February 2007 - Boost-users

threads - basic introduction
by Noah Roberts 16 Feb '07

16 Feb '07

looking through the boost.threads doc and there doesn't seem to be any "getting started" guide like entry. I often find these are easier to read and get to know the library a little bit before hitting the reference. Is there any out there somewhere?

4 5

[iterator || function_types] zip_iterator -> unpack and call?
by Sohail Somani 16 Feb '07

16 Feb '07

Hi, Say I have a function "int add(int, int)" and want to use it in a call to std::transform with zip_iterators. Is there a way to do it without writing an intermediate function object? If not, would it be sensible to write a generic one using the function_types library? I can't imagine this would be too hard but just wanted to know if there is anything I can leverage before writing it myself. Thanks in advance, Sohail

3 4

derived class and serialization
by Aljaz 16 Feb '07

16 Feb '07

Hello.. I use serialization to communicate betwen client and server.. Base and derived class for serialization (communication): class base { }; class derived : base { }; What would be the 'good' way to export/import serialized data? Can I do this with smart pointers? Is this okay: WRITING: derived dc(4, 5, "some data"); std::ostringstream archive_stream; boost::archive::text_oarchive archive(archive_stream); archive & dc; //send archive_stream.str() over network RECIVING: should I use base class pointer when deserializing? or can I do it with base class reference? For instance: base *bc; bc & in_data; ///////// I'm little confused about reciving data.. I tried using base class reference for deserializing derived class, but it doesnt work and it seems unlogical to do it this way.. I would be really glad if someone could explain.. Thanks a lot

2 5

clear_vertex and property_map
by Celso Cardoso 16 Feb '07

16 Feb '07

Hi, I'm removing edges and vertices from the graph using the calls: clear_vertex(source); remove_vertex(source); But i have a property map also: property_map<graph_t, edge_weight_t>::type weightmap; The properties of the removed edges are removed "automatically" or do i have to do something else? I couldn't found any docs about that. Thanks, Celso.

2 1

tokenizing a string
by abir basak 16 Feb '07

16 Feb '07

Hi, I want to tokenize a string based on one or multiple space or tab std::string s = " a b c d \te "; std::vector<std::string> tokens; boost::algorithm::split(tokens,s,boost::algorithm::is_any_of(" \t"),token_compress_on); copy(tokens.begin(),tokens.end(),ostream_iterator<string>(cout,"\n")); This gives more or less correct result, except it does not remove space/ tab from begin /end. It can be done like trim(s); and then call the above. But is there any other way it can be done in split itself ?

3 3

boost::thread_specific_ptr performance?
by Josef Grahn 16 Feb '07

16 Feb '07

Hi! What performance characteristics can be expected from a thread_specific_ptr on various platforms, as compared to other synchronizing primitives? Does accessing it cost the equivalent of locking a mutex, or is it generally faster than that? Specifically, would the cost of accessing it be noticeable if one were to use it (naively) to manage detailed per-thread logging, generating many thousands of access calls per second or more. Thanks in advance! Best regards, Josef Grahn

1 0

Re: [Boost-users] Help
by Sergey Ostrovskiy 16 Feb '07

16 Feb '07

"Peter Dimov" <pdimov(a)mmltd.net> wrote in message news:<005201c7515d$f2fbdff0$6607a8c0@pdimov2>... > Sergey Ostrovskiy wrote: > > > Now I want to write something more difficult: > > > > apReports.push_back(make_report(boost::lambda::bind(&CItem::GetItemCount)*boost::lambda::bind((&CItem::GetItemBytes))); > > boost::lambda::bind(&CItem::GetItemCount, _1) * > boost::lambda::bind((&CItem::GetItemBytes, _1) Unfortunately this code produces errors during the compiling: error C2678: binary '*' : no operator found which takes a left-hand operand of type 'const boost::lambda::lambda_functor<T>' (or there is no acceptable conversion)

1 0

[signals] Chaining slots
by Allan Douglas 16 Feb '07

16 Feb '07

Hello, I'm using Boost.Signals to implement Deferreds (http://twistedmatrix.com/projects/core/documentation/howto/defer.html) An important feature of Deferreds is the hability to chain callbacks. Speaking of Signals, I want a way to call slots using args from previous slots calls. For example: float square(float x) { return x*x; } float sum1(float x) { return x+1; } boost::signal<float (float x)> sig; /* do some magic here */ sig.connect(&square); sig.connect(&sum1); sig.connect(&square); // sig(2) should call square(sum1(square(2))) std::cout << sig(2) << std::endl; I want the example to output 25 ( (2**2)+1)**2) Any ideas on how to implement this? Thanks, Allan Douglas

3 4

[mem_fn] Using mem_fn with std::bind2nd
by Scott Meyers 16 Feb '07

16 Feb '07

The mem_fn documentation is clear that there may be problems using mem_fn with the standard binders: > mem_fn is not 100% compatible with the standard adaptors, although it comes pretty close. In particular, mem_fn does not return objects of type std::[const_]mem_fun[1][_ref]_t, as the standard adaptors do, and it is not possible to fully describe the type of the first argument using the standard argument_type and first_argument_type nested typedefs. Libraries that need adaptable function objects in order to function might not like mem_fn. Fair enough. The program below fails to compile with VC8 and g++ 4.1.1, and the VC8 error message (which I find much more readable than the g++ one) suggests that I'm running into the problem mentioned above. However, the program DOES compile using Comeau 4.3.8, and that's what confuses me. Does anybody have any insights into whether this code is supposed to compile? I'm especially interested in any remarks about whether it should compile using std::tr1::mem_fn, should there be any differences between that and boost::mem_fn. Thanks, Scott #include <vector> #include <algorithm> #include <functional> #include "boost/mem_fn.hpp" struct Point { void rotate(int degrees); }; int main() { std::vector<Point> vp; std::for_each(vp.begin(), vp.end(), std::bind2nd(boost::mem_fn(&Point::rotate), 10)); }

2 2

Boost: Regular Expression question/bug?
by Derrick Schommer 16 Feb '07

16 Feb '07

Hi, I'm trying to understand if a regular expression I've designed is working correctly. It seems that I can make the regular expression match work "off by one" character in certain circumstances. For instance: The pattern: ".{1,3}" finds 1 to 3 characters, however The pattern: ".[^b]{1,3}" finds 1 to 4 characters...? For some reason when I apply a "do not allow the character b" or any other bracket operator the {min,max} range value seems to allow for one additional character. I cannot imagine this being done on purpose. The work around is to allow one less then you really want (i.e. {1,2}) but I figured I'd post something and see if anyone has seen it or could explain what is happening. Maybe its just a bug. The code example: boost::wregex pattern( TEXT( ".{1,3}" ) ); boost::wcmatch matchResults; if( boost::regex_match( TEXT( "asd" ), matchResults, pattern ) ) { // Match } else { // No match } } catch( runtime_error &e ) { printf( "\n%%FAILURE: Regular expression error: %s.\n", e.what() ); } Yes, I'm running in a UTF-16 build environment in Microsoft Windows .NET using C++. If this isn't the correct venue for this message, sorry in advance. Thanks, Derrick

3 4