May 2013 - Boost-users - lists.boost.org

[Xpressive] Nested patterns and sub-matches' semantic values
by Simon 30 May '13

30 May '13

Hi, For one part of a large project, I need to parse math and turn the "mathematical objects" into a C++ abstraction of them. Is it possible that a nested pattern hold semantic values within itself and can be queried for that value somehow? As if a match were a function of its sub-patterns' matches, if that makes sense... Code might speak better than my english: class math_number {...}; class math_addition {...}; class math_arglist {...}; class math_funcdef {...}; sregex integer = (s1= +_d) [ return math_number( as<int>(s1) ) ]; sregex addition = (s1= integer) >> '+' >> (s2= integer) [ return math_addition( as<math_number>(s1), as<math_number>(s2) ) ]; sregex arglist = (s1= identifier) >> *( ',' >> (s2= arglist)) [ std::vector<math_id> ids( as<math_arglist>(s2).begin(), as<math_arglist>(s2).end() ); ids.push_back(as<math_id>(s1)); return ids; ] sregex funcdef = ((s1= identifier) >> '(' >> (s2= arglist) >> ")=" >> (s3= addition)) [ return math_funcdef( as<math_id>(s1), as<math_arglist>(s2), as<math_addition>(s3) ) ]; I just want to avoid the funcdef above to be responsible for the construction of the various math_* objects the funcdef will hold (some of which will be complex to construct). And if I really must store the objects in an external stack, how should the parent-objects find their child-objects? Thanks, Simon

2 2

Confused with BGL-- Six Degrees of Kevin Bacon example
by lizy10b 30 May '13

30 May '13

Hi there, I am new to BGL library. I am walking through the online tutorial and just finished the "10.2. Six Degrees of Kevin Bacon". As it is said at the end of the example: "Note that vertex descriptor objects can not always be used as indices into vectors or arrays such as bacon_number. This is valid with the adjacency_list class with VertexList=vecS, but not with other variations of adjacency_list. A more generic way to index based on vertices is to use the ID property map (vertex_index_t) in coordination with the iterator_property_map.". So I turn to have a look at the iterator_property_map page and its sample. First, this sample uses VertexList=vecS too. Second, I know this sample is using iterator_property_map to build a property map, but I think the property is still indexed by the edge descriptor objects. So what is the difference? Thanks. Zhiyu LI 2013-05-29 lizy10b

2 5

[program options] parsing std::string
by Michael Powell 30 May '13

30 May '13

Hello, I want to parse std::string, perhaps into a vector<string>, but I do not want the command line argument to be positional in nature (i.e. like the examples demonstrate), because there may be a handful of parameters I need to capture into string or vector<string>. I've been able to parse integral types like int, even vector<int>, but haven't had much success with string. Possible or is this a program options "feature" and/or bug? Regards, Michael Powell

2 1

Does the Proactor of boost uses edge-triggered epoll on linux?
by Journeyer J. Joh 30 May '13

30 May '13

Hello list, I have some questions about boost's async IO implementation for linux. 1. Does it use edge-triggered epoll to implement the Proactor (in terms of Reactor)? http://www.boost.org/doc/libs/1_53_0/doc/html/boost_asio/overview/core/asyn… The guide above says that boost's Proactor is implemented in terms of Reactor when it comes to Linux. If it is based on Reactor, I guess it should be level-triggered epoll. 2. Am I right? Then my question is 3. Does boost use edge-triggered epoll in any kind of async IO on linux? 4. Isn't edge-triggered epoll a good method to implement async IO on linux? Thank you very much. Sincerely Journeyer ---------------------------------------- Journeyer J. Joh o o s a p r o g r a m m e r a t g m a i l d o t c o m ----------------------------------------

3 5

Install Boost.Build: gcc command not found (Windows)
by Tom C 29 May '13

29 May '13

I'm trying to build boost from sources following this guide: "http://www.boost.org/doc/libs/1_53_0/more/getting_started/windows.html#or-b…" I got the following outputs: here is bootstrap.log: here is my C:[...]\boost_1_53_0\tools\build\v2\user-config.jam: here is my PATH: It seems like the user-config.jam is not read by the bootstrap command. -- View this message in context: http://boost.2283326.n4.nabble.com/Install-Boost-Build-gcc-command-not-foun… Sent from the Boost - Users mailing list archive at Nabble.com.

