[Tugindia] Indexing Sanskrit words with diacritical marks

[c puram]

I seek the help of those who are more knowledgeable.

I have English text with Sanskrit names in italics and would like to index
the Sanskrit words.  For this purpose I created a macro file containing
lines like

\newcommand{\bali}{\index{\ixbali}{\ixbali}}
\newcommand{\bana}{\index{\ixbana}{\ixbana}}

In the text wherever I want Bali (no accent, but with italics) I type
{\bali}.  Wherever I want Bana (with a bar over a and a dot under n - all in
italics) I type {\bana}.  Then I have a second file which converts the
macros {\ixbali} and {\ixbana} as shown below.

\newcommand{\ixbali}{\it Bali}
\newcommand{\ixbana}{\it B\=a\d{n}a}

Those familiar with Devanagari script know that Bana (with a bar over a)
should follow Bali.  It does not as can be seen (really?) in the following.

Index
*Akr¯ura, 6
Ambar¯is. a, 5
B¯an. a, 7
Bali, 5
Bhr. gu, 7
Brahma, 1, 5
Citraketu, 5
Daks. a, 5*

How can I achieve this?  To those who have written such text in English,
should I do that as the book is in English?

A second question that may be related to this is the following.

This index excerpt shown is according to the English alphabet.  Can it be
modified to follow Sanskrit alphabet?

Thanks.

Chith Puram