February 2009 - Boost-users

How to use boost::regex source code without using any *.lib file
by Jun Wan 02 Mar '09

02 Mar '09

---------- Forwarded message ---------- From: Jun Wan <junjun.wanjun(a)gmail.com> Date: 2009/2/17 Subject: How to use boost::regex source code without using any *.lib file To: boost-users(a)lists.boost.org Hello, Recently I used boost::regex in my project(VS2005 IDE). It works well with linking the lib file. But I want to use the source code of boost::regex.I include the regex.hpp in the project and used BOOST_REGEX_NO_LIB macro; But I got such error infomation when linking the code. error LNK2001: unresolved external symbol "private: class boost::basic_regex<char,struct boost::regex_traits<char,class boost::w32_regex_traits<char> > > & __thiscall boost::basic_regex<char,struct boost::regex_traits<char,class boost::w32_regex_traits<char> > >::do_assign(char const *,char const *,unsigned int)" (?do_assign@?$basic_regex@DU?$regex_traits@DV ?$w32_regex_traits@D@boost@@@boost@@@boost@@AAEAAV12@PBD0I@Z) error LNK2001: unresolved external symbol "private: void __thiscall boost::re_detail::perl_matcher<char const *,class std::allocator<struct boost::sub_match<char const *> >,struct boost::regex_traits<char,class boost::w32_regex_traits<char> > >::construct_init(class boost::basic_regex<char,struct boost::regex_traits<char,class boost::w32_regex_traits<char> > > const &,enum boost::regex_constants::_match_flags)" (?construct_init@?$perl_matcher@PBDV ?$allocator@U?$sub_match@PBD@boost@@@std@@U?$regex_traits@DV ?$w32_regex_traits@D@boost@@@boost@@@re_detail@boost @@AAEXABV?$basic_regex@DU?$regex_traits@DV?$w32_regex_traits@D@boost@ @@boost@@@3@W4_match_flags@regex_constants@3@@Z) error LNK2001: unresolved external symbol "public: bool __thiscall boost::re_detail::perl_matcher<char const *,class std::allocator<struct boost::sub_match<char const *> >,struct boost::regex_traits<char,class boost::w32_regex_traits<char> > >::find(void)" (?find@?$perl_matcher@PBDV ?$allocator@U?$sub_match@PBD@boost@@@std@@U?$regex_traits@DV ?$w32_regex_traits@D@boost@@@boost@@@re_detail@boost@@QAE_NXZ) I want to know why I can't use the source code of boost::regex. Is there any thing I need to do? Any help will be appreciated! Thank you! Jun

3 4

Fastest way of serializing a huge vector of ints into a human readable string
by Roland Bock 02 Mar '09

02 Mar '09

Hi, I have a program which produces a vector of integers (several million entries). I need to write that into a human-readable string of space separated numbers. I wonder, what would be the fastest way? My first attempt was stringstream resultStream; copy(integers.begin(), integers.end(), ostream_iterator<int>(resultStream, " ")); string result = resultStream.str(); But that requires copying the string. Thus I was wondering if I could use boost to write to the target string directly? boost::iostreams allow me to do the following: string result; boost::iostreams::filtering_ostream out(boost::iostreams::back_inserter(result)); copy(integers.begin(), integers.end(), ostream_iterator<uint64>(out, " ")); This is faster by almost a factor of 2. Any ideas how to increase speed even more? Thanks in advance, Roland

3 11

[Graph] Using a bit-based ajacency matrix.
by Thorsten Ottosen 01 Mar '09

01 Mar '09

Dear All, Does anybody have experience with the above. That is, to represent the graph as a bit-based ajacency matrix, and then use that together with Boost.Graph? I'm thinking that Boost.Graph will not exploit this representation optimally. Am I correct? Thanks in advance -Thorsten

3 8

newbie thread synchronization question
by Cory Riddell 01 Mar '09

01 Mar '09

I'm a total newbie with the Thread library, so please forgive me if this is a dumb question. Say I make a class called Shared that looks something like this: class Shared { private: boost::mutex monitor; int qty; public: Shared() : qty(0) { } void lock() { monitor.lock(); } void unlock() { monitor.unlock(); } int getQty() { return qty; } void setQty(int q) { qty = q; } }; If I use it in from multiple threads with code that looks like: // s is an instance of Shared s.lock(); int q = s.getQty(); ++q; s.setQty(q); s.unlock(); will it work reliably on a multi-core machine? When the mutex unlock() happens, does that flush the CPU cache and write the value of s.qty to main memory? Does the mutex lock() operation clear the CPU cache and force a reload of s from the main memory? I'm trying to understand how synchronization works among multiple cpus. I've only worked on multithreaded code in Java and there the memory model is well defined. AFAIK, that isn't the case in C++. Cory

