This repository was archived by the owner on Apr 12, 2019. It is now read-only.
forked from google/diff-match-patch
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathdiff_match_patch.lua
More file actions
2196 lines (2057 loc) · 72.1 KB
/
Copy pathdiff_match_patch.lua
File metadata and controls
2196 lines (2057 loc) · 72.1 KB
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
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
--[[
* Diff Match and Patch
* Copyright 2018 The diff-match-patch Authors.
* https://github.com/google/diff-match-patch
*
* Based on the JavaScript implementation by Neil Fraser.
* Ported to Lua by Duncan Cross.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
--]]
--[[
-- Lua 5.1 and earlier requires the external BitOp library.
-- This library is built-in from Lua 5.2 and later as 'bit32'.
require 'bit' -- <http://bitop.luajit.org/>
local band, bor, lshift
= bit.band, bit.bor, bit.lshift
--]]
local band, bor, lshift
= bit32.band, bit32.bor, bit32.lshift
local type, setmetatable, ipairs, select
= type, setmetatable, ipairs, select
local unpack, tonumber, error
= unpack, tonumber, error
local strsub, strbyte, strchar, gmatch, gsub
= string.sub, string.byte, string.char, string.gmatch, string.gsub
local strmatch, strfind, strformat
= string.match, string.find, string.format
local tinsert, tremove, tconcat
= table.insert, table.remove, table.concat
local max, min, floor, ceil, abs
= math.max, math.min, math.floor, math.ceil, math.abs
local clock = os.clock
-- Utility functions.
local percentEncode_pattern = '[^A-Za-z0-9%-=;\',./~!@#$%&*%(%)_%+ %?]'
local function percentEncode_replace(v)
return strformat('%%%02X', strbyte(v))
end
local function tsplice(t, idx, deletions, ...)
local insertions = select('#', ...)
for i = 1, deletions do
tremove(t, idx)
end
for i = insertions, 1, -1 do
-- do not remove parentheses around select
tinsert(t, idx, (select(i, ...)))
end
end
local function strelement(str, i)
return strsub(str, i, i)
end
local function indexOf(a, b, start)
if (#b == 0) then
return nil
end
return strfind(a, b, start, true)
end
local htmlEncode_pattern = '[&<>\n]'
local htmlEncode_replace = {
['&'] = '&', ['<'] = '<', ['>'] = '>', ['\n'] = '¶<br>'
}
-- Public API Functions
-- (Exported at the end of the script)
local diff_main,
diff_cleanupSemantic,
diff_cleanupEfficiency,
diff_levenshtein,
diff_prettyHtml
local match_main
local patch_make,
patch_toText,
patch_fromText,
patch_apply
--[[
* The data structure representing a diff is an array of tuples:
* {{DIFF_DELETE, 'Hello'}, {DIFF_INSERT, 'Goodbye'}, {DIFF_EQUAL, ' world.'}}
* which means: delete 'Hello', add 'Goodbye' and keep ' world.'
--]]
local DIFF_DELETE = -1
local DIFF_INSERT = 1
local DIFF_EQUAL = 0
-- Number of seconds to map a diff before giving up (0 for infinity).
local Diff_Timeout = 1.0
-- Cost of an empty edit operation in terms of edit characters.
local Diff_EditCost = 4
-- At what point is no match declared (0.0 = perfection, 1.0 = very loose).
local Match_Threshold = 0.5
-- How far to search for a match (0 = exact location, 1000+ = broad match).
-- A match this many characters away from the expected location will add
-- 1.0 to the score (0.0 is a perfect match).
local Match_Distance = 1000
-- When deleting a large block of text (over ~64 characters), how close do
-- the contents have to be to match the expected contents. (0.0 = perfection,
-- 1.0 = very loose). Note that Match_Threshold controls how closely the
-- end points of a delete need to match.
local Patch_DeleteThreshold = 0.5
-- Chunk size for context length.
local Patch_Margin = 4
-- The number of bits in an int.
