexOf("self.morphemeSegmentationKnowledgeBase") > -1) {
localStorage.removeItem(x);
}
}
try {
localStorage.setItem(self.dbname + "self.morphemeSegmentationKnowledgeBase", JSON.stringify(self.morphemeSegmentationKnowledgeBase));
} catch (error2) {
self.warn("Your lexicon is huge!", error2);
}
}
}
}
deferred.resolve(self.morphemeSegmentationKnowledgeBase);
}, function(reason) {
self.fetching = false;
self.debug("Error getting precedence self.morphemeSegmentationKnowledgeBase:", reason);
// self.bug(reason.userFriendlyErrors.join(" "));
deferred.reject(reason);
})
.fail(function(exception) {
self.fetching = false;
self.warn("Exception getting precedence self.morphemeSegmentationKnowledgeBase:", exception.stack);
self.bug("Exception getting precedence rules in the glosser.");
deferred.reject(exception);
});
return this.whenFetched;
}
},
morphemeSegmentationKnowledgeBase: {
get: function() {
Iif (!this._morphemeSegmentationKnowledgeBase && (!this.lastMorphemePrecedenceRelationsLookupTimestamp && Date.now() - this.lastMorphemePrecedenceRelationsLookupTimestamp > 30000) && this.corpus && this.corpus.this.dbname) {
try {
var value = localStorage.getItme(this.corpus.this.dbname + "morphemeSegmentationKnowledgeBase");
if (value) {
value = JSON.parse(value);
}
if (value) {
this.morphemeSegmentationKnowledgeBase = value;
}
this.lastMorphemePrecedenceRelationsLookupTimestamp = Date.now();
} catch (exception) {
console.warn("Glosser's morpheme segmentation knowledge cannot be persisted across page loads for offline use.", exception);
this.lastMorphemePrecedenceRelationsLookupTimestamp = Infinity;
}
}
return this._morphemeSegmentationKnowledgeBase;
},
set: function(value) {
if (value && value.rows && typeof value.rows.filter === "function") {
value = value.rows;
}
if (value && typeof value.filter === "function") {
// Reduce the rules such that rules which are found in multiple source
// words are only used/included once.
var uniqueRules = {};
var size = 0;
value = value.map(function(row) {
/* TODO handle lexicon precedecne rules instead of simple condensed rules */
if (row.key && typeof row.key.indexOf === "function" && row.key.indexOf("@-@") > -1) {
return;
}
// If all ready reduced, use the count as the value
Iif (typeof row.value === "number") {
uniqueRules[row.key] = row.value;
size++;
} else {
// If not ready reduced, add the value into an array which will serve as the count
if (!uniqueRules[row.key]) {
uniqueRules[row.key] = uniqueRules[row.key] || [];
size++;
}
uniqueRules[row.key].push(row.value);
}
});
uniqueRules.length = size;
value = uniqueRules;
}
this._morphemeSegmentationKnowledgeBase = value;
}
},
/**
* Takes in an utterance line and, based on our current set of precendence
* self.morphemeSegmentationKnowledgeBase, guesses what the morpheme line would be. The algorithm is
* very conservative.
*
* @param {String} unparsedUtterance The raw utterance line.
*
* @return {String} The guessed morphemes line.
*/
morphemefinder: {
value: function(datum, justCopyDontGuessIGT) {
Iif (!datum || !datum.utterance) {
return datum;
}
// Keep the current morphemes as an alternate, and copy the utterance into the morphemes if it was empty
Iif (datum.morphemes) {
datum.alternateMorphemeLines = datum.alternateMorphemeLines || [];
datum.alternateMorphemeLines.unshift(datum.morphemes);
Glosser.removeRedundantCopies(datum.alternateMorphemeLines);
} else {
datum.morphemes = datum.utterance + "";
}
// If the user doesnt want to guess on this call, this datum, or any glosser data, dont bother guessing segmentation
Iif (justCopyDontGuessIGT || this.justCopyDontGuessIGT || datum.justCopyDontGuessIGT) {
return datum;
}
// If there is no data, dont bother guessing segmentation
if (!this.morphemeSegmentationKnowledgeBase || !this.morphemeSegmentationKnowledgeBase.length) {
return datum;
}
// Divide the utterance line into words
datum.utteranceWithExplictWordBoundaries = "@" + datum.utterance.trim().toLocaleLowerCase().split(/\s+/).join("@") + "@";
datum.matchingRules = [];
this.findRelevantSegmentationContexts(datum);
this.debug(" Found matchingRules " + datum.utterance + " which conservatively yields exact segmentation matches " + datum.utteranceWithExplictWordBoundaries, datum.matchingRules);
if (this.conservative === false) {
