- Boost-users - lists.boost.org

[Test/Build] has there been added a feature to specify the working directory of a unit-test?
by Thorsten Ottosen 21 Sep '11

21 Sep '11

Hi Everybody, This problem was dicussed some time ago in this thread: http://boost.2283326.n4.nabble.com/Current-directory-in-unit-tests-td269491… Have a solution been implemented to this problem? regards Thorsten

1 0

Passing format strings to printing tuples
by Krzysztof Żelechowski 21 Sep '11

21 Sep '11

The current tuple output mechanism wants me to say cout << formatting_manipulator << tuple; The expression (cout << formatting_manipulator) return an "embossed" cout, but currently it returns just cout with internal settings modified. The call must change things allocated within cout, like the standard manipulators do; although it is supported by the standard output, I find this approach so CFront-ish. How about using the syntax (cout << format_specification (tuple)) instead? Or even (cout << tuple / format_specification)? Chris

1 0

Re: [Boost-users] [asio] Interval Timer
by Andrew Holden 21 Sep '11

21 Sep '11

On Wednesday, September 21, 2011 8:15 AM, asif saeed wrote: > I need an interval timer to continuously call my callback function > after every n milliseconds/seconds for ever once I set it up. I > also need the ability to kill/cancel the timer. Does Boost provide > such a facility now? You could assemble that timer using asio's deadline_timer http://www.boost.org/doc/libs/1_47_0/doc/html/boost_asio/reference/deadl ine_timer.html The deadline_timer calls a callback at a specified time. Your callback would need to calculate the time of the next callback and restart the timer. Perhaps the following pseudocode could get you started: Public data members: * deadline_timer::duration_type trigger_interval * bool is_killed Private data members: * deadline_timer::time_type next_trigger * deadline_timer the_timer Constructor: is_killed = false Set trigger_interval next_trigger = now + trigger_interval Initialize the_timer Call the_timer.expires_at Callback: if (! is_killed and no errors happened) { next_trigger += trigger_interval Call the_timer.expires_at Do your processing }

1 0

from a variant to class with members templated over the variant types
by Olivier Tournaire 21 Sep '11

21 Sep '11

Hi all, I am not sure if the title is self explanatory, so, let me show an example of the goal. I have: variant<type1, type2, type3> v; I want: struct { anobject<type1> _type1_object; anobject<type2> _type2_object; anobject<type3> _type3_object; } Is there a way to do such a thing with boost pp (or something else)? Regards, Olivier

2 1

Re: [Boost-users] [Timer] Using progress timers
by Alex Perry 21 Sep '11

21 Sep '11

On Wed, 21 Sep 2011 11:47:35 +0100 Robert Jones wrote:- > cat progress.cpp > > #include <boost/progress.hpp> > > int main( ) > { > boost::progress_timer t; > > sleep( 5 ); > > return 0; > } > > > g++ progress.cpp > > > ./a.out > 0.00 s > > I was expecting a number >= 5.00 s! What am I missing? Not sure if its any help but above does work on windows with appropriate change to sleep >type test.cpp #include <boost/progress.hpp> #include <windows.h> int main( ) { boost::progress_timer t; Sleep( 5000 ); return 0; } >cl test.cpp Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 16.00.30319.01 for 80x86 Copyright (C) Microsoft Corporation. All rights reserved. test.cpp Microsoft (R) Incremental Linker Version 10.00.30319.01 Copyright (C) Microsoft Corporation. All rights reserved. /out:test.exe test.obj C:\workspace\Test>test 5.00 s HTH Alex

1 0

[UTF] documentation of runtime configuration programmatically
by gast128 20 Sep '11

20 Sep '11

Hello all, the documentation of Boost.Test contains the following sentence "The UTF provides two ways to set a parameter at runtime: by specifying a command line argument and by setting an environment variable. The command line argument always overrides the corresponding environment variable." However it does not explain how. For example if I add this line to my code it crashes: boost::runtime::env::get<bool>("BOOST_TEST_AUTO_START_DBG"); Anyone?

