Spaces:
Sleeping
Sleeping
#!/usr/bin/env python3 | |
""" | |
Levenshtein distance calculation with wildcard support for both strings and bytes. | |
This module provides functions to calculate Levenshtein (edit) distance between | |
two sequences (strings or bytes) with support for wildcard positions. | |
""" | |
def ensure_same_type(seq1, seq2): | |
""" | |
Ensure both sequences are the same type (both str or both bytes). | |
Args: | |
seq1: First sequence (str or bytes) | |
seq2: Second sequence (str or bytes) | |
Returns: | |
Tuple of (seq1, seq2) with consistent types | |
""" | |
if isinstance(seq1, str) and isinstance(seq2, bytes): | |
seq2 = seq2.decode("utf-8", errors="replace") | |
elif isinstance(seq1, bytes) and isinstance(seq2, str): | |
seq2 = seq2.encode("utf-8", errors="replace") | |
return seq1, seq2 | |
def to_bytes(s): | |
""" | |
Convert a sequence to bytes if it's a string, otherwise return as is. | |
Args: | |
s: The sequence to convert (str or bytes) | |
Returns: | |
bytes: The input converted to bytes if it was a string | |
""" | |
return s.encode("utf-8", errors="replace") if isinstance(s, str) else s | |
def to_str(s): | |
""" | |
Convert a sequence to string if it's bytes, otherwise return as is. | |
Args: | |
s: The sequence to convert (str or bytes) | |
Returns: | |
str: The input converted to string if it was bytes | |
""" | |
return s.decode("utf-8", errors="replace") if isinstance(s, bytes) else s | |
def get_element_repr(element): | |
""" | |
Get a human-readable representation of a sequence element. | |
Args: | |
element: A single element from a sequence (byte or character) | |
Returns: | |
str: A printable representation of the element | |
""" | |
if isinstance(element, int): # For bytes objects | |
if 32 <= element <= 126: # Printable ASCII | |
return repr(chr(element)) | |
return f"0x{element:02x}" | |
return repr(element) # For str objects | |
def levenshtein_with_wildcard( | |
seq1, seq2, wildcard_offsets_seq1=None, wildcard_offsets_seq2=None, verbose=False | |
): | |
""" | |
Calculate the Levenshtein distance between two sequences with support for wildcards. | |
Works with both strings and bytes. | |
Args: | |
seq1: First sequence (str or bytes) | |
seq2: Second sequence (str or bytes) | |
wildcard_offsets_seq1 (iterable, optional): Indices in seq1 that are wildcards. Defaults to None. | |
wildcard_offsets_seq2 (iterable, optional): Indices in seq2 that are wildcards. Defaults to None. | |
verbose (bool, optional): If True, returns additional information about operations. Defaults to False. | |
Returns: | |
int: The Levenshtein distance between the two sequences. | |
list: If verbose=True, also returns a list of operations performed. | |
""" | |
# Initialize empty sets if None | |
wildcard_offsets_seq1 = set(wildcard_offsets_seq1 or []) | |
wildcard_offsets_seq2 = set(wildcard_offsets_seq2 or []) | |
m, n = len(seq1), len(seq2) | |
# Create a matrix of size (m+1) x (n+1) | |
dp = [[0] * (n + 1) for _ in range(m + 1)] | |
# Initialize the first row and column | |
for i in range(m + 1): | |
dp[i][0] = i | |
for j in range(n + 1): | |
dp[0][j] = j | |
# Fill the dp matrix | |
for i in range(1, m + 1): | |
for j in range(1, n + 1): | |
# Check if either position is a wildcard | |
is_seq1_wildcard = (i - 1) in wildcard_offsets_seq1 | |
is_seq2_wildcard = (j - 1) in wildcard_offsets_seq2 | |
# If either position is a wildcard, treat it as a match (cost = 0) | |
if is_seq1_wildcard or is_seq2_wildcard: | |
dp[i][j] = dp[i - 1][j - 1] # No cost for wildcard matches | |
else: | |
cost = 0 if seq1[i - 1] == seq2[j - 1] else 1 | |
dp[i][j] = min( | |
dp[i - 1][j] + 1, # deletion | |
dp[i][j - 1] + 1, # insertion | |
dp[i - 1][j - 1] + cost, # substitution | |
) | |
if verbose: | |
operations = explain_match( | |
seq1, seq2, dp, wildcard_offsets_seq1, wildcard_offsets_seq2 | |
) | |
return dp[m][n], operations | |
return dp[m][n] | |
def explain_match(seq1, seq2, dp, wildcard_offsets_seq1, wildcard_offsets_seq2): | |
