|
|
Boost : |
Subject: Re: [boost] [GSoC] Some new ideas for this year's GSoC
From: joel falcou (joel.falcou_at_[hidden])
Date: 2010-04-06 14:47:49
Topher Cooper wrote:
> Actually the difficulty is "easy". The basic precedence parser could be whipped out in an hour or so. Another day or a bit more to do the interfacing, define the expression data structures, and do the constant folding.
>
> Operators also need associativity, or, more expressively, two precedences -- a left and a right precedence.
>
> You don't really need to specify "allowed variables" any more than you need to specify allowed literals -- if the parser sees a variable it can use it.
>
> Personally I would include templates rather than just an operator string and then depend on the arity to determine the type. Then you can handle ternary and postfix operators.
>
> Topher
>
Isn't it somethign that cna be done in a few days using spirit anyway ?
-- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk