A tutorial on Fonts for Indian Languages:  Section-3

Problems in designing fonts for Indian languages

  In respect of Indian language text display and processing, we are confronted with the following issues. There is no simple solution to the problems faced. 
 

We cannot have standard fonts for each script as the number of glyphs required to form the aksharas  properly is many more than 256 for almost all the languages, except Tamil. Any attempt at standardization will necessarily have to compromise on the approach to displaying the conjuncts. Worse still, this will have to be done individually for each script.

It may be possible to agree on a compromise set of glyphs for each script and request font designers to place them at specified locations in the fonts. This way, some uniformity may be possible in displaying the text for that script. However, this may not be feasible in practice since printing requirements are fairly stringent on the conjuncts and require that many additional glyphs be included.

 An eight bit representation of the consonants and vowels may be feasible but one will have to live with multibyte codes for the aksharas. ISCII was an early attempt at achieving some uniformity but this was not sustainable across all the Indian languages. An eight bit representation will permit the use of an eight bit font and this will be useful since most applications today can handle eight bit fonts. However, coding only the basic vowels and consonants will restrict the display. A separate mechanism is required to go from a multibyte representation of a syllable to its shape. In reality, this mechanism will be  extremely complex.

 We need to support automatic transliteration to permit language independent information to be read in any script. Here one is confronted with the identification of a global set of aksharas. Fonts for different languages may require different number of glyphs and it will not be possible to achieve transliteration by changing the font face.

 A one to one mapping between an akshara and a glyph is ruled out if eight bit fonts are used. As of now, 16 bit fonts are not handled properly in most computer systems. 

 There is no clue to how many aksharas are actually needed in a language, for many new samyuktaksharas may be formed. the set of currently used syllables may eventually increase. It may be possible to design a font with about 300 glyphs in order to render most of the samyuktaksharas but not many applications can work with such fonts. Applications handling Unicode may be able to do this but Unicode for Indian languages also implies a multibyte scheme. Special Opentype fonts have been recommended for handling variable length codes but this does not work well in practice. Thus Unicode fonts will not really solve the problem.

 It is not merely the aksharas we must display. For educational needs, we also must display the symbols for the matras along with special letters and punctuation. These should also be coded into the character set and explicit glyphs for these must be provided. Due to the variable physical width of the aksharas themselves, more than one representation for a matra may be needed (see the glyphs of the Xdvng font).

 
Unicode Fonts for Indian languages

  Unicode has emerged as a meaningful standard for multilingual work and developers all over the world are taking advantage of the help provided by Operating systems to work with 16 bit codes. It turns out that Unicode does offer solutions for scripts which are based on syllables but the solutions depend on the availability of API to deal with Unicode strings. Unfortunately, the support for Unicode in respect of Indian languages has not been incorporated to any satisfactory extent in almost all the systems. The issues involved in using Unicode for computing with Indian languages are quite complex and cannot be discussed in this tutorial. This web site has a comprehensive set of pages devoted to this subject. Unicode for Indian languages is discussed in detail in the linked pages. The fonts issue is included as an important topic for discussion.

Back to Contents

The concept of the font

Font encoding

Fonts for Indian languages/scripts

Problems in designing fonts for our languages/scripts

Unicode fonts for Indian languages

Fonts as used in web pages

In summary

On account of this, it is not possible to define a character set and an accompanying encoding scheme that would apply to all the scripts uniformly. 

There is a need to look at this problem with some seriousness if we need to develop a standard scheme for displaying Indian scripts using eight bit fonts.

Unicode does not specify a meaningful character set for any of the Indian languages. The font for use with Unicode has to provide for additional flexibilities in being able to map a Unicode string for a syllable to the required shape of the syllable. Conventional fonts will not be able to satisfy this requirement. Opentype fonts are offered as the solution but these will not really provide an answer.

Acharya Logo
Text in Brahmi script at the Gate of the Great Stupa at Sanchi. The text records the donation of the pillar by a desciple of Arya Kshudra. The text reads "aya chuDasa atevAsino balamitasa dAnam thabho". More information about the Brahmi script is presented under Languages and Scripts.