nv_addtype() is never called with optstr set to NULL

[krader1961]

While working on issue #507 I noticed that nv_addtype() is only called from two places and neither passes a NULL pointer for the optstr parameter. So this block of code is bogus:

ast/src/cmd/ksh93/sh/nvtype.c

Lines 732 to 736 in c68faf0

	if (optstr) {
	cp->optstring = optstr;
	} else {
	cp->optstring = sh_opttype;
	}

The condition is always true, thus the else block will never be executed. This is also true in the ksh93u+ branch (ast.master) which includes one other call site that has since been removed.

[krader1961]

This is, in one sense, a trivial issue. Perhaps there was a time long ago (> 7 years) in a pre-ksh93u+ release where nv_addtype() was called with optstr == NULL (but I suspect that is not true). I opened this issue because I want to communicate to other people interested in this project that cleaning up cruft is a worthy goal in its own right. Every time irrelevant code is deleted or simplified the overall code quality is improved.