3 2

Re: How to Boost static lib build under Mountain Lion and XCODE 4.6
by B Hart 27 May '13

27 May '13

Hi, I'm trying to install and use Boost 1_53_0 on MAC OS X Mountain Lion and use it with the latest XCODE 4.6.x. I want to port a small MS VS MFC C++ app. I developed on Windows to MAC. Specifically I need to install and build boost as static libraries. I'm completely new to programming on the MAC side of things. I understand that XCODE now uses LLVM Clang, with libc++ and not GCC. I've used GCC on Linux. I started following the instructions at: http://www.boost.org/doc/libs/1_53_0/more/getting_started/unix-variants.html but quickly ran into trouble. Does anyone have detailed instructions for how to install and compile boost statically under Mountain Lion and XCode 4.6? Below is the terminal input and output showing how far I got. Thanks. Terminal IO: ./bootstrap.sh --with-bjam=BJAM --with-toolset=TOOLSET users-Mac-mini:boost_1_53_0 user$ ./bootstrap --with-bjam=BJAM --with-toolset=TOOLSET -bash: ./bootstrap: No such file or directory users-Mac-mini:boost_1_53_0 user$ ./bootstrap.sh --with-bjam=BJAM --with-toolset=TOOLSET -n Detecting Python version... 2.7 -n Detecting Python root... /System/Library/Frameworks/Python.framework/Versions/2.7 -n Unicode/ICU support for Boost.Regex?... not found. Generating Boost.Build configuration in project-config.jam... Bootstrapping is done. To build, run: ./b2 To adjust configuration, edit 'project-config.jam'. Further information: - Command line help: ./b2 --help - Getting started guide: http://www.boost.org/more/getting_started/unix-variants.html - Boost.Build documentation: http://www.boost.org/boost-build2/doc/html/index.html users-Mac-mini:boost_1_53_0 user$ ./b2 link=static threading=multi toolset=darwin cxxflags="-arch i386 -arch x86_64" macosx-version=10.6 stage -bash: ./b2: No such file or directory

2 3

Using boost.Test check_is_small for my own
by Olaf Peter 26 May '13

26 May '13

Hello, I try to use parts of boost.test lib not for unit testing self but inside an algorithm. Anyway, at the end of header http://www.boost.org/doc/libs/1_53_0/boost/test/floating_point_comparison.h… this is written inside anonymous namespace it works obviously, but not at me. What happened here and how to fix it? Did I not read the header carefully? Thanks, Olaf ---8<---- #include <boost/test/floating_point_comparison.hpp> namespace tools { using boost::test_tools::check_is_small_t; check_is_small_t const& check_is_small = unit_test::ut_detail::static_constant<check_is_small_t>::value; } int main() { bool ok = tools::check_is_small(900, 0.1); } --->8---- test.cpp:5:30: error: 'boost::test_tools::check_is_small_t' has not been declared test.cpp:7:5: error: 'check_is_small_t' does not name a type test.cpp: In function 'int main()': test.cpp:12:15: error: 'check_is_small' is not a member of 'tools'

2 1

variant visitor using SFINAE problem
by Olaf Peter 26 May '13

26 May '13

Hi, attached my code to show my problem. It seems to work (partially) so far, but not the vector<int> where I get a compile error. This is my first try with SFINAE in that way. Are the conditions correct and robust enough? How to compile it? Did I forgot one use case? Using this small helper would simplify some post processing. Thanks, Olaf

