September 2004 - Boost-users

Using Boost Tuple Library instead of variadic macros
by Barak Simon 23 Sep '04

23 Sep '04

Hi, I am trying to use the Tuple library in order to provide a logging interface which can get a variable number of parameters and handle them in a type-safe manner. However, though constructing a tuple by make_tuple seems easy enough (from a logging's user point of view) implementation wise, I am not sure how to actually iterate on the make_tuple_mapper which is being provided to my logging method, as generically getting the length and using invoking "get" on it does not work due to compile-time reasons (and there are no iterators available). I am sure it must be possible to do it somehow, or else the ostream serialisation capabilities would not have been possible to implement, but I guess I am just not sophisticated enough to fully grasp the MP intricacies... My logging method currently looks like that: template<class TheTuple> inline void logTuple(TheTuple theTuple) { } which might not be the correct way about it, but I was not sure how to explicitly provide a tuple type that's created by make_tuple, let alone how to iterate on the values. Any help with this would be greatly appreciated. Thanks, Barak _________________ Barak Simon GED IT Core Platform Deutsche Bank 190 George St. Sydney, 2000 Phone +61 2 925 85070 Fax +61 2 925 95050 -- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

2 1

Does BGL support hypergraph?
by Peng Yu 23 Sep '04

23 Sep '04

Hi, I notices that there is some discussion about hypergraph in the mailing list archives 2 years ago. Is hypergraph implementation in BGL now? Best wishes, Peng

2 1

RE: [Boost-users] Re: Bad interaction between boost::bindandBoost.Python?
by Nat Goodspeed 23 Sep '04

23 Sep '04

> -----Original Message----- > From: David Abrahams [mailto:dave@boost-consulting.com] > Sent: Thursday, September 23, 2004 8:24 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: Bad interaction between > boost::bindandBoost.Python? > > In principle your boost/python header has to be #included first [before boost/bind] > anyway, at least if you intend your code to run on POSIX. [Nat] ? I don't understand why that would be? Shouldn't each header #include everything to make it self-sufficient, rather than depending on the includer's prior #includes? > All I can > suggest is that you try the latest Boost CVS state, which doesn't have > this problem. A new release of Boost is imminent anyway. [Nat] That's fair, and thanks for testing it. But given the amount of time I've spent so far on this 1.31.0 upgrade, I'm going to lay low for a few days before announcing that I intend to upgrade to yet another version of Boost... ;-)

2 1

Re: A boost.graph edge problem
by Lars Kunert 23 Sep '04

23 Sep '04

Have a look at your compare operator! It orders two pairs of values (A,B)(C,D) like this: (A<B) || (!(A<B) && (C<D)) this expession equals: (A<B) || (C<D) lets consider the test case: (1,2) (2,1) Your operator will return return true in whatever order this pair is compared... I think an operator like (A<B) || ((A==B) && (C<D)) will help > ... > typedef multiset<DataEdge,ltSubgraph> Edges; > ... > struct ltSubgraph : public > std::binary_function<DataEdge,DataEdge,bool> > { > bool operator() (const DataEdge& s1, const DataEdge& s2) const > { > return s1.m_source < s2.m_source || > (!(s1.m_source < s2.m_source) && s1.m_target < s2.m_target); > } > };

1 0

A boost.graph edge problem
by Jeffrey Holle 22 Sep '04

22 Sep '04

I'm using boost 1.31.0 and gcc 3.4.2. While this problem is more of an application code problem, since I got no help on other lists, I trying this one with hopes of a response. I need to identify parallel edges from my graph, which is defined as: typedef subgraph<adjacency_list<vecS, vecS, bidirectionalS, property<vertex_index_t,int,VertexProperty>, EdgeProperty> > DataGraphT; typedef boost::graph_traits <DataGraphT>::edge_descriptor DataEdge; In the test case that I'm working on, I know I have parallel edges, actually 4 sets of 3. I'd like to use this code: DataGraph::handleParallelEdges(void) { typedef multiset<DataEdge,ltSubgraph> Edges; Edges edges; graph_traits<DataGraphT>::edge_iterator ei,ei_end; for (tie(ei,ei_end)=boost::edges(dataGraph_);ei!=ei_end;++ei) edges.insert(*ei); for (Edges::const_iterator iter = edges.begin(); iter != edges.end(); advance(iter,edges.count(*iter))) if (edges.count(*iter)>1) cout << "See duplicate" << endl; } However, when count never returns. The likely cause of this (I suspect) is the ltSubgraph functor. Its definition is: struct ltSubgraph : public std::binary_function<DataEdge,DataEdge,bool> { bool operator() (const DataEdge& s1, const DataEdge& s2) const { return s1.m_source < s2.m_source || (!(s1.m_source < s2.m_source) && s1.m_target < s2.m_target); } }; I say this because I can see that this functor is being continuously called in this infinite loop. Note that I use this functor in multiple other cases, but not with an associate container that allows duplicate keys. Is there a problem with this code?

