- Boost-users - lists.boost.org

How to make Boost.Format fault tolerant?
by John M. Dlugosz 11 Feb '12

11 Feb '12

A common case for using Boost.Format is in the production of error or warning messages. So, these don't get run on a normal use of the program, and problems with the arguments will go unnoticed until there _is_ an error, in which case I get (at best) no error message, or worse, turn a log-only warning statement into an exception. Does Format already have some features to make it not throw exceptions? Idealy, for messages that go into a log or are used to diagnose some program bug, any information is better than none and I'd like to see what it did get as best as it can, and not worry about how pretty it may be. So missing arguments use a default placeholder text instead, and extra arguments without specifiers in the format string just get appended with spaces in between, and any formatting spec error just goes back to default string, and any error during that gives a placeholder but the other arguments are processed. —John

3 4

Boost.Graph DFS and TerminatorFunc
by John Eddy 10 Feb '12

10 Feb '12

Is there a good reason that the TerminatorFunc template parameter of the depth_first_visit is not exposed through the DFS function? It would be easy to do and very helpful to me. I'd rather not modify my version of the boost code especially if this would be generally useful to others. Thanks! John

1 0

Tokenizer / offset_separator misbehaving - boost 1.47.0 / Windows 2003 Standard Server 32-bit / Visual C++ 2010 Express
by asif saeed 10 Feb '12

10 Feb '12

Hi, I wrote the following code to test offset_separator //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// int main (int argc, char **argv) { using namespace std; typedef boost::offset_separator sep_type_t; typedef boost::tokenizer<sep_type_t> tokenizer_t; typedef tokenizer_t::const_iterator const_iterator; typedef tokenizer_t::iterator iterator; const string sentence = "Marry had a little lamb!"; int endMarkers[] = {5,3,1,6,4}; sep_type_t sep(endMarkers,endMarkers+3); tokenizer_t t_(sentence,sep); for (tokenizer_t::iterator i=t_.begin();i!=t_.end();++i) cout << endl << *i; return 0; } //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// I get the following result: Marry ha d a li ttl e lamb ! Any idea how I can fix it? I want each token on a separate line. I am providing zero-based end indices to separator. Thanks in advance, -Asif

1 0

boost-introduced ambiguity in _1/_2 etc.
by Piotr Wyderski 10 Feb '12

10 Feb '12

Hello, I am refactoring our application which uses boost extensively in order to replace some boost-provided entities with their std counterparts defined in the C++11 standard. I've experienced a problem related to boost::bind placeholders defined in placeholders.hpp. The file begins with: namespace { static inline boost::arg<1> _1() { return boost::arg<1>(); } [...] unconditionally injecting the _N() functions into the global namespace scope. When I type "using namespace std::placeholders;" in order to use e.g. std::placeholders::_1, an unsolvable ambiguity occurs, as there is no way to transfer the boost-defined variants into an uncolliding namespace. Could you please add a macro in 1.49 which would allow me to override the default behaviour and define the functions in boost:: or boost::placeholders namespaces? #if defined(BOOST_USE_BOOST_PLACEHOLDERS_NAMESPACE) namespace boost { #else namespace { [...] Best regards, Piotr

1 0

Re: [Boost-users] [iostreams] multichar_output_filter works only the first time
by Steven Samuel Cole 10 Feb '12

10 Feb '12

i tried using a boost::iostreams::file_sink in the output filter instead of std::cout --> same result. i also tried running the code on the latest ubuntu instead of mac osx lion --> same result.

1 0

Using boost::asio with thousands of concurrent connections
by Gheorghe Marinca 09 Feb '12

09 Feb '12

Hi, I started to use boost::asio some time ago and want to use it for a windows commercial product that does web filtering and handles thousands of connections concurrently (proxy based product) I am wandering if anybody had experience using it at such a large scale, where besides handling many connections you have to handle inside a connection communication with databases, things that potentially incur delaying handling of subsequent requests. What kind of configuration would you advise in such a case (multiple I/O services per n threads, on IO service that handles all incoming connections and a thread pool that handles processing of async handlers - how one would assure in this case a connections is handled in the same thread then ..etc) Thanks for any input -Ghira