3 2

Some question about cross-compiling boost
by 邹磊 01 Mar '09

01 Mar '09

Hi, Volodya. Thanks for your reply. I have tried the method in the document ( http://www.boost.org/boost-build2/doc/html/bbv2/tasks/crosscompile.html ), there were still some problems. Some libraries has been built successfully, but many not (such as libboost_serialization, libboost_wave and so on). I got the information as following: >>...failed gcc.compile.c++ bin.v2/libs/math/build/gcc-3.8.0/release/link-static/threading-multi/sph_neumannf.o... >>.......... >>...skipped <p/usr/local/lib/>libboost_serialization-gcc38-mt.a for lack of<p/usr/local/lib>libboost_serialization-gcc38-mt-1_38.a... I didn't know why these libraries built failed. I wonder whether the problem is due to my gcc kernel of cross-compiler . Maybe it's version is too low to build that libraries? How to solve the problem? Thank you! Zou ___________________________________________________________ 好玩贺卡等你发，邮箱贺卡全新上线！ http://card.mail.cn.yahoo.com/

2 3

Having a tough time installing and setting up Boost and associated libraries
by ajay gopalakrishnan 01 Mar '09

01 Mar '09

Hi all, I am a newbie when it comes to using boost. My primary motivation behind using boost is that I want to use the Text processing, Numerical computing facilities together in one common language. Boost Regex, Spriti and The BLas Libraries would serve my purpose. However, I am having a tough time installing and running it. I am not able to find any good documentation for using this. My requirement is this: 1. I want to be able to install Boost locally in my Unix account. I don't have root privileges. 2. There is a wrapper called Boost Numerics Package written and hosted at http://mathema.tician.de/software/boost-bindings. I want to be able to use this also. This allows me to use ATLAS, BLAS in Boost. I would also like to know, how tough is it to interface C++ code with Python using Boost.Python. I am not an expert in Python. What is the procedure to do this? Your help will be highly appreciated. Thanks, Ajay G.

3 2

shared_ptr memory leak
by gast128 01 Mar '09

01 Mar '09

Dear all, we managed to create a shared_ptr memory leak, without making a cycle. Granted it is highly controversial case. One has to make call in the destructor of the owned object which loops back to the point of assignment. Is this something for the documentation, or may this indeed be flagged as a 'nut' case. Code below is a simplified example, the production case was with some kind of change manager which notification bounced back to the point of assignment. One has to split out header and cpp to get it compilable. void TestSharedPtrLeak() { //make the leak KTestBoostHlpSpLeakParent leak; leak.Reset(false); } //ktestboosthlpspleakparent.h struct KTestBoostHlpSpLeakParent { KTestBoostHlpSpLeakParent(); void Reset(bool b); boost::shared_ptr<KTestBoostHlpSpLeak> m_ptr; }; //ktestboosthlpspleakparent.cpp KTestBoostHlpSpLeakParent::KTestBoostHlpSpLeakParent() { Reset(true); } void KTestBoostHlpSpLeakParent::Reset(bool b) { boost::shared_ptr<KTestBoostHlpSpLeak> ptr(new KTestBoostHlpSpLeak(this, b)); m_ptr = ptr; //m_ptr.swap(ptr); //would solve it btw } //ktestboosthlpspleak.h struct KTestBoostHlpSpLeak { KTestBoostHlpSpLeak(KTestBoostHlpSpLeakParent* p, bool b); ~KTestBoostHlpSpLeak(); public: KTestBoostHlpSpLeakParent* m_p; bool m_b; }; //ktestboosthlpspleak.cpp KTestBoostHlpSpLeak::KTestBoostHlpSpLeak(KTestBoostHlpSpLeakParent* p, bool b) : m_p(p) , m_b(b) { } KTestBoostHlpSpLeak::~KTestBoostHlpSpLeak() { if (m_b) { //recursive call m_p->Reset(false); } }