local Match_MaxBits = 32
function settings(new)
if new then
Diff_Timeout = new.Diff_Timeout or Diff_Timeout
Diff_EditCost = new.Diff_EditCost or Diff_EditCost
Match_Threshold = new.Match_Threshold or Match_Threshold
Match_Distance = new.Match_Distance or Match_Distance
Patch_DeleteThreshold = new.Patch_DeleteThreshold or Patch_DeleteThreshold
Patch_Margin = new.Patch_Margin or Patch_Margin
Match_MaxBits = new.Match_MaxBits or Match_MaxBits
else
return {
Diff_Timeout = Diff_Timeout;
Diff_EditCost = Diff_EditCost;
Match_Threshold = Match_Threshold;
Match_Distance = Match_Distance;
Patch_DeleteThreshold = Patch_DeleteThreshold;
Patch_Margin = Patch_Margin;
Match_MaxBits = Match_MaxBits;
}
end
end
-- ---------------------------------------------------------------------------
-- DIFF API
-- ---------------------------------------------------------------------------
-- The private diff functions
local _diff_compute,
_diff_bisect,
_diff_halfMatchI,
_diff_halfMatch,
_diff_cleanupSemanticScore,
_diff_cleanupSemanticLossless,
_diff_cleanupMerge,
_diff_commonPrefix,
_diff_commonSuffix,
_diff_commonOverlap,
_diff_xIndex,
_diff_text1,
_diff_text2,
_diff_toDelta,
_diff_fromDelta
--[[
* Find the differences between two texts. Simplifies the problem by stripping
* any common prefix or suffix off the texts before diffing.
* @param {string} text1 Old string to be diffed.
* @param {string} text2 New string to be diffed.
* @param {boolean} opt_checklines Has no effect in Lua.
* @param {number} opt_deadline Optional time when the diff should be complete
* by. Used internally for recursive calls. Users should set DiffTimeout
* instead.
* @return {Array.<Array.<number|string>>} Array of diff tuples.
--]]
function diff_main(text1, text2, opt_checklines, opt_deadline)
-- Set a deadline by which time the diff must be complete.
if opt_deadline == nil then
if Diff_Timeout <= 0 then
opt_deadline = 2 ^ 31
else
opt_deadline = clock() + Diff_Timeout
end
end
local deadline = opt_deadline
-- Check for null inputs.
if text1 == nil or text1 == nil then
error('Null inputs. (diff_main)')
end
-- Check for equality (speedup).
if text1 == text2 then
if #text1 > 0 then
return {{DIFF_EQUAL, text1}}
end
return {}
end
-- LUANOTE: Due to the lack of Unicode support, Lua is incapable of
-- implementing the line-mode speedup.
local checklines = false
-- Trim off common prefix (speedup).
local commonlength = _diff_commonPrefix(text1, text2)
local commonprefix
if commonlength > 0 then
commonprefix = strsub(text1, 1, commonlength)
text1 = strsub(text1, commonlength + 1)
text2 = strsub(text2, commonlength + 1)
end
-- Trim off common suffix (speedup).
commonlength = _diff_commonSuffix(text1, text2)
local commonsuffix
if commonlength > 0 then
commonsuffix = strsub(text1, -commonlength)
text1 = strsub(text1, 1, -commonlength - 1)
text2 = strsub(text2, 1, -commonlength - 1)
end
-- Compute the diff on the middle block.
local diffs = _diff_compute(text1, text2, checklines, deadline)
-- Restore the prefix and suffix.
if commonprefix then
tinsert(diffs, 1, {DIFF_EQUAL, commonprefix})
end
if commonsuffix then
diffs[#diffs + 1] = {DIFF_EQUAL, commonsuffix}
end
_diff_cleanupMerge(diffs)
return diffs
end
--[[
* Reduce the number of edits by eliminating semantically trivial equalities.
* @param {Array.<Array.<number|string>>} diffs Array of diff tuples.
--]]
function diff_cleanupSemantic(diffs)
local changes = false
local equalities = {} -- Stack of indices where equalities are found.
local equalitiesLength = 0 -- Keeping our own length var is faster.
local lastequality = nil
-- Always equal to diffs[equalities[equalitiesLength]][2]
local pointer = 1 -- Index of current position.
-- Number of characters that changed prior to the equality.
local length_insertions1 = 0
local length_deletions1 = 0
-- Number of characters that changed after the equality.
local length_insertions2 = 0
local length_deletions2 = 0
while diffs[pointer] do
if diffs[pointer][1] == DIFF_EQUAL then -- Equality found.
equalitiesLength = equalitiesLength + 1
equalities[equalitiesLength] = pointer
length_insertions1 = length_insertions2
length_deletions1 = length_deletions2
length_insertions2 = 0
length_deletions2 = 0
lastequality = diffs[pointer][2]
else -- An insertion or deletion.
if diffs[pointer][1] == DIFF_INSERT then
length_insertions2 = length_insertions2 + #(diffs[pointer][2])
else
length_deletions2 = length_deletions2 + #(diffs[pointer][2])
end
-- Eliminate an equality that is smaller or equal to the edits on both
-- sides of it.
if lastequality
and (#lastequality <= max(length_insertions1, length_deletions1))
and (#lastequality <= max(length_insertions2, length_deletions2)) then
-- Duplicate record.
tinsert(diffs, equalities[equalitiesLength],
{DIFF_DELETE, lastequality})
-- Change second copy to insert.
diffs[equalities[equalitiesLength] + 1][1] = DIFF_INSERT
-- Throw away the equality we just deleted.
equalitiesLength = equalitiesLength - 1
-- Throw away the previous equality (it needs to be reevaluated).
equalitiesLength = equalitiesLength - 1
pointer = (equalitiesLength > 0) and equalities[equalitiesLength] or 0
length_insertions1, length_deletions1 = 0, 0 -- Reset the counters.
length_insertions2, length_deletions2 = 0, 0
lastequality = nil
changes = true
end
end
pointer = pointer + 1
end
-- Normalize the diff.
if changes then
_diff_cleanupMerge(diffs)
end
_diff_cleanupSemanticLossless(diffs)
-- Find any overlaps between deletions and insertions.
-- e.g: <del>abcxxx</del><ins>xxxdef</ins>
-- -> <del>abc</del>xxx<ins>def</ins>
-- e.g: <del>xxxabc</del><ins>defxxx</ins>
-- -> <ins>def</ins>xxx<del>abc</del>
-- Only extract an overlap if it is as big as the edit ahead or behind it.
pointer = 2
while diffs[pointer] do
if (diffs[pointer - 1][1] == DIFF_DELETE and
diffs[pointer][1] == DIFF_INSERT) then
local deletion = diffs[pointer - 1][2]
local insertion = diffs[pointer][2]
local overlap_length1 = _diff_commonOverlap(deletion, insertion)
local overlap_length2 = _diff_commonOverlap(insertion, deletion)
if (overlap_length1 >= overlap_length2) then
if (overlap_length1 >= #deletion / 2 or
overlap_length1 >= #insertion / 2) then
-- Overlap found. Insert an equality and trim the surrounding edits.
tinsert(diffs, pointer,
{DIFF_EQUAL, strsub(insertion, 1, overlap_length1)})
diffs[pointer - 1][2] =
strsub(deletion, 1, #deletion - overlap_length1)
diffs[pointer + 1][2] = strsub(insertion, overlap_length1 + 1)
pointer = pointer + 1
end
else
if (overlap_length2 >= #deletion / 2 or
overlap_length2 >= #insertion / 2) then
-- Reverse overlap found.
-- Insert an equality and swap and trim the surrounding edits.
tinsert(diffs, pointer,
{DIFF_EQUAL, strsub(deletion, 1, overlap_length2)})
diffs[pointer - 1] = {DIFF_INSERT,
strsub(insertion, 1, #insertion - overlap_length2)}
diffs[pointer + 1] = {DIFF_DELETE,
strsub(deletion, overlap_length2 + 1)}
pointer = pointer + 1
end
end
pointer = pointer + 1
end
pointer = pointer + 1
end
end
--[[
* Reduce the number of edits by eliminating operationally trivial equalities.
* @param {Array.<Array.<number|string>>} diffs Array of diff tuples.
--]]
function diff_cleanupEfficiency(diffs)
local changes = false
-- Stack of indices where equalities are found.
local equalities = {}
-- Keeping our own length var is faster.
local equalitiesLength = 0
-- Always equal to diffs[equalities[equalitiesLength]][2]
local lastequality = nil
-- Index of current position.
local pointer = 1
-- The following four are really booleans but are stored as numbers because
-- they are used at one point like this:
--
-- (pre_ins + pre_del + post_ins + post_del) == 3
--
-- ...i.e. checking that 3 of them are true and 1 of them is false.
-- Is there an insertion operation before the last equality.
local pre_ins = 0
-- Is there a deletion operation before the last equality.
local pre_del = 0
-- Is there an insertion operation after the last equality.
local post_ins = 0
-- Is there a deletion operation after the last equality.
local post_del = 0
while diffs[pointer] do
if diffs[pointer][1] == DIFF_EQUAL then -- Equality found.
local diffText = diffs[pointer][2]
if (#diffText < Diff_EditCost) and (post_ins == 1 or post_del == 1) then
-- Candidate found.
equalitiesLength = equalitiesLength + 1
equalities[equalitiesLength] = pointer
pre_ins, pre_del = post_ins, post_del
lastequality = diffText
else
-- Not a candidate, and can never become one.
equalitiesLength = 0
lastequality = nil
end
post_ins, post_del = 0, 0
else -- An insertion or deletion.
if diffs[pointer][1] == DIFF_DELETE then
post_del = 1
else
post_ins = 1
end
--[[
* Five types to be split:
* <ins>A</ins><del>B</del>XY<ins>C</ins><del>D</del>
* <ins>A</ins>X<ins>C</ins><del>D</del>
* <ins>A</ins><del>B</del>X<ins>C</ins>
* <ins>A</del>X<ins>C</ins><del>D</del>
* <ins>A</ins><del>B</del>X<del>C</del>
--]]
if lastequality and (
(pre_ins+pre_del+post_ins+post_del == 4)
or
(
(#lastequality < Diff_EditCost / 2)
and
(pre_ins+pre_del+post_ins+post_del == 3)
)) then
-- Duplicate record.
tinsert(diffs, equalities[equalitiesLength],
{DIFF_DELETE, lastequality})
-- Change second copy to insert.
diffs[equalities[equalitiesLength] + 1][1] = DIFF_INSERT
-- Throw away the equality we just deleted.
equalitiesLength = equalitiesLength - 1
lastequality = nil
if (pre_ins == 1) and (pre_del == 1) then
-- No changes made which could affect previous entry, keep going.
post_ins, post_del = 1, 1
equalitiesLength = 0
else
-- Throw away the previous equality.
equalitiesLength = equalitiesLength - 1
pointer = (equalitiesLength > 0) and equalities[equalitiesLength] or 0
post_ins, post_del = 0, 0
end
changes = true
end
end
pointer = pointer + 1
end
if changes then
_diff_cleanupMerge(diffs)
end
end
--[[
* Compute the Levenshtein distance; the number of inserted, deleted or
* substituted characters.
* @param {Array.<Array.<number|string>>} diffs Array of diff tuples.
* @return {number} Number of changes.
--]]
function diff_levenshtein(diffs)
local levenshtein = 0
local insertions, deletions = 0, 0
for x, diff in ipairs(diffs) do
local op, data = diff[1], diff[2]
if (op == DIFF_INSERT) then
insertions = insertions + #data
elseif (op == DIFF_DELETE) then
deletions = deletions + #data
elseif (op == DIFF_EQUAL) then
-- A deletion and an insertion is one substitution.
levenshtein = levenshtein + max(insertions, deletions)
insertions = 0
deletions = 0
end
end
levenshtein = levenshtein + max(insertions, deletions)
return levenshtein
end
--[[
* Convert a diff array into a pretty HTML report.
* @param {Array.<Array.<number|string>>} diffs Array of diff tuples.
* @return {string} HTML representation.
--]]
function diff_prettyHtml(diffs)
local html = {}
for x, diff in ipairs(diffs) do
local op = diff[1] -- Operation (insert, delete, equal)
local data = diff[2] -- Text of change.
local text = gsub(data, htmlEncode_pattern, htmlEncode_replace)
if op == DIFF_INSERT then
html[x] = '<ins style="background:#e6ffe6;">' .. text .. '</ins>'
elseif op == DIFF_DELETE then
html[x] = '<del style="background:#ffe6e6;">' .. text .. '</del>'
elseif op == DIFF_EQUAL then
html[x] = '<span>' .. text .. '</span>'
end
end
return tconcat(html)
end
-- ---------------------------------------------------------------------------
-- UNOFFICIAL/PRIVATE DIFF FUNCTIONS
-- ---------------------------------------------------------------------------
--[[
* Find the differences between two texts. Assumes that the texts do not
* have any common prefix or suffix.
