>Well, because it wasn't possible at the time, I assume ;) >Times have changed since then ...
Indeed. Looks like it. :) >> What I wanted to know: >> - can I pass a phoenix expression to a decltype / BOOST_TYPEOF and >> default-construct it? >In general, no, because of various terminals that need to be initialized. >You can however limit the phoenix grammar to only those expressions that can >be default constructed. If you mean all except, say, ints or chars, it's something I can live with because I already do ;-) >Why not copy construct them? I don't see how it would help. I mean, the transition table I'm generating with the grammar ist and mpl::vector< Row<...>, Row<...> >. Then run my metaprograms and I have nothing left to construct from. >Yes, the entry point is pretty high, but i hope its worth it ... If you come >to the conclusion that writing your own library is faster and easier phoenix >maybe isn't quite >ready for prime time with that respect. No, it's not what I mean. The library is already written, since about 1 year, it's inside MSM's eUML front-end (see for example http://www.boost.org/doc/libs/1_46_0/libs/msm/doc/HTML/ch03s04.html#d0e2059). But the implementation will not win a beauty contest ;-) So I'm rewriting it and the examples I presented in this thread are only a cleaned-up, smaller version to show what I'm looking for, with some stuff I added after I realized it was not usable only with MSM. Now that I tried to explain what I was looking for and after your answers and some time I just spent investigating phoenix, I think it's worth giving a try with phoenix. So if you accept helping me a bit to come to speed with phoenix internals, I think MSM can pretty fast be your first "client" ;-) If you don't mind, I offer to close this thread and start hacking code. I already have the first questions ready. >I am looking forward to improvments in the docs, or additions to code you >might have! Sure, we'll see on the way what I'm missing. Thanks, Christophe _______________________________________________ proto mailing list [email protected] http://lists.boost.org/mailman/listinfo.cgi/proto
