September 2008 - Boost - lists.preview.boost.org

Re: [boost] [switch] Is there any interest in library switch extension?
by Kasra Nassiri(Math & ComSci) 15 Sep '08

15 Sep '08

I am not sure how it could be performed on O(1), however, the current algorithm in O(n). Current algorithm is something like: BOOST_SWITCH( "hello world" ) { #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("invalid case") std::cout << "Reached invalid case!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("another string") std::cout << "Reached invalid case!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE( "hello world" ) std::cout << "Reached the correct code!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("hello world 1") std::cout << "Reached the correct code! (no breaks)" << std::endl; BOOST_BREAK; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("hello world") std::cout << "Should throw exception! (duplicated value)" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("another string 1") std::cout << "Reached invalid case! (broken)" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_DEFAULT std::cout << "Reached default Hello!" << std::endl; } BOOST_SWITCH_END It works fine but it is O(n) and uses hash tables. I want to get rid of "#include BOOST_PP_UPDATE_COUNTER()" somehow, but at this time I can't. Even when you use BOOST_BREAK at early statements it is still O(n), since it checks for duplicated values. But using break would change to O(m) : m < n. The 'default' statement at the moment must be at the end of all other cases, however, it will be updated so it could be anywhere with in the switch. By the way where is this switch_utility? Any other suggestions?

1 0

[BENUM] Any interest in auto-generated enum-to/from-string functions, without longer/more-complex usage?
by Tony D 15 Sep '08

15 Sep '08

Anyone think this functionality belongs in boost? (I have working code) ENUM(X, B = N, C, D = B + C }; // or whatever X x1 = C; cout << x1 << '\n'; // outputs "C\n" X x2("D"); // construct from string... if (x2 == D) { will do } Tony __________________________________________________________ Sent from Yahoo! Mail. A Smarter Email http://uk.docs.yahoo.com/nowyoucan.html

4 3

[boost.test] Customized test suite gives SIGSEGV when setup fails
by Markus Schöpflin 15 Sep '08

15 Sep '08

Hello, I'm getting a SIGSEGV when trying the following with Boost 1.35.0 and gcc 4.1.2: ---%<--- #define BOOST_TEST_ALTERNATIVE_INIT_API #include <boost/test/included/unit_test.hpp> using namespace boost::unit_test; struct my_test_suite : public test_suite { my_test_suite() : test_suite("my_test_suite") { throw std::runtime_error("runtime_error"); } }; bool init_unit_test() { test_suite *test = BOOST_TEST_SUITE(""); test->add(new my_test_suite()); framework::master_test_suite().add(test); return false; } ---%<--- When run in the gdb, I get the following output: ---%<--- (gdb) run Starting program: a.out Test setup error: std::runtime_error: runtime_error Program received signal SIGSEGV, Segmentation fault. 0x0806e2fa in ~framework_impl (this=0x8089740) at boost/test/impl/framework.ipp:126 126 delete (test_suite const*)tu.second; --->%--- The code in question has been ported from an old Boost.Test version to 1.35. When looking at the documentation for Boost.Test (1.36, as this seems more in sync with Boost.Test 1.35) I'm unsure that custom test suites are still supported in this way. What I need to do is to set up and tear down a database connection at the start/end of a test suite execution. Is there some other way for this? TIA, Markus

2 3

[Boost-Test] Patch xml_log_formatter.hpp for sun-5.9 on Linux
by K. Noel Belcourt 15 Sep '08

15 Sep '08

Hi, Attached is a patch to fix this problem with Boost.test trunk and sun-5.9 on Linux. sun.compile.c++ ../../../bin.v2/libs/test/build/sun-5.9/debug/link- static/unit_test_main.o "../../../boost/test/impl/unit_test_main.ipp", line 115: Error: Could not find a match for std::vector<boost::unit_test::test_case_filter::single_filter>::vector (boost::unit_test::basic_string_token_iterator<char, boost::unit_test::ut_detail::default_char_compare<char>>, boost::unit_test::basic_string_token_iterator<char, boost::unit_test::ut_detail::default_char_compare<char>>) needed in boost::unit_test::test_case_filter::test_case_filter (boost::unit_test::basic_cstring<const char>). "../../../boost/test/impl/unit_test_main.ipp", line 115: Error: Formal argument x of type const std::vector<boost::unit_test::test_case_filter::single_filter>& in call to std::vector<std::vector<boost::unit_test::test_case_filter::single_filte r> >::push_back(const std::vector<boost::unit_test::test_case_filter::single_filter>&) is being passed int. 2 Error(s) detected. "/home/sntools/extras/compilers/studio12/sunstudio12/bin/CC" +d - g -DBOOST_ALL_NO_LIB=1 -I"../../.." -c -o "../../../bin.v2/libs/test/ build/sun-5.9/debug/link-static/unit_test_main.o" "../../../libs/test/ src/unit_test_main.cpp" The patch adds a using declaration to bring the base class function into the overload set. Sun's interpretation of the standard seems to be correct, from what I can tell. Here's their rational: [ from their sun-5.9 FAQ ] Q. Why does the compiler say that a derived-class virtual function hides a base-class virtual function with a different signature? My other compiler doesn't complain about the code. A. The C++ rule is that overloading occurs only within one scope, never across scopes. A base class is considered to be in a scope that surrounds the scope of a derived class. Any name declared in a derived class therefore hides, and cannot overload, any function in a base class. This fundamental C++ rule predates the ARM. If another compiler does not complain, it is doing you a disservice, because the code will not behave as you probably expect. Our compiler issues a warning while accepting the code. (The code is legal, but probably does not do what you want.) If you wish to include base-class functions in an overloaded set, you must do something to bring the base-class functions into the current scope. If you are compiling in default standard mode, you can add a using-declaration: Okay to commit to trunk? -- Noel

