November 2009 - Boost - lists.preview.boost.org

[gil] Committing some code changes
by Christian Henning 10 Nov '09

10 Nov '09

Hi there, I'm in the process to add some non-invasive code changes for boost::gil. Since this is my first time to commit changes to boost I would like to make sure I'm doing it right. The changes would be added to the trunk. Is that the right place? I have done the obvious steps on my local machine. And besides adding comments when checking in my changes what other steps do I have to do? Thanks, Christian

2 3

Jamfile to control MSVC compiler options /Za and /Ze
by Paul A. Bristow 10 Nov '09

10 Nov '09

I have been taking my own medicine and trying to get Boost.math free of warnings with the /Za No MS extensions option. Mostly this is fine. project : requirements <toolset>msvc:<cxxflags>/Za # disable MS Extensions in Boost.Test. ... But a few tests *require* the extensions. I have tried to enable them for just one module thus: # pow test requires type_of that MS extensions. run pow_test.cpp ../../test/build//boost_test_exec_monitor : # command line : # input files : # requirements <toolset>msvc:<cxxflags>/Ze # Requires type_of which requires MS extensions. : test_pow ; but the compiler complains about deprecation (and most importantly, that later options do not override, so you can't specify /Za and them /Ze. compile-c-c++ ..\..\..\bin.v2\libs\math\test\test_pow.test\msvc-9.0\debug\asynch-exceptions-on \threading-multi\pow_test.obj cl : Command line warning D9035 : option 'Ze' has been deprecated and will be removed in a future release cl : Command line error D8016 : '/Za' and '/Ze' command-line options are incompatible Can any jam experts suggest a way round this (short of leaving the default enable extensions, and making an exception for all the vast majority of modules that are OK with /Za - too tedious to be attractive? Splitting into two classes of tests is also ugly.) Thanks. Paul --- Paul A. Bristow Prizet Farmhouse Kendal, UK LA8 8AB +44 1539 561830, mobile +44 7714330204 pbristow(a)hetp.u-net.com

2 1

[smart_ptr] make_shared_array/allocate_shared_array function templates
by Roman Kecher 09 Nov '09

09 Nov '09

Hi, I'm suggesting to add the following function templates to allow easy, exception-safe, creation of shared_array objects: - make_shared_array (very similar to make_shared which is used to create a shared_ptr) - allocate_shared_array (again, same functionality like allocate_shared) Is there any reason not to include these? Regards, Roman.

3 4

Time for another bug sprint (after 1.41 ships)?
by Marshall Clow 09 Nov '09

09 Nov '09

Earlier this year, we had a "bug sprint", where many people focused on fixing bugs in Boost. We managed to get the # of open tickets down from 800 to about 720. As of today, there are 820 open tickets in Trac. Is it time to focus on swotting bugs again? 1.41 is scheduled for release on 1-November; that's six weeks from yesterday. How about a bug sprint starting on 14-November, finishing on 22- November? That's two weekends - and finishes up before the American Thanksgiving holiday. -- Marshall P.S. This doesn't mean that people have to wait until November to fix bugs ;-) ;-)

12 25

[function] "paranoia" vs efficiency
by Domagoj Saric 09 Nov '09

09 Nov '09

As far as I can see boost::function<>::operator() performs: if (this->empty()) boost::throw_exception(bad_function_call()); before invoking the stored function (object)... My question is why or why only this "checked"/"safe" approach/method of invocation exists? Because this: - causes (extra) code branching - causes the compiler to insert exception throwing and handling code in surrounding/calling code (which need not otherwise be if the function stored in the boost function instance is itself nothrow) - inserts a call to the bad_function_call constructor - disables the compiler from inlining the function (MSVC atleast) - disables NVRO (MSVC atleast, http://msdn.microsoft.com/en-us/library/ms364057(VS.80).aspx#nrvo_cpp05_top… ) while at the same time I would most often consider calling a "null-function-pointer" a programmer error (that would therefor need to be "handled" with an assert not an exception), not a runtime 'unpredictable' exception... ...IMO the default should be not to include this check (i.e. use an assert) and provide a safeInvoke() type of method for those users and cases that actually require/want this type of check/behaviour... ...or atleast "it would be welcomed" if a non-checking unsafeInvoke() type of method would be provided for those that are 'sure' that the boost::function<> object is valid (for example in the line before i've just performed the same check to decide whether to call the function object or not)... -- "That men do not learn very much from the lessons of history is the most important of all the lessons of history." Aldous Huxley

