Hi. Sorry about my bad picture and unattractive name. :)
My methodology depends on the text, and will be optimized while under development
BUT
I would simply search for the first letter til i find it, then check all remaining letters(if they are alle correct), and if it fails I would search the next position.
So basically you need an incrementing start position and an offset. :)
I could also easily pick out the closest match this way.
But you would normally cut of the search for near matches after one or two missing letters to increase speed.
I have done eye-scanning in javascript, and string manipulation in python for use with autoconverted PDFs to make them readable by speechsynthesis programs.
I am sure someone can do this faster(coding time) than me.
I consider myself a slow coder, but my code usually performs pretty good.
For me this project would be just within range of my capability, so I would be motivated by a challenge that is not to hard and not to easy. :)
Would love to hear from you.
I also use skype: Espen Hæhre
Good luck with your project.
PS: I understand i totally misunderstood the assignment. LOL.
The result will vary with search string length, so the program should
generate a table for different string lengths I suppose.
Then present this table, maybe with highlights for spikes or big variations.