- Boost-users - lists.boost.org

[multi_index] REVISITED: hash_unique with differently produced hash codes
by Ovanes Markarian 26 Apr '07

26 Apr '07

Hello Joaquín, now I found the failure, to the problem described in message http://lists.boost.org/boost-users/2007/04/27211.php with std::string as a key not being found. The problem in my case was that I use Xalan to parse XML and have to convert string to single byte characters. Since std::string has some build in caching to improve performance and I decode double byte string (XalanDOMString) as follows: std::string get_string(XalanDOMString const& str) { XalanDOMString::CharVectorType v; str.transcode(v); return std::string(v.begin(), v.end()); } this code produces a string which still points to the double byte buffer and therefore this results in a different hash value for the returned string as the value would be for the basic_string<char> or C-String: An example: XalanDOMString str("Sint8"); std::string char_str(get_string(str)); Calling find in the multi-index results in a hash value 40. Calling find in the multi-index for const char* str="Sint8"; results in a hash value 13 The same result (13) can be achieved when calling on char_str.c_str(); c_str() cause the strings internal buffer to be rebuild. Since boost::hash is on the way to become a part of STL it should ensure to work with the correctly intended copy of string. What is your opinion on this issue? With Kind Regards, Ovanes

3 3

[BGL] Basic graph construction
by Alejandro Marcos Aragón 26 Apr '07

26 Apr '07

I was wondering if someone could give me a tip on how to construct the following graph. The problem is that the queue type definition does not know what "edge" is until the type definition for the graph is created. Is there a way to get around this? Thank you #include <boost/graph/adjacency_list.hpp> #include <queue> // Type definition for a priority queue of (double,edge) pairs typedef std::priority_queue<std::pair<double,int> > edge_queue; struct vertex_edge_priority_t { typedef vertex_property_tag kind; }; // Vertex properties: typedef property < vertex_edge_priority_t, edge_queue > domainVertexProperty; // Edge propeties: typedef property < edge_index_t, int, property < edge_weight_t, double > > domainEdgeProperty; // Graph used to represent possible vertices and edges within the domain. typedef adjacency_list< setS, vecS, undirectedS, domainVertexProperty, domainEdgeProperty > domainGraph; typedef graph_traits<domainGraph>::edge_descriptor edge;

2 2

[smart_ptr] Conflict with Maya?
by Paul Giaccone 26 Apr '07

26 Apr '07

Does anyone know of any evidence that the memory management in smart_ptr might be incompatible with Maya, in particular, in gcc 3.2.3? My supervisor, who is unfamiliar with Boost's smart_ptr library, but is familiar with Maya, tells me that Maya has its own memory management, and that Boost's smart pointers might somehow interfere with this. I've used Boost's smart pointers for long enough that I really, really, really don't want to go back to the horrors of having to delete memory that I allocate on the heap and spending hours tracking down memory leaks if I get it wrong. I'd really rather not do that, especially as it's a potential memory leak that I'm trying to fix (Maya plug-ins are created using a bare "new" which doesn't seem to require a corresponding delete, according to the examples I have seen). So someone please tell me that is couldn't possibly be true. Thanks.

5 16

[scoped_array] "optimization"
by Michael Marcin 26 Apr '07

26 Apr '07

Hello, I have a new programmer on my team who has been brought on as an optimization engineer. She has a lot more experience than I do but I fear some of that experience may have left impressions that have stuck with her for too long. On the first check in labeled spot optimizations there were many things changed that disturbed me. One of them was changing a class from essentially. #include <boost/scoped_array.hpp> class Context { public: Context( unsigned int width, unsigned int height ) : m_buffer( new unsigned int[width*height] ) { } unsigned int* GetBuffer() { return m_buffer.get(); } private: boost::scoped_array<unsigned int> m_buffer; }; to class Context { public: Context( unsigned int width, unsigned int height ) : m_buffer( new unsigned int[width*height] ) { } ~Context() { delete m_buffer; } inline unsigned int* GetBuffer() { return m_buffer; } private: unsigned int* m_buffer; }; First of all there is the defect of delete being called instead of delete[]. Other than that I was under the impression that there should be no difference between the two other than the former being more descriptive (it instantly tells me this is an array and the memory is owned by the instance of this class). Is this really an optimization? Thanks, Michael Marcin

9 11

[multi-index] hashed_unique and sequenced question
by Ovanes Markarian 26 Apr '07

26 Apr '07

Hello, Is there any way to find an element in the multi-index container by it's hash and proceed with all elements which were added later on in the sequence after it? With Kind Regards, Ovanes Markarian

