- Boost-users - lists.boost.org

[regex] why partial match / early break ?
by U.Mutlu 01 Nov '11

01 Nov '11

Hi, why is this regex const string sRe = "((([a-zA-Z]|([a-zA-Z][a-zA-Z0-9\\-]))+[a-zA-Z0-9])\\.)+" "((([a-zA-Z]|([a-zA-Z][a-zA-Z0-9\\-]))+[a-zA-Z0-9]))"; not matching this string wholly? "a1a.a2a.a3a.a4aaaa" It rather matches only this part: "a1a.a2a.a3a.a4" What's the problem here? (I know I can append a delimiter to solve the problem, but I think the regex should've match it wholly, shouldn't it?)

4 6

transfer_at_least_t
by Carla Strembicke 01 Nov '11

01 Nov '11

Hi, I have an object of transfer_at_least_t. I would like to find out what the minimum value is due to the fact that my app that set it is requiring so many bytes of unencrypted data, and the transfer is including the encryption. Is there any interface that can get this value or do I have to store it before I create the object transfer_at_least_t? Carla Strembicke Software Developer [cid:image015772.JPG@7ddd0ff8.4ca8ab2e] SUBNET Solutions Inc. www.SUBNET.com<http://www.SUBNET.com/> Making Substations More Intelligent T: 403.270.8885 F: 403.270.9631 #100, 4639 Manhattan Road SE, Calgary, Alberta, Canada; T2G 4B3 CONFIDENTIAL INFORMATION NOTICE: The information contained in this e-mail is privileged, confidential and intended solely for the use of the addressee named above. If the reader of this e-mail is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this e-mail is strictly prohibited. If you received this in error, please contact the sender and destroy all copies of this e-mail. Thank you.

2 1

using a boost::shared_mutex in a share_ptr
by Sean Farrow 01 Nov '11

01 Nov '11

Hi: I have the following variable declared in a class: Boost::shared_ptr<boost;:shared_mutex> If I have a standard boost::shared_mutex, I can pass this to a shared or unique lock. How do I pass a boost::shared_ptr<boost::shared_mutex> to a lock? This is probably really simple, but I carn't fathum it out! Cheers Sean.

2 1

serialization of a ptr
by Jonathan Bishop 01 Nov '11

01 Nov '11

How do I prevent boost::serialize from treating a pointer as an object and serializing the object. I have written the specialization of save/load for the object pointer. I get the following errors when I attempt to compile the program below... boost/serialization/access.hpp: In static member function 'static void boost::serialization::access::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = X]': boost/serialization/serialization.hpp:81: instantiated from 'void boost::serialization::serialize(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = X]' boost/serialization/serialization.hpp:140: instantiated from 'void boost::serialization::serialize_adl(Archive&, T&, unsigned int) [with Archive = boost::archive::binary_oarchive, T = X]' boost/archive/detail/oserializer.hpp:151: instantiated from 'void boost::archive::detail::oserializer<Archive, T>::save_object_data(boost::archive::detail::basic_oarchive&, const void*) const [with Archive = boost::archive::binary_oarchive, T = X]' s3.cpp:53: instantiated from here boost/serialization/access.hpp:109: error: 'class X' has no member named 'serialize' ============================================================================================== #include <fstream> #include <sstream> #include <cstddef> #include <boost/archive/binary_iarchive.hpp> #include <boost/archive/binary_oarchive.hpp> #include "boost/serialization/split_free.hpp" using namespace std; class X { X(const X& other); X& operator=(const X& other); public : static X* get(int id); X() : id(0) {} int id; }; // Non-intrusive serialization (pretend I am not able to modify class X) // I was hoping serialization would use this specialization // archive X by id template<class Archive> void save(Archive& ar, X*& x, unsigned int version) { int id = x->id; ar & id; } // load X by lookup template<class Archive, class T> void load(Archive& ar, X*& x, unsigned int version) { int id; ar & id; x = X::get(id); } template<class Archive, class T> inline void serialize(Archive& ar, X*& t, const unsigned int version) { split_free(ar, t, version); } int main() { X x; std::ofstream ofs("junk"); boost::archive::binary_oarchive oa(ofs); oa & x; ofs.close(); return 0; }

3 3

[test] boost.test bug report email address?
by legalize+jeeves＠mail.xmission.com 01 Nov '11