2 4

[thread] "almost works" clarification
by Michael Powell 26 May '13

26 May '13

Hello, Okay, so without starting a flame war, because that is not my intention... Also bear in mind, I am coming from an informed C# .NET frame of mind having worked with .NET threading, synchronization, events, and so forth. I ran across this article, which although a bit dated, is a pretty good sketch of what I would consider a fairly robust thread runner abstraction, which if you read past the frustration, makes good sense, at least in my opinion: http://peter.bourgon.org/blog/2008/12/12/boost-thread-kindof-sucks.html. Nothing against the author, I might turn that around and couch it instead: boost-thread-almost-works: in the sense that, while there are indeed a rich set of primitives provided to hide the details of pthread, Win32 threading, or what have you, it stops just short of providing that abstraction layer for what Peter and I (probably just he and I) consider a robust runner, monitor, and worker. The ingredients are there, though, and with a little bit of persuading, can compose a fairly robust little set of threaded runners. Okay, with that said, am I correct in understanding, the combination of condition and mutex facilitates the analog to C# .NET AutoResetEvent or ManualResetEvent (loosely, ResetEvent) type thing? In other words, thread keeps-on-a-runnin' until condition is notified, at which point thread loop(s) know to stop working when they check in with their conditioned mutex? In other words, mechanically speaking, a worker-bee that starts obtains a lock on the mutex. While the worker is running, on-lookers may poll the monitor for thread completion (which times out as long as the worker-bee is running and has a scoped lock), or otherwise signal the thread to complete through the notify mechanism (which tells the worker-bee to return, which clears the lock). The worker-bee runs until it is completed or its conditioned mutex has been notified (whichever comes first). In another rendition, I was actually using the thread-interruption mechanism, which I may still do, is another way of signaling a thread that it should stop what it's doing. However, seems a bit less graceful than routing requests through a formalized abstraction layer I think, except through creative use of interrupt disabled (which I was also using), but is something else to consider. Anywho, food for thought. Comments, critiques welcome. Thank you... Regards, Michael Powell

4 5

[ublas] Linkage problems in assignment.hpp
by Jérémy Coulon 25 May '13

25 May '13

Hello, Is there any chance for the following bug to be fixed in 1.54 ? https://svn.boost.org/trac/boost/ticket/6010 Thanks. Jérémy

1 0

boost 1.53 parallel graph samples build or test failure on windows
by lizy10b 25 May '13

25 May '13

Hi there, I am a windows user and facing the similar problem as decribled in an archived thread posted by Jeremiah Willcock in 2010, but with no practical solutions: "Result statuses for run failures of MPI tests" http://boost.2283326.n4.nabble.com/Result-statuses-for-run-failures-of-MPI-… My environment: Boost 1.53, Win7 x64, mpich2-1.3.2p1-win-ia32, Python 2.66 x86, VS2010 SP1 The content of the user-cofig.jam file used to compile boost is: using python : 2.6 : C:\\Python266 ; using mpi : : <find-static-library>mpi <library-path>"D:\\Program Files (x86)\\MPICH2\\lib" <include>"D:\\Program Files (x86)\\MPICH2\\include" : "\"D:\\Program Files (x86)\\MPICH2\\bin\\mpiexec\"" ; and the bjam command is: bjam --prefix=E:\boost_1_53_0\debug32 stage toolset=msvc-10.0 variant=debug link=static address-model=32 threading=multi I did not use the bjam self-test to test all the parallel graph samples, but only picked some samples and build them in VS2010 and try to run. The dependencies linked in VS2010 when build these samples are: mpi.lib libboost_graph-vc100-mt-gd-1_53.lib libboost_graph_parallel-vc100-mt-gd-1_53.lib libboost_mpi-vc100-mt-gd-1_53.lib libboost_mpi_python-vc100-mt-gd-1_53.lib mpi-vc100-mt-gd-1_53.lib Testing results: No. Sample Name Build 1 proc test 2 proc test -------------------------------------------------------------------- 1 adjlist_redist_test.cpp Passed No Error No Error ----------------------------------------------------------------------- 2 adjlist_build_test.cpp Passed Error Error Error Message: Building graph with add_edge from everywhere...application called MPI_Abort(MPI_COMM_WORLD, -1) - process 0 application called MPI_Abort(MPI_COMM_WORLD, -1) - process 1 job aborted: rank: node: exit code[: error message] 0: lee-win7-x64: -1: process 0 exited without calling finalize 1: lee-win7-x64: -1: process 1 exited without calling finalize ----------------------------------------------------------------------------- 3 breadth_first_search.cpp Error -- -- Error Message: error LNK2019: unresolved external symbol "void __cdecl boost::detail::bfs_helper<class boost::adjacency_list<struct boost::vecS... ------------------------------------------------------------------------------ 4 dijkstra_shortest_paths.cpp Passed No Error No Error ------------------------------------------------------------------------------- 5 distributed_csr_test.cpp 18 Errors -- -- Error Messags (3 of 18): Error 5 error C2039: 'key_type' : is not a member of 'boost::property_traits<PA>' e:\boost\boost_1_53_0\debug32\include\boost-1_53\boost\graph\distributed\compressed_sparse_row_graph.hpp 1904 1 boost_test_1 Error 6 error C2146: syntax error : missing ';' before identifier 'local_key_type' e:\boost\boost_1_53_0\debug32\include\boost-1_53\boost\graph\distributed\compressed_sparse_row_graph.hpp 1904 1 boost_test_1 Error 12 error C2602: 'boost::property_map<Graph,Property,Enable>::value_type' is not a member of a base class of 'boost::property_map<Graph,Property,Enable>' e:\boost\boost_1_53_0\debug32\include\boost-1_53\boost\graph\distributed\compressed_sparse_row_graph.hpp 1906 1 boost_test_1 ----------------------------------------------------------------------------------------- So where is the problem? Any help is appreciated. Thanks Zhiyu Li Student China 2013-05-25 lizy10b

