- Boost - lists.boost.org

Boost inspection notification (2007-06-11/RC_1_34_0) *X*
by Rene Rivera 11 Jun '07

11 Jun '07

Boost Inspection Report Run Date: 16:06:15 UTC, Monday 11 June 2007 An inspection program <http://www.boost.org/tools/inspect/index.html> checks each file in the current Boost CVS for various problems, generating this as output. Problems detected include tabs in files, missing copyrights, broken URL's, and similar misdemeanors. Totals: 11452 files scanned 909 directories scanned (including root) 163 problems reported Problem counts: 0 files with invalid line endings 0 bookmarks with invalid characters 3 invalid urls 71 broken links 21 unlinked files 22 file/directory names issues 2 files with tabs 9 violations of the Boost min/max guidelines 35 usages of unnamed namespaces in headers (including .ipp files) Summary: archive (3) bind (1) boost-root (1) build (1) date_time (1) doc (2) filesystem (3) graph (3) iostreams (2) lambda (3) math (8) more (4) mpl (4) multi_array (2) ptr_container (1) python (8) regex (1) regression (32) serialization (1) test (81) type_traits (1) Details: *R* invalid (cr only) line-ending *A* invalid bookmarks, invalid urls, broken links, unlinked files *N* file/directory names issues *T* tabs in file *M* uses of min or max that have not been protected from the min/max macros, or unallowed #undef-s *U* unnamed namespace in header |archive| boost/archive/basic_streambuf_locale_saver.hpp: *N* name exceeds 31 characters boost/archive/impl/xml_wiarchive_impl.ipp: *U* unnamed namespace at line 53 boost/archive/iterators/remove_whitespace.hpp: *U* unnamed namespace at line 57 |bind| boost/bind/placeholders.hpp: *U* unnamed namespace at line 25 |boost-root| index.htm: *A* broken link: doc/hmtl/date_time/details.html#changes |build| tools/build/v2/test/test_system.html: *A* unlinked file |date_time| libs/date_time/xmldoc/date_time_docs_howto.html: *A* unlinked file |doc| doc/html/boost_math/inverse_complex.html: *A* unlinked file doc/html/jam.html: *A* unlinked file |filesystem| libs/filesystem/doc/do-list.htm: *A* invalid URL (hardwired file): file://?/ *A* invalid URL (hardwired file): file://?/UNC/ libs/filesystem/doc/tr2_proposal.html: *A* invalid URL (hardwired file): file:///C|/boost/site/libs/filesystem/doc/operations.htm#complete_note |graph| boost/graph/relax.hpp: *T* libs/graph/example/file_dependencies.cpp: *M* violation of Boost min/max guidelines on line 139 libs/graph/test/serialize.cpp: *T* |iostreams| libs/iostreams/doc/concepts/multi-character.html: *A* unlinked file libs/iostreams/doc/installation.html: *A* broken link: ../../../tools/build/v1/build_system.htm |lambda| boost/lambda/core.hpp: *U* unnamed namespace at line 62 boost/lambda/detail/lambda_functors.hpp: *U* unnamed namespace at line 25 boost/lambda/exceptions.hpp: *U* unnamed namespace at line 24 |math| libs/math/test/common_factor_test.cpp: *M* violation of Boost min/max guidelines on line 146 *M* violation of Boost min/max guidelines on line 147 *M* violation of Boost min/max guidelines on line 193 *M* violation of Boost min/max guidelines on line 194 |more| more/cvs.html: *A* unlinked file more/getting_started/unix-variants.html: *A* broken link: ../../doc/html/thread/build.html#thread.build more/getting_started/windows.html: *A* broken link: ../../doc/html/thread/build.html#thread.build more/separate_compilation.html: *A* broken link: ../libs/config/test/link/test/Jamfile.v2 |mpl| boost/mpl/alias.hpp: *U* unnamed namespace at line 17 libs/mpl/doc/refmanual/for-each.html: *A* broken link: ./value-initialized.html |multi_array| boost/multi_array/base.hpp: *U* unnamed namespace at line 69 libs/multi_array/test/generative_tests.hpp: *U* unnamed namespace at line 57 |ptr_container| libs/ptr_container/doc/tutorial_example.html: *A* unlinked file |python| libs/python/doc/building.html: *A* broken link: ../index.htm libs/python/doc/tutorial/doc/html/python/hello.html: *A* broken link: ../../../../../example/tutorial/Jamrules libs/python/doc/tutorial/index.html: *A* broken link: ../../../LICENSE_1_0.txt libs/python/doc/v2/May2002.html: *A* broken link: ../../../../tools/build/v1/build_system.htm *A* broken link: ../../../../tools/build/v1/gen_aix_import_file.py libs/python/doc/v2/faq.html: *A* broken link: ../../../../tools/build/v1/build_system.htm |regex| libs/regex/performance/input.html: *A* unlinked file |regression| regression/.htaccess: *N* leading character of ".htaccess" is not alphabetic tools/regression/test/test-cases/incremental/bjam.log.1: *N* name contains more than one dot character ('.') tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/compile-fail.