[Fusion] Adapting read-only fields of ADTs - Boost-users - lists.preview.boost.org

newer
Cycle detection in a directed...

[Fusion] Adapting read-only fields of ADTs

older
a quick question regarding...

Sebastian Redl

11 Apr 2011 11 Apr '11

7:18 a.m.

Hi, How would I use BOOST_FUSION_ADAPT_ADT to turn an ADT that doesn't expose setters into a read-only sequence? In other words, what do I pass as set_exprN if there is no valid set expression? Sebastian

Reply

Sign in to reply online Use email software

Show replies by date

Tore Halvorsen

11 Apr 11 Apr

11:05 a.m.

On Mon, Apr 11, 2011 at 9:18 AM, Sebastian Redl wrote:

Hi,

How would I use BOOST_FUSION_ADAPT_ADT to turn an ADT that doesn't expose setters into a read-only sequence? In other words, what do I pass as set_exprN if there is no valid set expression?

Just pass it nothing. e.g. BOOST_FUSION_ADAPT_ADT( name, (std::string const&, std::string const&, obj.get(), ) ; -- Eld på åren og sol på eng gjer mannen fegen og fjåg. [Jøtul] <demo> 2011 Tore Halvorsen || +052 0553034554

Reply

Sign in to reply online Use email software

Joel de Guzman

12 Apr 12 Apr

3:15 p.m.

On 4/11/2011 7:05 PM, Tore Halvorsen wrote:

On Mon, Apr 11, 2011 at 9:18 AM, Sebastian Redl wrote:

...
Hi,

How would I use BOOST_FUSION_ADAPT_ADT to turn an ADT that doesn't expose setters into a read-only sequence? In other words, what do I pass as set_exprN if there is no valid set expression?

Just pass it nothing.

e.g.

BOOST_FUSION_ADAPT_ADT( name, (std::string const&, std::string const&, obj.get(), ) ;

That's a good workaround, but really, we ought to have read-only adaptors. Doing above would still allow elem = val; albeit will be a no-op. Please add a trac "feature request" ticket, if you will. Regards, -- Joel de Guzman http://www.boostpro.com http://boost-spirit.com

Reply

Sign in to reply online Use email software

Christopher Schmidt

5:18 p.m.

Joel de Guzman writes: [...]

That's a good workaround, but really, we ought to have read-only adaptors. Doing above would still allow elem = val; albeit will be a no-op.

Please add a trac "feature request" ticket, if you will.

Regards,

A simple workaround is to fall back to an invalid expression. --8<---------------cut here---------------start------------->8--- #include #include template<typename>struct dummy{}; template<typename T> typename dummy<T>::this_type_does_not_exist invalid_operation(T const&) {} class A{}; BOOST_FUSION_ADAPT_ADT(A,(int,int,1,invalid_operation(val))) int main() { A a; boost::fusion::front(a)=1; } --8<---------------cut here---------------end--------------->8--- This should give a nice and readable error message. -Christopher

Reply

Sign in to reply online Use email software

Joel de Guzman

11:51 p.m.

On 4/13/2011 1:18 AM, Christopher Schmidt wrote:

Joel de Guzman writes:

[...]

...
That's a good workaround, but really, we ought to have read-only adaptors. Doing above would still allow elem = val; albeit will be a no-op.

Please add a trac "feature request" ticket, if you will.

Regards,

A simple workaround is to fall back to an invalid expression.

[...] Good idea, Christopher! So perhaps this is just a matter of documentation? Regards, -- Joel de Guzman http://www.boostpro.com http://boost-spirit.com

Reply

Sign in to reply online Use email software

5066

Age (days ago)

5067

Last active (days ago)

Download

4 comments

4 participants

tags

participants (4)

Christopher Schmidt
Joel de Guzman
Sebastian Redl
Tore Halvorsen