text_alignment 0.2.8 → 0.3.13
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/bin/align_annotations +192 -39
- data/lib/text_alignment/anchor_finder.rb +146 -0
- data/lib/text_alignment/approximate_fit.rb +50 -49
- data/lib/text_alignment/find_divisions.rb +199 -101
- data/lib/text_alignment/glcs_alignment.rb +297 -297
- data/lib/text_alignment/glcs_alignment_fast.rb +94 -94
- data/lib/text_alignment/glcs_required.rb +50 -50
- data/lib/text_alignment/lcs_alignment.rb +115 -115
- data/lib/text_alignment/lcs_cdiff.rb +46 -48
- data/lib/text_alignment/lcs_comparison.rb +53 -53
- data/lib/text_alignment/lcs_min.rb +144 -138
- data/lib/text_alignment/mappings.rb +68 -69
- data/lib/text_alignment/mixed_alignment.rb +193 -0
- data/lib/text_alignment/text_alignment.rb +228 -174
- data/lib/text_alignment/version.rb +1 -1
- data/text_alignment.gemspec +1 -1
- metadata +5 -13
- data/spec/spec_helper.rb +0 -1
- data/spec/text_alignment/glcs_alignment_spec.rb +0 -302
- data/spec/text_alignment/lcs_alignment_spec.rb +0 -98
- data/spec/text_alignment/lcs_comparision_spec.rb +0 -322
- data/spec/text_alignment/text_alignment_spec.rb +0 -302
@@ -7,111 +7,209 @@ module TextAlignment; end unless defined? TextAlignment
|
|
7
7
|
# to work on the hash representation of denotations
|
8
8
|
# to assume that there is no bag representation to this method
|
9
9
|
|
10
|
-
|
11
|
-
TextAlignment::SIMILARITY_THRESHOLD = 0.8
|
12
|
-
end
|
10
|
+
TextAlignment::SIMILARITY_THRESHOLD = 0.7 unless defined? TextAlignment::SIMILARITY_THRESHOLD
|
13
11
|
|
14
12
|
class << TextAlignment
|
15
13
|
|
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
|
-
|
14
|
+
# It finds, among the targets, the right divisions for the taraget text to fit in.
|
15
|
+
def find_divisions(source, targets, mappings = [])
|
16
|
+
raise ArgumentError, "nil source" if source == nil
|
17
|
+
raise ArgumentError, "nil or empty targets" if targets == nil || targets.empty?
|
18
|
+
raise ArgumentError, "nil mappings" if mappings == nil
|
19
|
+
|
20
|
+
character_mappings = mappings.select{|m| m[0].length == 1 && m[1].length == 1}
|
21
|
+
mappings.delete_if{|m| m[0].length == 1 && m[1].length == 1}
|
22
|
+
characters_from = character_mappings.collect{|m| m[0]}.join
|
23
|
+
characters_to = character_mappings.collect{|m| m[1]}.join
|
24
|
+
characters_to.gsub!(/-/, '\-')
|
25
|
+
|
26
|
+
source.tr!(characters_from, characters_to)
|
27
|
+
targets.each{|target| target[:text].tr!(characters_from, characters_to)}
|
28
|
+
|
29
|
+
# to process smaller ones first
|
30
|
+
targets.sort!{|s1, s2| s1[:text].size <=> s2[:text].size}
|
31
|
+
|
32
|
+
TextAlignment._find_divisions(source, targets)
|
33
|
+
end
|
34
|
+
|
35
|
+
def _find_divisions(_source, _targets)
|
36
|
+
indice = []
|
37
|
+
history = []
|
38
|
+
cache = {}
|
39
|
+
source = _source.dup
|
40
|
+
targets = _targets.dup
|
41
|
+
until source.strip.empty? || targets.empty?
|
42
|
+
mode, cmp = nil, nil
|
43
|
+
candidates = []
|
44
|
+
targets.each_with_index do |target, i|
|
45
|
+
if source.size < target[:text].size
|
46
|
+
mode = :t_in_s
|
47
|
+
str1 = source
|
48
|
+
str2 = target[:text]
|
49
|
+
else
|
50
|
+
mode = :s_in_t
|
51
|
+
str1 = target[:text]
|
52
|
+
str2 = source
|
53
|
+
end
|
54
|
+
|
55
|
+
len1 = str1.length
|
56
|
+
len2 = str2.length
|
57
|
+
|
58
|
+
offset_begin, offset_end = if (len2 - len1) > len1 * (1 - TextAlignment::SIMILARITY_THRESHOLD)
|
59
|
+
approximate_fit(str1, str2)
|
60
|
+
else
|
61
|
+
# the whole source
|
62
|
+
[0, -1]
|
63
|
+
end
|
64
|
+
|
65
|
+
unless offset_begin.nil?
|
66
|
+
key = str1 + ' _:_ ' + str2[offset_begin .. offset_end]
|
67
|
+
cmp = if cache.has_key? key
|
68
|
+
cache[key]
|
69
|
+
else
|
70
|
+
cmp = TextAlignment::LCSComparison.new(str1, str2[offset_begin .. offset_end])
|
71
|
+
end
|
72
|
+
cache[key] = cmp
|
73
|
+
|
74
|
+
if (cmp.similarity > TextAlignment::SIMILARITY_THRESHOLD) && ((len1 - (cmp.str1_match_final - cmp.str1_match_initial + 1)) < len1 * (1 - TextAlignment::SIMILARITY_THRESHOLD))
|
75
|
+
candidates << {idx:i, offset:offset_begin, mode:mode, cmp:cmp}
|
76
|
+
end
|
77
|
+
end
|
78
|
+
end
|
79
|
+
|
80
|
+
# return remaining source and targets if m.nil?
|
81
|
+
break if candidates.empty?
|
82
|
+
|
83
|
+
choice = candidates.max{|a, b| a[:cmp].similarity <=> a[:cmp].similarity}
|
84
|
+
m = choice[:idx]
|
85
|
+
mode = choice[:mode]
|
86
|
+
|
87
|
+
index = if mode == :t_in_s
|
88
|
+
{divid:targets[m][:divid], region:[0, source.size]}
|
89
|
+
else # :s_in_t
|
90
|
+
cmp = choice[:cmp]
|
91
|
+
offset = choice[:offset]
|
92
|
+
{divid:targets[m][:divid], region:[cmp.str2_match_initial + offset, cmp.str2_match_final + offset + 1]}
|
93
|
+
end
|
94
|
+
|
95
|
+
source = source[0 ... index[:region][0]] + source[index[:region][1] .. -1]
|
96
|
+
history << index[:region].dup
|
97
|
+
|
98
|
+
before_begin = index[:region][0]
|
99
|
+
before_end = index[:region][1]
|
100
|
+
|
101
|
+
rhistory = history.reverse
|
102
|
+
rhistory.shift
|
103
|
+
rhistory.each do |h|
|
104
|
+
gap = h[1] - h[0]
|
105
|
+
index[:region][0] += gap if index[:region][0] >= h[0]
|
106
|
+
index[:region][1] += gap if index[:region][1] > h[0]
|
107
|
+
end
|
108
|
+
|
109
|
+
indice << index
|
110
|
+
|
111
|
+
targets.delete_at(m)
|
112
|
+
end
|
113
|
+
|
114
|
+
unless source.strip.empty? && targets.empty?
|
115
|
+
index = {divid:nil}
|
116
|
+
index[:remaining_source] = source unless source.strip.empty?
|
117
|
+
index[:remaining_targets] = targets.collect{|s| s[:divid]} unless targets.empty?
|
118
|
+
indice << index
|
119
|
+
end
|
120
|
+
|
121
|
+
indice
|
122
|
+
end
|
123
|
+
|
124
|
+
def _find_divisions_old(source, targets)
|
125
|
+
mode, m, c, offset_begin = nil, nil, nil, nil
|
126
|
+
|
127
|
+
targets.each_with_index do |target, i|
|
128
|
+
if source.size < target[:text].size
|
129
|
+
mode = :t_in_s
|
130
|
+
str1 = source
|
131
|
+
str2 = target[:text]
|
132
|
+
else
|
133
|
+
mode = :s_in_t
|
134
|
+
str1 = target[:text]
|
135
|
+
str2 = source
|
136
|
+
end
|
137
|
+
|
138
|
+
len1 = str1.length
|
139
|
+
len2 = str2.length
|
140
|
+
|
141
|
+
offset_begin, offset_end = 0, -1
|
142
|
+
offset_begin, offset_end = approximate_fit(str1, str2) if (len2 - len1) > len1 * (1 - TextAlignment::SIMILARITY_THRESHOLD)
|
143
|
+
|
144
|
+
unless offset_begin.nil?
