- Boost-users - lists.boost.org

Q: exceptions and -fPIC
by Andrey Nechypurenko 19 Sep '03

19 Sep '03

Hi, I am on the following platform: Linux 2.6.0-test5 #11 Wed Sep 10 20:28:31 CEST 2003 i686 unknown Reading specs from /usr/lib/gcc-lib/i386-slackware-linux/3.2.2/specs Configured with: ../gcc-3.2.2/configure --prefix=/usr --enable-shared --enable-threads=posix --enable-__cxa_atexit --disable-checking --with-gnu-ld --verbose --target=i386-slackware-linux --host=i386-slackware-linux Thread model: posix gcc version 3.2.2 Boost version 1.30.2 compiled with: bjam "-sTOOLS=gcc" I have a question regarding the following boost::filesystem use case: try { for(fs::directory_iterator dir_itr(full_path); dir_itr != end_iter; ++dir_itr ) { // use dir_itr } } catch(const std::exception &ex) { std::cout << ex.what() << std::endl; } In case of full_path pointed to the directory without read permissions the corresponding exception is thrown and catched in the catch block. My problem is that if I compile my application with -fPIC option, I am unable to catch the exception anymore and the program is simply aborted. I need to use -fPIC because of other library which has own build system and if I follow suggested guidlines for Makefile, I get -fPIC automatically. I would verry appreciate if somebody could tell me are there any ways to make the code mentioned above work when it is compiled with -fPIC? Should I recompile boost with -fPIC? If yes, what is the right place to set this parameter? BTW how it is supposed to write a shared library which includes boost code? As I understand, than my code must be compiled with -fPIC... Thanks, Andrey. P.S. Here is example I am using and the makefile: // ttt.cpp #include <iostream> #include <boost/filesystem/operations.hpp> #include <boost/filesystem/path.hpp> namespace fs = boost::filesystem; int main(int argc, char **argv) { try { fs::path full_path = fs::system_complete(fs::path(argv[1], fs::native)); fs::directory_iterator end_iter; for(fs::directory_iterator dir_itr(full_path); dir_itr != end_iter; ++dir_itr ) { std::cout << dir_itr->leaf() << std::endl; } } catch(const std::exception &ex) { std::cout << ex.what() << std::endl; } return 0; } Makefile: CPPFLAGS := -I/usr/local/src/boost LDFLAGS := -L/usr/local/src/boost/libs/filesystem/build/bin/libboost_filesystem.a/gcc/debug/runtime-link-dynamic/ LIBS := $(LIBS) -lboost_filesystem ttt: ttt.cpp c++ $(CPPFLAGS) -g -ftemplate-depth-128 -fPIC -Wall $(LDFLAGS) -o $@ ttt.cpp $(LIBS) __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com

1 0

Spirit library - Problem with the copy constructor of the class rule<> ?
by fritz.birnbacher＠t-online.de 18 Sep '03

18 Sep '03

Hello, I am new in using boost and I played a little bit with the spirit library. I tried following code using MSVC 7.1 and boost 1.30.2 #include"stdafx.h" #include<spirit.hpp> using namespace boost::spirit; rule<> GetRule() { rule<> r = anychar_p; return r; } void Parse1() { char s[] = "ABCD"; parse(s, GetRule()); } void Parse2() { rule<> r = anychar_p; char s[] = "ABCD"; parse(s,r); } int _tmain() { Parse1(); return 0; } Parse1() throws an exception (access violation) Parse2() works correct. Is it a compiler problem or a problem in spirit or my problem ? Thanks

2 1

may it change ?
by maxch 17 Sep '03

17 Sep '03

Hi All, I am a new user of Boost. but when I look the source of Boost, I found that almost all the implementations is put in head files together, not as usual to put into CPP file. I know, it must be intended, however, could it possible to move the implementations into cpp file in future? Sorry if this question has been posted. Thanks!