2 1

Boost with comeau c++
by Sean Farrow 14 Sep '08

14 Sep '08

Hi: Does anybody have any experience/information as to how well boost compiles with comeau c++? Any information would be gratefully apreciated. Cheers Sean. __________ Information from ESET NOD32 Antivirus, version of virus signature database 3440 (20080913) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

2 1

[switch] Is there any interest in library switch extention?
by Kasra (Math & ComSci) 14 Sep '08

14 Sep '08

Hello to all; I came up with an idea of extending the current switch functionality using library extensions. The problem is mostly when we have char[] (or any other object) in the switch statement: void do_switch(const char *cstr) { switch( cstr ) { case "Sunday": case "Saturday": ... break; case "Monday": ... }; } Obviously we can't do this using C++ alone, so if we have something like BOOST_SWITCH() which would accomplish this would be great. I have a runtime solution (with need of some attention) for this, however, a compile time solution is desirable. The library should do something like: BOOST_SWITCH( "hello world" ) { BOOST_CASE( "hello" ): BOOST_CASE( "world" ): ... break; BOOST_DEFAULT: ... }; (Please don't suggest 'enum' solutions since they are not as dynamic and are confined to very small domain) Any suggestions, ideas & etc? With best regards Kasra

2 1

[RangeEx] Range & RangeEx
by Robert Jones 14 Sep '08

14 Sep '08

RangeEx appears to have the functionality I'm seeking, which is the std algorithms overloaded for ranges, but there's few things I'm not clear about. I've downloaded the range_ex.zip archive from the vault, but I'm not clear if it replaces or extends Boost.Range. Am I now able to write things like #include <vector> #include <iostream> #include <iterator> #include <boost/range.hpp> void f( vector<int> const & v ) { using namespace std; using namespace boost; copy( v, ostream_iterator<int>( cout, " " ); }

6 12

[Xpressive] vs libpcre performance
by Sebastian Redl 14 Sep '08

14 Sep '08

Hi, A German journal recently published a small programming contest with a very simple text processing problem. The program needs to renumber and sort footnotes in a long text. I wrote a solution in C++ using a number of Boost libraries, among them Xpressive, but also Interprocess for memory-mapping the input file. Then I looked at the submissions page and found one C++ solution already there, an admirably short (but rather inflexible) program using libpcre. http://www.linux-magazin.de/static/listings/magazin/2008/10/leser/cpp/footn… Comparing the performance, my program was sightly but consistently slower than this small program, even when I felt that my usage of memory-mapping and then scanning the whole file instead of going line by line ought to give me a speed advantage. So to test the performance I took the existing submission and replaced libpcre with Xpressive (see attached file). I believe the solutions to be functionally equivalent. However, the original takes 6 seconds to process a 55MB file, whereas my variation takes ~15 seconds. That's on the second run of each program, meaning that the entire file is in the OS cache. This seems awfully slow. Has anyone done a proper performance comparison between Xpressive and libpcre? Boost.Regex's performance pages lists PCRE, but in version 4.1, where 7.8 is the most recent release. My system: Athlon 64 2000MHz (64-bit mode) 1GB RAM Linux 2.6.23, GCC 4.1.2 Boost trunk as of 2008-09-11 libpcre 7.7 Sebastian Redl

3 4

warnings in tests
by Emil Dotchevski 13 Sep '08

13 Sep '08

In Boost testing, wouldn't it be a good idea to have a script that scans the output generated from each successful test for the word "warning" and indicate its presence in the test matrix? Perhaps leave the box green but instead of "pass" have it say "pass/w" or something. Without this, platform-specific warnings can only be addressed after they're reported by the end users. Emil Dotchevski Reverge Studios, Inc. http://www.revergestudios.com/reblog/index.php?n=ReCode

2 2

[Dataflow] [Review] Dataflow review has ended
by jarvi＠cs.tamu.edu 13 Sep '08

13 Sep '08

The number of reviews for the dataflow library was very low (one, positive). Rather than extending the review period in hopes for more reviews, the review period of the data-flow library is now over, and at this time the library is not accepted as a Boost library. The discussions indicated that there's some confusion regarding the library's role; this concern needs to be addressed prior to moving forward with the library. Stjepan is content of the feedback received during the review, and will use it to clarify the library's documentation and further develop the library. He intends to bring the library for review again when the concerns raised have been addressed. Even though the review period is over, I encourage everyone to submit their comments, half-written reviews, reviews that you were planning to write, etc. as those will help Stjepan in his efforts in improving the library. Thank you for everyone for participating in the review, and thank you Stjepan for your work for Boost! Jaakko Järvi Review Manager

11 11