1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
"""
Operations on strings:
- calculate distance between two strings
- transform strings with transformation codes
"""
import unicodedata
from .char_player import distanceBetweenChars
#### N-GRAMS
def getNgrams (sWord, n=2):
"return a list of Ngrams strings"
return [ sWord[i:i+n] for i in range(len(sWord)-n+1) ]
|
>
|
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
"""
Operations on strings:
- calculate distance between two strings
- transform strings with transformation codes
"""
import unicodedata
import re
from .char_player import distanceBetweenChars, dDistanceBetweenChars
#### N-GRAMS
def getNgrams (sWord, n=2):
"return a list of Ngrams strings"
return [ sWord[i:i+n] for i in range(len(sWord)-n+1) ]
|
48
49
50
51
52
53
54
55
56
57
58
59
60
61
|
sWord = sWord.lower().translate(_xTransCharsForSimplification)
sNewWord = ""
for i, c in enumerate(sWord, 1):
if c != sWord[i:i+1] or (c == 'e' and sWord[i:i+2] != "ee"): # exception for <e> to avoid confusion between crée / créai
sNewWord += c
return sNewWord.replace("eau", "o").replace("au", "o").replace("ai", "éi").replace("ei", "é").replace("ph", "f")
_xTransNumbersToExponent = str.maketrans({
"0": "⁰", "1": "¹", "2": "²", "3": "³", "4": "⁴", "5": "⁵", "6": "⁶", "7": "⁷", "8": "⁸", "9": "⁹"
})
def numbersToExponent (sWord):
"convert numeral chars to exponant chars"
|
>
>
>
>
>
|
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
sWord = sWord.lower().translate(_xTransCharsForSimplification)
sNewWord = ""
for i, c in enumerate(sWord, 1):
if c != sWord[i:i+1] or (c == 'e' and sWord[i:i+2] != "ee"): # exception for <e> to avoid confusion between crée / créai
sNewWord += c
return sNewWord.replace("eau", "o").replace("au", "o").replace("ai", "éi").replace("ei", "é").replace("ph", "f")
def cleanWord (sWord):
"remove letters repeated more than 2 times"
return re.sub("(.)\\1{2,}", '\\1\\1', sWord)
_xTransNumbersToExponent = str.maketrans({
"0": "⁰", "1": "¹", "2": "²", "3": "³", "4": "⁴", "5": "⁵", "6": "⁶", "7": "⁷", "8": "⁸", "9": "⁹"
})
def numbersToExponent (sWord):
"convert numeral chars to exponant chars"
|
102
103
104
105
106
107
108
109
110
111
112
113
114
115
|
d[i, j-1] + 1, # Insertion
d[i-1, j-1] + nCost, # Substitution
)
if i and j and s1[i] == s2[j-1] and s1[i-1] == s2[j]:
d[i, j] = min(d[i, j], d[i-2, j-2] + nCost) # Transposition
return d[nLen1-1, nLen2-1]
def distanceSift4 (s1, s2, nMaxOffset=5):
"implementation of general Sift4."
# https://siderite.blogspot.com/2014/11/super-fast-and-accurate-string-distance.html
if not s1:
return len(s2)
if not s2:
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
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
|
d[i, j-1] + 1, # Insertion
d[i-1, j-1] + nCost, # Substitution
)
if i and j and s1[i] == s2[j-1] and s1[i-1] == s2[j]:
d[i, j] = min(d[i, j], d[i-2, j-2] + nCost) # Transposition
return d[nLen1-1, nLen2-1]
def distanceJaroWinkler (a, b, boost = .666):
# https://github.com/thsig/jaro-winkler-JS
#if (a == b): return 1.0
a_len = len(a)
b_len = len(b)
nMax = max(a_len, b_len)
a_flag = [None for _ in range(nMax)]
b_flag = [None for _ in range(nMax)]
search_range = (max(a_len, b_len) // 2) - 1
minv = min(a_len, b_len)
# Looking only within the search range, count and flag the matched pairs.
Num_com = 0
yl1 = b_len - 1
for i in range(a_len):
lowlim = i - search_range if i >= search_range else 0
hilim = i + search_range if (i + search_range) <= yl1 else yl1
for j in range(lowlim, hilim+1):
if b_flag[j] != 1 and a[j:j+1] == b[i:i+1]:
a_flag[j] = 1
b_flag[i] = 1
Num_com += 1
break
# Return if no characters in common
if Num_com == 0:
return 0.0
# Count the number of transpositions
k = 0
N_trans = 0
for i in range(a_len):
if a_flag[i] == 1:
for j in range(k, b_len):
if b_flag[j] == 1:
k = j + 1
break
if a[i] != b[j]:
N_trans += 1
N_trans = N_trans // 2
# Adjust for similarities in nonmatched characters
N_simi = 0
if minv > Num_com:
for i in range(a_len):
if not a_flag[i]:
for j in range(b_len):
if not b_flag[j]:
if a[i] in dDistanceBetweenChars and b[j] in dDistanceBetweenChars[a[i]]:
N_simi += dDistanceBetweenChars[a[i]][b[j]]
b_flag[j] = 2
break
Num_sim = (N_simi / 10.0) + Num_com
# Main weight computation
weight = Num_sim / a_len + Num_sim / b_len + (Num_com - N_trans) / Num_com
weight = weight / 3
# Continue to boost the weight if the strings are similar
if weight > boost:
# Adjust for having up to the first 4 characters in common
j = 4 if minv >= 4 else minv
i = 0
while i < j and a[i] == b[i]:
i += 1
if i:
weight += i * 0.1 * (1.0 - weight)
# Adjust for long strings.
# After agreeing beginning chars, at least two more must agree
# and the agreeing characters must be more than half of the
# remaining characters.
if minv > 4 and Num_com > i + 1 and 2 * Num_com >= minv + i:
weight += (1 - weight) * ((Num_com - i - 1) / (a_len * b_len - i*2 + 2))
return weight
def distanceSift4 (s1, s2, nMaxOffset=5):
"implementation of general Sift4."
# https://siderite.blogspot.com/2014/11/super-fast-and-accurate-string-distance.html
if not s1:
return len(s2)
if not s2:
|