2 2

Boost.Python - implementing slicing on extension container class
by Steve Horne 16 Sep '03

16 Sep '03

I'm currently writing a set of container classes using Boost.Python to wrap an existing C++ library. Progress has been very good - not even two days and I've wrapped all the existing C++ methods (at least 100). But of course a library designed for C++ is not a library designed for Python. Job 2 is to make it more Pythonic. The current problem is implementing slicing in the __getitem__, __setitem__ and __delitem__ methods. These are already working with single keys. To handle slicing, I need to identify in the key object is a slice object. This is the sticking point right now. I can access the type of the object using p.attr ("__class__") - but how do I get access to the slice type object from __builtins__ in order to compare? I could probably use a workaround such as extracting and checking a string representation from p.attr ("__class__").str (), but it seems to me that the ability to import symbols from another module (whether __builtins__ or some other Python or extension module). This could reduce to the ability to convert a PyObject* to an object instance - even more useful for using any C API functions that don't have a Boost.Python wrapper - but there seems to be no obvious way to do this. In addition, my containers are currently declared as follows... class_<classname, boost::noncopyable>("classname") ... ; The boost::noncopyable is there because the containers don't have copy constructors due to certain issues using the library in C++. Therefore, I'd like to better understand the implications of this. For instance, in Python the copy constructor is often used to make copy the entire container (as opposed to just rebinding names). Presumably Boost.Python supports this using the C++ classes copy constructor. This raises the questions... 1. How can I create an object to return to Python, containing an instance of the C++ container, without removing boost::noncopyable? (I need to do this to implement slicing too) This, I think, is roughly what is done in the Inheritance section of the tutorial so I probably need something like the 'factory' function. So, to reduce it to a simple copy-function example, the following should be about right... Base* container::make_copy () { container *temp = new container (); temp->do_copy (*this); return temp; } ... .def("make_copy", container::make_copy, return_value_policy<manage_new_object>()) I think this should be right, but confidence is low. 2. How can I support the Python copy constructor without having a C++ copy constructor. I'm using boost-1.30.2 with Python 2.2.3 and Visual Studio 7. Any advice would be appreciated. Thanks.

2 4

Web page errors
by Reed Hedges 16 Sep '03

16 Sep '03

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Looks like I just found the *real* list :) The web page (http://boost.org/more/mailing_lists.htm#users) still links to the Yahoo page, and seems to have some other problems (some repeated text below the horizontal line). Would it be possible to fix that soon so as not to confuse others like me too much? thanks reed - -- Reed Hedges reed(a)zerohour.net http://zerohour.net/~reed -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.1 (Darwin) iD8DBQE/ZHFjFK83gN8ItOQRApIQAJ9Mav7GFLBokSwHBb5s+3jzGF/UuwCeMkq7 FN80pQo0c7osCwWm0mNFEsk= =sLUi -----END PGP SIGNATURE-----

2 1

building Jam and Boost on Solaris
by banibrata dutta 16 Sep '03

16 Sep '03

hi, is it possible to build Jam and Boost on Solaris. going thru the Boost.Jam documentation, for building Jam from scratch, i figured that the latest Jam tarball, doesn't seem to have a Makefile. in that case how do i build Jam, and how do i build Boost? any help would be appreciated. thanks, bdutta. _________________________________________________________________ Interact with peers. Learn from experts. http://server1.msn.co.in/sp03/teched/index.asp Be tech-empowered!