|
145
|
+
c = TextAlignment::LCSComparison.new(str1, str2[offset_begin .. offset_end])
|
146
|
+
if (c.similarity > TextAlignment::SIMILARITY_THRESHOLD) && ((len1 - (c.str1_match_final - c.str1_match_initial + 1)) < len1 * (1 - TextAlignment::SIMILARITY_THRESHOLD))
|
147
|
+
m = i
|
148
|
+
break
|
149
|
+
end
|
150
|
+
end
|
151
|
+
end
|
152
|
+
|
153
|
+
# return remaining source and targets if m.nil?
|
154
|
+
return [[-1, [source, targets.collect{|s| s[:divid]}]]] if m.nil?
|
155
|
+
|
156
|
+
index = if mode == :t_in_s
|
157
|
+
[targets[m][:divid], [0, source.size]]
|
158
|
+
else # :s_in_t
|
159
|
+
[targets[m][:divid], [c.str2_match_initial + offset_begin, c.str2_match_final + offset_begin + 1]]
|
160
|
+
end
|
161
|
+
|
162
|
+
next_source = source[0 ... index[1][0]] + source[index[1][1] .. -1]
|
163
|
+
targets.delete_at(m)
|
164
|
+
|
165
|
+
if next_source.strip.empty? || targets.empty?
|
166
|
+
return [index]
|
167
|
+
else
|
168
|
+
more_index = _find_divisions(next_source, targets)
|
169
|
+
gap = index[1][1] - index[1][0]
|
170
|
+
more_index.each do |i|
|
171
|
+
if (i[0] > -1)
|
172
|
+
i[1][0] += gap if i[1][0] >= index[1][0]
|
173
|
+
i[1][1] += gap if i[1][1] > index[1][0]
|
174
|
+
end
|
175
|
+
end
|
176
|
+
return [index] + more_index
|
177
|
+
end
|
178
|
+
end
|
179
|
+
|
91
180
|
end
|
92
181
|
|
93
182
|
if __FILE__ == $0
|
94
|
-
|
95
|
-
|
96
|
-
|
97
|
-
|
98
|
-
|
99
|
-
|
100
|
-
|
101
|
-
|
102
|
-
|
103
|
-
|
104
|
-
|
105
|
-
|
106
|
-
|
107
|
-
|
108
|
-
|
109
|
-
|
110
|
-
|
111
|
-
|
112
|
-
|
113
|
-
|
114
|
-
|
115
|
-
|
116
|
-
|
183
|
+
require 'json'
|
184
|
+
if ARGV.length == 2
|
185
|
+
source = JSON.parse File.read(ARGV[0]), :symbolize_names => true
|
186
|
+
source_text = source[:text].strip
|
187
|
+
|
188
|
+
targets = JSON.parse File.read(ARGV[1]), :symbolize_names => true
|
189
|
+
div_index = TextAlignment::find_divisions(source_text, targets)
