Using the result of a Phoenix function in the parse

I have a Phoenix function that returns bool and I'd like to use this
result in the parse; i.e., fail if the function returns false. The
techniques I'm familiar with all involve assigning the result to a
variable and then introducing another function object to test the value.
Is there a less verbose way of doing this? (That is, a way that doesn't
involve writing another function object just to test the value.)

--

-- 
Braden McDaniel                           e-mail: <braden <at> endoframe.com>
<http://endoframe.com>                    Jabber: <braden <at> jabber.org>

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Problem with wave in boost 1_34_0

Hi, Hartmut and wave experts,

I had a program using wave in boost_33_1 that worked perfect. It used wave
library as a preprocessor to a script parser, and it worked with boost_33_1.
But when I was trying to compile it with wave in boost 1_34_0, I got error.
I think it might be a known problem to you guys and you might know the
solution to the problem. 

The program code that give error was:
    lexer_type it = lexer_type(instr.begin(), instr.end(), pos,
boost::wave::language_support(
boost::wave::support_cpp|boost::wave::support_long_long));

And the error was:

C2039: 'support_long_long' : is not a member of 'boost::wave'
1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2065:
'support_long_long' : undeclared identifier
1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2514:
'boost::wave::language_support' : class has no constructors
1>
c:\users\tools\boost\boost_1_34_0\boost\wave\language_support.hpp(25) : see
declaration of 'boost::wave::language_support'
1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2661:
'boost::wave::cpplexer::lex_iterator<TokenT>::lex_iterator' : no overloaded
function takes 4 arguments

Can you please point me to the right direction or give me a tip to resolve
this problem?

Re: Problem with wave in boost 1_34_0

Fanz, 

> I had a program using wave in boost_33_1 that worked perfect. 
> It used wave library as a preprocessor to a script parser, 
> and it worked with boost_33_1.
> But when I was trying to compile it with wave in boost 
> 1_34_0, I got error.
> I think it might be a known problem to you guys and you might 
> know the solution to the problem. 
> 
> The program code that give error was:
>     lexer_type it = lexer_type(instr.begin(), instr.end(), 
> pos, boost::wave::language_support( 
> boost::wave::support_cpp|boost::wave::support_long_long));
>    
> And the error was:
> 
> C2039: 'support_long_long' : is not a member of 'boost::wave'
> 1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2065:
> 'support_long_long' : undeclared identifier
> 1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2514:
> 'boost::wave::language_support' : class has no constructors
> 1>
> c:\users\tools\boost\boost_1_34_0\boost\wave\language_support.
hpp(25) : see declaration of 'boost::wave::language_support'
> 1>c:\users\diamond\src\scriptpreproc.cpp(80) : error C2661:
> 'boost::wave::cpplexer::lex_iterator<TokenT>::lex_iterator' : 
> no overloaded function takes 4 arguments
> 
> Can you please point me to the right direction or give me a 

Re: Range checking problem with uint_parser?

> > Actually, I ended up digging into this and fixing it 
> myself, so here goes.
> > 
> > The implementation I've attached works much the same way as 
> in Spirit2. 
> > The key is to do the overflow/underflow tests before actually doing 
> > the calculation, not afterwards.
> 
> Oh I see. Yeah. Spirit2 fixes this problem alright. I did 
> have a patch that someone submitted. I have it somewhere. It 
> has some optimizations to for typical cases (e.g. not doing 
> the checks when unnecessary). I'll look it up and apply it to 
> Spirit1 when I get back home. In the meantime, your fix will 
> surely suffice.
> 
> > I don't know what format you use for diff's so I'm just 
> attaching the 
> > entire file (the fix is against spirit 1.8.4 (shipping with 
> boost 1.34).
> > 
> > If you'd prefer the diffs, please let me know how to 
> produce them and 
> > I'll post these instead (would 'svn diff' work for you?)
> > 
> > I've tested against MSVC8, gcc-3.4.6 and gcc-4.1.2 which are the 
> > compilers I have access to right now.
> 
> Can we wait before I go back? I do not have access to all my 
> files. It'll be another 2 weeks though :(, so if someone 

Re: Range checking problem with uint_parser?

