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

Building on AIX?
by scottnesbitt 05 Jul '02

05 Jul '02

Am new to boost, trying to build on AIX. Here is what I get: $ l total 96 -rw-r--r-- 1 jsn mup 1461 Feb 7 10:33 Jamfile -rw-r--r-- 1 jsn mup 91700 Feb 7 10:33 threads.mcp $ pwd /usr/local/src/boost_1_27_0/libs/thread/build $ uname -a AIX io 3 4 00028AAF4C00 $ set | grep BOOST BOOST_ROOT=/usr/local/src/boost_1_27_0 $ jam +: line 1304: yacc stack overflow at keyword } don't know how to make all ...found 1 target... ...can't find 1 target... $ Are there other args I need to give to jam? Read the build instructions, it seemed ambiguous about AIX support. Thanks for any tips, Scott Nesbitt

3 2

boost.python installation problems
by Christos Nicolaou 05 Jul '02

05 Jul '02

Hey Seems that I've run into a pretty common problem people have when they try to install Boost.Python: I get those error messages saying that Python.h and patchlevel.h Details: - NT system, installed mingw-2.95.2 (exe in my path), boost_1_28_0. - within f:\boost_1_28_0 directory I type: bjam -sBOOST_ROOT=. -sTOOLS="gcc" -sPYTHON_ROOT=c:\python21 - I get the missing header files errors. Lots of them. Any clues? There seem to be postings about this problem but I couldn't find a solution. Thanks in advance Christos

3 2

Python boost inheritance documentation error
by J.Flenter 05 Jul '02

05 Jul '02

Hello, This is my first post here, but I found some code that generated an error using Visual Studio 6.0 (with Service Pack 5): The documentation (url: http://www.boost.org/libs/python/doc/inheritance.html) said: [snip] base_class.def(python::constructor<void>()); [snip] when I use void at that location the compiler generates an error (although it does work without specifying the keyword void). The error is: c:\BOOST\boost/python/detail/init_function.hpp(319) : error C2440: 'type cast' : cannot convert from 'void' to 'class boost::python::detail::reference_parameter<void>' Expressions of type void cannot be converted to other types c:\BOOST\boost/python/detail/init_function.hpp(314) : while compiling class-template member function 'class boost::python::detail::instance_holder_base *__thiscall boost::python::detail::init1<class boost::python::detail::instance_value_hold I do use a different class (included as a zip file), but in this case I don't think that will generate the compiler error. I hope this makes it clear... Jacco Flenter begin 666 sourceFiles.zip M4$L#! H```````F%Y"P````````````````)````<V5T=&EN9W,O4$L#!!0` M```(`(][XBQLJ6V_? ```"D!```3````<V5T=&EN9W,O<V5T=&EN9W,N:./E M4DY)3<O,2U4(B SQ\/?CY>+ETM=7"*DL2%4`2V269.;G%5OQ<I4`A8 B"J5Y MQ9GI>:DI"LD9B44*I1;66*2*,_*+2A1*#<VP2>;DYZ4KE!H;68/L@LDBFUF, M;":*B<4H)B*;5XQF'JHKDRI+4G&[,Z,\OR@%ITNAD@!02P,$% ````@`V9[D M+/TW#9R&`0``I0,``!(```!R969E<F5N8V5#;W5N=&5R+FBM4MMJPD 0?4X@ M_S#5%_7!K"U:+U (,6*AU>*EM% 88K*)"S&19),7L=_>W=5ZQ3<#@67V[#EG MYHQI0DH#&GO43O*8T[2^[ *3A\#U* 1)"GQ)+S#(a)16Z6U0W=-.5_C\_0#;W, M`GCP:<!BZE>LP1=.G($SLAU[/!_-G D.$9O$)JVGUB.V+6)AH]%OH64Y%A+2 M) /[T2:=SC/BZ\A^F_>=/E8-72OO&.$NA-*F)GWB^]3&3V<"+] @A(BJ5EZG M;KAR(1&#DB@:^P)HFM=8Q1%[4>Y3*&64<Q:'V<__H;XL*8@8;4UJ"?_P\3T; MCD=2)797-%O+;#:2VRH2YHL\<BYR$:]56F&4+-P(#M#ZP4Y-#5I3^:E0T]-8 M05QMI,@Z7T3,Z\JC)I8!0LHG9^"*E*WV%."2IN(E<<8A;P-#;U>:[EN[]6)? MEU*N[Q^D3NLI724%O;PJ6,IST>KO#<IUR@J7TV,GR/ <NE,09O'*;4_-:MN3 M0:CM/$]B*VIJJ(8NQWK,^^X[_ =02P,$% ````@`V9[D+ )810E+`0``E 8` M`!0```!R969E<F5N8V5#;W5N=&5R+F-P<,53T6J#,!1]KN _7+H7I;3IWD;% MO71?T'U D7A;`YI(3#K8Z+Y]B;9E-<I*29D@Z+TGG'//N2$$).Y0(J>X%IHK ME M:URM@55UBA5QEB@D.8@>J0 <+M,R:9A$&A-C7QQ,&8?#$."UUCC!UQ!53 M"_!%10BL!6^4U-2.2=[P\NUSGOX4JU6_$E$K`_0+L"WM2N^H%./[)@Z#KS"8 MJ((U\]>MTX;4/9+\QE]3&?BS:1]OE'40+/];P/)NPF^'L6,['? 5@<EYCZI= M8:JEX5.]589#5FKTF3ES..R\1L7FJGB:=R)1:<FC$1/CQ+\A69Z#X A*#%SM MQSMAZ"].1/T5NX;/9LF_."2Q$@=L3;J0/=Z8CK7OS;@[\WG2MHW@32?8`N&C M0#X0+)C6)TIA#K#=F)EINHQM'B?*@4MJ`SDGLFS_CF<1775PIV[+\ =02P,$ M% ````(a)`EZ#D+.C8_7"8`0``2@0```T```!H;VQY1W)A:6PN8W!PM9-=2\,P M%(:O-]A_.%203J2]E3H'.H<*ZD3GA2"$M#E=(VU2\B$.\;^;MNN(TPMO=E%( MWO/QG/.&QC%<S\\OYX]/$8R&<0Q<9*5E"*9 T&@,%RL-5#!(J48HD#)4>C0\ MZ/,FJ93:Q/7:%%+$64FU)JGEI4N+BKJ>>JE!W^^U/T1%X,<5YJA09#B35IBF M@0LW"3G#'!Y>EM>+^]%P(&B%NJ890D>%,VB'2)+N?MH4#2X6BZ<EZ8K(W>+R M^79.;NYOEF$AR_65HKP<NZQ/]PW<VC.%U*!;%&3ZAID!A;5"C:*9TYG!->"' M0:&YXU62V1*CIK0C)DDG]9NW!:33PF +#,;=: ZXX34N/W1+F'6-/3R7"J15 M'K(U-CJ*?>8/LR>[YDUA5R%M0>C-=OS+<F_$<\: 0FW3DF=0896ZO7(K,N/F M:7?_NW_DWBH\W(TER0K-XP_Q&()?FH>'EB^ $"ZX(>3?[*T]4FBC;&:DFK1G ML"<P#<=[0FPZ[\$YRMC6)6>:?PWVA%18R7?TJ3O*YJ&^W-^)@O%\-/P&4$L! M`A0`"@``````"87D+ ````````````````D````````````0`,!!`````'-E M='1I;F=S+U!+`0(4`!0````(`(][XBQLJ6V_? ```"D!```3``````````$` M( " @2<```!S971T:6YG<R]S971T:6YG<RYH4$L!`A0`% ````@`V9[D+/TW M#9R&`0``I0,``!(``````````0`@`("!U ```')E9F5R96YC94-O=6YT97(N M:%!+`0(4`!0````(`-F>Y"P"6$4)2P$``)0&```4``````````$`( " @8H" M``!R969E<F5N8V5#;W5N=&5R+F-P<%!+`0(4`!0````(`)>@Y"SHV/UPF $` M`$H$```-``````````$`( " @0<$``!H;VQY1W)A:6PN8W!P4$L%!@`````% -``4`-0$``,H%```````` ` end

