Develop a stand-alone (USB portable) application to parse and modify an RTF document

Organisation: Starving writer :-)
Goals: More effective pre-editing of book manuscripts. More specifically, the stage where over-used and commonly used words are assessed and/ or reduced.

Available budget: Dependant on quotes - but I'm still a starving writer :-)

Rights:
As an end product, I want exclusive ownership rights, including the right to sell or give away the application.

Purpose:
Standalone application, capable of being run from a USB key.
Application will process two RTF files. One will be a book manuscript. The second will be a reference file.

The Reference file:
User editable. Line delimited list of character strings. The character strings may be a word or phrase, and may include spaces. Each line in the reference file will be highlighted with a color of the user’s choice. An example might be:

that -- Highlit in red
he whispered --- Highlight in green
very --- Highlight in blue

The application will read the reference file and the manuscript and create a copy of the manuscript. In the copy, every occurrence of the line text from the reference file (case insensitive) would be highlit with the same color that text had in the reference file. For example, using the reference file above:

It was raining. That, Dirk thought, was just typical. “Very funny.” he whispered to the gods he didn't believe in. “Every bloody time.” He tucked the delivery under his arm. Time to go.

The copy file would have 'That' highlit in red, 'he whispered' highlit in green and 'Very' highlit in blue. But the 'very' in 'Every' and 'delivery' would not be highlit as it is part of different words. The application should ignore punctuation, so:

. Very
, very
.Very
,very
“Very
(very

And similar structures would be highlit.
The reference file should accomodate wild card elements. For instance:

*ly --- All words ending in 'ly'
comb* --- All words beginning 'comb'
*tin* --- all words that have the string 'tin' in them