Hartmut Kaiser wrote:

>>> Incidentally, the bug I fixed revealed a bug in the test suite: In 
>>> numeric_tests.cpp, line 119, max_octal should be "37777777777", not 
>>> "77777777777".
>> Yeah, I noticed that too when I wrote the Spirit2 numerics.
> 
> I applied Martins patch to the Boost CVS::HEAD (it hit me while
> preprocessing certain constructs in Wave). Should we try to get this into
> 1.34.1 as well?

Yes. It is a bug.

Regards,
--

-- 
Joel de Guzman
http://www.boost-consulting.com
http://spirit.sf.net

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[Phoenix2] lambda an let tests fail to compile

Hello,
I'am playing with Phoenix 2 for a few days and finds it very usefull  
(could be "promoted" just as Fusion 2 was IMO).

I'm increasing the complexity of my use cases and I'm now facing a problem  
with let and lambda scope utility.
My toolset is MSVC8 SP1 and I follow Harmut advices for the code : a fresh  
checkout from boost CVS and from spirit SVN (spirit/trunk/final). The  
headers are included in the right order.

The tests that fail are those with let/lambda and local variables. Without  
local variables it's fine. One of the failling test is :

     {
         int x = 1, y = 10, z = 13;
         BOOST_TEST(
             lambda(_a = _1, _b = _2)
             [
                 _1 + _a + _b
             ]
             (x, z)(y) == x + y + z
         );
     }

They are (all) failling in the same place, that is in the nested result  
meta-function of the initialize_local class where result is instantiated  
(Actor) with a function type instead of a class type. The compiler output  
stands bellow.

Re: [Phoenix2] lambda an let tests fail to compile

Samuel DEBIONNE wrote:
> Hello,
> I'am playing with Phoenix 2 for a few days and finds it very usefull  
> (could be "promoted" just as Fusion 2 was IMO).
> 
> I'm increasing the complexity of my use cases and I'm now facing a problem  
> with let and lambda scope utility.
> My toolset is MSVC8 SP1 and I follow Harmut advices for the code : a fresh  
> checkout from boost CVS and from spirit SVN (spirit/trunk/final). The  
> headers are included in the right order.
> 
> The tests that fail are those with let/lambda and local variables. Without  
> local variables it's fine. One of the failling test is :
> 
>      {
>          int x = 1, y = 10, z = 13;
>          BOOST_TEST(
>              lambda(_a = _1, _b = _2)
>              [
>                  _1 + _a + _b
>              ]
>              (x, z)(y) == x + y + z
>          );
>      }
> 
> They are (all) failling in the same place, that is in the nested result  
> meta-function of the initialize_local class where result is instantiated  
> (Actor) with a function type instead of a class type. The compiler output  
> stands bellow.
> 

Learning Spirit2

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

_______________________________________________
Spirit-general mailing list
Spirit-general <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/spirit-general

Re: Learning Spirit2

Hi,

> Is there anywhere(other than the source) we can find more 
> information about spirit2 than what's found in the 
> presentation slides in the doc folder? I find it somewhat 
> hard to piece together from the samples and source 
> alone(although I kinda managed so far)... Is there access to 
> a (possibly incomplete) documentation for it? 

Nope, sorry...

> Failing that, can someone answer my current question, which 
> is as simple as how to achieve the same functionality in 
> spirit2 that was obtained from chset in spirit1.8?

Chset isn't implemented in spirit2 yet. 

> Also, would you say it's a better idea to try to reuse some 
> spirit1.8 grammar code(that use closures and phoenix 
> functions) to convert it to spirit2 or should I just start 
> rewriting it from the original ebnf spec?

Converting from the 1.8 grammar shouldn't be too hard. The required changes
should be straightforward.

Regards Hartmut

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Is permutation parser available in spirit-2.0

Hi,

Is permutation parser available in boost-spirit 2.0 ? If yes, can any one point
me towards its docs?

Regards,
Reetesh Mukul

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/