2 1

Re: ScopeGuard Implementation
by ye_feng 05 Jul '02

05 Jul '02

don't forget to make "bool dismissed_" mutable, so that dismiss() can change it for "typedef const scoped_guard_impl& scoped_guard" Regards, yefeng --- In Boost-Users@y..., Jon Wang <webmaster@b...> wrote: > > >This solution won't work, because you are storing a reference to a > >temporary object, which will cease to exist after the constructor call. > >Then the destructor is accessing memory that has already been destructed > >and probably reused. > > Sorry! A silly mistake here:-) > Well, what I meant is > > typedef const scoped_guard_impl& scoped_guard; > // not "typedef scoped_guard_impl& scoped_guard;" > > Then "a const reference initialized with a temporary value makes that temporary value live for the lifetime of the reference itself". Is this fine? > > > > Jon

1 0

Re: [Boost-Users] ScopeGuard Implementation
by Jon Wang 04 Jul '02

04 Jul '02

>This solution won't work, because you are storing a reference to a >temporary object, which will cease to exist after the constructor call. >Then the destructor is accessing memory that has already been destructed >and probably reused. Sorry! A silly mistake here:-) Well, what I meant is typedef const scoped_guard_impl& scoped_guard; // not "typedef scoped_guard_impl& scoped_guard;" Then "a const reference initialized with a temporary value makes that temporary value live for the lifetime of the reference itself". Is this fine? Jon