8 18

[shared_ptr] Set pointer to null
by Mateusz Loskot 09 Nov '09

09 Nov '09

Hi, I'd like to make a custom deleter resetting pointer to 0 or other value (e.g. for debugging purposes) after resource has been deallocated. Would anyone suggest better option to this presented below? struct DeleterT { void operator()(::T* ptr) { ::DestroyT(ptr); // uses Visual C++ specific integer type *reinterpret_cast<__int64*>(&ptr) = 0; } }; Best regards, -- Mateusz Loskot, http://mateusz.loskot.net

2 2

[iostreams] building zlib from source
by Vladimir Prus 09 Nov '09

09 Nov '09

I am working on improving detection of external dependencies (for 1.42), and I've noticed that the iostreams library is rather smart when it comes to building zlib (and bz2). It allows the user to specify either: - location of headers and libraries files, or - location of sources for zlib In the latter case, the sources will be compiled automatically. How important is the second use case? It's not very hard to support, and appears handy for Windows users, but if in practice, nobody cares, I can spend the time on other matters. Thanks, Volodya

2 3

[thread] Is interrupting a non boost.thread possible?
by Christoph Mathys 09 Nov '09

09 Nov '09

Hello I've a question about boost.threads: Is it possible to interrupt a thread for which I don't have a thread-object but which is using boost synchronization primitives (mutex, condition)? Thanks! Christoph

2 1

[1.41.0] Beta 2 plans
by Beman Dawes 08 Nov '09

08 Nov '09

Due to the issues exposed in the "Official warnings policy?" thread, there will be a second 1.41.0 beta, tentatively scheduled for Sunday, November 15th. Until Saturday, November 14th, the release branch is open for merges of low-risk bug fixes. In particular, fixes that clear compiler warnings are encouraged. Please do, however, make sure any changes are stable on trunk before merging them to release. I'll be traveling until Monday, with little or no email connectivity. So if questions arise, it would be great if the other release managers could step in with responses. Thanks, --Beman

5 5

ppc64 and OSX Snow Leopard
by Vladimir Prus 08 Nov '09

08 Nov '09

Folks, it was reported that in Snow Leopard, Apple has removed support for ppc64, and folks trying to use that receive fairly obscure errors. Some of Boost users were already confused by this. Now, this is of course more like Apple's problem, but here's a patch, against current trunk, to do a nice diagnostic of this in Boost.Build. The intention is that if you try to build just ppc64, or a two-way fat binary including ppc64 you get an error message, and if you try tobuild 4-way fat binary, ppc64 is skipped and you get 3-way binary. Can some of OSX users apply this patch to trunk and give it a try? If it works, I would like to merge to release branch. Thanks, Volodya

2 1

[1.41.0] Beta 1 release candidate
by Beman Dawes 07 Nov '09

07 Nov '09

1.41.0 beta 1 release candidates are available at http://boost.cowic.de/rc/ Before pushing the beta out to SourceForge, I'd appreciate it if several people would try the release candidate and report success or problems. Thanks, --Beman

