Would the team please consider adding some type of Fuzzy Search option?
General description:
The end user would hit the search box, which would include a new option in the field list just below "All", e.g. named "All (fuzzy)".
Entering a search phrase would produce matches that don't require the user to be accurate with their spelling – ignoring punctuation, symbols, vowels, mistaken double-letters, etc.
-------------------------
The algorithm isn't at all complex, I have written something very much along the following lines for a search facility in a music catalogue system, loosely as follows:
Database would store a new redundant internal 'fuzzy' text field for each track, populated by a concatenation of filtered characters from pertinent text fields, inc.Track Title, Artist, Album. The fuzzy algorithm/subroutine would filter-in only consonant, numeric digit characters, (either rejecting vowels or normalising 'AEIOU' characters to a token), rejecting consecutively repeating characters, and (like iTunes allows for) normalising/converting accented character variants to their bare non-accented counterpart.
Optional enhancement to normalise similar sounding consonants, e.g. M and N, which sound similar (and are neighbouring keys on an english keyboard).
Examples :
Input Album "Café Del Mar Volumen Dos", Artist: "Mark's & Henry's", Track Title "(The Making Of...) Jill"
Output excluding vowels (for loose results): "CFDLMRVLMNDSMRKSHNRYSTHMKNGFJL"
Output including tokenised vowels (for tighter results): "C#F#D#LM#RV#L#M#ND#SM#RKSH#NR#STH#M#K#N#G#FJ#L"
Happy to discuss further :)