PHP Reference ยท String Functions
PHP levenshtein() function
Reference for PHP levenshtein(): edit distance between two strings, custom insert, replace and delete costs, typo suggestions and byte-based limits.
levenshtein() returns the Levenshtein distance between two strings: the smallest number of single-character insertions, replacements and deletions needed to turn the first string into the second. A distance of 0 means the strings are identical.
It is the standard tool for "did you mean" suggestions and for fuzzy matching of short words.
<?php echo levenshtein("kitten", "sitting"); // 3
Try it
Press Run (or Ctrl + Enter) to see the result here.
Program input ยท what you'd type while the program runs
Type the answers before you press Run, one per line. Each time the program reads a line (with fgets(STDIN)), it gets the next line from this box.
Press Run (or Ctrl + Enter) to see the result here.
Program input ยท what you'd type while the program runs
Type the answers before you press Run, one per line. Each time the program reads a line (with fgets(STDIN)), it gets the next line from this box.
Syntax
levenshtein( string $string1, string $string2, int $insertion_cost = 1, int $replacement_cost = 1, int $deletion_cost = 1 ): int
| Name | Description |
|---|---|
$string1 |
The string to start from. |
$string2 |
The string to reach. |
$insertion_cost |
Cost of inserting a character into $string1. Default 1. |
$replacement_cost |
Cost of replacing a character. Default 1. |
$deletion_cost |
Cost of deleting a character. Default 1. |
Return value: the total cost of the cheapest sequence of edits.
Usage notes
- No length limit since PHP 8.0. Older versions returned
-1for strings longer than 255 characters. - Case-sensitive. Lowercase both strings for case-insensitive matching.
- Bytes, not characters. An accented UTF-8 letter counts as two bytes, so distances for non-ASCII text are inflated. Convert to single-byte characters (for example via
mb_str_split()and your own function) when it matters. - Cost grows with length product. Comparing two strings of lengths m and n takes about m times n steps; it is fine for words, not for whole documents.
- Swapping two neighbors costs 2, since there is no transposition operation (that would be Damerau-Levenshtein).
- Choosing a threshold: a distance of 1 or 2 is typical for typo suggestions on words under ten letters; scale with word length for longer text.
- Sound-alike matching is a different task; see
soundex()andmetaphone().