Larry Evans

> > mentioning customization points. I'd suggest adding a topic entitled "Customization Points" which highlights all places where the user has the ability to customize. > > In addition, there...

> You just need forward declarations to break the cycle, no? You'd need to forward declare the result type of the as_spirit_parser customization, which is the type of some, presumably,...

To further clarify what the above code shows, I've shown the output with and without defined(AS_SPIRIT_PARSER_FWD). defined(AS_SPIRIT_PARSER_FWD): ``` -*- mode: compilation; default-directory: "~/prog_dev/boost.org/boost.exploratory/boost.replacements/as_spirit_parser/test/x3/" -*- Compilation started at Sun Dec 11...

Thanks @MatthewSuttles; I cloned the tkieley fork and used `-I` in the ruby command where `` is where the lib subdirectory is in the downloaded the fork and it no...

> I ported some weeks ago as an exercise the spirit::x3 example calc2.cpp augmented with building the AST and it's evaluation. It's still using integers as input (though returning double)...

> > ``` > > , "2+(3*4)" > > ``` > > @cppljevans Why did you put parentheses around `3*4`? Operator precedence (multiplication binds tighter than addition) should have been...

@LegalizeAdulthood, there's no need for the added complexity of vn_stack as demonstrated by: ```c++ /*============================================================================= Copyright (c) 2001-2014 Joel de Guzman Distributed under the Boost Software License, Version 1.0. (See...

> @cppljevans wrote: > > > Here's corrected code (with slight change in uint parser): > > ...but now it has `2*(3+4)` and is missing `2+3*4` which would demonstrate the...

> Zach, > > AFAICT the [BOOST_PARSER_DEFINE_IMP_macro](https://github.com/tzlaine/parser/blob/develop/include/boost/parser/parser.hpp#L5741) performs, "essentially" the same purpose as the [BOOST_SPIRIT_DEFINE_ macro](https://github.com/boostorg/spirit/blob/develop/include/boost/spirit/home/x3/nonterminal/rule.hpp#L228). > > Now, again, AFAICT, the purpose of those macros is to "associate" the...

> This doesn't seem any simpler to me. Here, we name the expected attribute type at the call site, but that may or may not be returned, because you changed...