February 2007 - Boost-users

sub_range, assign and permutation_iterator related problem
by abir basak 22 Feb '07

22 Feb '07

Hi, It is an extension to one my previous post, where I needed help while giving access to a sub_range for of an container. In that problem , Mr Thorsten Ottosen, provided answer to all of my queries (Many thanks to him ). This time I am interested to provide indexed access to a vector with a sub_range ,using permutation_iterator. Here some codes are creating problem, while I am using sub_range with permutation iterator. To specify, I am providing a small test code. typedef std::pair<std::size_t,std::size_t> pair_t; template<typename T> class vector_e : private std::vector<T>{ public: typedef vector_e<T> self_type; typedef typename std::vector<T>::iterator iterator; typedef typename std::vector<T>::const_iterator const_iterator; typedef boost::sub_range<self_type> range; typedef boost::sub_range<const self_type> const_range; typedef typename std::vector<T>::size_type size_type; typedef std::vector<int> index_type; typedef typename boost::permutation_iterator<typename iterator,index_type::const_iterator> index_iterator; typedef typename boost::permutation_iterator<typename const_iterator,index_type::const_iterator> const_index_iterator; vector_e() : std::vector<T>(){} using std::vector<T>::begin; using std::vector<T>::end; using std::vector<T>::operator []; using std::vector<T>::insert; index_iterator begin_i(const index_type& i){ return boost::make_permutation_iterator(begin(),i.begin()); } const_index_iterator begin_i(const index_type& i)const{ return boost::make_permutation_iterator(begin(),i.begin()); } index_iterator end_i(const index_type& i){ return boost::make_permutation_iterator(begin(),i.end()); } const_index_iterator end_i(const index_type& i)const{ return boost::make_permutation_iterator(begin(),i.end()); } range get_range(const pair_t& range){ return boost::make_iterator_range(begin()+range.first,begin()+range.second); } const_range get_range(const pair_t& range)const{ return boost::make_iterator_range(begin()+range.first,begin()+range.second); } range get_range(const index_type& i){ return boost::make_iterator_range(boost::make_permutation_iterator(begin(),i.begin()), boost::make_permutation_iterator(begin(),i.end())); } }; int main(int argc,char** argv) { using namespace boost; using namespace boost::assign; using namespace boost::lambda; using boost::lambda::_1; typedef vector_e<int> elem_t; typedef std::vector<int> index_t; elem_t v; int a[] = {00,10,20,30,40,50,60,70,80,90,100}; v.insert(v.end(),a,a+11); const elem_t& cv = const_cast<elem_t&>(v); //v+= 00,10,20,30,40,50,60,70,80,90,100; std::cout<<"original array : "; std::for_each(v.begin(),v.end(),std::cout<<_1<<" "); ///For continuous ranges. elem_t::range r = v.get_range(std::make_pair(1,5)); std::cout<<std::endl<<" range (1,5) : "; ///working fine std::for_each(r.begin(),r.end(),std::cout<<_1<<" "); ///working fine elem_t::const_range cr = v.get_range(std::make_pair(1,5)); std::cout<<std::endl<<" c_range (1,5) : "; std::for_each(cr.begin(),cr.end(),std::cout<<_1<<" "); ///working fine elem_t::const_range cr1 = cv.get_range(std::make_pair(1,5)); std::cout<<std::endl<<" c_range1 (1,5) : "; std::for_each(cr1.begin(),cr1.end(),std::cout<<_1<<" "); index_t i; i+=0,2,6,4,9; std::cout<<std::endl<<" index : "; std::for_each(i.begin(),i.end(),std::cout<<_1<<" "); std::cout<<std::endl<<"index range "; //working fine std::for_each(v.begin_i(i),v.end_i(i),std::cout<<_1 << " "); std::cout<<std::endl<<"c index range "; ///working fine std::for_each(cv.begin_i(i),cv.end_i(i),std::cout<<_1 << " "); ///not working (and its const counter part ) elem_t::range ir = v.get_range(i); } The error it is giving in visual studio 7.1 for sub_range of permutation_iterator is, E:\boost_1_33_1\boost\range\iterator_range.hpp(60): error C2440: 'type cast' : cannot convert from 'boost::range_iterator<C>::type' to 'boost::range_iterator<C>::type' with [ C=const boost::iterator_range<vector_e<int>::index_iterator> ] and [ C=vector_e<int>::self_type ] The questions are 1) Can anyone point out the problem ? (Is sub_range is typed by iterator by any chance, or typed by container only ? ) 2) This is aux question. How to make assign library work for my inherited container (vector_e here ), like the commented line ? any help is appreciated, and thanks to the community . abir -- Abir Basak, Member IEEE Software Engineer, Read Ink Technologies B. Tech, IIT Kharagpur email: abir(a)abirbasak.com homepage: www.abirbasak.com

1 1

[BGL] Dijkstra visitor for examine edge event
by Alejandro Marcos Aragón 22 Feb '07