16 38

compile problem with thread without rvalue references
by Bob Walters 07 Nov '09

07 Nov '09

I've just attempted a build of the current (1_40_0) boost distribution on a Solaris 10 machine, using the SunStudio12 compiler (cc). There seems to be a problem with boost::thread which is related to this compiler's lack of rvalue references. My build process has been: tar xzf boost_1_40_0.tgz cd boost_1_40_0 export BOOST_ROOT=`pwd` ./bootstrap.sh ./bjam toolset=sun address-model=64 The error I'm receiving: "./boost/thread/detail/thread.hpp", line 344: Error: boost::thread::thread(boost ::thread&) is not accessible from boost::move(boost::detail::thread_move_t<boost ::thread>). 1 Error(s) detected. "CC" -library=stlport4 -xO4 -mt -erroff=%none -m64 -KPIC -DBOOST_ALL_NO_LIB= 1 -DBOOST_THREAD_BUILD_DLL=1 -DBOOST_THREAD_POSIX -DNDEBUG -I"." -c -o "bin.v2/l ibs/thread/build/sun/release/address-model-64/stdlib-sun-stlport/threading-multi /pthread/thread.o" "libs/thread/src/pthread/thread.cpp" move() seems to rely on a copy constructor for its returned thread when rvalue references aren't supported. There is no such constructor. Aside from changing compilers, is there any work around for this?

1 0

[system][filesystem v3] Question about error_code arguments
by Beman Dawes 07 Nov '09

07 Nov '09

A POSIX sub-group is looking at C++ bindings for POSIX. They are talking about a filesystem library binding at least loosely based on Boost.System, Boost.Filesystem, and C++ TR2. In looking at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2009/n2838.html, "Library Support for Hybrid Error Handling (Rev 2)", they are concerned about specifying hybrid error handling like this: void f(error_code& ec=throws()); They would rather see it specified like this: void f(error_code* ec=0); One particular concern is that as currently worded, the N2838 proposal requires throws() return a null reference, which they see as relying on undefined behavior. I wasn't at the meeting where the LWG asked that a null reference be mandated; my reading of the standard is that it is unclear if returning a null reference is or isn't undefined behavior. Some of the core folks seem to think it is undefined, but compilers accept it. Although the reference form is usual for C++, the use of anything other than the default is limited, and usually in low-level code at that. Although I could personally live with either form, I'm leaning slightly toward the pointer form. Regardless, I'd like to hear Boosters views on the two interfaces above. Which do you prefer? Why? --Beman

24 102

[wave][function] Suppressing gcc-4.4.1 warnings
by John Maddock 07 Nov '09

07 Nov '09

Looks like the new type-punning warnings from gcc-4.4.1 are false positives (but I'd like someone else to confirm that!). The fixes below illustrate how to suppress them simply by declaring a temporary variable. HTH, John. Index: boost/wave/util/flex_string.hpp =================================================================== --- boost/wave/util/flex_string.hpp (revision 57230) +++ boost/wave/util/flex_string.hpp (working copy) @@ -1211,7 +1211,10 @@ }; Storage& Data() const - { return *reinterpret_cast<Storage*>(buf_); } + { + Storage* p = reinterpret_cast<Storage*>(buf_); + return *p; + } RefCountType GetRefs() const { Index: boost/function/function_base.hpp =================================================================== --- boost/function/function_base.hpp (revision 57230) +++ boost/function/function_base.hpp (working copy) @@ -316,11 +316,13 @@ new ((void*)&out_buffer.data) functor_type(*in_functor); if (op == move_functor_tag) { - reinterpret_cast<functor_type*>(&in_buffer.data)->~Functor(); + functor_type* f = reinterpret_cast<functor_type*>(&in_buffer.data); + f->~Functor(); } } else if (op == destroy_functor_tag) { // Some compilers (Borland, vc6, ...) are unhappy with ~functor_type. - reinterpret_cast<functor_type*>(&out_buffer.data)->~Functor(); + functor_type* f = reinterpret_cast<functor_type*>(&out_buffer.data); + f->~Functor(); } else if (op == check_functor_type_tag) { const BOOST_FUNCTION_STD_NS::type_info& check_type = *out_buffer.type.type;