01 Nov '11

[Please do not mail me a copy of your followup] The docs say I can report a bug to the email address <boost-test =at= emailaccount =dot= com> but this is undeliverable for 7 days. -- "The Direct3D Graphics Pipeline" -- DirectX 9 version available for download <http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/> Legalize Adulthood! <http://legalizeadulthood.wordpress.com>

2 1

Common parameter on command line and in a config file with sections
by Gruhn, Dan 01 Nov '11

01 Nov '11

Greetings, I would like to have a parameter "expire-days" that has a default and can be set via the command line --expire-days|-e <days> and a config file [foo] expire-days = <days> It would be most elegant to be able to specify things like this (notice the automatic storage to the expireDays variable): namespace bpo = boost::program_options; int expireDays = 30; bpo::options_description desc("Allowed options"); desc.add_options() ("expire-days,e", bpo::value<int>(&expireDays)->default_value(expireDays), "Number of days before files expire") ; bpo::options_description configFileDesc; configFileDesc.add_options() ("foo.expire-days", bpo::value<int>(&expireDays), "") ; bpo::variables_map varMap; try { // Attempt to parse the configuration file. std::ifstream cfg("/tmp/all.cfg"); if (cfg.is_open()) { std::cout << "Opened /tmp/all.cfg" << std::endl; static const bool allowUnregistered = true; bpo::store(bpo::parse_config_file(cfg , configFileDesc, allowUnregistered), varMap); } // Parse the command line bpo::store(bpo::parse_command_line(argc, argv, desc), varMap); bpo::notify(varMap); } std::cout << "Expire days: " << expireDays << std::endl; My code may be strange but I am trying to find the right order and combination of adding options, parsing, storing, and notifying that gives the command line priority over the config file which in turn has priority over the 30 day default. Oddly enough, it doesn't seem to make much difference which order I use, the config file seems to have precedence This is not what I expected so perhaps I'm doing something wrong but I can't find it. I know that I can accomplish this via extra code on my part, but I'm trying to come up with a general scheme that would work for multiple programs with multiple variables using a common config file and the simpler the better. The documentation mentions using sections in config files but doesn't give much detail. Can anyone help me here or at least describe for me which of add_options, parse_*, store, and notify does the following: 1) Sets the default value 2) Marks that a value has been set from the command line/config file. 3) Actually does the store to the given variable. 4) What does notify do? I've tried (and continue to try) looking through the code but am having problems find the right place examine. I'll take pointers in that area as well. This does not seem like an exotic use case. Has anyone else accomplished this or something like it? As always, any help of any kind will be greatly appreciated. Dan

3 2

How to build the bjam tool on Windows?
by Congmin min 31 Oct '11

31 Oct '11

Hello. I just downloaded the boost library. I want to build bjam tool on my Windows machine. I need to get boost work for my eclipse IDE. But why I can't find the jam directory under boost/tools? How can I do that? Thanks.

3 2

Re: [Boost-users] [Tokenizer] Compile error with lots of missing binary operator statments
by Collin Day 31 Oct '11

31 Oct '11

Previously, I had posted what I thought was a bug. I did not think to do this originally, but I just tried a simple Hello World program that included boost/tokenizer.hpp. It compiled fine. I then looked at where the sample code was pulling headers from, and it appears that the download included old boost headers it was including instead of 1.47.0. I completley missed this. I apologize - and take back the error - there is none :) Thanks!

1 0

boost random uniform real infinite loop
by Peter Man 31 Oct '11

31 Oct '11

Hi, Recently, I've encountered a problem with the generate_uniform_real function (the is_integral version) in boost/random/uniform_real_distribution.hpp. A bug in my code meant that I was trying to generate a uniform random number where the min and max bounds were exactly equal. However, this meant that when on entering the function stated earlier, an infinite loop is never exited from! Is this a bug in boost::random, or have I missed something? Thanks Peter

2 1

sending large serialized objects over network using MPI
by Jonathan Bishop 31 Oct '11

31 Oct '11

HI, I am involved in a project which requires the transport of large C++ instances over MPI communication channels. I am using boost serialization to pack and unpack the objects into a buffer. The whole buffer is sent in one shot. Of course for large objects this can be problematic. I have two questions... 1) is there any way to tell beforehand how large the buffer needs to be? In other words, can you run serialization in a counting only mode? 2) is there an archive implemented which streams across MPI communication channels (not sure this is the right terminology.)? Seems like it might be useful. This might be a better question for the boost MPI team, though. Thanks, Jon

2 1