def search_editdist_zalg(text, pat):
pat_len = len(pat)
txt_len = len(text)
first_string = pat + "$" + text
# create the first string by concatenating pattern, "$", and text
str_len = len(first_string)
first_lis = calculateZ(first_string)
# calculate the z values for the first string
second_string = pat[::-1] + "$" + text[::-1]
# create the second string by concatenating reversed pattern, "$", and reversed text
second_lis = calculateZ(second_string)
# calculate the z values for the second string
result = []
for i in range(pat_len + 1, str_len - pat_len + 2):
substitute_sum = first_lis[i] + second_lis[txt_len + pat_len + 2 - i]
# sum the z value of the first character of the pattern in the first string, and the z value of the
# last character in the second string. Case substitution
delete_sum = first_lis[i] + second_lis[txt_len + pat_len + 2 - i - 1]
# sum the z value of the first character of the pattern in the first string, and the z value of the
# one before the last character in the second string so that the corresponding substring with length
# of len(pat) + 1. Case insertion
insert_sum = first_lis[i] + second_lis[txt_len + pat_len + 2 - i + 1]
# sum the z value of the first character of the pattern in the first string, and the z value of the
# second last character in the second string so that the corresponding substring with length
# of len(pat) - 1. Case deletion
if substitute_sum == pat_len * 2:
print("match", i, substitute_sum)
print(first_string[i:i + 5])
result.append([i - len(pat), 0])
# if they exactly match
elif substitute_sum == pat_len - 1:
print("sub", i, substitute_sum)
print(first_string[i:i + 5])
result.append([i - len(pat), 1])
# if they match with hamming distance = 1
elif insert_sum == pat_len - 1:
print("insert", i, insert_sum)
print(first_string[i:i + 5])
result.append([i - len(pat), 1])
# if they match by deleting one character from pattern
elif delete_sum == pat_len:
# if they match by inserting one character into pattern
if first_lis[i + 1] + second_lis[
txt_len + pat_len + 1 -
i] != 2 * pat_len: #删除的时候，防止下一位开始正好完全匹配
print("delete", i, delete_sum)
print(first_string[i:i + 5])
result.append([i - len(pat), 1])
# 1. This conditional statement will check cases in order, if any cases matched, it will not consider
# the other cases. For example, if xyz matches xyz, then xy matches xyz with edit distance = 1 will
# not be included
# 2. When insertion case happens at i, the position will be stored only if there is not any exactly match at
# i+1
# 3. When there is an exactly match at i, the deletion case at i+1 will not be considered because
# its corresponding z_value in the second string will be the length of the pattern, exceeding the len -1
return result
def goodsufix(pat):
m = len(pat)
z = list(reversed(calculateZ(pat[::-1])))
good_suffix = [-1 for _ in range(m + 1)]
for p in range(m):
j = m - z[p]
good_suffix[j] = p
good_suffix.pop()
return good_suffix
def computeSP(pat):
m = len(pat)
z = calculateZ(pat)
SP = [0 for _ in range(m)]
for j in range(m - 1, -1, -1):
i = j + z[j] - 1
if i == -1:
break
SP[i] = z[j]
return SP
def edit_distance(text, pat):
str = pat + "$" + text
z1 = calculateZ(str)
reverse_str = pat[::-1] + "$" + text[::-1]
z2 = calculateZ(reverse_str)
str_len = len(str)
m = len(pat)
target = []
for i in range(m + 1, str_len - m + 2):
if str_len - i + 2 <= str_len - 1:
# sum the z value of the first character and the z value of the last character in current pattern
substitution = z1[i] + z2[str_len - i + 1]
#sum the z value of the first character and the z value of the next character of the last character in current pattern
deletion = z1[i] + z2[str_len - i]
# sum the z value of the first character and the z value of the character that before the last character in current pattern
# if str_len-i+2<=str_len-1:
insertion = z1[i] + z2[str_len - i + 2]
else:
continue
if z1[i] == m:
print(str[i:i + 5])
print("match", i)
target.append([i - m, 0])
# the extra case 'i<=str_len-m 'make sure if the length of remain text < the length of pattern, then it should not use substitution
elif substitution == m - 1 and i <= str_len - m:
print(str[i:i + 5])
print("sub", i, substitution)
target.append([i - m, 1])
elif insertion == m - 1 and i <= str_len - m + 1:
print(str[i:i + 5])
print("insert", i, insertion)
target.append([i - m, 1])
elif deletion == m and i <= str_len - m - 1:
print(str[i:i + 5])
if z1[i +
1] != m: # if the pattern that start from i+1 matches the pat,ignore the case that start from i as redundant
print("delete", i, deletion)
target.append([i - m, 1])
return target
def matchedprefix(pat):
m = len(pat)
matched_prefix = [-1 for _ in range(m)]
z = calculateZ(pat)
for i in range(m - 1, -1, -1):
if i + z[i] - 1 == m - 1:
matched_prefix[i] = z[i]
elif i == m - 1: # 最后一位且不匹配
matched_prefix[i] = 0
else:
matched_prefix[i] = matched_prefix[i + 1]
matched_prefix[0] = m - 1
return matched_prefix
def computeSPx(pat):
m = len(pat)
z = calculateZ(pat)
#The size of Spi is m*75, each sublist records the length of the longest proper suffix of pat[1...i] that matches its prefix, with the extra condition
# that pat[spi(x)+1]=x
SP = [[0 for _ in range(CHARACTER)] for _ in range(m)]
for j in range(m - 1, -1, -1):
#for each j,compute its i will record spi with the longest when multiple z_box with different start point j but same end point i
i = j + z[j] - 1
if i == -1: # if i=-1 sp[i] will be recognised as the last value of SP
break
SP[i][ord(pat[z[j]]) -
ord('0')] = z[j] #recording spi and the next character
return SP