6 9

[Boost] Introduction of numeric_adaptor
by Bruno Lalande 07 Nov '09

07 Nov '09

Hi, I would like to introduce the numeric_adaptor library that Barend and I have developed to address a recurring problem: integration of big numbers. We already talked about it in the past, when discussions about geometry would lead to robustness. We're introducing it now because of the upcoming formal review of GGL. The purpose of this library is to turn big-numbers proposed by the most well-known big number libraries into value-types, (almost) usable like any number. This way, a library developer who has a robustness sensible function to write can take numbers as template parameters, and use them in such a way that they can be a fundamental type or one of the value-types provided by numeric_adaptor. This is basically done by: - defining in all the big-number types all the operators needed to manipulate them like fundamental types - defining some free functions in the boost namespace for other operations (sin, cos, hypot, conversions...), which have an default overload for fundamental types and a specific overload for each big-number type proposed Example: template <class T> T my_robustness_sensitive_function(T value) { value /= 2; T value2 = boost::to<T>("123.0123456789"); value += boost::sqrt(value2); return value; }; This function is numeric_adaptor-ready. If T is a fundamental type, boost::to() will forward to boost::lexical_cast and boost::sqrt() will forward to std::sqrt(). If T is one of the big number types provided by the library, specific overloads will be used. The precision of the function will thus entirely depend on the type passed. If the user chooses to sacrifice precision and use fundamental types, there won't be any run-time overhead. If the user wants something more precise and uses a big-number type, the precision of the result will be the one of the chosen type. The library is extensible. Anyone can add a new big number type with all needed operators, and overload the free functions (or specialize a dedicated traits class - see boost::to()). Note: For conversions, we would have liked to extend boost::lexical_cast directly, but the library is not designed in that sense. That's why we wrapped it into a boost::to() function which is explicitly designed to be extended, via traits. Last week on this list list a generic convert_to was discussed and that system could be used alternatively. The library is in the Boost sandbox: https://svn.boost.org/svn/boost/sandbox/numeric_adaptor/ The library currently proposes 2 big number types: CLN and GMP. This is not a formal review request, rather a preview and a reference that we will possibly use during the review of GGL. However, we would really like to have your feedback and to know if such a library would make sense. Regards Bruno

7 13

boost::noncopyable errors are difficult to diagnose on MSVC
by Zachary Turner 07 Nov '09

07 Nov '09

Is there anything that can be done to improve the error messages generated by the compiler when a copy fails to be made due to a boost::noncopyable? I had just finished a sprint where I had typed close to 1,000 lines of code with tons of new classes and objects before trying to compile, and one of the classes I had entered was similar to the following: class foo { public: foo(boost::asio::ip::tcp::socket& socket) : socket_(socket) { } private: boost::asio::ip::tcp::socket socket_; }; Problem here is that I forgot to add a reference to the class member, it was by value so the compiler is going to generate an error since a socket cannot be copied. The error this generated from the compiler was: 1>Q:\include\boost\1.39.0\boost/asio/basic_io_object.hpp(92) : error C2248: 'boost::noncopyable_::noncopyable::noncopyable' : cannot access private member declared in class 'boost::noncopyable_::noncopyable' 1> Q:\include\boost\1.39.0\boost/noncopyable.hpp(27) : see declaration of 'boost::noncopyable_::noncopyable::noncopyable' 1> Q:\include\boost\1.39.0\boost/noncopyable.hpp(22) : see declaration of 'boost::noncopyable_::noncopyable' 1> This diagnostic occurred in the compiler generated function 'boost::asio::basic_io_object<IoObjectService>::basic_io_object(const boost::asio::basic_io_object<IoObjectService> &)' 1> with 1> [ 1> IoObjectService=boost::asio::stream_socket_service<boost::asio::ip::tcp> 1> ] The most I can possibly figure out from this is that it's related to a socket. but this hardly helped as I had been using socket stuff all over the place. It took a long time to track this down because there was zero information about even what source file the offending code was in. I know this is sort of catering to the compiler, but is there any trickery that can be done so that at least the compiler generates an error related to the source file the object was declared in?

4 4

ASIO
by Mohammed Rasmy 07 Nov '09

07 Nov '09

Hello, I am trying to do a POST request using Boost::Asio, of a variable x that holds an email address, and a variable y that holds the contents of a zip file. I couldn't find samples floating around and I'm unsure of how to proceed. I've come up with the following code: boost::asio::streambuf request; std::ostream request_stream(&request); wxString buffer = wxT("POST "); buffer += wxT("http://serveraddress.com"); buffer += wxT(" HTTP/1.1\r\n"); buffer += wxT("Host: serveraddress.com\r\n"); buffer += wxT("Accept: */*\r\n"); buffer += wxT("Connection: close\r\n\r\n"); buffer += wxT("email_address=test(a)test.com"); buffer += wxT("zip_file=")/*actual zip file data here?*/; request_stream << buffer.mb_str(); Note that I am using wxWidgets, wxT is nothing more than a way of defining strings there, and wxString is, well, a string. The above code compiles fine, but I am not sure how to add the zip file? Also would the above code work? I know it is working for a GET request but that's it. Please advise. Thanks for your time, Mohammed _________________________________________________________________ Keep your friends updated—even when you’re not signed in. http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/so…

