A translation of Siderite Zackwehdex's Sift4 algorithm for ranking how close one string is to another, which is a few orders of magnitude faster than calculating the Levenshtein distance
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
// | |
// Super Fast and Accurate string distance algorithm: Sift4 | |
// http://siderite.blogspot.com/2014/11/super-fast-and-accurate-string-distance.html | |
// | |
public func sift4(firstString: String, secondString: String, maxOffset: Int = 5, maxDistance: Int? = nil) -> Int? { | |
if firstString.isEmpty { | |
return secondString.isEmpty ? 0 : count(secondString) | |
} else if secondString.isEmpty { | |
return count(firstString) | |
} | |
let s1 = firstString as NSString | |
let s2 = secondString as NSString | |
let l1 = s1.length; | |
let l2 = s2.length; | |
var c1 = 0 // cursor for string 1 | |
var c2 = 0 // cursor for string 2 | |
var lcss = 0 // largest common subsequence | |
var local_cs = 0 // local common substring | |
var numberOfTranspositions = 0 // e.g. 'ab' vs 'ba' | |
typealias Offset = (c1: Int, c2: Int, isTransposition: Bool) | |
var offsets: [Offset] = []; // offset pair array, for computing the transpositions | |
while ((c1 < l1) && (c2 < l2)) { | |
if (s1.characterAtIndex(c1) == s2.characterAtIndex(c2)) { | |
local_cs++; | |
var isTransposition = false | |
//see if current match is a transposition | |
var i = 0; | |
while i < count(offsets) { | |
var offset = offsets[i] | |
if c1 <= offset.c1 || c2 <= offset.c2 { | |
// when two matches cross, the one considered a transposition is the one with the largest difference in offsets | |
isTransposition = abs(c2 - c1) >= abs(offset.c2 - offset.c1) | |
if isTransposition { | |
numberOfTranspositions++ | |
} else if !offset.isTransposition { | |
offset.isTransposition = true; | |
numberOfTranspositions++; | |
} | |
offsets[i] = offset | |
break | |
} else { | |
if c1 > offset.c2 && c2 > offset.c1 { | |
offsets.removeAtIndex(i) | |
} else { | |
i++ | |
} | |
} | |
} | |
offsets.append((c1, c2, isTransposition)) | |
} else { | |
lcss+=local_cs | |
local_cs = 0 | |
if c1 != c2 { | |
let min_c = min(c1, c2) //using min allows the computation of transpositions | |
c1 = min_c | |
c2 = min_c | |
} | |
//if matching characters are found, remove 1 from both cursors (they get incremented at the end of the loop) | |
//so that we can have only one code block handling matches | |
for var i = 0; i < maxOffset && (c1+i < l1 || c2+i < l2); i++ { | |
if (c1 + i < l1) && (s1.characterAtIndex(c1 + i) == s2.characterAtIndex(c2)) { | |
c1 += i-1 | |
c2-- | |
break | |
} | |
if (c2 + i < l2) && (s1.characterAtIndex(c1) == s2.characterAtIndex(c2 + i)) { | |
c1-- | |
c2 += i - 1 | |
break | |
} | |
} | |
} | |
c1++ | |
c2++ | |
if let max_d = maxDistance where (max(c1, c2) - lcss + numberOfTranspositions) > max_d { | |
return nil | |
} | |
// this covers the case where the last match is on the last token in list, so that it can compute transpositions correctly | |
if (c1 >= l1) || (c2 >= l2) { | |
lcss += local_cs | |
local_cs = 0 | |
let min_c = min(c1,c2) | |
c1 = min_c | |
c2 = min_c | |
} | |
} | |
lcss += local_cs; | |
return max(l1, l2) - lcss + numberOfTranspositions //add the cost of transpositions to the final result | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi, and thank you for using my algorithm.
First of all, I would like to update the URL of the Sift4 algorithm blog post. It is no longer located at https://siderite.blogspot.com/, instead it can be found here: https://siderite.dev/blog/super-fast-and-accurate-string-distance.html
Next, there is a bug that I have fixed today, related to maxDistance, and that has crept into this adaptation as well. The fix is:
Thanks!