1 0

[Test] How to span tests into multiply files
by Daniele Barzotti 24 May '13

24 May '13

Hi, I want to divide my test suites into multiply files. I follow this example: http://stackoverflow.com/questions/2906095/boost-test-looking-for-a-working… testmain.cpp ----------------- #define BOOST_TEST_DYN_LINK #define BOOST_TEST_MODULE Main #include <boost/test/unit_test.hpp> testConfigClass.cpp ----------------- BOOST_AUTO_TEST_SUITE(ConfigClass_TestSuite) BOOST_AUTO_TEST_CASE(testConfigClass) { ... } BOOST_AUTO_TEST_SUITE_END() testDispatcherClass.cpp ----------------- BOOST_AUTO_TEST_SUITE(DispatcherClass_TestSuite) BOOST_AUTO_TEST_CASE(testDispatcherClass) { ... } BOOST_AUTO_TEST_SUITE_END() Maybe in the testConfigClass I have some problems, which I cannot figure out. When the test ends, the program terminate with: Exception first-chance a 0x7569c41f in TestSuite.exe: Exception of Microsoft C++: boost::execution_exception into memory position 0x0034dda4.. In this scenario, where can I found the 'main' to debug deeply? BR, Daniele.

2 1

[thread] SEVERE bug on packaged_task constructor in C++11 mode.
by Vicente J. Botet Escriba 23 May '13

23 May '13

Hi, There is a severe bug in packaged_task constructor when compiling in C++11 mode and giving a copyable functor as parameter, as reported in https://svn.boost.org/trac/boost/ticket/8596. I believe that this is not a regression, this have never worked. Until I find a fix, the packaged_task could be used only with free functions and with movable functors when compiling in C++11 mode. Fortunately there is no regression on C++98 compilers (at least not identified yet). Sorry for the disagreement, Vicente

2 1

[SmartPtr] shared_ptr as a default value in function template
by Igor R 23 May '13

23 May '13

Hello, In the wake of one of recent SO posts... The following doesn't compile in MSVC10 (Boost 1.53): #include <boost/shared_ptr.hpp> #include <boost/make_shared.hpp> template<class T> void f(T item, boost::shared_ptr<int> name = boost::make_shared<int>()) { } int main() { f(0); } Is it a bug in make_shared or the above violates some Standard paragraph? Thanks.