1 0

Re: [Boost-Users] ScopeGuard Implementation
by Jon Wang 04 Jul '02

04 Jul '02

>Looks fine to me. To rein in the compile times a little bit, you can pull in >just the 0-argument version of function, e.g., > #include <boost/function/function0.hpp> // (not <boost/function.hpp>) >and use boost::function0<void> instead of boost::function<void>. Under MSVC, >you might even get fewer senseless ICEs :) Oh, yes, function0 is enough. Thanks very much, Doug. Well, could the copy constructing from a boost.bind result be avoided? How about the following solution? Thanks. #include <boost/function/function0.hpp> #include <boost/bind.hpp> #include <boost/utility.hpp> #define ON_BLOCK_EXIT(x) scoped_guard guard##__LINE__ = make_guard(::boost::bind##x);guard##__LINE__; class scoped_guard_impl: ::boost::noncopyable { private: typedef const ::boost::function0<void>& f_type; bool dismissed_; f_type f_; public: scoped_guard_impl(f_type v): f_(v), dismissed_(false) {} ~scoped_guard_impl() throw() { if(!dismissed_) try {f_();} catch(...) {} } void dismiss() throw() { dismissed_ = true; } }; typedef scoped_guard_impl& scoped_guard; typedef scoped_guard_impl make_guard; #include <stdio.h> int main() { FILE* fp = fopen("hello.txt", "r"); ON_BLOCK_EXIT((fclose, fp)); //scoped_guard guard = make_guard(boost::bind(fclose, fp)); return 0; } Jon

1 0

Re: [Boost-Users] ScopeGuard Implementation
by Douglas Paul Gregor 04 Jul '02

04 Jul '02

On Thu, 4 Jul 2002, Jon Wang wrote: > Oh, yes, function0 is enough. Thanks very much, Doug. > Well, could the copy constructing from a boost.bind result be avoided? > How about the following solution? Thanks. This solution won't work, because you are storing a reference to a temporary object, which will cease to exist after the constructor call. Then the destructor is accessing memory that has already been destructed and probably reused. Doug

1 0

Re: [Boost-Users] ScopeGuard Implementation
by Douglas Gregor 03 Jul '02

03 Jul '02

On Wednesday 03 July 2002 12:30 am, Jon Wang wrote: > Hi, > > Recently I've tried to implement Andrei's ScopeGuard (see > http://www.cuj.com/experts/1812/alexandr.htm?topic=experts for details) > with boost.function and boost.bind. Is there anything wrong with it? Thanks > for reviewing. Looks fine to me. To rein in the compile times a little bit, you can pull in just the 0-argument version of function, e.g., #include <boost/function/function0.hpp> // (not <boost/function.hpp>) and use boost::function0<void> instead of boost::function<void>. Under MSVC, you might even get fewer senseless ICEs :) Doug

1 0

ScopeGuard Implementation
by Jon Wang 03 Jul '02

03 Jul '02

Hi, Recently I've tried to implement Andrei's ScopeGuard (see http://www.cuj.com/experts/1812/alexandr.htm?topic=experts for details) with boost.function and boost.bind. Is there anything wrong with it? Thanks for reviewing. #include <boost/function.hpp> #include <boost/bind.hpp> #include <boost/utility.hpp> #define ON_BLOCK_EXIT(x) scoped_guard guard##__LINE__(::boost::bind##x) class scoped_guard: ::boost::noncopyable { private: typedef const ::boost::function<void> f_type; // reference might cause crash bool dismissed_; f_type f_; public: scoped_guard(f_type& v): f_(v), dismissed_(false) {} ~scoped_guard() throw() { if(!dismissed_) try {f_();} catch(...) {} } void dismiss() throw() { dismissed_ = true; } }; With the macro ON_BLOCK_EXIT, we can easily use it like this: // ... #include <stdio.h> int main() { FILE* fp = fopen("hello.txt", "r"); ON_BLOCK_EXIT((fclose, fp)); //scoped_guard guard(boost::bind(fclose, fp)); return 0; }　　　　　　　　　　　　　　　　Jon Wang babysloth(a)c-view.org 　　　　　　　　　　　　2002-07-03

1 0

regex with double byte character sets
by bthiesfield 03 Jul '02

03 Jul '02

Hi all, I am currently trying to use the boost regex library with Japanese language strings. It appears like DBCS is not supported. For example, using the following code (with compile definition of BOOST_REGEX_USE_C_LOCALE) I get the output strings as 0 = "。" 1 = "English" Instead of the expected: 0 = "やゆよわをー。" 1 = "English" This is due to the fact that the Japanese (SJIS encoding) for one of these characters uses the [ character as one of the characters in the encoding. setlocale( LC_COLLATE, "Japanese" ); setlocale( LC_CTYPE, "Japanese" ); char * pszText = "やゆよわをー。 [english]", char * pszRule ="([^\\[]*)\\[([[:word:]]*)\\]"; // split the string into it's components std::vector<std::string> vPart; boost::regex eParseExpr( pszRule, boost::regbase::normal | boost::regbase::icase ); boost::regex_split( std::back_inserter(vPart), std::string(pszText), eParseExpr ); Is there some what to modify the library to enable DBCS? For example, can the char_traits be modified to enable DBCS processing? (Keeping in mind that the biggest problem with DBCS is that a single character may consist of 2 bytes which tends to blow out all assumptions about the size of characters). Brodie.

4 4