* @param {string} text1 Old string to be diffed.
* @param {string} text2 New string to be diffed.
* @param {boolean} checklines Has no effect in Lua.
* @param {number} deadline Time when the diff should be complete by.
* @return {Array.<Array.<number|string>>} Array of diff tuples.
* @private
--]]
function _diff_compute(text1, text2, checklines, deadline)
if #text1 == 0 then
-- Just add some text (speedup).
return {{DIFF_INSERT, text2}}
end
if #text2 == 0 then
-- Just delete some text (speedup).
return {{DIFF_DELETE, text1}}
end
local diffs
local longtext = (#text1 > #text2) and text1 or text2
local shorttext = (#text1 > #text2) and text2 or text1
local i = indexOf(longtext, shorttext)
if i ~= nil then
-- Shorter text is inside the longer text (speedup).
diffs = {
{DIFF_INSERT, strsub(longtext, 1, i - 1)},
{DIFF_EQUAL, shorttext},
{DIFF_INSERT, strsub(longtext, i + #shorttext)}
}
-- Swap insertions for deletions if diff is reversed.
if #text1 > #text2 then
diffs[1][1], diffs[3][1] = DIFF_DELETE, DIFF_DELETE
end
return diffs
end
if #shorttext == 1 then
-- Single character string.
-- After the previous speedup, the character can't be an equality.
return {{DIFF_DELETE, text1}, {DIFF_INSERT, text2}}
end
-- Check to see if the problem can be split in two.
do
local
text1_a, text1_b,
text2_a, text2_b,
mid_common = _diff_halfMatch(text1, text2)
if text1_a then
-- A half-match was found, sort out the return data.
-- Send both pairs off for separate processing.
local diffs_a = diff_main(text1_a, text2_a, checklines, deadline)
local diffs_b = diff_main(text1_b, text2_b, checklines, deadline)
-- Merge the results.
local diffs_a_len = #diffs_a
diffs = diffs_a
diffs[diffs_a_len + 1] = {DIFF_EQUAL, mid_common}
for i, b_diff in ipairs(diffs_b) do
diffs[diffs_a_len + 1 + i] = b_diff
end
return diffs
end
end
return _diff_bisect(text1, text2, deadline)
end
--[[
* Find the 'middle snake' of a diff, split the problem in two
* and return the recursively constructed diff.
* See Myers 1986 paper: An O(ND) Difference Algorithm and Its Variations.
* @param {string} text1 Old string to be diffed.
* @param {string} text2 New string to be diffed.
* @param {number} deadline Time at which to bail if not yet complete.
* @return {Array.<Array.<number|string>>} Array of diff tuples.
* @private
--]]
function _diff_bisect(text1, text2, deadline)
-- Cache the text lengths to prevent multiple calls.
local text1_length = #text1
local text2_length = #text2
local _sub, _element
local max_d = ceil((text1_length + text2_length) / 2)
local v_offset = max_d
local v_length = 2 * max_d
local v1 = {}
local v2 = {}
-- Setting all elements to -1 is faster in Lua than mixing integers and nil.
for x = 0, v_length - 1 do
v1[x] = -1
v2[x] = -1
end
v1[v_offset + 1] = 0
v2[v_offset + 1] = 0
local delta = text1_length - text2_length
-- If the total number of characters is odd, then
-- the front path will collide with the reverse path.
local front = (delta % 2 ~= 0)
-- Offsets for start and end of k loop.
-- Prevents mapping of space beyond the grid.
local k1start = 0
local k1end = 0
local k2start = 0
local k2end = 0
for d = 0, max_d - 1 do
-- Bail out if deadline is reached.
if clock() > deadline then
break
end
-- Walk the front path one step.
for k1 = -d + k1start, d - k1end, 2 do
local k1_offset = v_offset + k1
local x1
if (k1 == -d) or ((k1 ~= d) and
(v1[k1_offset - 1] < v1[k1_offset + 1])) then
x1 = v1[k1_offset + 1]
else
x1 = v1[k1_offset - 1] + 1
end
local y1 = x1 - k1
while (x1 <= text1_length) and (y1 <= text2_length)
and (strelement(text1, x1) == strelement(text2, y1)) do
x1 = x1 + 1
y1 = y1 + 1
end
v1[k1_offset] = x1
if x1 > text1_length + 1 then
-- Ran off the right of the graph.