2 2

error when using boost with visual c++
by Sean Farrow 01 Mar '09

01 Mar '09

Hi; When using boost with visual c++ 2005, I'm getting the following two error: c:\program files\microsoft visual studio 8\vc\include\intrin.h(198) : error C2373: '_InterlockedExchange' : redefinition; different type modifiers c:\program files\boost\boost\detail\interlocked.hpp(60) : see declaration of '_InterlockedExchange' Has anyone got any ideas what is causing this? I'm including the following files in the file that is indicating the error: #include <vector> #include <boost/algorithm/string.hpp> #include <boost/array.hpp> #include <boost/filesystem/path.hpp> #include <boost/foreach.hpp> #include <windows.h> #include <psapi.h> #include <ole2.h> #include <exdisp.h> #include <mshtml.h> #include "O2ServerImpl.h" #include "Globals.h" //Crypto++ remove as necessary. #include <default.h> #include <filters.h> #include <hex.h> #include <md5.h> #include <ripemd.h> #include <rng.h> #include <gzip.h> #include <randpool.h> #include <ida.h> #include <base64.h> #include <factory.h> #include <whrlpool.h> #include <tiger.h> any help appreciated. Sean.

2 3

[Proto] Compialtion-time of user-defined functions : terminals vs make_expr
by Joel Falcou 28 Feb '09

28 Feb '09

As stated in a previous thread, the question was open on what to choose to build large number of functions inside Proto and Eric mentionned being interested by the results .. so ;) The choices were : - using a terminal containing the Polymorphic Function Object of the function and statcially instanciate it bp::terminal< my_func_pfo > my_func = {{}}; - using make_expr in a template function Here is the breakdown of measures I run this day using g++-4.3 and latest proto. I used Boost::Preprocessor to generate all the code and measured the compiel time on a streak of 100 compialtion for eahc version. Naked main : 0.05s Naked main w/ proto : 1.50s => overhead of proto include = 1.45s Without Call to the actual function: Main with 1 proto term : 1.52s => overhead = 0.02s = 0.020s/term Main with 10 proto term : 1.55s => overhead = 0.05s = 0.005s/term Main with 100 proto term : 1.99s => overhead = 0.49s = 0.005s/term Main with 150 proto term : 2.55s => overhead = 1.05s = 0.007s/term Main with 200 proto term : 3.48s => overhead = 1.98s = 0.009S/term Main with 256 proto term : 4.80s => overhead = 3.30s = 0.013s/term Main with 1 proto func : 1.52s => overhead = 0.02s for one function = 0.0200s/func Main with 10 proto func : 1.53s => overhead = 0.03s for one function = 0.0030s/func Main with 100 proto func : 1.53s => overhead = 0.03s for one function = 0.00030s/func Main with 150 proto func : 1.55s => overhead = 0.05s for one function = 0.00033s/func Main with 200 proto func : 1.57s => overhead = 1.98s for one function = 0.00035s/func Main with 256 proto func : 1.61s => overhead = 3.30s for one function = 0.00043s/func With Call to the defined functions in sequence : func behaves like term without call (aka compilation time between 1.5s and 4.8s) term behaves like term without call with a linear overhead of 0.017s/call (aka compilation time between 4.5s and 9.1s) Other measures : executable size skyrocket with term even without any functions being called. 256 functions instanciated yields a 7.3kb binary vs a 1.4kb binary for make_expr func so an overhead of 23,6 bytes/functions in the case of terminal object The scaling seems fine and make_expr is, as expected, the best contender for large number of user-defined expression generator. -- ___________________________________________ Joel Falcou - Assistant Professor PARALL Team - LRI - Universite Paris Sud XI Tel : (+33)1 69 15 66 35