1 1

boost graph meant to contain classes and functions?
by Stephen torri 22 Sep '04

22 Sep '04

I see that the boost graph library is used to contain data but can it be used to contain a class which a visitor moves along the graph per a preset algorithm calling interfacing functions as it moves along? Here is an example. class Color_Base { virtual ~Color_Base(){} virtual std::string get_Name () = 0; }; class Red : Color_Base { virtual ~Red (){} virtual std::string get_Name () { return std::string("Red"); } }; class Yellow : Color_Base { virtual ~Yellow (){} virtual std::string get_Name () { return std::string("Yellow"); } }; class Blue : Color_Base { virtual ~Blue (){} virtual std::string get_Name () { return std::string("Blue"); } }; So we could have a graph like | Blue -- Red | / | / | / Yellow I have a visitor called Color_Visitor which has the following pseudo code: for each vertice in graph visit vertice call get_Name() print result Stephen -- Email: storri(a)torri.org

3 4

No defines using bjam on Windows with msvc
by Phillip Seaver 21 Sep '04

21 Sep '04

As I said in an earlier post, I'm trying to start using jam for our projects. I'm using Boost.Jam ver 3.1.4. When I build on Windows with TOOLS=msvc and use "bjam -d2", the command-line doesn't have any define flags on it. Several of our source files rely on "WIN32", e.g. to determine the platform. Has anyone seen this before? I feel like I'm overlooking something obvious, but I haven't been able to figure it out. Here's the beginning of the output from "bjam -d2": ***** ...found 156 targets... ...updating 11 targets... vc-Cc ..\..\libs\ETSopt\bin\libetsopt.lib\msvc\debug\runtime-link-dynamic\etsopt .obj "cl" /Zm800 -nologo -c /Z7 /Od /Ob0 /GX /GR /MDd -I"..\..\libs\ETSopt" -I"C:\PROGRA~1\MICROS~2\VC98\include" -Fo"..\..\libs\ETSopt\bin\libetsopt.lib\ msvc\debug\runtime-link-dynamic\etsopt.obj" "etsopt.c" etsopt.c ***** Thanks in advance, Phillip

2 1

Re: [Boost-users] [Linker error] undefined reference to`boost::gregorian::greg_month::as_short_string() const'
by aqsis＠mail.tele2.co.uk 21 Sep '04

21 Sep '04

I understand what you mean now. In that case you would need to pass the filter name into the DSO function as a parameter. So your DSO declaration would look like... SHADEOP_TABLE(shadeopname) = { { "float shadeopname_fs(float, string)", "", "" }, . . }; And the argv[] argument passed to the method would contain the filter name as it's second value, i.e. SHADEOP(shadeopname_fs) { float *result = (float *)argv[0]; float x = *(float *)argv[1]; STRING_DESC *filt = (STRING_DESC *)argv[1]; if(strcmp(filt->s, "box")==0) { /* Filter with a box filter */ . . . Hope this helps. Paul G Original Message: ----------------- From: Jeff Garland jeff(a)crystalclearsoftware.com Date: Tue, 21 Sep 2004 05:15:39 -0700 To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] [Linker error] undefined reference to`boost::gregorian::greg_month::as_short_string() const' On Mon, 20 Sep 2004 20:33:51 -0700 (PDT), Zhoujie Mao wrote I'm going to presume the built library file libboost_date_time-gcc-1_31.a is in D:/boost/lib. Please double check that. > Below is part of my make file > CPP = g++.exe > CC = gcc.exe > WINDRES = windres.exe > RES = > OBJ = main.o $(RES) > LINKOBJ = main.o $(RES) > LIBS = -L"D:/Dev-Cpp499/lib" -L"D:/boost/lib" > -L"D:/boost/lib/libboost_date_time-gcc-1_31.a" ^^^^^^^^^^^^ this looks wrong ^^^^^^^^^^^ change it to the following: -lboost_date_time-gcc-1_31 so your LIBS line looks like: LIBS = -L"D:/Dev-Cpp499/lib" -L"D:/boost/lib" -lboost_date_time-gcc-1_31 > I did include the libboost_date_time-gcc-1_31 but it > did not work. Well, not actually because the -L controls the search path and you can't point it directly to the library.... HTH, Jeff _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users -------------------------------------------------------------------- mail2web - Check your email from the web at http://mail2web.com/ .