k1end = k1end + 2
elseif y1 > text2_length + 1 then
-- Ran off the bottom of the graph.
k1start = k1start + 2
elseif front then
local k2_offset = v_offset + delta - k1
if k2_offset >= 0 and k2_offset < v_length and v2[k2_offset] ~= -1 then
-- Mirror x2 onto top-left coordinate system.
local x2 = text1_length - v2[k2_offset] + 1
if x1 > x2 then
-- Overlap detected.
return _diff_bisectSplit(text1, text2, x1, y1, deadline)
end
end
end
end
-- Walk the reverse path one step.
for k2 = -d + k2start, d - k2end, 2 do
local k2_offset = v_offset + k2
local x2
if (k2 == -d) or ((k2 ~= d) and
(v2[k2_offset - 1] < v2[k2_offset + 1])) then
x2 = v2[k2_offset + 1]
else
x2 = v2[k2_offset - 1] + 1
end
local y2 = x2 - k2
while (x2 <= text1_length) and (y2 <= text2_length)
and (strelement(text1, -x2) == strelement(text2, -y2)) do
x2 = x2 + 1
y2 = y2 + 1
end
v2[k2_offset] = x2
if x2 > text1_length + 1 then
-- Ran off the left of the graph.
k2end = k2end + 2
elseif y2 > text2_length + 1 then
-- Ran off the top of the graph.
k2start = k2start + 2
elseif not front then
local k1_offset = v_offset + delta - k2
if k1_offset >= 0 and k1_offset < v_length and v1[k1_offset] ~= -1 then
local x1 = v1[k1_offset]
local y1 = v_offset + x1 - k1_offset
-- Mirror x2 onto top-left coordinate system.
x2 = text1_length - x2 + 1
if x1 > x2 then
-- Overlap detected.
return _diff_bisectSplit(text1, text2, x1, y1, deadline)
end
end
end
end
end
-- Diff took too long and hit the deadline or
-- number of diffs equals number of characters, no commonality at all.
return {{DIFF_DELETE, text1}, {DIFF_INSERT, text2}}
end
--[[
* Given the location of the 'middle snake', split the diff in two parts
* and recurse.
* @param {string} text1 Old string to be diffed.
* @param {string} text2 New string to be diffed.
* @param {number} x Index of split point in text1.
* @param {number} y Index of split point in text2.
* @param {number} deadline Time at which to bail if not yet complete.
* @return {Array.<Array.<number|string>>} Array of diff tuples.
* @private
--]]
function _diff_bisectSplit(text1, text2, x, y, deadline)
local text1a = strsub(text1, 1, x - 1)
local text2a = strsub(text2, 1, y - 1)
local text1b = strsub(text1, x)
local text2b = strsub(text2, y)
-- Compute both diffs serially.
local diffs = diff_main(text1a, text2a, false, deadline)
local diffsb = diff_main(text1b, text2b, false, deadline)
local diffs_len = #diffs
for i, v in ipairs(diffsb) do
diffs[diffs_len + i] = v
end
return diffs
end
--[[
* Determine the common prefix of two strings.
* @param {string} text1 First string.
* @param {string} text2 Second string.
* @return {number} The number of characters common to the start of each
* string.
--]]
function _diff_commonPrefix(text1, text2)
-- Quick check for common null cases.
if (#text1 == 0) or (#text2 == 0) or (strbyte(text1, 1) ~= strbyte(text2, 1))
then
return 0
end
-- Binary search.
-- Performance analysis: http://neil.fraser.name/news/2007/10/09/
local pointermin = 1
local pointermax = min(#text1, #text2)
local pointermid = pointermax
local pointerstart = 1
while (pointermin < pointermid) do
if (strsub(text1, pointerstart, pointermid)
== strsub(text2, pointerstart, pointermid)) then
pointermin = pointermid
pointerstart = pointermin
else
pointermax = pointermid
end
pointermid = floor(pointermin + (pointermax - pointermin) / 2)
end
return pointermid
end
--[[
* Determine the common suffix of two strings.
* @param {string} text1 First string.
* @param {string} text2 Second string.
* @return {number} The number of characters common to the end of each string.
