mathdatech/Python
1
0
Fork 0
mirror of https://github.com/TheAlgorithms/Python.git synced 2025-02-07 10:00:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

suffix_array_lcp.py

Browse Source
This commit is contained in:
Putul Singh 2024-10-19 13:49:17 +05:30 committed by GitHub
parent 1b37c1c127
commit a4073ca188
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 16 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
divide_and_conquer/suffix_array_lcp.py
View File

@ -2,7 +2,6 @@
import doctest import doctest
def build_suffix_array(input_string: str) -> list[int]: def build_suffix_array(input_string: str) -> list[int]:
""" """
Build the suffix array for the given string. Build the suffix array for the given string.
@ -14,7 +13,7 @@ def build_suffix_array(input_string: str) -> list[int]:
list[int]: The suffix array (a list of starting indices of list[int]: The suffix array (a list of starting indices of
suffixes in sorted order). suffixes in sorted order).
Example: Examples:
>>> build_suffix_array("banana") >>> build_suffix_array("banana")
[5, 3, 1, 0, 4, 2] [5, 3, 1, 0, 4, 2]
""" """
@ -23,7 +22,6 @@ def build_suffix_array(input_string: str) -> list[int]:
suffix_array = [suffix[1] for suffix in suffixes] suffix_array = [suffix[1] for suffix in suffixes]
return suffix_array return suffix_array
def build_lcp_array(input_string: str, suffix_array: list[int]) -> list[int]: def build_lcp_array(input_string: str, suffix_array: list[int]) -> list[int]:
""" """
Build the LCP array for the given string and suffix array. Build the LCP array for the given string and suffix array.
@ -35,9 +33,9 @@ def build_lcp_array(input_string: str, suffix_array: list[int]) -> list[int]:
Returns: Returns:
list[int]: The LCP array. list[int]: The LCP array.
Example: Examples:
>>> suffix_arr = build_suffix_array("banana") >>> suffix_array = build_suffix_array("banana")
>>> build_lcp_array("banana", suffix_arr) >>> build_lcp_array("banana", suffix_array)
[0, 1, 3, 0, 0, 2] [0, 1, 3, 0, 0, 2]
""" """
n = len(input_string) n = len(input_string)
@ -53,30 +51,29 @@ def build_lcp_array(input_string: str, suffix_array: list[int]) -> list[int]:
for i in range(n): for i in range(n):
if rank[i] > 0: if rank[i] > 0:
j = suffix_array[rank[i] - 1] j = suffix_array[rank[i] - 1]
while ( while (i + h < n) and (j + h < n) and (input_string[i + h] == input_string[j + h]):
(i + h < n)
and (j + h < n)
and (input_string[i + h] == input_string[j + h])
):
h += 1 h += 1
lcp[rank[i]] = h lcp[rank[i]] = h
if h > 0: if h > 0:
h -= 1 # Decrease h for the next suffix h -= 1 # Decrease h for the next suffix
return lcp return lcp
# Example usage # Example usage
if __name__ == "__main__": if __name__ == "__main__":
test_string = "banana" s = "banana"
suffix_array = build_suffix_array(test_string) suffix_array = build_suffix_array(s)
lcp_array = build_lcp_array(test_string, suffix_array) lcp_array = build_lcp_array(s, suffix_array)
print("Suffix Array:") print("Suffix Array:")
for i in range(len(suffix_array)): for i in range(len(suffix_array)):
print(f"{suffix_array[i]}: {test_string[suffix_array[i]:]}") print(f"{suffix_array[i]}: {s[suffix_array[i]:]}")
print("\nLCP Array:") print("\nLCP Array:")
for i in range(1, len(lcp_array)): for i in range(1, len(lcp_array)):
print( lcp_info = (f"LCP between {s[suffix_array[i - 1]:]} and "
f"LCP between {test_string[suffix_array[i - 1]:]} and {test_string[suffix_array[i]]}: {lcp_array[i]}" f"{s[suffix_array[i]]}: {lcp_array[i]}")
) print(lcp_info)
# Run doctests
if __name__ == "__main__":
doctest.testmod()