|
|
Boost : |
Subject: Re: [boost] [yap] review part 3: tests + misc + summary
From: Zach Laine (whatwasthataddress_at_[hidden])
Date: 2018-02-21 21:48:54
On Wed, Feb 21, 2018 at 11:31 AM, Steven Watanabe via Boost <
boost_at_[hidden]> wrote:
> AMDG
>
> On 02/21/2018 10:11 AM, Zach Laine via Boost wrote:
> > On Wed, Feb 21, 2018 at 10:28 AM, Steven Watanabe via Boost <
> > boost_at_[hidden]> wrote:
> >
> >> On 02/20/2018 11:16 PM, Zach Laine via Boost wrote:
> >>> On Tue, Feb 20, 2018 at 9:29 AM, Steven Watanabe via Boost <
> >>> boost_at_[hidden]> wrote:
> >>> <snip>
> >>> That looks like a great candidate for an example, so I made one out of
> >> it:
> >>>
> >>> https://github.com/tzlaine/yap/commit/4b383f9343a2a8affaf132c5be1eeb
> >> 99a56e58df
> >>>
> >>> <snip>
> >>> [snip]
> >>>
> >>
> >> evaluate(let(_a = 1_p << 3) [
> >> _a << "1", _a << "2"
> >> ], std::cout); // prints 3132, but should print 312
> >>
> >
> > Why should that print out 312? Isn't equivalent to:
> >
> > std::cout << 3 << "1", std::cout << 3 << "2"
> >
> > ? If not, why not?
> >
>
> The expected behavior is that _a is
> a local variable, not macro-like substitution.
>
> let(_a=<expr>) [ <body> ]
>
> should be equivalent to
>
> auto _a = <expr>;
> {
> <body>;
> }
>
> rather than:
>
> #define _a <expr>
> <body>;
>
> >
> >> Also,
> >> let(_a=_a+_a)[let(_a=_a+_a)[let(_a=_a+_a)[...]]]
> >> has exponential cost.
> >
> >
> > Sure. It's also not allowed, though. From the let docs:
> >
>
> Right. Assume the whole thing is wrapped in let(_a=1_p)[]
> and then it is legal. (each _a refers to the _a in the
> outer scope.)
I think I get it now. It seems evaluation *must* happen outer-to-inner,
not inner-to-outer, and that you need substitutions for things like
placeholders so you can do partial evaluations in order to figure out what
type a let-placeholder will be. This is the first really compelling use
case I have seen for why transform_evaluate() is necessary.
Zach
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk