Sinhala Font Design Simplified: Analyzing Legacy Fonts to Identify the Minimum Glyph Resembling Shapes and Automating Unicode Compliant Font Generation with FontForge

Abstract

ABSTRACT This thesis focusses on analyzing the ASCII based legacy fonts to identify common visual features of different glyphs in order to identify the minimum number of glyphs needed to be designed to create a Unicode compliant Sinhala font. Due to the lack of codepoints availability, most of the times, legacy Sinhala fonts consist only the glyphs which cannot be created by combining multiple glyphs. And since they are in common use among Sri Lankan people for almost 3 decades, this supports the hypothesis that the glyphs in these fonts can support most if not all Sinhala words in common use. We identified that with 594 glyphs we can design a Unicode compliant Sinhala font that would facilitate the common Sinhala writing. In order to find the minimum glyph count needed to design these 594 glyphs we have suggested 6 main groups that the glyphs in a legacy font can be divided into. And we also suggested that glyphs in 2 of these groups can also be grouped into 8 sub-groups depending on their visual similarities. Finally, we have identified that the minimum number of glyphs needed to be designed is 167 glyphs. And we also have suggested a way to reduce the number of glyphs to be designed to 81 by using glyphs from an open licensed font to generate glyphs for the codepoints in basic Latin character set. With identifying the minimum number of glyphs to be designed we move to the second part of this research which is to generate a Unicode compliant font using the minimum glyphs designed by utilizing the python scripting capabilities of FontForge software. Here we have identified that there are 2 ways of doing this by either creating everything including lookup tables programmatically using python or using an existing font to create a new font by replacing its glyphs and other general information.