3 4

Is boost::this_thread::sleep_for interruption point?
by Serg Gulko 23 May '13

23 May '13

Hello! According to documentation(v 1.50.0) call boost::this_thread::sleep_for can be interrupted. For some reason this functionality not works for me. Here is code sample: //Simple thread function void Test::worker(int delay) { try { boost::this_thread::sleep_for(boost::chrono::milliseconds(delay)); cout << "Doing some work" << endl; } catch (boost::thread_interrupted const& e) { cout << "INTERRUPTED" << endl; } } //Far in the code boost::thread workerThread(boost::bind(&Test::worker, this, 1200)); boost::this_thread::sleep(boost::posix_time::milliseconds(30)); workerThread.interrupt(); boost::this_thread::sleep(boost::posix_time::milliseconds(1500)); I always getting 'Doing some work'... -- View this message in context: http://boost.2283326.n4.nabble.com/Is-boost-this-thread-sleep-for-interrupt… Sent from the Boost - Users mailing list archive at Nabble.com.

3 3

[msm] Manually setting next state
by samriti katoch 22 May '13

22 May '13

Thanks for the reply Christophe. I had considered this alternative, but adding a transition in every state to the initial state, when Error -> Ok defeats the purpose of having an orthogonal region, which I added to basically reduce transitions on an Error. -Samriti

1 0

Compile errors including boost/regex.hpp
by Jordan H. 22 May '13

22 May '13

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 When I try to compile a file that includes <boost/regex.hpp> I get the following error: # /home/me/Projects/myproject> make # /usr/bin/make all-am # make[1]: Entering directory `/home/me/Projects/myproject' # compiling main.cpp (g++) # In file included from /usr/local/include/boost/regex/regex_traits.hpp:27:0, # from /usr/local/include/boost/regex/v4/regex.hpp:39, # from /usr/local/include/boost/regex.hpp:31, # [ ... more traceback ... ] # # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: macro "test" passed 2 arguments, but # takes just 1 # In file included from /usr/local/include/boost/regex/v4/regex.hpp:79:0, # from /usr/local/include/boost/regex.hpp:31, # [ ... more traceback ... ] # # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: macro "test" passed 2 arguments, but # takes just 1 # In file included from /usr/local/include/boost/regex/regex_traits.hpp:27:0, # from /usr/local/include/boost/regex/v4/regex.hpp:39, # from /usr/local/include/boost/regex.hpp:31, # [ ... more traceback ... ] # # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: template declaration of ‘char (& # boost::re_detail::test)[2]’ # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: expected unqualified-id before ‘...’ # token # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: expected ‘)’ before ‘...’ token # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: expected ‘)’ before ‘...’ token # /usr/local/include/boost/regex/v4/regex_traits.hpp:87:1: error: expected unqualified-id before ‘(’ # token # In file included from /usr/local/include/boost/regex/v4/regex.hpp:79:0, # from /usr/local/include/boost/regex.hpp:31, # [ ... more traceback ... ] # # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: template declaration of ‘char (& # boost::re_detail::test)[2]’ # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: expected unqualified-id before ‘...’ # token # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: expected ‘)’ before ‘...’ token # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: expected ‘)’ before ‘...’ token # /usr/local/include/boost/regex/v4/regex_format.hpp:849:1: error: expected unqualified-id before ‘(’ # token # make[1]: *** [myproject-main.o] Error 1 # make[1]: Leaving directory `/home/me/Projects/myproject' # make: *** [all] Error 2 # *** Failure: Exit code 2 *** This is how I'm compiling: # g++ -DHAVE_CONFIG_H -I. `mysql_config --include` `mysql_config --cflags` -std=c++0x -g -O2 -MT # myproject-main.o -MD -MP -MF .deps/myproject-main.Tpo -c -o myproject-main.o `test -f 'src/main.cpp' # || echo './'`src/main.cpp Basically, the compile errors trace back to when I include boost/regex.hpp: | #include <boost/regex.hpp> Any ideas why this occurs? Maybe it's because I'm not including -lbooost_regex as a compile option? I specified it as my "myproject_LDADD" macro in Makefile.am so I assume g++ knows when to use it. Thanks. Sorry for the lengthy email. -----BEGIN PGP SIGNATURE----- iQEcBAEBCgAGBQJRnN/nAAoJEOT99oqiSuRNim0H/1U+xgx+TCwOtDiGfC5FH6yR 1HwZYwTb0F/7CwldOBsPoDMfjSJmptH4a4s84PR0xDHFrhvgDkjDvzYkoWPLR7gf eSAMHzQG3kvJawvHiJNSu/eTZivVfkCE14+zoE+cVriCJicJUJg1g8hp5Cydk6D3 OM9WtOY/GCVHD8lK84UX5HMqnjeX4nPLEA7nJGd1u2+5PkHxpQMF2RO+EWnlNm0e eTUdZBlGLc/RGHMgl2uF1q3b+qCRrC/Wwn/4OtkDUn0Y3V879LbkYej06oB94zWD 9kKHAnTFOisQJDUmkHr9BiEGgJ9UNO2JCVkHlzrjVgfwKe1r6FIaZmTGNpIme0c= =Av+g -----END PGP SIGNATURE-----