1 0

setting thread priority?
by Simmons, Aaron 21 Sep '04

21 Sep '04

In Boost.Threads, are there plans to allow adjustment of a thread's priority? Thanks, aaron

2 1

sockets/pipes
by Simmons, Aaron 20 Sep '04

20 Sep '04

Does Boost have named-pipes and/or sockets classes? If no... Are any in the works? Does anyone know of a good C++ pipes and/or sockets wrapper classes? Thanks, Aaron Simmons

2 1

RE: [Boost-users] sockets/pipes
by Richard Reitmeyer 20 Sep '04

20 Sep '04

I'd suggest ACE. Look for Douglas Schmidt's stuff at wustl. > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Simmons, Aaron > Sent: Monday, September 20, 2004 2:12 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] sockets/pipes > > > Does Boost have named-pipes and/or sockets classes? > > If no... Are any in the works? Does anyone know of a good C++ pipes > and/or sockets wrapper classes? > > > > Thanks, > Aaron Simmons > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

RE: [Boost-users] Can boost::tokenizer tokenize 2byte character s tring?
by Rozental, Gennadiy 20 Sep '04

20 Sep '04

If it 2byte why are you using boost::char_separator<char>? Shouldn't you be using boost::char_separator<wchar_t>? Gennadiy. > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Lee, Joo-Young > Sent: Saturday, September 18, 2004 12:39 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Can boost::tokenizer tokenize 2byte > character string? > > > Hi. > > I try to use 'boost::tokenizer<boost::char_separator<char> >' > to separate 2byte character string like Korean, Japanese, or Chinese. > > But, I found that it does not works correctly. > > Is there a solution? > > Thanks for the help, > > Lee Joo-Young > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

2 1

RE: [Boost-users] Can boost::tokenizer tokenize 2byte character string?
by Ben Hutchings 20 Sep '04

20 Sep '04

Gennadiy Rozental <gennadiy.rozental(a)thomson.com> wrote: > If it 2byte why are you using boost::char_separator<char>? > > Shouldn't you be using boost::char_separator<wchar_t>? I think Lee Joo-Young is talking about multibyte strings in an encoding such as GB2312 and Shift_JIS that uses 1 or 2 bytes per character rather than wide strings that use a fixed 2 bytes per character on certain platforms (or rather, used to until UTF-16 came along). Ben.

1 0

RE: [Boost-users] Can boost::tokenizer tokenize 2byte character string?
by Ben Hutchings 20 Sep '04

20 Sep '04

Joo-Young Lee <trowind(a)gmail.com> wrote: > I try to use 'boost::tokenizer<boost::char_separator<char> >' > to separate 2byte character string like Korean, Japanese, or > Chinese. > > But, I found that it does not works correctly. > > Is there a solution? You should probably convert them to wide strings and use the wchar_t version instead. The standard C++ library and Boost libraries tend not to work well with multibyte encodings.

1 0

boost::serialization problem, derived class through shared_ptr
by Jedediah Smith 20 Sep '04

20 Sep '04

I'm trying to serialize a derived class through a shared_ptr<> to the base class. I'm getting an unregistered class exception even though my derived class is registered with register_type<>. It seems the exception is being thrown while trying to serialize boost::detail::sp_counted_base which I believe is part of shared_ptr. Could this be an internal error in the serialization library?

2 1

[boost-users][MPL] integral_c<long long,...>
by Cromwell Enage 19 Sep '04

19 Sep '04

The attached program fails to compile if BOOST_HAS_LONG_LONG is defined. The error is mainly about integral_rank<long long int> not having a static member named 'value'. Is integral_c not meant to be used with long long? Cromwell Enage __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail

