Devanagari Text

The second part of the input file on which the preprocessor acts is actual DN text. The remainder of the input file (text in latin script, TEX-commands, etc. ) is left untouched. Of course the preprocessor must be told where DN text starts or ends. You can delimit DN text in two ways. Firstly, DN text can be surrounded by {\dn and } . ( means a space character). I.e. {\dn acchaa} produces aQCA. Secondly, if the preprocessor command @dollars is given, DN text can also be delimited by dollars: $acchaa$. The second method is convenient when you have to switch often between DN text and text in latin script. For large pieces of text in DN however, the first method is to be preferred, because this way it is easier to determine where DN text ends and where it begins.

What can be typed between DN text-delimiters?

From the above and section 2 it follows that the following characters are illegal between DN-delimiters:
^ < > x w F N Q S V W X Y Z and the following characters when they are not used in a combination mentioned in the table on page 2: " . and ~