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/completed.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/compile-fail.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/completed.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/lib.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/misfire.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/no-run.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/run_pyd.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/expected/test-case.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/test-case.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_is_test_log_complete/test-driver.xsl: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_re_match/expected/test_re_match.xml: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ tools/regression/xsl_reports/test/test-components/test_test_case_status/explicit-failures-markup.xml.test: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ *N* name contains more than one dot character ('.') *N* name exceeds 31 characters tools/regression/xsl_reports/test/test-components/test_test_structure/explicit-failures-markup.xml.test: *N* file path will exceed 100 characters in a directory tree with a root of the form boost_X_XX_X/ *N* name contains more than one dot character ('.') *N* name exceeds 31 characters tools/regression/xsl_reports/xsl/html/issues_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/html/library_developer_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/html/library_user_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/html/make_tinyurl.html: *A* unlinked file tools/regression/xsl_reports/xsl/html/summary_developer_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/html/summary_user_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/issues_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/library_developer_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/library_user_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/make_tinyurl.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/summary_developer_legend.html: *A* unlinked file tools/regression/xsl_reports/xsl/v2/html/summary_user_legend.html: *A* unlinked file |serialization| libs/serialization/src/basic_xml_grammar.ipp: *U* unnamed namespace at line 43 |test| boost/test/floating_point_comparison.hpp: *U* unnamed namespace at line 206 *U* unnamed namespace at line 228 boost/test/impl/cpp_main.ipp: *U* unnamed namespace at line 42 boost/test/impl/exception_safety.ipp: *U* unnamed namespace at line 400 boost/test/impl/framework.ipp: *U* unnamed namespace at line 199 boost/test/impl/plain_report_formatter.ipp: *U* unnamed namespace at line 45 boost/test/impl/progress_monitor.ipp: *U* unnamed namespace at line 38 boost/test/impl/results_collector.ipp: *U* unnamed namespace at line 106 boost/test/impl/results_reporter.ipp: *U* unnamed namespace at line 48 boost/test/impl/unit_test_log.ipp: *U* unnamed namespace at line 79 boost/test/impl/unit_test_monitor.ipp: *U* unnamed namespace at line 35 boost/test/impl/unit_test_parameters.ipp: *U* unnamed namespace at line 50 boost/test/results_collector.hpp: *U* unnamed namespace at line 40 boost/test/test_tools.hpp: *U* unnamed namespace at line 262 boost/test/utils/iterator/token_iterator.hpp: *U* unnamed namespace at line 166 boost/test/utils/named_params.hpp: *U* unnamed namespace at line 216 boost/test/utils/runtime/cla/dual_name_parameter.ipp: *U* unnamed namespace at line 43 boost/test/utils/runtime/cla/modifier.hpp: *U* unnamed namespace at line 34 boost/test/utils/runtime/env/modifier.hpp: *U* unnamed namespace at line 34 boost/test/utils/runtime/file/config_file.hpp: *U* unnamed namespace at line 169 *U* unnamed namespace at line 64 *U* unnamed namespace at line 74 boost/test/utils/runtime/file/config_file_iterator.hpp: *U* unnamed namespace at line 68 boost/test/utils/trivial_singleton.hpp: *U* unnamed namespace at line 52 *U* unnamed namespace at line 61 libs/test/build/msvc71_proj/config_file_iterator_test.vcproj: *N* name exceeds 31 characters