""" | |
Traces the optimal alignment path and explains each step of the matching process. | |
Args: | |
seq1: First sequence (str or bytes) | |
seq2: Second sequence (str or bytes) | |
dp (list): The dynamic programming matrix. | |
wildcard_offsets_seq1 (set): Indices in seq1 that are wildcards. | |
wildcard_offsets_seq2 (set): Indices in seq2 that are wildcards. | |
Returns: | |
list: A list of explanation strings for each operation performed. | |
""" | |
m, n = len(seq1), len(seq2) | |
operations = [] | |
# Find the optimal path | |
i, j = m, n | |
path = [] | |
while i > 0 or j > 0: | |
path.append((i, j)) | |
if i == 0: | |
j -= 1 | |
elif j == 0: | |
i -= 1 | |
else: | |
substitution_cost = dp[i - 1][j - 1] | |
deletion_cost = dp[i - 1][j] | |
insertion_cost = dp[i][j - 1] | |
min_cost = min(substitution_cost, deletion_cost, insertion_cost) | |
if min_cost == substitution_cost: | |
i -= 1 | |
j -= 1 | |
elif min_cost == deletion_cost: | |
i -= 1 | |
else: | |
j -= 1 | |
path.append((0, 0)) | |
path.reverse() | |
# Generate explanations for each step | |
for idx in range(1, len(path)): | |
prev_i, prev_j = path[idx - 1] | |
curr_i, curr_j = path[idx] | |
# Diagonal move (match or substitution) | |
if curr_i > prev_i and curr_j > prev_j: | |
char1_idx = curr_i - 1 | |
char2_idx = curr_j - 1 | |
char1 = seq1[char1_idx] | |
char2 = seq2[char2_idx] | |
is_seq1_wildcard = char1_idx in wildcard_offsets_seq1 | |
is_seq2_wildcard = char2_idx in wildcard_offsets_seq2 | |
char1_repr = get_element_repr(char1) | |
char2_repr = get_element_repr(char2) | |
if is_seq1_wildcard and is_seq2_wildcard: | |
operations.append( | |
f"Double wildcard: Position {char1_idx} in seq1 and position {char2_idx} in seq2 are both wildcards" | |
) | |
elif is_seq1_wildcard: | |
operations.append( | |
f"Wildcard match: Position {char1_idx} in seq1 is a wildcard, matches {char2_repr} at position {char2_idx} in seq2" | |
) | |
elif is_seq2_wildcard: | |
operations.append( | |
f"Wildcard match: Position {char2_idx} in seq2 is a wildcard, matches {char1_repr} at position {char1_idx} in seq1" | |
) | |
elif char1 == char2: | |
operations.append( | |
f"Match: {char1_repr} at position {char1_idx} matches {char2_repr} at position {char2_idx}" | |
) | |
else: | |
operations.append( | |
f"Substitution: Replace {char1_repr} at position {char1_idx} with {char2_repr} at position {char2_idx}" | |
) | |
# Horizontal move (insertion) | |
elif curr_i == prev_i and curr_j > prev_j: | |
char_idx = curr_j - 1 | |
char_repr = get_element_repr(seq2[char_idx]) | |
operations.append( | |
f"Insertion: Insert {char_repr} at position {char_idx} in seq2" | |
) | |
# Vertical move (deletion) | |
elif curr_i > prev_i and curr_j == prev_j: | |
char_idx = curr_i - 1 | |
char_repr = get_element_repr(seq1[char_idx]) | |
operations.append( | |
f"Deletion: Delete {char_repr} at position {char_idx} in seq1" | |
) | |
return operations | |
def create_gap_element(sequence): | |
""" | |
Create a gap element compatible with the sequence type. | |
Args: | |
sequence: The sequence (str or bytes) to create a gap for | |
Returns: | |
The appropriate gap element for the sequence type | |
""" | |
if isinstance(sequence, bytes): | |
return b"-" | |
else: | |
return "-" | |
def print_match_summary( | |
seq1, seq2, wildcard_offsets_seq1=None, wildcard_offsets_seq2=None | |
): | |
""" | |
Prints a summary of the match between two sequences, highlighting wildcards by their offsets. | |
Works with both strings and bytes. | |
Args: | |
seq1: First sequence (str or bytes) | |
seq2: Second sequence (str or bytes) | |
wildcard_offsets_seq1 (iterable, optional): Indices in seq1 that are wildcards. Defaults to None. | |
wildcard_offsets_seq2 (iterable, optional): Indices in seq2 that are wildcards. Defaults to None. | |
Returns: | |
tuple: (distance, operations) The edit distance and list of operations | |
""" | |
# Ensure sequences are of the same type for comparison | |
seq1, seq2 = ensure_same_type(seq1, seq2) | |