|
190
|
+
pp div_index
|
191
|
+
|
192
|
+
# str1 = File.read(ARGV[0]).strip
|
193
|
+
# str2 = File.read(ARGV[1]).strip
|
194
|
+
# div_index = TextAlignment::find_divisions(str1, [str2])
|
195
|
+
|
196
|
+
# puts "source length: #{source_text.length}"
|
197
|
+
# div_index.each do |i|
|
198
|
+
# unless i[:divid].nil?
|
199
|
+
# puts "[Div: #{i[:divid]}] (#{i[:region][0]}, #{i[:region][1]})"
|
200
|
+
# puts source_text[i[:region][0] ... i[:region][1]]
|
201
|
+
# puts "=========="
|
202
|
+
# else
|
203
|
+
# p i
|
204
|
+
# end
|
205
|
+
|
206
|
+
# # if i[0] >= 0
|
207
|
+
# # puts "[Div: #{i[0]}] (#{i[1][0]}, #{i[1][1]})"
|
208
|
+
# # puts source_text[i[1][0] ... i[1][1]]
|
209
|
+
# # puts "=========="
|
210
|
+
# # else
|
211
|
+
# # p i
|
212
|
+
# # end
|
213
|
+
# end
|
214
|
+
end
|
117
215
|
end
|
@@ -6,306 +6,306 @@ module TextAlignment; end unless defined? TextAlignment
|
|
6
6
|
# An instance of this class holds the results of generalized LCS computation for the two strings str1 and str2.
|
7
7
|
# an optional dictionary is used for generalized suffix comparision.
|
8
8
|
class TextAlignment::GLCSAlignment
|
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
|
-
|
9
|
+
# The mapping function from str1 to str2
|
10
|
+
attr_reader :position_map_begin, :position_map_end
|
11
|
+
|
12
|
+
# The position initial and final position of matching on str1 and str2
|
13
|
+
attr_reader :str1_match_begin, :str1_match_end, :str2_match_begin, :str2_match_end
|
14
|
+
|
15
|
+
# The length of GLCS
|
16
|
+
attr_reader :length
|
17
|
+
|
18
|
+
# the elements that are common in the two strings, str1 and str2
|
19
|
+
attr_reader :common_elements
|
20
|
+
|
21
|
+
# the elements that are mapped to each other in the two strings, str1 and str2
|
22
|
+
attr_reader :mapped_elements
|
23
|
+
|
24
|
+
# the string of non-mapped characters
|
25
|
+
attr_reader :diff_strings
|
26
|
+
|
27
|
+
attr_reader :similarity
|
28
|
+
|
29
|
+
# It initializes the GLCS table for the given two strings, str1 and str2.
|
30
|
+
# When the array, mappings, is given, general suffix comparision is performed based on the mappings.
|
31
|
+
# Exception is raised when nil given passed to either str1, str2 or dictionary
|
32
|
+
def initialize(str1, str2, mappings = [])
|
33
|
+
raise ArgumentError, "nil string" if str1 == nil || str2 == nil
|
34
|
+
raise ArgumentError, "nil dictionary" if mappings == nil
|
35
|
+
|
36
|
+
# index the mappings in hash.
|
37
|
+
@dic = (mappings + mappings.map{|e| e.reverse}).to_h
|
38
|
+
|
39
|
+
# prefix dictionary
|
40
|
+
@pdic = Dictionary.new(mappings.flatten)
|
41
|
+
|
42
|
+
@len1 = str1.length
|
43
|
+
@len2 = str2.length
|
44
|
+
|
45
|
+
# add a final marker to the end of the strings
|
46
|
+
@str1 = str1 + '_'
|
47
|
+
@str2 = str2 + '_'
|
48
|
+
|
49
|
+
# compute the GLCS table
|
50
|
+
@glcs = _compute_glcs_table
|
51
|
+
@length = @glcs[0][0]
|
52
|
+
|
53
|
+
_trace_glcs_table
|
54
|
+
end
|
55
|
+
|
56
|
+
# Prints the GLCS table
|
57
|
+
def show_glcs
|
58
|
+
puts "\t\t" + @str2.split(//).join("\t")
|
59
|
+
@glcs.each_with_index do |row, i|
|
60
|
+
h = (@str1[i].nil?)? '' : @str1[i]
|
61
|
+
puts i.to_s + "\t" + h + "\t" + row.join("\t")
|
62
|
+
end
|
63
|
+
end
|
64
|
+
|
65
|
+
# Returns the character-by-character difference
|
66
|
+
def cdiff
|
67
|
+
cdiff1, cdiff2 = '', ''
|
68
|
+
p1, p2 = 0, 0
|
69
|
+
begin
|
70
|
+
s1, s2 = _prefix_eq(@str1[p1...@len1], @str2[p2...@len2])
|
71
|
+
if s1 != nil
|
72
|
+
l1, l2 = s1.length, s2.length
|
73
|
+
|
74
|
+
cdiff1 += s1; cdiff2 += s2
|
75
|
+
if l1 > l2 then cdiff2 += ' ' * (l1 - l2) else cdiff1 += ' ' * (l2 - l1) end
|
76
|
+