2 1

[BLL] Right usage of he placeholder _1 ?
by pheres 26 Apr '07

26 Apr '07

Hello, I've two problems so far with a simple for_each: 1. vector<string*> v2(10); fill(v2.begin(), v2.end(), new string("hehe")); for_each(v2.begin(), v2.end(), cout << *_1 << " "); // works for_each(v2.begin(), v2.end(), delete _1 ); // error // error: type ‘const struct // boost::lambda::lambda_functor<boost::lambda::placeholder<1> // >’ argument given to ‘delete’, expected pointer Shouldn't the placeholder _1 actually become a (string-)pointer as it obviously does in the expression above ( cout << *_1 << " " )? 2. Is it possible to call member functions from an placeholder of a boost-like lambda expression? vector<string> v3(10); fill(v3.begin(), v3.end(), "hoho"); for_each(v3.begin(), v3.end(), cout << _1->c_str() << " "); // error // error: base operand of ‘->’ has non-pointer type ‘const // boost::lambda::lambda_functor<boost::lambda::placeholder<1> >’ Thank you for any input!

2 1

[program_options] Problem with custom parser and boolean options
by Tarjei Knapstad 26 Apr '07

26 Apr '07

I'm trying to support non-conventional syntax on the form "option=value", where boolean options are given as "option=on" or "option=off". For this I've implemented a custom parser that splits a string on the '=' character and returns a std::pair<std::string, std::string> containing the option and it's value. This works fine for most things except I can't figure out how to handle boolean options. I've tried the following which just makes the option "disappear" if it's set to "off" and returns a pair with the option and an empty string as value if it's set to "on". It works for "off", but for "on" I get an exception stating "extra parameter in 'option'". Is there any way to get around this? Here's my custom parser: //---------------- START: code std::pair<std::string, std::string> custom_parser(std::string const & s) { std::pair<std::string, std::string> option; if (s.find("=") != std::string::npos) { std::vector<std::string> slist; boost::split(slist, s, boost::is_any_of("=")); if (slist[1] == "on") { option.first = slist[0]; } else if (slist[1] == "off") { // Return an empty option if the feature is turned off } else { option.first = slist[0]; option.second = slist[1]; } } return option; } //-------------- END: code Any help appreciated! Regards, -- Tarjei

1 0

[multi_index] inherit from a multi_index
by Guillaume Lazzara 26 Apr '07

26 Apr '07

Hi, I would like to know if it is safe to create a struct and make it inherit from a multi_index. Actually I already tried but I got this error: /usr/include/boost/multi_index_container.hpp:204: warning: base class 'struct boost::multi_index::detail::header_holder<boost::multi_index::detail::ordered_index_node<boost::multi_index::detail::index_node_base<vcsn::InitialValue<vcsn::rat::exp<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, float> > > >, boost::multi_index::multi_index_container< [...] >, boost::multi_index::indexed_by<boost::multi_index::ordered_unique<boost::multi_index::member<vcsn::InitialValue<vcsn::rat::exp<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, float> >, [...] >' should be explicitly initialized in the copy constructor Is it an issue due to Boost or because I'm doing something really bad? Thanks, -- Guillaume Lazzara Epita CSI 2008

2 2

regex_replace -- format_default
by ivan＠0x4849.net 26 Apr '07

26 Apr '07

Hi I am trying to use boost::regex_replace but I am having trouble understanding the format to use for the the format string. The Ecma-262.pdf rules for how to create a format string seem quite terse. Does anyone know of a reference that has some more examples? Here is the code I am trying to write: #include <boost/regex.hpp> #include <string> #include <iostream> int main(int argc, char** argv) { std::string str; boost::regex expression("abc"); std::string fmt(""); // WHAT GOES HERE? fn = "abcabcabc"; boost::regex_replace(str, expression, fmt); std::cout << str << std::endl; return 0; } Thank you, Ivan Novick http://www.0x4849.net

2 1

boost::random::normal_distribution
by Jeffrey Brent McBeth 26 Apr '07

26 Apr '07

#include <boost/random.hpp> #include <iostream> int main(int argc, char *argv[]) { boost::mt19937 eng; boost::normal_distribution<> norm(0,1); std::cout << norm(eng) << std::endl; } Gives me a NaN everytime. Looking at the code, it doesn't really surprise me, as _r1 = eng(); _r2 = eng(); _cached_rho = sqrt(-result_type(2)*log(result_type(1)-_r2)); would only work if eng were in the range (0,1) with 0 being degenerate, and 1 being undefined. Looking through the mt19937, that is returning a full UInt32. Perhaps those two eng() should be replaced by uniform_01? Jeffrey McBeth -- ---------------------------------------------------------------------------- Hofstadter's Law states: It always takes longer than you expect, even when you take into account Hofstadter's Law. ----------------------------------------------------------------------------

2 1