console.warn(" Continuing to look harder for segmentation. Currently " + datum.utterance + " directly segments as " + datum.utteranceWithExplictWordBoundaries);
} else {
this.debug(" Not looking harder for segmentation. Currently " + datum.utterance + " directly segments as " + datum.utteranceWithExplictWordBoundaries);
datum.morphemes = datum.utteranceWithExplictWordBoundaries.replace(/-?@-?/g, " ").trim();
return datum;
}
this.findaAllPossibleMorphemeLines(datum);
return datum;
}
},
/**
* Find the this.morphemeSegmentationKnowledgeBase which match in local precedence
* @param {Object} options [description]
* @return {Object} options after processing [description]
*/
findRelevantSegmentationContexts: {
value: function(options) {
Iif (!options.utteranceWithExplictWordBoundaries) {
return options;
}
this.debug("Looking for matching rules " + options.utteranceWithExplictWordBoundaries);
Iif (!options.matchingRules) {
options.matchingRules = [];
}
var originalUtteranceWithWordBoundaries = options.utteranceWithExplictWordBoundaries + "";
for (var segmentation in this.morphemeSegmentationKnowledgeBase) {
if (!this.morphemeSegmentationKnowledgeBase.hasOwnProperty(segmentation) || segmentation === "length") {
continue;
}
if (originalUtteranceWithWordBoundaries.indexOf(segmentation.replace(Glosser.morphemeBoundaryRegEX, "")) >= 0) {
this.debug(" matching rule " + originalUtteranceWithWordBoundaries + " <-" + segmentation);
// If this has enough context then consider it a correct segmentation
if (segmentation.split(Glosser.morphemeBoundaryRegEX).length > 2) {
options.utteranceWithExplictWordBoundaries = options.utteranceWithExplictWordBoundaries.replace(segmentation.replace(Glosser.morphemeBoundaryRegEX, ""), segmentation);
}
this.debug(" now " + options.utteranceWithExplictWordBoundaries);
options.matchingRules.push({
segmentation: segmentation,
contexts: this.morphemeSegmentationKnowledgeBase[segmentation],
morphemes: segmentation.split(Glosser.morphemeBoundaryRegEX)
});
} else {
this.debug(" not matching rule " + originalUtteranceWithWordBoundaries + " |- " + segmentation);
}
}
return options;
}
},
findaAllPossibleMorphemeLines: {
value: function(options) {
// Attempt to find the longest template which the matching this.morphemeSegmentationKnowledgeBase can
// generate from start to end
options.alternateMorphemeLines = options.alternateMorphemeLines || [];
options.usedRules = options.usedRules || [];
Iif (!options.utteranceWithExplictWordBoundaries) {
return options;
}
options.utterance = options.utterance || Glosser.removeWordBoundaryPlaceholders(options.utteranceWithExplictWordBoundaries);
options.alternateMorphemeLines.push(options.utterance);
this.debug("alternateMorphemeLines " + options.alternateMorphemeLines);
if (!options.matchingRules || !options.matchingRules.length) {
this.findRelevantSegmentationContexts(options);
options.alternateMorphemeLines.push(Glosser.removeWordBoundaryPlaceholders(options.utteranceWithExplictWordBoundaries));
}
Eif (!options.alternateMorphemeSegmentationsByWord) {
options.alternateMorphemeSegmentationsByWord = {};
options.utteranceWithExplictWordBoundaries.split(/-?@-?/).map(function(word) {
if (word) {
options.alternateMorphemeSegmentationsByWord[word.replace(Glosser.morphemeBoundaryRegEX, "")] = {
fullParses: [word],
prefixTemplates: [],
suffixTemplates: []
};
}
});
}
var match,
matchIndex;
for (var word in options.alternateMorphemeSegmentationsByWord) {
this.debug("\nFinding prefixtemplate " + word);
var prefixtemplate = ["@"];
// up to 10 morphemes long
for (var i = 0; i < 10; i++) {
if (prefixtemplate[i] === undefined) {
break;
}
// Look at each rule
for (matchIndex = 0; matchIndex < options.matchingRules.length; matchIndex++) {
match = options.matchingRules[matchIndex];
if (prefixtemplate.length > 2 && prefixtemplate[prefixtemplate.length - 1] === "@") {
continue; //we found the end already
}
// this morpheme matches, and the following morpheme can be found somewhere later in the word
if (prefixtemplate[i] === match.morphemes[0] && (match.morphemes[1] === "@" || word.indexOf(match.morphemes[1]) > -1)) {
options.usedRules.push(match);
Iif (prefixtemplate[i + 1]) { // ambiguity (two potential following morphemes)