4 4

[smart_ptr] Add make/allocate functions for boost::shared_array
by Roman Kecher 06 Nov '09

06 Nov '09

Hi, I was wondering if there's any reason not to include make_shared_array() / allocate_shared_array() in boost/make_shared.hpp (or maybe add boost/make_shared_array.hpp), along with the already existing make_shared() and allocate_shared() functions? Regards, Roman.

1 0

[swap] ruminations
by Jeffrey Hellrung 06 Nov '09

06 Nov '09

1) Would it be a good idea (aside from breaking backwards compatibility) to insulate boost::swap from being found via ADL? Consider another library, tsoob, which also has a swap intended to be used in the same manner as boost::swap (i.e., it finds a free function swap via ADL, falling back to std::swap or maybe some custom swap or whatever). Now tsoob::swap and boost::swap cannot be used in the same program (ambiguous call to overloaded function swap) if the associated namespaces of the type being swapped includes both boost and tsoob (certainly possible). What I'm thinking about is the possibility of creating a namespace boost::adl which contains (what would have previously been called) boost::swap, as well as any other similar free functions. So the only way to invoke boost::adl::swap is to make it explicit; it will never be found via ADL. Plus it helps document the underlying mechanism (ADL) of boost::swap. 2) Is there some way for boost::swap and tsoob::swap to coexist peacefully while retaining backwards compatibility for both? - Jeff

2 3

[graph] Usage (and possible removal) of LEDA and Stanford GraphBase adaptors for Boost.Graph
by Jeremiah Willcock 06 Nov '09

06 Nov '09

Are there any users of the LEDA and Stanford GraphBase adaptors for Boost.Graph? They are no longer actively maintained and there have been bug reports that the LEDA adaptor does not work with current versions of LEDA. If they are no longer used, Andrew Sutton and I will consider deprecating and/or removing them. Please reply if you would like these components kept in Boost.Graph. -- Jeremiah Willcock

1 0

[random] Seeding again
by Steven Watanabe 06 Nov '09

06 Nov '09

AMDG The attached patch allows random number generators to be seeded with any value. Currently linear_congruential and a few other generators can assert for some seeds. (For instance minstd_rand cannot be seeded with 0). For linear_congruential I am using the algorithm described in n2960, section 26.5.3.1/4. For linear_feedback_shift I use a similar idea, adding 2^(w-k) if the seed is less than 2^(w-k). The behavior of any code that was valid before, should be unaffected by this change. The patch includes test and documentation updates. The tests pass with gcc, msvc, and como. I think it's ready to go now. Okay to commit? In Christ, Steven Watanabe

1 0

New boost packaging suggestion for windows
by Julian Bangert 06 Nov '09

06 Nov '09

Hello, yesterday I tried to download the boostpro binaries for windows and I had to download over a gigabyte to get all configurations for VS 2008. Sadly the version also was outdated and in general not that great ( e.g. Installer quite bad to handle and you have to register to download ) . So I decided to build from source and package it for a few friends who also needed these binaries. I managed to compress boost_1_40 ( no source except headers, no python AFAIK , but everything else you get with a standard build ) from over 900MB down to 37MB by using 7z. I uploaded it here https://obliviononline.com/pub/boost/boost_1_40_0.7z . Should boost use this form of releasing binaries in the future for windows ? Alternatively, you could endorse this form, like boostpro is endorsed right now and I could host it on google code, etc . An easy to set up ( just unzip this file ) binary distribution would make compiling boost-dependent software much easier. I would be willing to distribute future releases compatible with VS 2005 and 2008 in this form. Yours, Julian Bangert __________ Information from ESET NOD32 Antivirus, version of virus signature database 4551 (20091028) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