--]]
function _diff_commonSuffix(text1, text2)
-- Quick check for common null cases.
if (#text1 == 0) or (#text2 == 0)
or (strbyte(text1, -1) ~= strbyte(text2, -1)) then
return 0
end
-- Binary search.
-- Performance analysis: http://neil.fraser.name/news/2007/10/09/
local pointermin = 1
local pointermax = min(#text1, #text2)
local pointermid = pointermax
local pointerend = 1
while (pointermin < pointermid) do
if (strsub(text1, -pointermid, -pointerend)
== strsub(text2, -pointermid, -pointerend)) then
pointermin = pointermid
pointerend = pointermin
else
pointermax = pointermid
end
pointermid = floor(pointermin + (pointermax - pointermin) / 2)
end
return pointermid
end
--[[
* Determine if the suffix of one string is the prefix of another.
* @param {string} text1 First string.
* @param {string} text2 Second string.
* @return {number} The number of characters common to the end of the first
* string and the start of the second string.
* @private
--]]
function _diff_commonOverlap(text1, text2)
-- Cache the text lengths to prevent multiple calls.
local text1_length = #text1
local text2_length = #text2
-- Eliminate the null case.
if text1_length == 0 or text2_length == 0 then
return 0
end
-- Truncate the longer string.
if text1_length > text2_length then
text1 = strsub(text1, text1_length - text2_length + 1)
elseif text1_length < text2_length then
text2 = strsub(text2, 1, text1_length)
end
local text_length = min(text1_length, text2_length)
-- Quick check for the worst case.
if text1 == text2 then
return text_length
end
-- Start by looking for a single character match
-- and increase length until no match is found.
-- Performance analysis: http://neil.fraser.name/news/2010/11/04/
local best = 0
local length = 1
while true do
local pattern = strsub(text1, text_length - length + 1)
local found = strfind(text2, pattern, 1, true)
if found == nil then
return best
end
length = length + found - 1
if found == 1 or strsub(text1, text_length - length + 1) ==
strsub(text2, 1, length) then
best = length
length = length + 1
end
end
end
--[[
* Does a substring of shorttext exist within longtext such that the substring
* is at least half the length of longtext?
* This speedup can produce non-minimal diffs.
* Closure, but does not reference any external variables.
* @param {string} longtext Longer string.
* @param {string} shorttext Shorter string.
* @param {number} i Start index of quarter length substring within longtext.
* @return {?Array.<string>} Five element Array, containing the prefix of
* longtext, the suffix of longtext, the prefix of shorttext, the suffix
* of shorttext and the common middle. Or nil if there was no match.
* @private
--]]
function _diff_halfMatchI(longtext, shorttext, i)
-- Start with a 1/4 length substring at position i as a seed.
local seed = strsub(longtext, i, i + floor(#longtext / 4))
local j = 0 -- LUANOTE: do not change to 1, was originally -1
local best_common = ''
local best_longtext_a, best_longtext_b, best_shorttext_a, best_shorttext_b
while true do
j = indexOf(shorttext, seed, j + 1)
if (j == nil) then
break
end
local prefixLength = _diff_commonPrefix(strsub(longtext, i),
strsub(shorttext, j))
local suffixLength = _diff_commonSuffix(strsub(longtext, 1, i - 1),
strsub(shorttext, 1, j - 1))
if #best_common < suffixLength + prefixLength then
best_common = strsub(shorttext, j - suffixLength, j - 1)
.. strsub(shorttext, j, j + prefixLength - 1)
best_longtext_a = strsub(longtext, 1, i - suffixLength - 1)
best_longtext_b = strsub(longtext, i + prefixLength)
best_shorttext_a = strsub(shorttext, 1, j - suffixLength - 1)
best_shorttext_b = strsub(shorttext, j + prefixLength)
end
end
if #best_common * 2 >= #longtext then
return {best_longtext_a, best_longtext_b,
best_shorttext_a, best_shorttext_b, best_common}
else
return nil
end
end
--[[
* Do the two texts share a substring which is at least half the length of the
* longer text?
* @param {string} text1 First string.
* @param {string} text2 Second string.
* @return {?Array.<string>} Five element Array, containing the prefix of
* text1, the suffix of text1, the prefix of text2, the suffix of
* text2 and the common middle. Or nil if there was no match.