2 2

Regex Libraries giving errors
by Corey M. Ellis 19 Sep '04

19 Sep '04

Hi, I am trying to use the regex library (regex_match in particular) I included "boost/regex.hpp". I get the following errors when compiling my code. g++ -W -Wall -Wpointer-arith -O3 -g -pipe -D_REENTRANT -DACE_HAS_AIO_CALLS -D_GNU_SOURCE -I/usr/local/ACE_wrappers -DACE_HAS_EXCEPTIONS -D__ACE_INLINE__ -I/usr/local/ACE_wrappers -I/usr/local/include -I/usr/local/include/boost-1_31 -DACE_NTRACE=1 -D__USE_REG_EXPR -c -o .obj/FTP_Task.o FTP_Task.cpp In file included from /usr/local/include/boost-1_31/boost/cregex.hpp:30, from /usr/local/include/boost-1_31/boost/regex/v4/regex.hpp:25, from /usr/local/include/boost-1_31/boost/regex.hpp:34, from FTP_Task.h:19, from FTP_Task.cpp:1: /usr/local/include/boost-1_31/boost/regex/v4/cregex.hpp:75: error: syntax error before numeric constant /usr/local/include/boost-1_31/boost/regex/v4/cregex.hpp:96: error: ISO C++ forbids declaration of `reg_comp_flags' with no type /usr/local/include/boost-1_31/boost/regex/v4/cregex.hpp:100: error: syntax error before numeric constant /usr/local/include/boost-1_31/boost/regex/v4/cregex.hpp:103: error: ISO C++ forbids declaration of `reg_exec_flags' with no type make[1]: *** [.obj/FTP_Task.o] Error 1 make[1]: Leaving directory `/home/bipuser/nBIP-9.0/src' make: *** [NBIP] Error 2 Can sme please shed some light on the problem. I am only using the regex_match function. I create an expression object, and use regex_match to match against it. Thanks in advance for any help. Corey

2 5

Re: [Boost-users] Regex Libraries giving errors
by corey.m.ellis＠comcast.net 18 Sep '04

18 Sep '04

Thanks for the quick response. Do you know of a workaround that is being used for version 1_31. All the examples say include "boost/regex.hpp". Is there some way I can include the headers so the problem will go away. What's really strange is that I compiled the entire library using the Jam method, and it I got these regex errors. I then compiled the regex libraries using the gcc make file, and use the static libraries when I compiled. It worked. But after I made some more code changes (not regex changes) to my application and recompiled, I started getting the errors again. I thought the gcc compilation fixed it, but it didn't. Anyway, if you could tell me a workaround that would be great. Is it just a matter of trying to include the regex.hpp header first before the regex.h header gets included, have other people seen this problem in the regex library? Corey -------------- Original message -------------- > > I am trying to use the regex library (regex_match in particular) > > I included "boost/regex.hpp". I get the following errors when compiling > > my code. > > > > Can sme please shed some light on the problem. I am only using the > > regex_match function. I create an expression object, and use regex_match > > to match against it. > > > > Thanks in advance for any help. > > I suspect that you are including the platforms (possibly > indirectly) before including . That currently causes > conflicts: there is a fix in the works, but it involves a complete > restructuring of Boost.Regex that won't be released until Boost-1.33 (that's > the release after next). > > John. > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

RE: [Boost-users] Boost problems
by Puverle, Tomas (IT) 18 Sep '04

18 Sep '04

> <snip> > Done, apologies for the delay, I didn't notice your message > first time around, I should have made the topic more obvious, like "Bug report: ... " or something. Would it have been better or made it easier for you if I had posted the original report to the boost developer forum? > Thanks for the bug report, > > John. Thank you very much John. Also, regarding the last issue mentioned in my original email (Boost.Build), should I cross post it to the Boost.Build forum? Thanks, Tom -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited.

2 1

RE: [Boost-users] Re: Boost problems
by Puverle, Tomas (IT) 18 Sep '04

18 Sep '04

> That's been cleaned up in the latest CVS. Great! So that's 2 problems down, 2 to go. Do you think you could comment on the other 2 unresolved issues from my original email? (the defines in regex and Boost.build problem/feature request?) Thanks a lot. Tom -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited.

