A Finite State Transducer (FST) based Font Converter

This paper describes the rule based approach towards the development of an Oriya Font Converter that effectively converts the SAMBAD and AKRUTI proprietary font to standardize Unicode font. This can be very much helpful towards electronic storage of information in the native language itself, proper search and retrieval. Our approach mainly involves the Apertium machine translation tool that uses Finite State Transducers for conversion of symbolic data to standardized Unicode Oriya font. To do so it requires a map table mapping the commonly used Oriya syllables in Proprietary font to its corresponding font code and the dictionary specifying the rules for mapping the proprietary font code to Unicode font. Further some unhandled symbols that appear in the intermediate converted file are rectified by Flex scanner tool. The converted text thus obtained is in standard Unicode font and remains unchanged as Unicode font is supported by almost all the platforms.