12 20

Boost.Python release testing failures
by Steve M. Robbins 06 Nov '09

06 Nov '09

Hi, I'm running Boost nightly regression tests on a 64-bit Debian linux platform for quite some time. I've recently added gcc 4.4 and Boost.Python is failing miserably; c.f. http://tinyurl.com/yl7yjkv If I click on a test result URL, such as for "andreas_beyer" [http://tinyurl.com/yft8k3o] I'm redirected ("See boost/bin.v2/libs/python/build/gcc-4.4/debug") which shows that the compile succeeded, but nothing more [http://tinyurl.com/yft8k3o] Digging into the bjam.log file, it turns out these tests are skipped because the Boost.Python library failed to build. It would be nice to improve the error reporting to make this clear. The compilation error is: "g++-4.4" -ftemplate-depth-128 -std=c++0x -O0 -fno-inline -w -g -fPIC -DBOOST_ALL_NO_LIB=1 -DBOOST_PYTHON_SOURCE -I".." -I"/usr/include/python2.5" -c -o "/home/steve/Packages/boost/upstream/testing/release/results/boost/bin.v2/libs/python/build/gcc-4.4/debug/object/function_doc_signature.o" "../libs/python/src/object/function_doc_signature.cpp" ../libs/python/src/object/function_doc_signature.cpp: In static member function 'static boost::python::str boost::python::objects::function_doc_signature_generator::pretty_signature(const boost::python::objects::function*, size_t, bool)': ../libs/python/src/object/function_doc_signature.cpp:238: error: no match for 'operator%' in '"%s %s(%s%s%s%s)" % std::make_tuple(_Elements&& ...) [with _Elements = boost::python::str&, const boost::python::api::object&, boost::python::str, boost::python::str, boost::python::str, std::string](((const boost::python::api::object&)(& f->boost::python::objects::function::m_name)), ((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = int, V = size_t, U = boost::python::api::object](((const int&)((const int*)(&0))), ((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads))))))))))), ((boost::python::str&&)(&((n_overloads != 0u) ? ((((size_t)arity) != n_overloads) ? boost::python::str(((const char*)" [,")) : boost::python::str(((const char*)"[ "))) : boost::python::str()))), ((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = size_t, V = unsigned int, U = boost::python::api::object](((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads)))), ((const unsigned int&)((const unsigned int*)(& arity)))))))))), ((std::string&&)(& std::basic_string<char, std::char_traits<char>, std::allocator<char> >(n_overloads, 93, ((const std::allocator<char>&)((const std::allocator<char>*)(& std::allocator<char>())))))))' ../boost/python/object_operators.hpp:104: note: candidates are: boost::python::api::object boost::python::api::operator%(const boost::python::api::object&, const boost::python::api::object&) ../libs/python/src/object/function_doc_signature.cpp:249: error: no match for 'operator%' in '"%s(%s%s%s%s) -> %s" % std::make_tuple(_Elements&& ...) [with _Elements = const boost::python::api::object&, boost::python::str, boost::python::str, boost::python::str, std::string, boost::python::str&](((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = int, V = size_t, U = boost::python::api::object](((const int&)((const int*)(&0))), ((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads))))))))))), ((boost::python::str&&)(&((n_overloads != 0u) ? ((((size_t)arity) != n_overloads) ? boost::python::str(((const char*)" [,")) : boost::python::str(((const char*)"[ "))) : boost::python::str()))), ((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = size_t, V = unsigned int, U = boost::python::api::object](((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads)))), ((const unsigned int&)((const unsigned int*)(& arity)))))))))), ((std::string&&)(& std::basic_string<char, std::char_traits<char>, std::allocator<char> >(n_overloads, 93, ((const std::allocator<char>&)((const std::allocator<char>*)(& std::allocator<char>())))))), ((boost::python::str&)(& ret_type)))' ../boost/python/object_operators.hpp:104: note: candidates are: boost::python::api::object boost::python::api::operator%(const boost::python::api::object&, const boost::python::api::object&) ../libs/python/src/object/function_doc_signature.cpp:262: error: no match for 'operator%' in '"%s %s(%s%s%s%s) %s" % std::make_tuple(_Elements&& ...) [with _Elements = boost::python::str, const boost::python::api::object&, boost::python::str, boost::python::str, boost::python::str, std::string, boost::python::str](((const boost::python::api::object&)(& f->boost::python::objects::function::m_name)), ((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = int, V = size_t, U = boost::python::api::object](((const int&)((const int*)(&0))), ((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads))))))))))), ((boost::python::str&&)(&((n_overloads != 0u) ? ((((size_t)arity) != n_overloads) ? boost::python::str(((const char*)" [,")) : boost::python::str(((const char*)"[ "))) : boost::python::str()))), ((boost::python::str&&)(& boost::python::str::join(const T&) const [with T = boost::python::api::object_slice](((const boost::python::api::object_slice&)((const boost::python::api::object_slice*)(& boost::python::api::object_operators::slice(const T&, const V&) [with T = size_t, V = unsigned int, U = boost::python::api::object](((const size_t&)((const size_t*)(&(((size_t)arity) - n_overloads)))), ((const unsigned int&)((const unsigned int*)(& arity)))))))))), ((std::string&&)(& std::basic_string<char, std::char_traits<char>, std::allocator<char> >(n_overloads, 93, ((const std::allocator<char>&)((const std::allocator<char>*)(& std::allocator<char>())))))), ((boost::python::str&&)(&(cpp_types ? boost::python::str(((const char*)"")) : boost::python::str(((const boost::python::str&)((const boost::python::str*)(& ret_type))))))))' ../boost/python/object_operators.hpp:104: note: candidates are: boost::python::api::object boost::python::api::operator%(const boost::python::api::object&, const boost::python::api::object&) ...failed gcc.compile.c++ /home/steve/Packages/boost/upstream/testing/release/results/boost/bin.v2/libs/python/build/gcc-4.4/debug/object/function_doc_signature.o... I hope this helps the Boost.Python maintainers to fix it. Let me know if you need more info. Thanks, -Steve

