[Grace-core] Comment proposal

[Andrew P Black]

On 4 Feb 2015, at 12:46 , Michael Homer <mwh at ecs.vuw.ac.nz> wrote:

> In particular, if you have to understand what a "syntactic unit" is
> and is not, it's not predictable. Novice programmers are not
> linguists. 

This is a good point.   A layout rule like Fortran's (C in column 6) or a lexical rule like Algol's (; comment ... ; can appear where a ; could appear) can be applied without understanding the language structure.

Unfortunately,  such rules don't tell us what we need to know: where in the parse tree the comments belong.  They also don't let the programmer 
put short, partial-line comments close to the things that they wish to comment upon — which is, I believe, why modern languages use more flexible rules.

I like Lex's proposal of specifying where comments can go, and what they mean.   Something like:

	• at the end of a line, (but not at the end of a continuation that is continued on the following line).  
	  Such comments attach to the first declaration or expression on that line.
	• Following an opening { or (, that is, separated from the opening bracket only by newlines and spaces.
	  Such comments attach to the whole unit surrounded by the { or ( and the matching closing bracket.

These may be all that we need.  Of course, interpreting such rules still means that the reader has to understand a what a declaration or expression is.  Hence my contention that we should work on a publication grammar first, and then worry about comments. 

	Andrew