* @private
--]]
function _diff_halfMatch(text1, text2)
if Diff_Timeout <= 0 then
-- Don't risk returning a non-optimal diff if we have unlimited time.
return nil
end
local longtext = (#text1 > #text2) and text1 or text2
local shorttext = (#text1 > #text2) and text2 or text1
if (#longtext < 4) or (#shorttext * 2 < #longtext) then
return nil -- Pointless.
end
-- First check if the second quarter is the seed for a half-match.
local hm1 = _diff_halfMatchI(longtext, shorttext, ceil(#longtext / 4))
-- Check again based on the third quarter.
local hm2 = _diff_halfMatchI(longtext, shorttext, ceil(#longtext / 2))
local hm
if not hm1 and not hm2 then
return nil
elseif not hm2 then
hm = hm1
elseif not hm1 then
hm = hm2
else
-- Both matched. Select the longest.
hm = (#hm1[5] > #hm2[5]) and hm1 or hm2
end
-- A half-match was found, sort out the return data.
local text1_a, text1_b, text2_a, text2_b
if (#text1 > #text2) then
text1_a, text1_b = hm[1], hm[2]
text2_a, text2_b = hm[3], hm[4]
else
text2_a, text2_b = hm[1], hm[2]
text1_a, text1_b = hm[3], hm[4]
end
local mid_common = hm[5]
return text1_a, text1_b, text2_a, text2_b, mid_common
end
--[[
* Given two strings, compute a score representing whether the internal
* boundary falls on logical boundaries.
* Scores range from 6 (best) to 0 (worst).
* @param {string} one First string.
* @param {string} two Second string.
* @return {number} The score.
* @private
--]]
function _diff_cleanupSemanticScore(one, two)
if (#one == 0) or (#two == 0) then
-- Edges are the best.
return 6
end
-- Each port of this function behaves slightly differently due to
-- subtle differences in each language's definition of things like
-- 'whitespace'. Since this function's purpose is largely cosmetic,
-- the choice has been made to use each language's native features
-- rather than force total conformity.
local char1 = strsub(one, -1)
local char2 = strsub(two, 1, 1)
local nonAlphaNumeric1 = strmatch(char1, '%W')
local nonAlphaNumeric2 = strmatch(char2, '%W')
local whitespace1 = nonAlphaNumeric1 and strmatch(char1, '%s')
local whitespace2 = nonAlphaNumeric2 and strmatch(char2, '%s')
local lineBreak1 = whitespace1 and strmatch(char1, '%c')
local lineBreak2 = whitespace2 and strmatch(char2, '%c')
local blankLine1 = lineBreak1 and strmatch(one, '\n\r?\n$')
local blankLine2 = lineBreak2 and strmatch(two, '^\r?\n\r?\n')
if blankLine1 or blankLine2 then
-- Five points for blank lines.
return 5
elseif lineBreak1 or lineBreak2 then
-- Four points for line breaks.
return 4
elseif nonAlphaNumeric1 and not whitespace1 and whitespace2 then
-- Three points for end of sentences.
return 3
elseif whitespace1 or whitespace2 then
-- Two points for whitespace.
return 2
elseif nonAlphaNumeric1 or nonAlphaNumeric2 then
-- One point for non-alphanumeric.
return 1
end
return 0
end
--[[
* Look for single edits surrounded on both sides by equalities
* which can be shifted sideways to align the edit to a word boundary.
* e.g: The c<ins>at c</ins>ame. -> The <ins>cat </ins>came.
* @param {Array.<Array.<number|string>>} diffs Array of diff tuples.
--]]
function _diff_cleanupSemanticLossless(diffs)
local pointer = 2
-- Intentionally ignore the first and last element (don't need checking).
while diffs[pointer + 1] do
local prevDiff, nextDiff = diffs[pointer - 1], diffs[pointer + 1]
if (prevDiff[1] == DIFF_EQUAL) and (nextDiff[1] == DIFF_EQUAL) then
-- This is a single edit surrounded by equalities.
local diff = diffs[pointer]
local equality1 = prevDiff[2]
local edit = diff[2]
local equality2 = nextDiff[2]
-- First, shift the edit as far left as possible.
local commonOffset = _diff_commonSuffix(equality1, edit)
if commonOffset > 0 then
local commonString = strsub(edit, -commonOffset)
equality1 = strsub(equality1, 1, -commonOffset - 1)