Unified table approach for typographic rendering

Abstract

As digital documents become more widely available, efficient text rendering techniques will be needed to display them in an aesthetically pleasing form. In conventional rendering algorithms, separate data structures (tables) and subroutines are used to store information relevant to each typographical feature. This incurs a substantial computational overhead due to redundant computation, especially for large character sets such as Unicode. This paper presents a new optimized approach to text rendering based on the concept of a generic unified table that incorporates data structures for all typographic features, and thus allows rendering to be performed with fewer table lookups. This approach has the potential to bring major efficiency improvements in text rendering and alignment, and is expected to enhance current document viewers and Web browsers significantly.