3 6

[msm] Manually setting next state
by samriti katoch 22 May '13

22 May '13

Hi, Is there a way to manually set the next state for the state machine. I have an orthogonal region defined for handling errors, where Error is an interruptible state. ---------- ------------- | | ----------->| Error | | Ok | | | ---------- ------------- This is similar to the example here : http://svn.boost.org/svn/boost/trunk/libs/msm/doc/HTML/examples/Orthogonal-… Once the state machine recovers from the error, I would like to start from the initial state again in the main orthogonal region. Is there a way to do this ? Thanks! Samriti

2 1

[atomic] When to load/store
by Michael Powell 22 May '13

22 May '13

Hello, I have some thread synchronization to do, probably will end up using some sort of internally synchronized cancelation or other state to message between parent and child thread. If the state uses boost::atomic<bool> _canceled, for instance, how do I go about setting/getting the value? I am assuming that we do something like this: //To get the value. return _canceled.load(boost::memory_order::memory_order_acq_rel); //To set the value. _canceled.store(canceled, boost::memory_order::memory_order_acq_rel); Using acquire/release to get in, do what we need with the value, and get out, type thing. Also bearing in mind this is targeting ARM architecture (ARM5 I think). Also, how complex a value can be protected? Are we talking primitive types only? Basic data types like bool, int, float, that sort of thing? Thank you... Regards, Michael Powell

2 1

[Atomic] / PPC / error: impossible constraint in 'asm'
by nicolas.danet＠free.fr 22 May '13

22 May '13

Hello, I get the error below compiling with (Xcode 2.5 / OS X 10.4.11) only for PPC arch. boost_1_53_0/boost/atomic/detail/gcc-ppc.hpp:1204: error: impossible constraint in 'asm' The related code is: static boost::atomic<long> messageCount_(0); number_ = messageCount_.fetch_add(1, boost::memory_order_relaxed); Is this a known bug ? Is there any way to fix it ? Thanks.

1 0

[accumulator] with complex domain model
by Michael Powell 21 May '13

21 May '13

Hello, We've got a complex domain model which must be accumulated and eventually averaged together. I am already overloading the operators for that model. We could roll our own with just the operators we'll need to accommodate, like +, +=, /, etc, but would like to take a gander at Boost.Accumulator first. Would be much better, obviously, rather than having to reinvent the wheel. How doable would it be to accumulate an arbitrarily complex domain model. That's assuming that it is sensible to add, average, and so on, the model, which it is. Regards, Michael Powell

2 2

endless loop and/or crash in boost::interprocess
by Charanga 21 May '13

21 May '13

How do I debug a potential endless loop and heap corruption issue involving boost::interprocess "new managed_shared_memory" freezing forever or crashing? I posted this on stackoverflow but I am hopeful that a real expert in Boost::Interprocess may have seen this freeze-up and crash before. Outside the debugger it manifests as a freeze-up, inside the debugger it manifests as a series of exceptions inside an endless try-catch-ignore loop which is coded into the boost implementation code. http://stackoverflow.com/questions/16651878/how-do-i-debug-a-potential-endl… Warren