3 3

[rdb] 0.2.00 - bind parameter and result + dynamic sql
by Jean-Louis Leroy 05 Nov '09

05 Nov '09

I have uploaded version 0.2.01 to the vault. It contains two new major features. The first is the ability to bind parameters to variables allocated outside the prepared statement, e.g.: // in the following examples the database contains: // id first_name name age // 1 Homer Simpson 37 // 2 Marge Simpson 34 BOOST_FIXTURE_TEST_CASE(prepared_select_bind_integer_param, springfield_fixture) { person p; BOOST_AUTO(st, db.prepare(select(p.first_name).from(p).where(p.id == _))); integer id_param; st.bind_parameters(id_param); id_param = 1; BOOST_RDB_CHECK_SELECT_RESULTS(st.execute(), "((Homer))"); id_param = 2; BOOST_RDB_CHECK_SELECT_RESULTS(st.execute(), "((Marge))"); } The possibility of passing parameters to execute() remains but this is more efficient, as execute(params) binds the parameters each time it is called. Likewise, it is now possible to bind results, e.g.: BOOST_FIXTURE_TEST_CASE(prepared_select_bind_results, springfield_fixture) { person p; BOOST_AUTO(st, db.prepare(select(p.id, p.first_name).from(p))); integer id; varchar<30> first_name; st.bind_results(id, first_name); BOOST_AUTO(results, st.execute()); results.fetch(); BOOST_CHECK(!id.is_null()); BOOST_CHECK_EQUAL(id.value(), 1); BOOST_CHECK(!first_name.is_null()); BOOST_CHECK_EQUAL(string(first_name), "Homer"); results.fetch(); BOOST_CHECK(!id.is_null()); BOOST_CHECK_EQUAL(id.value(), 2); BOOST_CHECK(!first_name.is_null()); BOOST_CHECK_EQUAL(string(first_name), "Marge"); } The second feature is dynamic creation of statements. A fully static system doesn't make it possible e.g. to create a search form, where you enter (optional) values and the program returns a list of matching values. Dynamic statement parts (lists of expressions, columns, tables) are created via the make_dynamic() function. It returns a "stable" type, which does not incorporate the full type information for the encapsulated sql code. This makes it possible to create dynamic collections of sql constructs and pass them to a statement. Static and dynamic parts can be miexed freely. For example: BOOST_FIXTURE_TEST_CASE(prepared_select_bind_dynamic_integer_param, springfield_fixture) { person p; dynamic_boolean predicate = make_dynamic(p.id == _); BOOST_AUTO(st, db.prepare(select(p.first_name).from(p).where(predicate))); integer id_param; dynamic_values params; params.push_back(make_dynamic(id_param)); st.bind_parameters(params); id_param = 1; BOOST_RDB_CHECK_SELECT_RESULTS(st.execute(), "((Homer))"); id_param = 2; BOOST_RDB_CHECK_SELECT_RESULTS(st.execute(), "((Marge))"); } Here a dynamic boolean expression is used in the where clause. It contains a parameter so it is necessary to pass a collection of dynamic values when executing the statement. Here is a more complex example: BOOST_FIXTURE_TEST_CASE(prepared_select_dynamic_tables, springfield_fixture) { person p("husband"); person spouse("wife"); partner::qualified link; dynamic_expressions exprs; exprs.push_back(make_dynamic(spouse.first_name)); dynamic_tables