January 2010 - Boost-users - lists.preview.boost.org

[Filesystem] directory iteration throws
by gast128 06 Jan '10

06 Jan '10

Hello all, I was writing a simple application to enumerate files sorted on their size. I used the recursive_directory_iterator for this task. However on my Windows XP machine the directory iteration stops because of a throwing constructor. It is easy reproducable, by creating a directory with contents and remove all security attributes (except for the system account, leave there all permissions enabled) and use that directory for iteration, i.e. my c:\temp\bla directory contained 1 file and I removed the security attributes for the directory Bla. Tracing thru the source code of Boost.FileSystem the directory is opened with a '::FindFirstFileA' call which returns an 'INVALID_HANDLE_VALUE', resulting in a throw. This stops further iteration of the directory. Is there a way to work around this?

1 0

Return value of bool boost::interprocess::shared_memory_object::remove(const char * name)
by Jarno 06 Jan '10

06 Jan '10

Hello, It seems to me that the static function bool remove(const char * name) in class boost::interprocess::shared_memory_object returns false when it should return true and vice versa. The documentation on boost interprocess says: "The remove operation might fail returning false if the shared memory does not exist, the file is open or the file is still memory mapped by other processes" However, the following simple test program for example prints "should not print" on Windows XP 64. I'm compiling with MSVC 2008 and using boost 1.41.0. /* Test program */ #include <boost/interprocess/shared_memory_object.hpp> #include <iostream> int main(int argc, char* argv[]) { using namespace boost::interprocess; using namespace std; { shared_memory_object shm(open_or_create, "foo", read_write); } if(shared_memory_object::remove("foo")) cout << "should print" << endl; if(shared_memory_object::remove("foo")) cout << "should not print" << endl; return 0; } /* Test program ends */ Am I interpreting the documentation correctly? Regards, Jarno

2 1

Thread is not interrupted
by Lloyd 06 Jan '10

06 Jan '10

Hi, I am using boost::asio to create a server. When a new connection is established (inside the asio read handle) a new thread (boost::thread) is created. The thread object pointer is kept in a static vector. When the next connection is established, the thread pointer available in the static vector is read and interrupted. But the thread is not interrupted! The thread is just an infinite loop, in which interruption is enabled. I have wrttien the exception handler also. Not able to understand why this is not working in this situation. Otherwise this works. To make this clear given below are the code snippets... //ASIO based read handler void TCPConnection::HandleRead(const boost::system::error_code& Error) { TCPConnection::Threadvector[0]->interrupt(); //Threadvector is static variable in the class, for easy reading written like this boost::thread *th=new boost::thread(boost::bind(&TCPConnection::ExecutionRoutine,shared_from_this(),Command)); TCPConnection::Threadvector.push_back(th); } //ExecutionRoutine looks like this int TCPConnection::ExecutionRoutine(string Command) { try { while(1) { boost::this_thread::interruption_point(); } } catch(boost::thread_interrupted const&) { return 0; } return 0; } Please share some idea... Thanks, Lloyd ___________________ Happy New Year 2010