libs/test/doc/components/prg_exec_monitor/compilation.html: *A* broken link: ../../../build/Jamfile libs/test/doc/components/prg_exec_monitor/index.html: *A* broken link: ../../../../../boost/test/cpp_main.hpp libs/test/doc/components/test_tools/index.html: *A* broken link: ../../tests/boost_check_equal_str.html libs/test/doc/components/test_tools/reference/BOOST_CHECK_CLOSE.html: *A* broken link: BOOST_CHECK_CLOSE_FRACTION.html libs/test/doc/components/test_tools/reference/BOOST_CHECK_MESSAGE.html: *A* broken link: BOOST_MESSAGE.html libs/test/doc/components/test_tools/reference/BOOST_CHECK_SMALL.html: *A* broken link: BOOST_CHECK_CLOSE_FRACTION.html libs/test/doc/components/test_tools/reference/tools_list.html: *A* broken link: ../../btl1.gif *A* broken link: BOOST_CHECK_CLOSE_FRACTION.html libs/test/doc/components/utf/compilation.html: *A* broken link: ../../../build/Jamfile libs/test/doc/components/utf/components/index.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/components/test_case/abstract_interface.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/auto_register_facility.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/boost_function_tc.html: *A* broken link: ../../../../../../../boost/test/unit_test_suite_ex.hpp *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/class_tc.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/function_tc.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/index.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/param_boost_function_tc.html: *A* broken link: ../../../../../../../boost/test/unit_test_suite_ex.hpp *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/param_class_tc.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/param_function_tc.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_case/tc_template.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_log/custom_log_formatter.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_log/index.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_result/index.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/components/test_suite/index.html: *A* broken link: ../../../btl1.gif libs/test/doc/components/utf/index.html: *A* broken link: getting_started/index.html libs/test/doc/components/utf/parameters/build_info.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/catch_system_errors.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/detect_memory_leaks.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/index.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/log_format.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/log_level.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/no_result_code.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/output_format.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/random.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/report_format.html: *A* broken link: ../../../../../LICENSE_1_0.txt *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/report_level.html: *A* broken link: ../../btl1.gif libs/test/doc/components/utf/parameters/show_progress.html: *A* broken link: ../../btl1.gif libs/test/doc/examples/unit_test_example1.html: *A* broken link: ../../example/unit_test_example1.cpp libs/test/doc/examples/unit_test_example2.html: *A* broken link: ../../example/unit_test_example2.cpp libs/test/doc/examples/unit_test_example3.html: *A* broken link: ../../example/unit_test_example3.cpp libs/test/doc/examples/unit_test_example4.html: *A* broken link: ../../example/unit_test_example4.cpp libs/test/doc/examples/unit_test_example5.html: *A* broken link: ../../example/unit_test_example5.cpp *A* broken link: ../../example/unit_test_example5.input libs/test/doc/tests/auto_unit_test_test.html: *A* broken link: ../../test/auto_unit_test_test.cpp libs/test/doc/tests/auto_unit_test_test_mult.html: *A* broken link: ../../test/auto_unit_test_test_mult1.cpp *A* broken link: ../../test/auto_unit_test_test_mult2.cpp libs/test/doc/tests/unit_test_suite_ex_test.html: *A* broken link: ../../test/unit_test_suite_ex_test.cpp libs/test/doc/tutorials/hello_the_testing_world.html: *A* broken link: ../../../../../LICENSE_1_0.txt *A* broken link: ../execution_monitor/index.html libs/test/doc/tutorials/new_year_resolution.html: *A* broken link: ../../../../../../LICENSE_1_0.txt |type_traits| libs/type_traits/cxx_type_traits.htm: *A* unlinked file

