Offers keyboard aware text matching functions and data structures.
Typing filr, a typo of file.
A traditional fuzzy text matching function like Levenshtein distance simply counts the number of insertions, deletions, and subtitutions need to be made in order to turn a source word into a destination word. In this example, filr and file are of distance 1 from each other. Only the r needs to be substituted for an e.
Unfortunately, other words are also distance of 1 from filr, like fill and film. All of these words can be considered equal distance and equally likely to be the correct word.
Looking at a keyboard, it becomes obvious which of the three is the closest match to the misspelled word.
By asking not just "String Distance" but also physical distance, one can get a much more accurate answer for closet match.