3 2

Bug in boost::format with 0 valued integer
by Simon Clift 15 Sep '03

15 Sep '03

Hi folks, There appears to be a bug in boost::format, an unitialized value problem spotted when I was using valgrind on a program of mine. The following test code illustrates the (admittedly minor?) problem: #include <boost/format.hpp> #include <iostream> using namespace std; int main() { int dummy = 5; cout << boost::format( "%05d" ) % dummy << endl; dummy = 0; cout << boost::format( "%05d" ) % dummy << endl; }; I compiled this with -g and when run with valgrind it seems that the put_last function in feed_args.hpp sends an uninitialized value into the standard library <<(int) operator when the integer to be printed is 0 and leading zeros are specified. Valgrind reports (clipped for size) on the second ::format call (but not the first) of the test code: ==28879== Conditional jump or move depends on uninitialised value(s) ==28879== at 0x40289470: (within /usr/lib/libstdc++.so.5.0.3) ... ==28879== by 0x4028F36B: std::ostream::operator<<(int) (in /usr/lib/libstdc++.so.5.0.3) ==28879== by 0x804CF8D: void boost::io::detail::(anonymous namespace)::put_last<char, std::char_traits<char>, int>(std::basic_ostream<char, std::char_traits<char> >&, int&) (feed_args.hpp:114) ==28879== by 0x804C9A2: _ZN5boost2io6detail31_GLOBAL__N_test00Fill.cppfx7Zjb3putIcSt11char_traitsIcERiEEvT1_RKNS1_11format_itemIT_T0_EERSbIS9_SA_SaIS9_EERSt19basic_ostringstreamIS9_SA_SE_E (feed_args.hpp:176) ==28879== by 0x804BF44: void boost::io::detail::distribute<char, std::char_traits<char>, int&>(boost::basic_format<char, std::char_traits<char> >&, int&) (feed_args.hpp:237) ==28879== by 0x804ABB6: _ZN5boost2io6detail4feedIcSt11char_traitsIcERiEERNS_12basic_formatIT_T0_EESA_T1_ (feed_args.hpp:246) ==28879== by 0x804A14B: boost::basic_format<char, std::char_traits<char> >& boost::basic_format<char, std::char_traits<char> >::operator%<int>(int&) (format_class.hpp:69) ... Hope this helps. Otherwise I'm enjoying using boost! Best regards --- Simon -- Simon Clift E-mail: ssclift at math dot uwaterloo dot ca Ph.D. Student Scientific Computation Group University of Waterloo Waterloo, Ontario, Canada N2L 3G1

2 2

re: copy_if
by Witz 15 Sep '03

15 Sep '03

> > template<class II, class OI, class PRED> > > std::pair<II,OI> copy_if(II ibegin, II iend, OI obegin, OI oend, > > PRED p) > > { > > for(; ibegin != iend; ++ibegin) > > { > > if(p(*ibegin)) > > { > > *obegin = *ibegin; > > if(++obegin == oend) > > break; > > } > > } > > return std::make_pair(ibegin, obegin); > > } > > > > This should check whether obegin == end before copying, in case obegin > == oend initially. I don't know whether it should stop as soon as it > finds obegin == oend or wait until it knows there's another item to be > copied. The former behaviour saves a little time in the case that no > more items are wanted once the output buffer is full. The latter > behaviour avoids the occasional need for another pass that copies > nothing, if copy_if is called repeatedly to fill an output buffer which > is then flushed. > Ben. I've tended to rely on the common sense of the user (after all i cannot detect other possible errors such as 'oend < obegin') but i suppose the following addition at the start would seem sensible: ASSERT(obegin != oend); As for your second point - the amount of time saved by 'the former behaviour' depends on how expensive the predicate is and how many times it will be called if no more satisfactory items are available. As the entries in the output buffer are overwritten as and when required there is no need in any general sense to clear it between calls and this surely obviates any advantage your proposed approach may have. Witz

1 0

Re: Boost.Python - implementing slicing on extension container class
by Steve Horne 15 Sep '03

15 Sep '03

I have solved the issues I was having from my previous email. The cause of my problem was mainly that the information I needed was in the 'Embedding' section of the Boost.Python tutorial - I am not doing embedding so it did not occur to me to look there.

1 0

How do I browse the current 'boost-users' list?
by Andrew Bache 15 Sep '03

15 Sep '03

Is it possible to browse items on this list? I posted this via email having got to this page ... http://lists.boost.org/mailman/listinfo.cgi/boost-users What is not immediatly obvious (to me) is how I browse the current status of the list. Is there a URL that displays headers of the most recent posts? Can someone please point me at it? (There is a URL on the above page for an archive). Email reply please - for obvious reasons. Thanks Andy

2 1