1 1

[Serialization]Serialization of nested stl object
by Kangkook Jee 20 Sep '11

20 Sep '11

Hi, all A member of a target class, where one STL object (map) includes another (set), prevent the code from being built properly and the error message urges me to implement *serialize* method for pair<string*, set<unsigned>>. What is the typical way of implementing serialization using boost for this kind of data structures? * Code snippet class B { … map<const string*, set<unsigned> > lines; .. template<class Archive> void serialize (Archive & ar, const unsigned int version) { ar & lines; … * Build error message /opt/local/include/boost/serialization/access.hpp:118: error: ‘class std::map<const std::string*, std::set<unsigned int, std::less<unsigned int>, std::allocator<unsigned int> >, std::less<const std::string*>, std::allocator<std::pair<const std::string* const, std::set<unsigned int, std::less<unsigned int>, std::allocator<unsigned int> > > > >’ has no member named ‘serialize’ Thanks for your help in advance, /Kangkook

2 4

shared_ptr design question
by John M. Dlugosz 20 Sep '11

20 Sep '11

I'm wondering why there is not an efficient built-in way to check whether two things (e.g. a shared_ptr and a weak_ptr, but any combination should work) refer to the same object. In my implementation, I found that a handy thing to have. In changing code that uses a different home-brew smart pointer, I have to achieve this. The only way I see is to first resolve the weak pointer, and then compare the get() from each. That needlessly yoyo's the reference count, which is a horribly expensive atomic operation. I find that in the new C++ standard it refers to a owner_before ordering function, which is a bit more obscure than just comparing the pointers. So I'm wondering why this is not provided. Is there some reason why the feature is obscured, for portability? —John

7 8

gdb pretty printers
by Jonathan Brannan 20 Sep '11

20 Sep '11

Are there any pretty printers for the more complex boost types? Specifically I'm looking for boost::multi_index printers as looking at in debug seems unhelpful.

2 2

Flow rate dimensional analysis
by Michael Powell 19 Sep '11

19 Sep '11

Boost::units users, I am trying to perform a flow rate through an orifice calculation (Volume per Time or L^3T^-1) using the following dimensional analysis: Q = Cf * A * sqrt( 2 * abs(P1-P2) ) On paper the dimensional analysis works out correctly. I know we should be receiving a L^3T^-1 consequent to the calculation. If I just take this on one line, it seems to compile: __C_flow * __a_orifice * boost::units::sqrt(__two * __p_diff / __d); Where __a_orifice is the area of the orifice, __p_diff is the pressure differential (absolute value), and __d is the mass density. __C_flow (coefficient of flow through orifice) and __two (2.0) are both dimensionless. In terms of units quantities, these are all defined as quantity<si::area>, quantity<si::pressure>, quantity<si::mass_density>, and quantity<si::dimensionless>, to start with. I think my flow-rate dimension and/or unit must be incorrect, but I am failing to see why in order to know what to do to correct the error. typedef boost::units::derived_dimension<boost::units::length_base_dimension,3 ,boost::units::time_base_dimension,-1>::type flow_rate_dimension; typedef boost::units::unit<cs::units::flow_rate_dimension,boost::units::si::system> flow_rate; When I take: si::flow_rate __fr = __C_flow * __a_orifice * boost::units::sqrt(__two * __p_diff / __d); I am receiving the following compiler error as my clue... Error 1 error C2440: 'initializing' : cannot convert from 'boost::units::quantity<Unit,Y>' to 'boost::units::quantity<Unit,Y>' c:\Source\Simulations\CS.Calculations.Managed\Calculators\Calculator.cpp 461 CS.Calculations.Managed I'd like to set this up in as base a dimension as possible because later on we'll be introducing unit views, converting from base units to a view unit, or unit system, along these lines, but if we have to specify this in specific unit terms, so be it. We'll worry about unit conversions later on. Best regards, Michael Powell

2 6