this.debug("Ambiguity point for prefixes " + word + " " + prefixtemplate[i + 1], options.alternateMorphemeSegmentationsByWord[word]);
options.alternateMorphemeSegmentationsByWord[word].prefixTemplates.push(prefixtemplate.pop());
break;
} else {
prefixtemplate[i + 1] = match.morphemes[1];
this.debug(" prefixtemplate " + prefixtemplate.join("-"));
}
}
}
}
// If prefix template matches the whole word (ie we have seen it before, return the prefix template exactly)
if (prefixtemplate.length > 2 && prefixtemplate[0] === "@" && prefixtemplate[prefixtemplate.length - 1] === "@") {
this.debug("Prefix matches entire word. " + prefixtemplate.join("-"));
prefixtemplate.pop(); //remove final @
prefixtemplate.shift(); //remove intitial @
options.alternateMorphemeSegmentationsByWord[word].fullParses.push(prefixtemplate.join("-"));
prefixtemplate = options.alternateMorphemeSegmentationsByWord[word].prefixTemplates.pop() || ["@"];
this.debug("Considering another parse " + prefixtemplate.join("-"));
} else {
prefixtemplate = options.alternateMorphemeSegmentationsByWord[word].prefixTemplates.pop() || ["@"];
this.debug("Considering another parse " + prefixtemplate.join("-"));
}
// If the prefix template hit ambiguity in the middle, try from the suffix
// in until it hits ambiguity
var suffixtemplate = [];
Eif (prefixtemplate[prefixtemplate.length - 1] !== "@" || prefixtemplate.length === 1) {
// Suffix:
suffixtemplate.push("@");
this.debug(" suffixtemplate " + suffixtemplate.join("-"));
for (var ii = 0; ii < 10; ii++) {
if (suffixtemplate[ii] === undefined) {
break;
}
for (var suffixMatchIndex = 0; suffixMatchIndex < options.matchingRules.length; suffixMatchIndex++) {
var suffixmatch = options.matchingRules[suffixMatchIndex];
// this morpheme matches, and the following morpheme can be found somewhere later in the word
if (suffixtemplate[ii] === suffixmatch.morphemes[1] && (suffixmatch.morphemes[0] === "@" || word.indexOf(suffixmatch.morphemes[0]) > -1)) {
options.usedRules.push(suffixmatch);
if (suffixtemplate[ii + 1]) { // ambiguity (two potential
// following morphemes)
this.debug("Ambiguity point for suffixes " + word + " " + suffixtemplate[ii + 1], options.alternateMorphemeSegmentationsByWord[word]);
options.alternateMorphemeSegmentationsByWord[word].suffixTemplates.push(suffixtemplate.pop());
break;
} else {
suffixtemplate[ii + 1] = suffixmatch.morphemes[0];
this.debug(" suffixtemplate " + suffixtemplate.join("-"));
}
}
}
}
} else {
this.debug("what is this condition");
}
// Combine prefix and suffix templates into one regular expression which
// can be tested against the word to find a potential parse.
// Regular expressions will look something like
// (@)(.*)(hall)(.*)(o)(.*)(wa)(.*)(n)(.*)(@)
var template = [];
template = prefixtemplate.concat(suffixtemplate.reverse());
for (var slot in template) {
template[slot] = "(" + template[slot] + ")";
}
var regex = new RegExp(template.join("(.*)"), "");
this.debug("Considering regex ", regex);
// Use the regular expression to find a guessed morphemes line
var potentialParse = word
.replace(regex, "$1-$2-$3-$4-$5-$6-$7-$8-$9") // Use backreferences to parse into morphemes
.replace(/\$[0-9]/g, "") // Remove any backreferences that weren't used
.replace(/@/g, "") // Remove the start/end-of-line symbol
.replace(/--+/g, "-") // Ensure that there is only ever one "-" in a row
.replace(/^-/, "") // Remove "-" at the start of the word
.replace(/-$/, ""); // Remove "-" at the end of the word
this.debug("Potential parse of " + word.replace(/@/g, "") + " is " + potentialParse);
options.alternateMorphemeSegmentationsByWord[word].fullParses.push(potentialParse);
}
options.sort = Glosser.sortAlternatesByMoreSegmentation;
Glosser.calculateAllAlternativeCombinations(options);
// options.alternateMorphemeLines.push(fullParse);
// options.alternateMorphemeLines.push(parsedWords.join(" "));
options.morphemes = options.alternateMorphemeLines[0];
return options;
}
},
contextSensitiveGlossFinder: {
value: function(fields, justCopyDontGuessIGT) {
var morphemesLine = fields.morphemes;
Iif (!morphemesLine) {
return "";
}
if (justCopyDontGuessIGT || !this.lexicon) {
var justQuestionMarks = morphemesLine.trim().replace(/[^ =-]+/g, "?");
fields.gloss = justQuestionMarks;
} else {
fields = this.lexicon.guessContextSensitiveGlosses(fields);