22 Feb '07

Hi everyone, I posted this message but I haven't received any response so I decided to post it again. I am trying to store all the vertices with their respective distances from a source within the Dijkstra shortest path algorithm. For that I plan to use a std::vector<std::queue > >. I was able to have the visitor working but I don't know how to access to those values that are computed automatically by the algorithm (d_v, d_u, for example). The code I wrote is as follows: template <class Tag> struct vis : public default_dijkstra_visitor { vis() { } template <class Edge, class Graph> void edge_relaxed(Edge e, Graph& g){ cout<<"VISITOR!!!"<<endl; cout<<"Edge -> "<<e<<endl; } private: std::vector<std::queue<Vertex> > storage; }; template <class Tag> vis<Tag> target_visit(Tag) { return vis<Tag>(); } void Dijkstra() { // maps needed by the algorithm IndexMap indexmap = get(vertex_index, *gPtr); WeightMap weightmap = get(edge_weight, *gPtr); // define containers for the predecessor vertices and // for distances std::vector<Vertex> pred(num_vertices(*gPtr)); std::vector<double> dist(num_vertices(*gPtr)); // vertex to compute Dijkstra's algorithm on Vertex source = vertex(get_source(), *gPtr); // Dijkstra's algorithm dijkstra_shortest_paths(*gPtr, source, predecessor_map(&pred[0]).distance_map(&dist[0]).visitor(target_visit(on_examine_edge()))); ... Any help will be appreciated. a^2

2 3

[BGL] merger parallel edges
by Shufei Fan 22 Feb '07

22 Feb '07

Hi, I am new to Boost Graph Library. I have two questions. I defined my graph like this: struct Node{ int label; int mode[3]; //...;}; struct Edge{ double nCommBord;}; typedef boost::adjacency_list< boost::vecS, boost::vecS, boost::undirectedS, Node, Edge> Graph;Somewhere in my program, I construct it using add_edge() method. What I want is, I allow parallel edges, but I all parallel ones being merged(weight summed together). Can I add edges one by one, and use an general method to 'magically' merge them?(is there a function doing merging?) My second question is, where should I find a complete list of usage of all the library methods(say, manual for add_edge) Thanks, Regards, Shufei _________________________________________________________________ Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! http://spaces.live.com/spacesapi.aspx?wx_action=create&wx_url=/friends.aspx…

2 1

[signals] connection::add_tracked() proposal
by Frank Mori Hess 21 Feb '07

21 Feb '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'd like to propose we add an add_tracked() function to the connection object. This would allow an arbitrary shared_ptr/weak_ptr to be added to the list of objects tracked for the slot. The justification is to allow for situations where an object needed by the slot is not directly owned by a shared_ptr, or doesn't directly correspond to an argument of the bound function. For example, the bound argument might be contained inside another object which is in turn managed by a shared_ptr. We could actually dump boost::track() entirely if we wanted, if we have add_tracked(). - -- Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFF3Gxq5vihyNWuA4URAmS+AJkB3tgE/8p7GlCY+xq+UwfTFjZwigCgoQh0 fgQdVgH73Oh3j1g34HNy+JY= =3UVL -----END PGP SIGNATURE-----

4 9

Serialization newbie question
by Surya Kiran Gullapalli 21 Feb '07

21 Feb '07

Hello all, I'm a newbie to boost serialization and i need some help. I'm using boost serialization, to serialize a class, which contains 3 std::string members. After i serialize i add it to a database. The database I'm using is berkeley-db. The database API needs key/data pair, both in void* type. Now for my problem, the key is std::string, so I'm passing in key as (void *)key.c_str(), I know its crude, but i'm just starting. Now for the data part, I've to serialize the complete class. As, boost::archive::binary_oarchive, takes ostream& as argument for constructor, I've done like this. std::ostringstream os ; boost::archive::binary_oarchive oa (os) ; oa << class_obj ; so the class_obj has been written to os. so i'm sending data to db api as (void *)os.str().c_str(). To retrieve data, I'm using something like this data=db.get_data() ; istringstream is((char *)data) ; boost::archive::binary_iarchive (is) ; The code compiled fine and at run time i'm getting exception basic::string::resize during construction of binary_iarchive. Obviously i was doing wrong. My question is how to serialize data and write it to memory than to a file? and how do i access it again? Thanks in advance, Surya

2 3

Boost Serialization library with Berkeley-DB
by Surya Kiran Gullapalli 21 Feb '07

21 Feb '07

Hello all, Has any body ever used Boost serialization library with Berkeley-DB database to store/retrieve data. I'd like some pointers on how to use it?. Sample code would be most appreciated. Thanks, Surya

1 0

How to build guid lib?
by Christian Henning 21 Feb '07

21 Feb '07

Hi there, it's not as easy as I thought it would be. I copied the guid files in the appropriate folders in my boost/ directory. And then I tried: bjam -sTOOLS=vc-7_1 --with-guid But that doesn't work. So, how do I do that? I really have no clue about bjam, etc. Thanks in advance, Christian