2 2

asio Socket disconnection not working in 1.38?
by Michael Dehmlow 28 Feb '09

28 Feb '09

I'm trying to detect that a peer has been disconnected using the asio socket library. Below is a source listing of a unit test that I have created (which is one of several that I have tried; ask for more listings if nesasary) to figure out a reliable method to do so. I've also tried using (and have unit test listings for) the availble(errorCode) call the sock1.remote_endpoint(errorCode); calls. I need a way to detect that a client has been disconnected from my server without blocking. I've even tried using the read_some call with the io_control non_blocking option set but the then read some returns an error every time read_some is called. [code] for(int i=0;i<1;i++) { boost::asio::io_service ioServ1; boost::asio::io_service ioServ2; boost::asio::io_service ioServ3; boost::asio::io_service ioServ4; boost::asio::ip::tcp::socket sock1(ioServ1); boost::asio::ip::tcp::socket sock3(ioServ3); SocketAcceptor sA1(ioServ1,sock1,std::string("localHost"),8321); SocketAcceptor sA2(ioServ3,sock3,std::string("localHost"),8322); // The socket acceptor class simple starts a thread and and accept(_mySocket,ec) // with a boost::asio::ip::tcp::acceptor. If thread does not completed in the time passed in //the milliseconds then the call then the the accept call will return. sA1.accept(10); sA2.accept(10); boost::asio::ip::tcp::socket sock2(ioServ2); boost::asio::ip::tcp::socket sock4(ioServ4); boost::asio::ip::tcp::resolver resolver1(ioServ2); boost::asio::ip::tcp::resolver::query query1(boost::asio::ip::tcp::v4(), "localhost", (boost::lexical_cast<std::string>(8321)).c_str()); boost::asio::ip::tcp::resolver::iterator iterator1 = resolver1.resolve(query1); boost::asio::ip::tcp::resolver resolver2(ioServ4); boost::asio::ip::tcp::resolver::query query2(boost::asio::ip::tcp::v4(), "localhost", (boost::lexical_cast<std::string>(8322)).c_str()); boost::asio::ip::tcp::resolver::iterator iterator2 = resolver2.resolve(query2); sock2.connect(*iterator1); sock4.connect(*iterator2); TS_ASSERT_EQUALS(sA1.accept(200),true); TS_ASSERT_EQUALS(sA2.accept(200),true); boost::system::error_code errorCode; //std::cout<<"about to read some from socket"<<std::endl; //boost::asio::socket_base::non_blocking_io command(true); //sock1.io_control(command); //TS_ASSERT_EQUALS(sock1.read_some(boost::asio::buffer(data, length), errorCode),0); //std::cout<<"Hopefully that didn't take long"<<std::endl; sock2.close(); sock4.close(); //just a boost wrapped sleep call Thread::sleep(2000); //this assertion will always fail just used to see what is in the errorCode object. TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); sock1.available(errorCode); TS_ASSERT(errorCode); TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); sock2.available(errorCode); TS_ASSERT(errorCode); TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); //TS_ASSERT_EQUALS(sock3.read_some(boost::asio::buffer(data, length), errorCode),0); //TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); // TS_ASSERT(errorCode); sock3.available(errorCode); TS_ASSERT(errorCode); TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); sock3.available(errorCode); TS_ASSERT(errorCode); TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); //TS_ASSERT_EQUALS(sock2.read_some(boost::asio::buffer(data, length), errorCode),0); //TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); //TS_ASSERT(errorCode); sock4.available(errorCode); TS_ASSERT(errorCode); TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); //TS_ASSERT_EQUALS(sock4.read_some(boost::asio::buffer(data, length), errorCode),0); //TS_ASSERT_EQUALS(boost::asio::error::eof,errorCode); //TS_ASSERT(errorCode); TS_ASSERT(!sock1.is_open()); TS_ASSERT(!sock3.is_open()); } [/code] The output: //expected failure Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 00 00 00 00 38 A4 63 00 }) //sock1 no errorCode Error: Assertion failed: errorCode //sock1 no errorCode Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 00 00 00 00 38 A4 63 00 }) //-------------------------------- //sock2 ALWAYS returns an errorcode //-------------------------------- //sock2 some other errorCode Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 19 27 00 00 38 A4 63 00 }) //sock3 no errorCode Error: Assertion failed: errorCode Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 00 00 00 00 38 A4 63 00 }) //sock4 no errorCode Error: Assertion failed: errorCode Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 00 00 00 00 38 A4 63 00 }) //-------------------------------- //sock4 ALWAYS returns an errorcode //-------------------------------- //sock4 some other errorCode Error: Expected (boost::asio::error::eof == errorCode), found ({ 02 00 00 00 } != { 19 27 00 00 38 A4 63 00 }) //is_open doesn't seem all that helpful either Error: Assertion failed: !sock1.is_open() Error: Assertion failed: !sock3.is_open() -- View this message in context: http://www.nabble.com/asio-Socket-disconnection-not-working-in-1.38--tp2223… Sent from the Boost - Users mailing list archive at Nabble.com.

2 3