3 2

XML and Boost Property Tree
by Boost Users List 09 Feb '12

09 Feb '12

Hey Boost users! I'm iterating over an XML document using boost property tree and storing the results in a struct. The issue I have is that I can only get to the first "item" nodes and can't access the second "item" nodes. I was hoping someone would point out where I've made a mistake. My program output looks like this (you can see items are missing.. there is no cookie2, candy2 or chocolate2 items shown): jar : snAcks snack : coOkie item : cooKie1 snack : canDy item : caNdy1 snack : cHocolate item : choColate1 Here's the xml file: <root> <jar name="snAcks"> <snack name="coOkie"> <item name="cooKie1"></item> <item name="cookIe2"></item> </snack> <snack name="canDy"> <item name="caNdy1"></item> <item name="candY2"></item> </snack> <snack name="cHocolate"> <item name="choColate1"></item> <item name="chocOlate2"></item> </snack> </jar> </root> Here's the source code: void parse_xml( boost::property_tree::iptree const& pt ) { BOOST_FOREACH( boost::property_tree::iptree::value_type const& v, pt.get_child("root.jar") ) { // Show jar if ( boost::iequals( v.first, "<xmlattr>" ) ) { std::cout << "jar : " << v.second.get<std::string>("NaME") << std::endl; } if ( boost::iequals( v.first, "snack" ) ) { // Show snack std::cout << "snack : " << v.second.get<std::string>("<xmlattr>.name") << std::endl; try { BOOST_FOREACH( boost::property_tree::iptree::value_type const& val, v.second.get_child("item") ) { if ( boost::iequals( val.first, "<xmlattr>" ) ) { // Show item std::cout << "item : " << val.second.get<std::string>("nAmE") << std::endl; } } } catch (boost::property_tree::ptree_bad_path& e) { // Show what caused exception std::cout << "Exception: " << e.what() << std::endl; } } } } Thank you for taking time to read this. I think I've made a simple mistake, but cannot understand where.

1 0

[iostreams] multichar_output_filter works only the first time
by Steven Samuel Cole 09 Feb '12

09 Feb '12

hello, to get started with iostreams, i'm trying to write a multichar output filter that adds a string to the beginning and the end of whatever comes through. my code builds with a warning about unused parameter n and works - but only the first time, the second time, i get no output. i thought at first sending '\0' would trigger the stream's fail-bit or so, but out.good() returns true. any chance someone takes a look at my code and hazards a guess where the problem might be ? disclaimer: i was actually born and bred in C++ country long time ago, but i've been coding in python for a couple of years and got back only recently, so there's a good chance what i'm doing wrong is something really simple. guess you can tell from the code, it's been a while. good to be back though... :-) #include <iostream> #include <sstream> #include <boost/iostreams/concepts.hpp> #include <boost/iostreams/filtering_stream.hpp> #include <boost/iostreams/operations.hpp> using std::cin; using std::cout; using std::endl; using std::streamsize; using std::string; using std::stringstream; class add_string_output_filter : public boost::iostreams::multichar_output_filter { public: template<typename Sink> streamsize write(Sink& sink, const char* s, streamsize n) { // remove trailing '\0' to prevent line break // TODO: there's probably a better way string tmp_string = string(s); string out_string = tmp_string.substr(0, tmp_string.size()-1); string pre_string("prepended string - "); string app_string(" - appended string"); stringstream sstrm; sstrm << pre_string << out_string << app_string << endl; // TODO: char* to string, back to char* ?!? return boost::iostreams::write(sink, sstrm.str().c_str(), sstrm.str().length()); } }; int main() { boost::iostreams::filtering_ostream out; out.push(add_string_output_filter()); out.push(cout); cout << "prompt: "; string str; while (getline(cin, str)) { out << str << endl; cout << "prompt: "; } } # sample session: prompt: some string prepended string - some string - appended string prompt: another string prompt: why no output ? prompt: # ssc

1 0

Compiler wants to link to threading library all of a sudden
by Crispin Cooper 09 Feb '12

09 Feb '12

Hello Boost users I wonder if anyone can help me please. I have just updated my boost install to svn head to take advantage of some bugfixes. Now the compiler complains that it can't find libboost_thread-vc90-mt-sgd-1_49.lib. But I am not trying to use the threading libraries - my boost includes are: #include <boost/shared_ptr.hpp> #include <boost/pool/pool_alloc.hpp> #include <boost/pool/object_pool.hpp> #include <boost/static_assert.hpp> #include <boost/geometry.hpp> #include <boost/geometry/geometries/point_xy.hpp> #include <boost/geometry/geometries/polygon.hpp> #include <boost/geometry/multi/geometries/multi_point.hpp> Have any of these changed such that they require prebuilt libraries? Or is including boost from the head revision somehow different to a released version in this manner? Many thanks in advance Crispin -- Dr Crispin Cooper Sustainable Places Research Institute Cardiff University

4 5

retrieving section names from a boost::proper_tree
by Sean Farrow 09 Feb '12

09 Feb '12

Hi All: I have parsed a .ini file in to a boost::property_tree. I now would like to retrieve all the section names and place them in a vector. Am I needing the key_type or the value_type. I am suspecting the former! Also is there a quick way of outputting the section names without having to use a boost_foreach or similar loop construct? Any herlp appreciated. Regards Sean.

2 1