p1 += s1.length; p2 += s2.length
|
77
|
+
elsif p2 < @len2 && (p1 == @len1 or @glcs[p1][p2 + 1] > @glcs[p1 + 1][p2])
|
78
|
+
cdiff1 += ' '
|
79
|
+
cdiff2 += @str2[p2]
|
80
|
+
p2 += 1
|
81
|
+
elsif p1 < @len1 && (p2 == @len2 or @glcs[p1][p2 + 1] <= @glcs[p1 + 1][p2])
|
82
|
+
cdiff1 += @str1[p1]
|
83
|
+
cdiff2 += ' '
|
84
|
+
p1 += 1
|
85
|
+
end
|
86
|
+
end until p1 == @len1 && p2 == @len2
|
87
|
+
|
88
|
+
return [cdiff1, cdiff2]
|
89
|
+
end
|
90
|
+
|
91
|
+
|
92
|
+
# Computes the similarity of the two strings
|
93
|
+
def similarity(cut = false)
|
94
|
+
c = @length
|
95
|
+
|
96
|
+
l1 = c + @diff_strings[0].length
|
97
|
+
l2 = c + @diff_strings[1].length
|
98
|
+
|
99
|
+
if cut
|
100
|
+
l1 -= front_overflow if front_overflow > 0
|
101
|
+
l1 -= rear_overflow if rear_overflow > 0
|
102
|
+
l1 += front_overflow if front_overflow < 0
|
103
|
+
l1 += rear_overflow if rear_overflow < 0
|
104
|
+
end
|
105
|
+
|
106
|
+
similarity = 2 * c / (l1 + l2).to_f
|
107
|
+
end
|
108
|
+
|
109
|
+
def transform_a_span(span)
|
110
|
+
{:begin=>@position_map_begin[span[:begin]], :end=>@position_map_end[span[:end]]}
|
111
|
+
end
|
112
|
+
|
113
|
+
def transform_spans(spans)
|
114
|
+
spans.map{|span| transform_a_span(span)}
|
115
|
+
end
|
116
|
+
|
117
|
+
|
118
|
+
private
|
119
|
+
|
120
|
+
# Computes the GLCS table for the two strings, @str1 and @str2.
|
121
|
+
# Unlike normal LCS algorithms, the computation is performed from the end to the beginning of the strings.
|
122
|
+
def _compute_glcs_table
|
123
|
+
glcs = Array.new(@len1 + 1) { Array.new(@len2 + 1) }
|
124
|
+
|
125
|
+
# initialize the final row and the final column
|
126
|
+
(0..@len1).each {|p| glcs[p][@len2] = 0}
|
127
|
+
(0..@len2).each {|p| glcs[@len1][p] = 0}
|
128
|
+
|
129
|
+
# compute the GLCS table
|
130
|
+
str1_reverse_iteration = (0...@len1).to_a.reverse
|
131
|
+
str2_reverse_iteration = (0...@len2).to_a.reverse
|
132
|
+
|
133
|
+
str1_reverse_iteration.each do |p1|
|
134
|
+
str2_reverse_iteration.each do |p2|
|
135
|
+
s1, s2 = _prefix_eq(@str1[p1...@len1], @str2[p2...@len2])
|
136
|
+
unless s1 == nil
|
137
|
+
glcs[p1][p2] = glcs[p1 + s1.length][p2 + s2.length] + 1
|
138
|
+
else
|
139
|
+
glcs[p1][p2] = (glcs[p1][p2 + 1] > glcs[p1 + 1][p2])? glcs[p1][p2 + 1] : glcs[p1 + 1][p2]
|
140
|
+
end
|
141
|
+
end
|
142
|
+
end
|
143
|
+
|
144
|
+
glcs
|
145
|
+
end
|
146
|
+
|
147
|
+
# Backtrace the GLCS table, computing the mapping function from str1 to str2
|
148
|
+
# As its side effect, it updates four global variables
|
149
|
+
# * front_overflow: the length of the front part of str1 that cannot fit in str2.
|
150
|
+
# * rear_overflow: the length of the rear part of str1 that cannot fit in str2.
|
151
|
+
# * common_elements: an array which stores the common elements in the two strings.
|
152
|
+
# * mapped_elements: an array which stores the mapped elements in the two strings.
|
153
|
+
def _trace_glcs_table
|
154
|
+
@front_overflow, @rear_overflow = 0, 0
|
155
|
+
@common_elements, @mapped_elements = [], []
|
156
|
+
diff_string1, diff_string2 = '', ''
|
157
|
+
|
158
|
+
@position_map_begin, @position_map_end = {}, {}
|
159
|
+
addition, deletion = [], []
|
160
|
+
p1, p2 = 0, 0
|
161
|
+
|
162
|
+
while p1 <= @len1 && p2 <= @len2
|
163
|
+
s1, s2 = _prefix_eq(@str1[p1..@len1], @str2[p2..@len2])
|
164
|
+
if s1 != nil
|
165
|
+
l1, l2 = s1.length, s2.length
|
166
|
+
|
167
|
+
@position_map_begin[p1], @position_map_end[p1] = p2, p2
|
168
|
+
(p1 + 1 ... p1 + l1).each{|i| @position_map_begin[i], @position_map_end[i] = nil, nil}
|
169
|
+
|
170
|
+
@common_elements << [s1, s2]
|
171
|
+
|
172
|
+
if !addition.empty? && deletion.empty?