1 0

[GIL] Packing to 6 bytes
by Eloi Du Bois 06 Jan '10

06 Jan '10

Hi, I have a question: I need to create a view that hold 12 bit channel data packed as follow: 0000R0000G0000B0000;0000R0000G0000B0000;0000R0000G0000B0000;.... Meaning that each channel is shifted by 4 bits. The pixel is holden on 48 bits = 6 bytes. I have created my view like as follow: /// My 12 bits special rgb packed to 6 bytes typedef const packed_channel_reference<uint64_t, 04, 12, true> rgb12_packed_channel0_t; typedef const packed_channel_reference<uint64_t, 20, 12, true> rgb12_packed_channel1_t; typedef const packed_channel_reference<uint64_t, 36, 12, true> rgb12_packed_channel2_t; typedef mpl::vector3<rgb12_packed_channel0_t, rgb12_packed_channel1_t, rgb12_packed_channel2_t> rgb12_packed_channels_t; typedef packed_pixel<packed_channel_value<48>, rgb12_packed_channels_t, rgb_layout_t> rgb12_packed_pixel_t; typedef view_type_from_pixel<rgb12_packed_pixel_t>::type rgb12_packed_view_t; The thing is that I would now like to copy data to another rgb16_view_t... But this is not working because, in my opinion, 48 bits is between 32 and 64. I have been thinking of the solution that Christian gave me and that consist on doing something like this: T p( pData, 0 ); for( typename DST_V::y_coord_t y = 0; y < height; ++y ) { typename DST_V::x_iterator it = dst.row_begin( y ); for( typename DST_V::x_coord_t x = 0; x < width; ++x ) { color_convert( *p, *it ); ++p; ++it; } } where T is a bit_aligned_pixel_iterator and DST_V=rgb16_view_t but this can't be working on my case because channels are shifted by 4 bits ! If you could help me, I would really appreciate. I think all those cases are a good robustness test for gil. Regards, Eloi.

3 3

Trying to build docs....
by Marshall Clow 05 Jan '10

05 Jan '10

I'm trying to generate the docs for boost. I'm looking at the instructions at <http://www.boost.org/doc/libs/1_41_0/doc/html/boostbook/getting/started.html> I've installed: docbook-xml-4.2 docbook-xsl-175.2 fop-0.95 Jimi 1.0 (and done the 'move and rename' bit into fop) I believe that I've set up my user-config.jam correctly (attached). And when I try to build something, I get errors about "unable to external entity <a sourceforge URL>". If I take the <a sourceforge URL> and try to curl it, I get something that looks reasonable. Any suggestions? [ Oh, yeah: Mac OS X 10.6, gcc 4.2.1, from the trunk. ] -- Marshall > $ bjam --v2 onehtml > ...patience... > ...found 1251 targets... > ...updating 1 target... > xslt-xsltproc ../../../bin.v2/libs/fusion/doc/darwin-4.2.1/debug/fusion.html > warning: failed to load external entity "http://docbook.sourceforge.net/release/xsl/current/html/docbook.xsl" > compilation error: file /Users/marshall/Sources/boost/tools/boostbook/xsl/html-single.xsl line 15 element import > xsl:import : unable to load http://docbook.sourceforge.net/release/xsl/current/html/docbook.xsl > warning: failed to load external entity "http://docbook.sourceforge.net/release/xsl/current/html/admon.xsl" > compilation error: file /Users/marshall/Sources/boost/tools/boostbook/xsl/admon.xsl line 12 element import > xsl:import : unable to load http://docbook.sourceforge.net/release/xsl/current/html/admon.xsl > warning: failed to load external entity "http://docbook.sourceforge.net/release/xsl/current/lib/lib.xsl" > compilation error: file /Users/marshall/Sources/boost/tools/boostbook/xsl/relative-href.xsl line 12 element import > xsl:import : unable to load http://docbook.sourceforge.net/release/xsl/current/lib/lib.xsl > > XML_CATALOG_FILES="../../../bin.v2/boostbook_catalog.xml" > export XML_CATALOG_FILES > "xsltproc" --stringparam boost.libraries "../../../libraries.htm" --stringparam boost.root "../../../.." --stringparam chunk.first.sections "1" --stringparam chunk.section.depth "4" --stringparam generate.section.toc.level "4" --stringparam html.stylesheet "../../../../doc/html/boostbook.css" --stringparam toc.max.depth "3" --stringparam toc.section.depth "3" --xinclude -o "../../../bin.v2/libs/fusion/doc/darwin-4.2.1/debug/fusion.html" "/Users/marshall/Sources/boost/tools/boostbook/xsl/html-single.xsl" "../../../bin.v2/libs/fusion/doc/darwin-4.2.1/debug/fusion.docbook" > > ...failed xslt-xsltproc ../../../bin.v2/libs/fusion/doc/darwin-4.2.1/debug/fusion.html... > ...failed updating 1 target... > > mclow-mac:doc marshall$ curl http://docbook.sourceforge.net/release/xsl/current/lib/lib.xsl | bbedit > % Total % Received % Xferd Average Speed Time Time Time Current > Dload Upload Total Spent Left Speed > 100 18500 100 18500 0 0 76291 0 --:--:-- --:--:-- --:--:-- 141k >