2 1

[SmartPtr] enable_shared_from_raw / weak_from_raw remains not associated
by Slava 20 May '13

20 May '13

I'm converting existing code with parent-child object tree to boost shared_ptr/weak_ptr with parents having [now shared] ownership over the children and children having weak reference to the parent. In the existing code a parent usually becomes the children as constructor parameters, and thus I have to extract the weak_ptr to "this" in constructor, which does not work with current boost::enable_shared_from_this implementation. So I am looking to enable_shared_from_raw alternative, which is promising, but unfortunately does not quite work for my case. Here is a small example demonstrating the issue: test.cpp: #include <boost/smart_ptr/enable_shared_from_raw.hpp> #include <iostream> struct Parent; struct Child { Child() { std::cout << "Construct Child" << std::endl; } ~Child() { std::cout << "Destruct Child" << std::endl; } boost::weak_ptr<Parent> parent; }; struct Parent : public boost::enable_shared_from_raw { Parent(boost::shared_ptr<Child> _child) { std::cout << "Construct Parent" << std::endl; this->child = _child; this->child->parent = boost::weak_from_raw(this); } ~Parent() { std::cout << "Destruct Parent" << std::endl; } boost::shared_ptr<Child> child; }; int main() { boost::shared_ptr<Child> child(new Child()); boost::shared_ptr<Parent> parent(new Parent(child)); std::cout << "child->parent is " << (child->parent.lock() ? "not null" : "null") << std::endl; parent.reset(); std::cout << "child must be alone" << std::endl; std::cout << "child->parent is " << (child->parent.lock() ? "not null" : "null") << std::endl; return 0; } `g++ -g -O0 test.cpp && valgrind a.out` produces: ==8467== Memcheck, a memory error detector ==8467== Copyright (C) 2002-2012, and GNU GPL'd, by Julian Seward et al. ==8467== Using Valgrind-3.8.1 and LibVEX; rerun with -h for copyright info ==8467== Command: a.out ==8467== Construct Child Construct Parent child->parent is null Destruct Parent child must be alone child->parent is null Destruct Child ==8467== ==8467== HEAP SUMMARY: ==8467== in use at exit: 0 bytes in 0 blocks ==8467== total heap usage: 4 allocs, 4 frees, 112 bytes allocated ==8467== ==8467== All heap blocks were freed -- no leaks are possible ==8467== ==8467== For counts of detected and suppressed errors, rerun with: -v ==8467== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 2 from 2) The problem is that the weak_ptr to parent in child does not get associated properly, when shared_ptr to parent is finally complete. Do I expect too much from enable_shared_from_raw or did I hit some bug in the implementation? I'm using boost 1.53.0-2 on Archlinux. -- Slava

2 3

[boost][thread] Future's scheduler
by Klaim - Joël Lamotte 20 May '13

20 May '13

I'm using boost.thread V4 in my projects because I would like to use futures. I understand that boost 1.53 implementation of future.then() is incomplete but I think it's ok for my current use. However I think there might be something missing. I'm in a case where the user of my library provide his own task scheduler (or executor) which means that when I have functions that return futures, they will use that scheduler. I would like to limit the boilerplate necessary for the user to schedule tasks properly when one task is finished, so I assume the user will use then() with just lambdas or function pointers. The documentation of then() says: - When the scheduler or launch policy is not provided the continuation inherits the parent's launch policy or scheduler. Which is perfect to me as I just have to associate the first future with the scheduler, the user that will use .then() will not have to specify which scheduler to use. My problem is that I can't find how to associate a scheduler to a future (without using .then()). There is apparently no way to do this in boost::future or boost::promise interfaces. Am I missing something? I see that I could use a dummy initial future just to build another future with .then() which would associate the second future with the scheduler, but this is clearly just ductapping. Joel Lamotte

1 1

64 bit or 32 bit?
by Scott Alexander 20 May '13

20 May '13

Hi, I have a Windows 8 64 bit OS and would like to use Visual Studio Express 2012 for Windows 8. Should I install a 64 bit or 32 bit version of Boost? Thanks, Scott

4 4