2 1

Boost problems
by Tom Puverle 18 Sep '04

18 Sep '04

Hey boosters, We have found some problems with the following files: boost_1_31_0/libs/regex/src/cpp_regex_traits.cpp boost_1_31_0/libs/test/src/test_tools.cpp They use definitions from <ios> but never actually include it - they are relying on the fact that in some STLs the headers are pulled in by other header files. Unfortunately, under our STL (RW STL 6.0) this is not the case and hence the files fail to compile. Add the following include line to both to fix the problem: #include <ios> There is another STL related problem in boost_1_31_0/boost/regex/config/cwchar.hpp which manipulates some macros such as iswalnum etc. These are originally pulled in from <wctype.h> and <wchar.h> under /usr/include. With gcc, <cwctype> and <cwchar> undefine these macros as the first thing, however, our STL's headers don't. That leads to compilation errors from the above header. To fix this, add #undef iswalnum #undef iswalpha #undef iswblank #undef iswcntrl #undef iswdigit #undef iswgraph #undef iswlower #undef iswprint #undef iswprint #undef iswpunct #undef iswspace #undef iswupper #undef iswxdigit #undef iswctype #undef towlower #undef towupper #undef towctrans #undef wctrans #undef wctype when RW STL 6.0 is being used. The final problem is more of a feature request: In our environment, we build all shared objects with all external references unresolved, so that a .so A never depends on a shared library B. The symbols get only resolved once the final executable is linked. This causes us problems with boost when building the regression tests, as the final Jam link rule adds all additional libraries *before* the final libboost_*. That way, we end up with unresolved symbols to e.g. our STL. Is there any way in which this requirement could be incorporated into Boost.Build? I guess a simple solution would be to have another variable called e.g. ADDITIONAL_LIBS_LAST which would be passed as the last thing on linker line in the link rule. HW & SW details: OS: Red Hat Enterprise Linux 2.1 & 3.0 HW: various Compiler: gcc 3.2.3-20 STL: RogueWave 6.0 Boost: 1.31.0 Thanks, Tom _________________________________________________________________ It's fast, it's easy and it's free. Get MSN Messenger today! http://www.msn.co.uk/messenger

2 1

BOOST 1.32 soon?
by Daniel L Elliott 18 Sep '04

18 Sep '04

Hello, Will there be a BOOST 1.32 tagged soon? Thank you, to everyone, for this contribution! -- dan ******************* Dan Elliott *******************

4 7

Can boost::tokenizer tokenize 2byte character string?
by Lee, Joo-Young 17 Sep '04

17 Sep '04

Hi. I try to use 'boost::tokenizer<boost::char_separator<char> >' to separate 2byte character string like Korean, Japanese, or Chinese. But, I found that it does not works correctly. Is there a solution? Thanks for the help, Lee Joo-Young

1 0

Re: Newbie help in compiling boost
by Jeff Flinn 17 Sep '04

17 Sep '04

"David Abrahams" <dave(a)boost-consulting.com> wrote in message news:uzn3p2g31.fsf__38100.3229799264$1095428766$gmane$org@boost-consulting.com... > Murali Krishnan Ganapathy <gmkrishn(a)cs.uchicago.edu> writes: > > > When boost.jam tries to make a directory it tries to access %TEMP%. > > Machine complains that there is no "C;\DOCUME~1\...." (it is not > > checking the 8.3 version I guess) > > and the mkdir1 fails. Is the ';' here a typo, or is that what TEMP and TMP are really defined as? This is the 8.3 version, so I don't understand what your saying here. > > The work around is to have a simple %TEMP% and %TMP%. > > I reset %TEMP% and %TMP% to "C:\Temp" (no spaces and less than 8 > > chars), now boost is > > building as I write this. > > > > I guess this should be considered as a minor bug in bjam? > > Maybe so. Any Win/DOS experts out there who can suggest a correct > fix? NT, 2K and XP maintain "C:\Documents and Settings\<user-name>\Local Settings\Temp" directories for each user. Changing this to C:\Temp could cause problems for users w/out admin privileges. I've no problems with vc7.1 builds, and I imagine a lot of others utilize the same platform and can build as well. So it's not a general problem. The possible sources of this problem could be: - some odd access settings on the directory ( can even occur with admin privilege) - a lot of files in the temp directory from other misbehaving app's - environment variable definitions were mucked up by manual editing - generally mucking about with these dir's by the user can cause problems Jeff F

5 4

RE: [Boost-users] Boost problems
by Puverle, Tomas (IT) 17 Sep '04

17 Sep '04

On a similar note, I also noticed the following: iterator/detail/facade_iterator_category.hpp:# include <boost/python/detail/indirect_traits.hpp> iterator/indirect_iterator.hpp:#include <boost/python/detail/indirect_traits.hpp> It doesn't seem right that iterator should be including headers from python/detail. Tom -------------------------------------------------------- NOTICE: If received in error, please destroy and notify sender. Sender does not waive confidentiality or privilege, and use is prohibited.

2 1