2 1

Re: [Boost-users] [interprocess] Not implemented?
by Pascal Kesseli 05 Jan '10

05 Jan '10

Thanks for your reply. The complete system version goes as follows: Linux Moxa 2.6.23.1 #1266 Fri Nov 14 19:05:06 EST 2008 armv5teb unknown UC-8410 version 1.0 It is an embedded MOXA computer. Does this mean that shared memory objects are not fully implemented on this operating system? On some of this device's predecessors, I used standard Unix semaphores for this task, before I switched to the new boost version when the interprocess library was introduced in version 1.39.0. Would returning to this manual solution work or do semaphores not function at all on this machine? Thanks again and best regards Pascal

2 1

mpl: generate kind of permutations
by Hicham Mouline 05 Jan '10

05 Jan '10

Hello, I have 2 strictly ordered sequences of dates t1... tn, and td1...tdm, where m and n may or may not be different. We always have: t1<...<tn and td1<...<tdm. I am trying to place the (t1...tn) sequence over the (td1...tdm) sequence, and generate all these "permutations" ie . all ti < td1 or. all ti < tn = td1 or. all ti except tn < td1 and (tn >td1 or tn=td2 or td3>tn>td2 or tn=td3 or ....tn>tdm-1 or tn=tdm or tn>tdm) or. t1...tn-2 < td1 and tn-1=td1 and (td1<tn<td2 or tn=td2 or tn>td2 or tn=td3 or ....tn>tdm-1 or tn=tdm or tn>tdm) or. t1...tn-2 < td1 and td1<tn-1<td2 and (the same) or. t1...tn-2 < td1 and td2=tn-1 and (the same) ...etc...etc or. t1 < td1 and distribute the rest of t_i_s or. t1 = td1 and distribute the rest of t_i_s or. td2> t1 > td1 and distribute the rest of t_i_s etc or. tdm< t1< t2< ...<tn I hope the description is clear enough even though it is not formal at all. The problem is like distributing n balls over m-1 buckets or to the left or to the right of all buckets or exactly at bucket borders (with never more than 1 t_i on any bucket order) In the runtime world, the pseudo-code would look like: calling i0: number of t_is in the 0th bucket (the one to the left) calling i1: number of t_is in the 1th bucket (between td1 and td2) ... calling im-1: number of t_is in the (m-1)th bucket (between tdm-1 and tdm) calling im: number of t_is in the mth bucket (>tdm) j1 : number of t_is = td1 (0 or 1) ... jm: number of t_is = tdm (0 or 1) for (i0 = n; i0>=0; --i0) for (i1 = n-i0; i1>=0; --i1) for (i2 = n-i0-i1; i2>=0; --i2) ... for (im = n-i0-i1-im_1; im>=0; --im) ... 0. The pseudo code is incorrect here. I need to account for the j_s as well. 1. I would normally think of using recursion to generate the serie of inner-loops. 2. I believe there must be a purely iterative solution. 3. This problem seems familiar in that perhaps one of the STL algorithm does something similar 4. n and m are never bigger than 4 or 5. Though already, that yields a large number of permutations. I haven't figured out yet the exact number. So I think compilation will be slow but not unbearably (it's a guess). I am interested in doing this in compile-time though. I encode dates as an ints. I would start with the td dates the mpl::vector_c< 20100109, 20100125, 20100322, 201004228 > which gives m=4. Then I would set n to 3 for e.g. Then I would ask the mpl metafunction to return all the possible combinations. That would be a vector or vectors? vector_c< permut1, permumt2, ... permut_last > For each vector of the permut_s, I would have a runtime function to run (chooses dates and distribute them as provided by the vector). Regards,

