Re: question about linker

On 12/3/2024 1:27 PM, Janis Papanagnou wrote:

On 03.12.2024 19:57, BGB wrote:

On 12/2/2024 12:13 PM, Scott Lurndal wrote:

>
Indeed.  One wonders at Bart's familiarity with formal grammars.

>
In my case, personally I haven't encountered much that doesn't work well
enough with a recursive-descent parser.

 Is that meant as a contradiction? - If so, how?
 

Formal grammars and parser generators aren't usually necessary IME, since recursive descent can deal with most everything (and is generally more flexible than what one can deal with in a formal grammar).
Though, it seems I didn't read enough, they were debating about syntactic use of semicolons as a separator rather than parser writing...
But, yeah, in that context I think it is more a tradeoff between semicolons and syntactically significant line-breaks.
Can go either way, where line-breaks make more sense if one is wanting to design a parser like:
   Read one line of input;
   Break into tokens;
   Parse a statement.
But then the language generally needs an explicit line continuation character, say:
   c=func(a, \
     b, c)
And, debate whether it is better to have a semicolon on every line, or a line continuation to break up long lines.
Separators can serve to help avoiding expressions merging in unintended ways, but could be avoided or left optional if one is going for a simplistic syntax, say:
   proc A B C
   dst = func A B C
Probably with the language disallowing the creation of any complex expressions (and/or, one goes down the path that leads to either S-expressions or Smalltalk syntax).

Janis