# Initialize empty sets if None | |
wildcard_offsets_seq1 = set(wildcard_offsets_seq1 or []) | |
wildcard_offsets_seq2 = set(wildcard_offsets_seq2 or []) | |
distance, operations = levenshtein_with_wildcard( | |
seq1, seq2, wildcard_offsets_seq1, wildcard_offsets_seq2, verbose=True | |
) | |
# For reporting, convert to a human-readable representation if needed | |
seq1_repr = repr(seq1) | |
seq2_repr = repr(seq2) | |
print(f"Comparing {seq1_repr} and {seq2_repr}") | |
print(f"Wildcards in seq1: {sorted(wildcard_offsets_seq1)}") | |
print(f"Wildcards in seq2: {sorted(wildcard_offsets_seq2)}") | |
print(f"Edit distance: {distance}") | |
print("\nMatch process:") | |
for i, op in enumerate(operations): | |
print(f"Step {i+1}: {op}") | |
# Visual representation of the alignment | |
i, j = 0, 0 | |
is_bytes = isinstance(seq1, bytes) | |
if is_bytes: | |
aligned_seq1 = bytearray() | |
aligned_seq2 = bytearray() | |
gap = ord("-") | |
else: | |
aligned_seq1 = "" | |
aligned_seq2 = "" | |
gap = "-" | |
match_indicators = "" | |
for op in operations: | |
if ( | |
"Match:" in op | |
or "Substitution:" in op | |
or "Wildcard match:" in op | |
or "Double wildcard:" in op | |
): | |
if is_bytes: | |
aligned_seq1.append(seq1[i]) | |
aligned_seq2.append(seq2[j]) | |
else: | |
aligned_seq1 += seq1[i] | |
aligned_seq2 += seq2[j] | |
# Determine match indicator | |
if "Wildcard match:" in op or "Double wildcard:" in op: | |
match_indicators += "*" # Wildcard match | |
elif "Match:" in op: | |
match_indicators += "|" # Exact match | |
else: | |
match_indicators += "X" # Substitution | |
i += 1 | |
j += 1 | |
elif "Insertion:" in op: | |
if is_bytes: | |
aligned_seq1.append(gap) | |
aligned_seq2.append(seq2[j]) | |
else: | |
aligned_seq1 += gap | |
aligned_seq2 += seq2[j] | |
match_indicators += " " | |
j += 1 | |
elif "Deletion:" in op: | |
if is_bytes: | |
aligned_seq1.append(seq1[i]) | |
aligned_seq2.append(gap) | |
else: | |
aligned_seq1 += seq1[i] | |
aligned_seq2 += gap | |
match_indicators += " " | |
i += 1 | |
print("\nAlignment:") | |
if is_bytes: | |
aligned_seq1 = bytes(aligned_seq1) | |
aligned_seq2 = bytes(aligned_seq2) | |
print(repr(aligned_seq1)) | |
print(match_indicators) | |
print(repr(aligned_seq2)) | |
print("\nLegend:") | |
print( | |
"| = exact match, * = wildcard match, X = substitution, - = gap (insertion/deletion)" | |
) | |
# Summary of wildcard matches | |
wildcard_matches = [ | |
op for op in operations if "Wildcard match:" in op or "Double wildcard:" in op | |
] | |
if wildcard_matches: | |
print("\nWildcard matches:") | |
for match in wildcard_matches: | |
print(f"- {match}") | |
return distance, operations | |
# Example usage | |
if __name__ == "__main__": | |
print("\n--- String Examples ---") | |
# Example 1: "hello" vs "hello" with no wildcards | |
print_match_summary("hello", "hello") | |
# Example 2: "hello" vs "hallo" with no wildcards - expect distance of 1 | |
print_match_summary("hello", "hallo") | |
# Example 3: "hello" with 3rd position (index 2) as wildcard vs "hallo" - expect distance of 0 | |
print_match_summary("hello", "hallo", wildcard_offsets_seq1=[2]) | |
# Example 4: "hello" vs "hillo" with 2nd position (index 1) as wildcard in seq2 - expect distance of 0 | |
print_match_summary("hello", "hillo", wildcard_offsets_seq2=[1]) | |
# Example 5: Multiple wildcards in seq1 | |
print_match_summary("hello", "haxyz", wildcard_offsets_seq1=[2, 3, 4]) | |
print("\n--- Bytes Examples ---") | |
# Example 6: Working with bytes | |
print_match_summary(b"hello", b"hallo") | |
# Example 7: Working with bytes with wildcard | |
print_match_summary(b"hello", b"hallo", wildcard_offsets_seq1=[2]) | |
# Example 8: Mixed types (bytes and string) | |
print_match_summary(b"hello", "hallo", wildcard_offsets_seq1=[2]) | |
# Example 9: Non-printable bytes example | |
print_match_summary( | |
b"\x01\x02\x03\x04", b"\x01\x05\x03\x04", wildcard_offsets_seq1=[1] | |
) | |