}
return fields;
}
},
// simplisticGlossFinder: {
// value: function(morphemesLine, justCopyDontGuessIGT) {
// if (!morphemesLine) {
// return "";
// }
// if (justCopyDontGuessIGT) {
// var justQuestionMarks = morphemesLine.trim().replace(/[^ -]+/g, "?");
// return justQuestionMarks;
// }
// if (!this.dbname) {
// if (!this.application || !this.application.get("corpus")) {
// throw "Glosser can't be guessed, there is no app so the URL must be defined.";
// }
// this.dbname = this.application.get("corpus").get("couchConnection").this.dbname;
// }
// var lexiconNodes = localStorage.getItem(this.dbname + "lexiconResults");
// if (!lexiconNodes) {
// return "";
// }
// lexiconNodes = JSON.parse(lexiconNodes);
// var glossGroups = [];
// var morphemeGroup = morphemesLine.split(/\s+/);
// for (var group in morphemeGroup) {
// var morphemes = morphemeGroup[group].split("-");
// var glosses = [];
// for (var m in morphemes) {
// var matchingNodes = [];
// var node = lexiconNodes[morphemes[m]];
// if (node) {
// matchingNodes.push(node);
// }
// node = lexiconNodes[morphemes[m].toLowerCase()];
// if (node) {
// matchingNodes.push(node);
// }
// var gloss = "?"; // If there's no matching gloss, use question marks
// if (matchingNodes.length > 0) {
// // Take the first gloss for this morpheme
// this.debug("Glosses which match: ", matchingNodes);
// try {
// gloss = matchingNodes[0][0].gloss;
// } catch (e) {
// this.debug(matchingNodes);
// }
// }
// glosses.push(gloss);
// }
// glossGroups.push(glosses.join("-"));
// }
// // Replace the gloss line with the guessed glosses
// return glossGroups.join(" ");
// }
// },
guessUtteranceFromMorphemes: {
value: function(fields) {
if (!fields.utterance && fields.morphemes) {
fields.utterance = fields.morphemes.replace(/[-.]/g, "");
}
return fields;
}
},
guessMorphemesFromUtterance: {
value: function(fields, justCopyDontGuessIGT) {
Iif (fields.morphemes) {
return fields;
}
this.morphemefinder(fields, justCopyDontGuessIGT);
Eif (!fields.gloss) {
fields = this.contextSensitiveGlossFinder(fields, justCopyDontGuessIGT);
}
return fields;
}
},
guessGlossFromMorphemes: {
value: function(fields, justCopyDontGuessIGT) {
Iif (fields.gloss) {
return fields;
}
Iif (!fields.morphemes) {
fields.morphemes = fields.utterance;
}
fields = this.contextSensitiveGlossFinder(fields, justCopyDontGuessIGT);
return fields;
}
},
// glossFinder: {
// value: function(morphemesLine) {
// //Guess a gloss
// var morphemeGroup = morphemesLine.split(/\s+/);
// var glossGroups = [];
// if (!this.lexicon) {
// var corpusSize = 31; //TODO get corpus size another way. // app.get("corpus").datalists.models[app.get("corpus").datalists.models.length-1].get("datumIds").length;
// if (corpusSize > 30 && !Glosser.toastedUserToSync) {
// Glosser.toastedUserToSync = true;
// this.popup("You probably have enough data to train an autoglosser for your corpus.\n\nIf you sync your data with the team server then editing the morphemes will automatically run the auto glosser.", "alert-success", "Sync to train your auto-glosser:");
// } else {
// Glosser.toastedUserToImport++;
// if (Glosser.toastedUserToImport % 10 === 1 && corpusSize < 30) {
// this.popup("You have roughly " + corpusSize + " datum saved in your pouch, if you have around 30 datum, then you have enough data to train an autoglosser for your corpus.", "alert-info", "AutoGlosser:");
// }
// }
// }
// for (var group in morphemeGroup) {
// var morphemes = morphemeGroup[group].split("-");
// var glosses = [];
// for (var m in morphemes) {
// if (!this.lexicon) {
// glosses.push("?");
// continue;
// }
// console.log("looking for morpheme " + morphemes[m]);
// // Take the first gloss for this morpheme
// var matchingNode = this.lexicon.find("morphemes", morphemes[m]);
// var gloss = "?"; // If there"s no matching gloss, use question marks
// this.warn(matchingNode);
// if (matchingNode.length) {
// gloss = matchingNode[0].gloss;
// }
// glosses.push(gloss);
// }
// glossGroups.push(glosses.join("-"));
// }
// // Replace the gloss line with the guessed glosses
// return glossGroups.join(" ");
// }
// },
/**
* Takes as a parameters an array of self.morphemeSegmentationKnowledgeBase which came from CouchDB precedence rule query.