|
173
|
+
# If an addition is found in the front or the rear, it is a case of underflow
|
174
|
+
@str2_match_begin = addition.length if p1 == 0
|
175
|
+
@str2_match_end = l2 - addition.length if p1 == @len1
|
176
|
+
|
177
|
+
if p1 == 0
|
178
|
+
# leave as it is
|
179
|
+
elsif p1 == @len1
|
180
|
+
# retract from the end
|
181
|
+
@position_map_begin[p1] = p2 - addition.length
|
182
|
+
@position_map_end[p1] = @position_map_begin[p1]
|
183
|
+
else
|
184
|
+
# correct the position for end
|
185
|
+
@position_map_end[p1] = p2 - addition.length
|
186
|
+
end
|
187
|
+
elsif addition.empty? && !deletion.empty?
|
188
|
+
# If a deletion is found in the front or the rear, it is a case of overflow
|
189
|
+
@str1_match_begin = deletion.length if p1 == deletion.length
|
190
|
+
@str1_match_end = l1 - deletion.length if p1 == @len1
|
191
|
+
|
192
|
+
deletion.each{|p| @position_map_begin[p], @position_map_end[p] = p2, p2}
|
193
|
+
elsif !addition.empty? && !deletion.empty?
|
194
|
+
# If an addition and a deletion are both found in the front or the rear,
|
195
|
+
# the overflow/underflow is approximated to the difference.
|
196
|
+
al, dl = addition.length, deletion.length
|
197
|
+
@front_overflow = dl - al if p1 == dl
|
198
|
+
@rear_overflow = dl - al if p1 == @len1
|
199
|
+
|
200
|
+
@mapped_elements << [@str1[deletion[0], dl], @str2[addition[0], al]]
|
201
|
+
|
202
|
+
@position_map_begin[deletion[0]], @position_map_end[deletion[0]] = addition[0], addition[0]
|
203
|
+
deletion[1..-1].each{|p| @position_map_begin[p], @position_map_end[p] = nil, nil}
|
204
|
+
end
|
205
|
+
|
206
|
+
addition.clear; deletion.clear
|
207
|
+
p1 += l1; p2 += l2
|
208
|
+
|
209
|
+
elsif p2 < @len2 && (p1 == @len1 || @glcs[p1][p2 + 1] > @glcs[p1 + 1][p2])
|
210
|
+
diff_string2 += @str2[p2]
|
211
|
+
|
212
|
+
addition << p2
|
213
|
+
p2 += 1
|
214
|
+
elsif p1 < @len1 && (p2 == @len2 || @glcs[p1][p2 + 1] <= @glcs[p1 + 1][p2])
|
215
|
+
diff_string1 += @str1[p1]
|
216
|
+
|
217
|
+
deletion << p1
|
218
|
+
p1 += 1
|
219
|
+
end
|
220
|
+
end
|
221
|
+
|
222
|
+
@common_elements.pop
|
223
|
+
@diff_strings = [diff_string1, diff_string2]
|
224
|
+
end
|
225
|
+
|
226
|
+
# General prefix comparison is performed based on the dictionary.
|
227
|
+
# The pair of matched suffixes are returned when found.
|
228
|
+
# Otherwise, the pair of nil values are returned.
|
229
|
+
def _prefix_eq(str1, str2)
|
230
|
+
return nil, nil if str1.empty? || str2.empty?