3 3

[interprocess] Not implemented?
by Pascal Kesseli 05 Jan '10

05 Jan '10

Hi folks Using boost 1.39.0, I receive the following error message when running my program on the target system: terminate called after throwing an instance of 'boost::interprocess::interprocess_exception' what(): Function not implemented Aborted My program uses boost named mutexes. How do I have to interpret this exception? The target is a MOXA embedded UC-8410 computer, featuring a GCC 4.2 platform. Thanky for any suggestions and best regards Pascal _________________________________________________________________ Samichlaus du liebe Maa, hesch dis Hotmail hüt scho gha? Gratis Geschenk runterladen! http://www.microsoft.com/switzerland/windows/de/windowslive/hotmail_bl1/hot…

2 1

Re: [Boost-users] [unit_test_framework] checkpoints for BOOST_CHECK*?
by rozelak＠volny.cz 04 Jan '10

04 Jan '10

> > rozelak(a)volny.cz wrote: > >> rozelak(a)volny.cz wrote: > >> > >>> Is there a way how to achieve what I am looking > >>> for? > >>> >>> > >> Use BOOST_TEST_MESSAGE and set log level to message > >> >> > > > > Yes, I have read about this option and it works - > > thank you. I consider > > > it as not perfect, but working solution. > > > > However, I want to ask, if the idea of checkpoints > > could be generalised > > > and used during check failure printing as well (as > > has been suggested > > > in the example in my first post). > > I don't know boost source in depth, but it seems > > that BOOST_TEST_CHECKPOINT > > > just calls boost::unit_test::unit_test_log.set_checkpoint(...), > > > > > > > > which > > > stores the checkpoint info in unit_test_log (which > > is supposed to be > > > singleton). > > Would, therefore, be possible to print the checkpoint > > stored also when > > > test failure is printed? May I be helpfull somehow > > to do this? > > > > > Or is it against the idea of the checkpoints? > > > > I think the issue is that BOOST_CHECK and friends set > a checkpoint. > I see, you are right. It is set by BOOST_TEST_PASSPOINT() macro. Well, do you think, that it would be beneficial to define new macro, e.g. BOOST_TEST_FAILRECORD (please suggest a better name, if you have one)? The macro would be implemented in the way very similar to BOOST_TEST_CHECKPOINT, but the data "recorded" by the macro would be printed only when a testcase fails or an unexpected exception occurs. It also has the advantage, that when not used, it does not affect the current behaviour of boost. Is there a change to accept it in the boost test framework, if I try to implement it? Thank you for suggestions/comments, Dan

3 7

Using atan2 in boost::units
by Brandon Olivares 04 Jan '10

04 Jan '10

Hi, I need to use the boost::units::atan2() function in boost/units/cmath.hpp, but am having trouble figuring out how. Do the units have to be dimensionless, or can they be any unit? I have two variables, x_ and y_ which are of a scaled unit type, called light_second_unit. I'm calling it like this: boost::units::atan2(y_, x_); I get this error: ./physics/cartesian_coordinate_system/point.cpp:85: error: no matching function for call to `atan2(boost::units::quantity<boost::units::astronomical::light_second_unit, double>, boost::units::quantity<boost::units::astronomical::light_second_unit, double>)' If it has to be dimensionless, how do I convert them to a dimensionless quantity? Thanks, Brandon -- www.creatorspiritus.com Blog and forum discussing living a Christ-centered life Follow me on Twitter: http://twitter.com/devbanana Follow creatorspiritus.com on Twitter: http://twitter.com/creatorspiritus

3 6