Typing Sanskrit

[Peter J. Claus]

Although I, too, am not a Sanskritist and I try to avoid the task of 
writing ANY SA language in Romanized script (for the languages I work in, 
if the reader knows the language, they know the script), it seems to me 
that a strong advantage of a transcription method for computers that uses as 
few 'search and replace' variables is preferable over one which requires 
many.  Someone else mentioned using macros. A macros devised to search 
and replace, say, '@' throughout the text could rely on a simple 'case 
statement' to change a@, i@, u@ ... d@ t@ (retroflexes) ... etc.  The 
addition of only one or two others could be used to handle languages with 
three or four variants of an ascii character (eg. s, s@ [sh], s# [sya]). 
Furthermore, making a number of variant transcription styles using the 
same simple macro, changing only a couple of line, also would be 
facilitated.

Eg., in editing contributions to an encyclopedia, I have found it 
necessary to send the press one form of transcription 
(K<r/.bl><s/.bl><n/.bl>a  is how they want Kr at s@n at a) and to send the 
contributors their edited copy in hardcopy with the diacritics in place, 
above or below the line.  The two different versions of an entire 
manuscript with 100 transliterated terms can be done in a matter of two or 
three seconds. 

Finally, with the same set of @s, #s, and maybe ~s, one can do all SA and 
Middle Eastern scripts (given, say, LC equivalency charts).

The system is not particularly pretty, but can be done on a typewriter, 
too.  It is not, however, MY system, but one some body has already 
suggested to the LIST.