|
231
|
+
prefixes1 = @pdic.prefixes(str1)
|
232
|
+
prefixes1.each {|p1| p2 = @dic[p1]; return p1, p2 if str2.start_with?(p2)}
|
233
|
+
return str1[0], str2[0] if (str1[0] == str2[0])
|
234
|
+
return nil, nil
|
235
|
+
end
|
236
236
|
|
237
237
|
end
|
238
238
|
|
239
239
|
if __FILE__ == $0
|
240
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
|
-
|
241
|
+
dictionary = [
|
242
|
+
["×", "x"], #U+00D7 (multiplication sign)
|
243
|
+
["•", "*"], #U+2022 (bullet)
|
244
|
+
["Δ", "delta"], #U+0394 (greek capital letter delta)
|
245
|
+
["Φ", "phi"], #U+03A6 (greek capital letter phi)
|
246
|
+
["α", "alpha"], #U+03B1 (greek small letter alpha)
|
247
|
+
["β", "beta"], #U+03B2 (greek small letter beta)
|
248
|
+
["γ", "gamma"], #U+03B3 (greek small letter gamma)
|
249
|
+
["δ", "delta"], #U+03B4 (greek small letter delta)
|
250
|
+
["ε", "epsilon"], #U+03B5 (greek small letter epsilon)
|
251
|
+
["κ", "kappa"], #U+03BA (greek small letter kappa)
|
252
|
+
["λ", "lambda"], #U+03BB (greek small letter lambda)
|
253
|
+
["μ", "mu"], #U+03BC (greek small letter mu)
|
254
|
+
["χ", "chi"], #U+03C7 (greek small letter chi)
|
255
|
+
["ϕ", "phi"], #U+03D5 (greek phi symbol)
|
256
|
+
[" ", " "], #U+2009 (thin space)
|
257
|
+
[" ", " "], #U+200A (hair space)
|
258
|
+
[" ", " "], #U+00A0 (no-break space)
|
259
|
+
[" ", " "], #U+3000 (ideographic space)
|
260
|
+
["−", "-"], #U+2212 (minus sign)
|
261
|
+
["–", "-"], #U+2013 (en dash)
|
262
|
+
["′", "'"], #U+2032 (prime)
|
263
|
+
["‘", "'"], #U+2018 (left single quotation mark)
|
264
|
+
["’", "'"], #U+2019 (right single quotation mark)
|
265
|
+
["“", '"'], #U+201C (left double quotation mark)
|
266
|
+
["”", '"'] #U+201D (right double quotation mark)
|
267
|
+
]
|
268
|
+
|
269
|
+
# str1 = "-betakappaxyz-"
|
270
|
+
# str2 = "-ijkβκ-"
|
271
|
+
|
272
|
+
# str1 = "-βκ-β-z-xy"
|
273
|
+
# str2 = "abc-betakappa-beta-z"
|
274
|
+
|
275
|
+
# str1 = "-βκ-z-xy"
|
276
|
+
# str2 = "abc-betakappa-z"
|
277
|
+
|
278
|
+
# str1 = "abc-βκ-β-z"
|
279
|
+
# str2 = "-betakappa-beta-z-xyz"
|
280
|
+
|
281
|
+
# str1 = "-β-"
|
282
|
+
# str2 = "-beta-"
|
283
|
+
|
284
|
+
# str1 = "-κ-"
|
285
|
+
# str2 = "-kappa-"
|
286
|
+
|
287
|
+
# str1 = File.read(ARGV[0]).strip
|
288
|
+
# str2 = File.read(ARGV[1]).strip
|
289
|
+
|
290
|
+
str1 = "beta"
|
291
|
+
str2 = "β***"
|
292
|
+
|
293
|
+
# puts "str1: #{str1}"
|
294
|
+
# puts "str2: #{str2}"
|
295
|
+
sa = TextAlignment::GLCSAlignment.new(str1, str2, dictionary)
|
296
|
+
sa.position_map_begin.each {|h| p h}
|
297
|
+
puts '-----'
|
298
|
+
sa.position_map_end.each {|h| p h}
|
299
|
+
puts '-----'
|
300
|
+
puts "common_elements: #{sa.common_elements}"
|
301
|
+
puts '-----'
|
302
|
+
puts "mapped_elements: #{sa.mapped_elements}"
|
303
|
+
puts '-----'
|
304
|
+
# puts "diff_string1: #{sa.diff_strings[0]}"
|
305
|
+
# puts "diff_string2: #{sa.diff_strings[1]}"
|
306
|
+
puts "front_overflow: #{sa.front_overflow}"
|
307
|
+
puts "rear_overflow : #{sa.rear_overflow}"
|
308
|
+
puts '-----'
|
309
|
+
puts "similarity : #{sa.similarity}"
|
310
|
+
puts "similarity(cut): #{sa.similarity(true)}"
|
311
311
|
end
|