- From: Simon Sapin <simon.sapin@exyr.org>
- Date: Thu, 24 Apr 2014 21:35:37 +0100
- To: "Tab Atkins Jr." <jackalmage@gmail.com>
- CC: Anne van Kesteren <annevk@annevk.nl>, "www-style@w3.org" <www-style@w3.org>
On 24/04/2014 20:39, Tab Atkins Jr. wrote: > On Thu, Apr 24, 2014 at 9:22 AM, Simon Sapin <simon.sapin@exyr.org> wrote: >> On 18/04/2014 19:09, Tab Atkins Jr. wrote: >>> I plan to rewrite the parsing section into a Syntax-style parser, at >>> which point it would have the same structure as Syntax wrt "parse a >>> foo" algorithms. Would that work? >> >> Why is this desirable? >> >> As far as I can tell, the reason Syntax moved away from a grammar was to be >> able to precisely define non-trivial error handling behaviors. But the error >> handling of Selectors *is* trivial: if the input does not match the grammar, >> it’s invalid. Done. >> >> When it makes sense and is sufficiently precise, I’d prefer to have a >> concise grammar than an algorithm written in English. > > Yeah, I thought about this more, and decided that just using CSS-style > grammar instead of Bison-style would be sufficient. Sounds good. -- Simon Sapin
Received on Thursday, 24 April 2014 20:36:03 UTC