tables; tables.push_back(make_dynamic(spouse)); tables.push_back(make_dynamic(link)); dynamic_boolean predicate = make_dynamic(p.id == link.husband && link.wife == spouse.id); BOOST_AUTO(st, db.prepare(select(p.first_name, exprs).from(p, tables).where(predicate))); varchar<30> him; dynamic_values results; varchar<30> her; results.push_back(make_dynamic(her)); st.bind_results(him, results); BOOST_AUTO(cursor, st.execute()); cursor.fetch(); BOOST_CHECK(!him.is_null()); BOOST_CHECK_EQUAL(string(him), "Homer"); BOOST_CHECK(!her.is_null()); BOOST_CHECK_EQUAL(string(her), "Marge"); } Here the projection is partly dynamic, and so is the table list and the where clause. Static entities are checked at compile-time as usual, and dynamic entities at run-time - and thus may fail with an exception. However, the system is designed to (attempt to) minimize the risk of type mismatch in the dynamic parts - no casts are required in client code (nor in the implementation ftm). The test suite contains several other examples of dynamic statements (see test_odbc_dynamic.cpp). At this point I feel that all the major features in the scope of this project have been addressed. Okay a lot is still missing - e.g. more types (currently only integer and varchar are supported), constraints, count() and such functions, group/order by... - but all these should be easy to implement in the existing framework. And the thing needs cleaning up and documentation, plus testing on Unix. J-L

1 0

[filesystem][system] Usage guideline request
by Andrey Semashev 05 Nov '09

05 Nov '09

Hi, I'm updating our software from Boost 1.33 to 1.40 and one of the major issues I encountered is using exceptions from Boost.Filesystem. The exception basic_filesystem_error has changed its interface in a breaking way, as it now relies on Boost.System error codes. I'm not aware of whether this change has been announced properly and the API deprecation period has been offered. And that is actually not what bothers me mostly. The 1.33 version of Boost.Filesystem offered a number of unified error codes that our code used to react accordingly. For example, there were error codes out_of_space_error, already_exists_error, which were used on any platform supported by Boost.Filesystem. Now, with Boost.System in place, there are at least three sets of errors: system::errc, system::windows_error and system::linux_error, although, I assume, the latter is an addition to system::errc. In order to check for the older out_of_space_error I now have to write something like this: catch (basic_filesystem_error& e) { if ( e.code() == system::errc::make_error_code( system::errc::no_space_on_device) #if defined(WIN32) || e.code() == system::windows_error::make_error_code( system::windows_error::disk_full) #endif ) { // out_of_space_error } } Is this the right way to do this? Is there a portable way to do it? I tried to find a better way in the Boost.System documentation, but was highly disappointed as there is virtually no documentation at all. Also, no examples available for this library. I wonder how this library passed the review in the first place.

2 2