4 10

[graph] Crash in breadth_first_search
by Jens Müller 21 Feb '07

21 Feb '07

Can someone help me with the following program? It crashes with Boost 1.33.1 on gcc34 (GCC) 3.4.3 20050113 (Red Hat 3.4.3-16): $ ./1 make_map_as_colormap entered writing something to MapAsColorMap Speicherzugriffsfehler I have tracked down the crash to this line in <boost/graph/breadth_first_search.hpp>: for (tie(ei, ei_end) = out_edges(u, g); ei != ei_end; ++ei) { But I don't know what happens there ... I'm quite sure the bug is NOT in my color map proxy ... Thanks a lot in advance! Cheers, Jens // //======================================================================= // Copyright 2007 University of Karlsruhe // Author: Jens Mueller // // Distributed under the Boost Software License, Version 1.0. (See // accompanying file LICENSE_1_0.txt or copy at // http://www.boost.org/LICENSE_1_0.txt) //======================================================================= // #ifndef BFS_DISTMAP_8737 #define BFS_DISTMAP_8737 /* Use distance map (or anything else which suits the following requirements) as a color map. Write operations are just ignored. Read operations work as follows: When the value in the underlying map equals a specified value, e.g. -1 in a distance map, or the null pointer in a predecessor map, the 'White' value is returned, otherwise the 'Black' value is returned. Note: This only is usable for application where there is no need to distinguish gray and black non-tree edge targets. Note: Setting the value in the underlying map so that it no longer corresponds to a 'White' value will have to be done by the visitor on the discover_vertex event. */ /* Template parameters: - ColorValue (default value used by helper function: default_color_type) - ReadablePropertyMap (underlying map) */ /* Runtime parameters: - ReadablePropertyMap pm - ReadablePropertyMap::value_type white (must be == comparable) */ #include <boost/property_map.hpp> #include <iostream> template<typename ReadablePropertyMap, typename ColorValue> class MapAsColorMap { private: typename boost::property_traits<ReadablePropertyMap>::value_type white_; const ReadablePropertyMap& pm_; public: typedef typename boost::property_traits<ReadablePropertyMap>::key_type key_type; typedef ColorValue value_type; typedef ColorValue reference; typedef boost::read_write_property_map_tag category; ColorValue get(key_type key); MapAsColorMap(const ReadablePropertyMap& pm, typename boost::property_traits<ReadablePropertyMap>::value_type white): white_(white), pm_(pm) { /* nothing */ }; }; template<typename ReadablePropertyMap> MapAsColorMap<ReadablePropertyMap, boost::default_color_type> make_map_as_colormap(const ReadablePropertyMap& pm, typename boost::property_traits<ReadablePropertyMap>::value_type white) { std::cout << "make_map_as_colormap entered" << std::endl; return MapAsColorMap<ReadablePropertyMap, boost::default_color_type>(pm, white); } template<typename ReadablePropertyMap, typename ColorValue> ColorValue MapAsColorMap<ReadablePropertyMap, ColorValue>::get(key_type key) { std::cout << "get entered" << std::endl; std::cout << key << " " << boost::get(pm_, key) << std::endl; return (boost::get(pm_, key)==white_) ? boost::color_traits<ColorValue>::white() : boost::color_traits<ColorValue>::black(); } template<typename ReadablePropertyMap, typename ColorValue> ColorValue get(MapAsColorMap<ReadablePropertyMap, ColorValue> pmap, typename boost::property_traits<ReadablePropertyMap>::key_type key) { return pmap.get(key); } template<typename ReadablePropertyMap, typename ColorValue> void put(MapAsColorMap<ReadablePropertyMap, ColorValue> pmap, typename boost::property_traits<ReadablePropertyMap>::key_type key, typename boost::property_traits<ReadablePropertyMap>::value_type value) { std::cout << "writing something to MapAsColorMap" << std::endl; /* nothing */ } #endif

3 7

local variable shadow
by Chris Saunders 21 Feb '07

21 Feb '07

I find local variable shadow warnings helpful, but when I include various boost libraries, I have to remove a lot of shadowing within the boost functions themselves. Note that I'm not refering to name conflicts with my own code. How can I submit patches to remove these type of warnings? Thanks, -Chris

3 3

[boost::program_options] got stuck with some options scenario
by Ovanes Markarian 21 Feb '07

21 Feb '07

I have some requirement for input parameters: Some of my input parameters represent a pair of files. Is there any way to parse this kind of a form: -I "file1" "file1_related_file" Trying to use the construct with: po::value< std::pair<std::string,std::string> >() produces a compilation error within a lexical cast, that there is no operator >> defined. I know I could pass smth like: -I "fiel1+related_file", but then I would need to split the string, what I would like to do, since I see this as a task of program_options lib. With Kind Regards, Ovanes Markarian

1 0