Boost logo

Boost :

From: andy.elvey_at_[hidden]
Date: 2001-05-25 02:26:08


Hi all.

 I've been looking through the discussion of Spirit's syntax - very
enlightening it is , too ... :-)

 Although I'm a newcomer to the parsing area, I thought I'd add my 2c
worth anyway (on the theory that (hopefully!) a fresh and new pair of
eyes from outside may be useful .

 I've found the current syntax of Spirit to be not confusing at all.
Given that some with vastly more experience than me *may* find it
confusing, I was wondering - would it be possible to have (say) a
*flag* of some kind , so that users could have a *choice* of syntax?

 By choice, I mean only two choices - the current syntax , or
something else ....

 For example , if I (as a newbie) wanted to use Spirit with its
current syntax , I could put "Spirit -c" ( for "classic" syntax -
or "newbie" syntax, if you like) .

   Someone else may find the current syntax confusing. In that case,
they could put "Spirit -a" (for "alternative" syntax).

 I know the above probably sounds *really* silly ( :-) ) , but I have
to say that (as a newcomer), I'd been poking around and looking at
EBNF syntax here and there, and I *immediately* fell in love with
Spirit *as it is* (with the syntax it has now). I'm just trying to
put forward the possibility of having the option (of syntax "as it is
now" , or "other" , for those who'd prefer another syntax).

 Oh - there *is* another way around this - you could have a "newbie"
version of Spirit (with syntax as now) , and an "advanced" version of
it (with syntax as put forward elsewhere on this list). Just
another possiblity. I'm just saying that as a newbie, Spirit is
***way*** easier to understand than *anything* else I've come across
in the parser area. Would like to see it able to be used by all
skill levels, not only the gurus! :-))

  The above would no doubt add to the complexity of things, but I
thought I'd mention it anyway ..... :-) Looking forward to hearing
your comments - thanks for your time !

   
  


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk