[Boost-users] [Spirit] uint_parser inconsistent results

Subject: [Boost-users] [Spirit] uint_parser inconsistent results
From: Michael Powell (mwpowellhtx_at_[hidden])
Date: 2018-11-20 16:47:18

Hello,

I am trying to figure out why the uint_parser yields what looks like
raw characters in one instance, yet in another instance yields the
original given phrase. What do I mean by that?

I have a string-literal rule, subset of the Google protobuf v2
grammar, and the octal-escaped rule yields a raw character, but the
"same" hex-escaped rule yields the given input string.

Both rules are of type qi::rule<It, char()>. And the str_lit rule is
qi::rule<It, std::string()>, basically.

oct_esc %= '\\' >> uint_parser<unsigned char, 8, 3, 3>{};
hex_esc %= no_case["\\x"] >> uint_parser<unsigned char, 16, 2, 2>{};
char_val %= hex_esc | oct_esc | char_esc | ~char_("\0\n\\");
str_lit %= ("'" >> *(char_val - "'") >> "'")
    | ('"' >> *(char_val - '"') >> '"')
    ;

struct escapes_t : qi::symbols<char, char> {
    escapes_t() {
        this->add("\\a", '\a')
            ("\\b", '\b')
            ("\\f", '\f')
            ("\\n", '\n')
            ("\\r", '\r')
            ("\\t", '\t')
            ("\\v", '\v')
            ("\\\\", '\\')
            ("\\'", '\'')
            ("\\\"", '"')
            ;
    }
} char_esc;

The only thing I can figure is that perhaps it has something to do
with no-case having a secondary effect on the rule?

I do not care which way it lands, per se, in either raw or original
string, although landing in the raw, that is, unescaped, has some
advantages, I suppose.

However, it is a mystery to me how there are two different outcomes.

Any thoughts? Suggestions?

Thanks!

Michael Powell