* Example Rule: {"key":{"x":"@","relation":"precedes","y":"aqtu","context":"aqtu-nay-wa-n"},"value":2}
*/
morphemeSegmentationKnowledgeBaseGraph: {
get: function() {
if (this._morphemeSegmentationKnowledgeBaseGraph) {
return this._morphemeSegmentationKnowledgeBaseGraph;
}
if (!this.morphemeSegmentationKnowledgeBase) {
return;
}
/*
* Cycle through the precedence this.morphemeSegmentationKnowledgeBase, convert them into graph edges with the morpheme index in the morpheme array as the source/target values
*/
var morphemeLinks = [];
var morphemes = [];
for (var i in this.morphemeSegmentationKnowledgeBase) {
/* make the @ more like what a linguist recognizes for word boundaries */
if (this.morphemeSegmentationKnowledgeBase[i].key.x === "@") {
this.morphemeSegmentationKnowledgeBase[i].key.x = "#_";
}
if (this.morphemeSegmentationKnowledgeBase[i].key.y === "@") {
this.morphemeSegmentationKnowledgeBase[i].key.y = "_#";
}
var xpos = morphemes.indexOf(this.morphemeSegmentationKnowledgeBase[i].key.x);
if (xpos < 0) {
morphemes.push(this.morphemeSegmentationKnowledgeBase[i].key.x);
xpos = morphemes.length - 1;
}
var ypos = morphemes.indexOf(this.morphemeSegmentationKnowledgeBase[i].key.y);
if (ypos < 0) {
morphemes.push(this.morphemeSegmentationKnowledgeBase[i].key.y);
ypos = morphemes.length - 1;
}
//To avoid loops?
if (this.morphemeSegmentationKnowledgeBase[i].key.y.indexOf("@") === -1) {
morphemeLinks.push({
source: xpos,
target: ypos,
value: 1 //TODO use the context counting to get a weight measure
});
}
}
/*
* Build the morphemes into nodes and color them by their morpheme length, could be a good measure of outliers
*/
var morphemenodes = [];
for (var m in morphemes) {
morphemenodes.push({
name: morphemes[m],
length: morphemes[m].length
});
}
/*
* Create the JSON required by D3
*/
this.morphemeSegmentationKnowledgeBaseGraph = {};
this.morphemeSegmentationKnowledgeBaseGraph.links = morphemeLinks;
this.morphemeSegmentationKnowledgeBaseGraph.nodes = morphemenodes;
return this.morphemeSegmentationKnowledgeBaseGraph;
}
},
saveAndInterConnectInApp: {
value: function(callback) {
if (typeof callback === "function") {
callback();
}
}
},
visualizeMorphemesAsForceDirectedGraph: {
value: function() {
this.warn("visualizeMorphemesAsForceDirectedGraph is deprecated use visualizePrecedenceRelationsAsForceDirectedGraph instead");
}
},
/*
* Show precedes relations in a connected graph
*
* http://bl.ocks.org/mbostock/1153292
* http://alignedleft.com/tutorials/d3/binding-data
*
* @param {Object} element Obligatory element where the render should go
* @return {Glosser} returns itself for chaining calls
*/
visualizePrecedenceRelationsAsForceDirectedGraph: {
value: function(options) {
options = options || {};
if (!this.lexicon || !this.lexicon.length) {
this.warn("Cannot visualize an empty lexicon.");
return this;
}
Iif (!this.lexicon.connectedGraph && !this.lexicon.entryRelations && this.morphemeSegmentationKnowledgeBase) {
this.lexicon.entryRelations = this.morphemeSegmentationKnowledgeBase;
}
this.lexicon.visualizeAsForceDirectedGraph(options);
return this;
}
},
render: {
value: function(options) {
return this.visualizePrecedenceRelationsAsForceDirectedGraph(options);
}
},
isWithinConfidenceRange: {
value: function(morpheme, confidenceRange) {
if (!confidenceRange) {
return true;
}
return morpheme.confidence <= confidenceRange.max && morpheme.confidence >= confidenceRange.min;
}
}
});
exports.Glosser = Glosser;
|