1 0

boost-docs list down?
by Matias Capeletto 11 Jun '07

11 Jun '07

I am not receiving posts from the boost-docs list from two days ago, it seems to be down. Can anybody take a look? Thanks Matias

3 3

[system] patch for error_code.cpp
by Boris Gubenko 11 Jun '07

11 Jun '07

Module libs/system/src/error_code.cpp has conditionalization which ensures that strerror_r() is called only on the platforms where this function is availble. The condition does not take into account Tru64 and VMS and, as a result, the module does not compile on these platforms. Attached patch fixes this problem. Ok to commit to HEAD? Thanks, Boris

1 1

bug found in boost::intrusive::detail::ihashtable<...>::rehash
by Yoni Lavi 11 Jun '07

11 Jun '07

I hope I'm posting it to the correct newsgroup. I wanted to mail this directly to the developer of boost::intrusive (Ion Gazta?aga), but I can't find his e-mail anywhere, or any way to directly report the bug. I am not a boost developer myself. Position of a contained element modulo m (new_buckets_len) is known to remain equal to its previous position modulo n (old_buckets_len) ONLY IF m < n AND n %m == 0, but the latter condition isn't being checked. The code assumes it can skip rehashing of m first buckets if m < n: //If we are shrinking the bucket array, just rehash the last nodes if(same_buffer && (old_buckets_len > new_buckets_len)){ n = new_buckets_len; } //Iterate through nodes for(; n < old_buckets_len; ++n){ ... } Also, the following comment: //If this is a buffer expansion don't move if it's not necessary if(same_buffer && new_n == n){ ++before_i; ... Needs to be changed simply to "don't move if it's not necessary" - it needs to be checked/may happen when shrinking the hash table also. Here's a suggested bugfix: - check if (old_bucket_len > new_buckets_len && 0 == old_bucket_len % new_buckets_len). - if true, i'th bucket needs to store the contents of all previous buckets at positions i + j*new_buckets_len for j in [0, old_buckets_len / new_buckets_len) spliced one after another. There is no need to check on a per-element basis. I believe this optimization should be done even when the rehash isn't in-place. A special case where this optimization is useful is when using power-of-2 num_buckets. - if false, buckets have to be re-scanned, using the existing code in the body of the existing loop, but starting from n = 0 always: for(; n < old_buckets_len; ++n){ ... } Please correct me if I'm wrong on this.. but just to be sure, I made a test program that verified rehash() can break the container's invariants when shrinking the container in place, which was indeed found to be the case. As a totally unrelated note, I don't understand why iunordered_set doesn't support a default constructor (so that buckets = 0 and num_buckets = 0, given the existence of a rehash() function. I see it like the reset() function in a scoped_ptr. If a "reset" style function is supported, the class shouldn't force you to initialize it at construction time. Then again, it's a matter of taste in design more than anything...

2 1

[fusion] transitive const / mutable views?
by Martin Weiser 11 Jun '07

11 Jun '07

Dear all, while I appreciate boost::fusion very much, I've recently stumbled across a point that I don't understand. It appears to me that - const is transitive, i.e. dereferencing an iterator into a const vector yields a reference to a const element and - transformation algorithms work on const sequences. Combined, these two imply that an in-place modification of a sequence through a view generated by a transformation algorithm is impossible. While this is perfectly reasonable for transform, it seems quite restrictive for filter or erase. After all, filter_view accepts a non-const input sequence. My point is probably best illustrated by the following code snippet: #include <iostream> #include <boost/type_traits/is_same.hpp> #include <boost/fusion/algorithm/iteration/for_each.hpp> #include <boost/fusion/algorithm/transformation/filter.hpp> #include <boost/fusion/sequence/container/vector.hpp> #include <boost/fusion/sequence/io/out.hpp> #include <boost/fusion/sequence/view/filter_view.hpp> struct SetToZero { template <typename X> void operator()(X& x) const { x = 0; } }; int main(void) { using namespace boost; using namespace boost::fusion; typedef vector<int,char,double,int> Vec; Vec vec(3,'.',1.0,4); std::cout << vec << '\n'; // This does not compile (error: assignment of read-only reference »x«) for_each(filter<int>(vec),SetToZero()); // This works... for_each(filter_view<Vec,is_same<mpl::_,int> >(vec), SetToZero()); std::cout << vec << '\n'; return 0; } So I'm wondering if this is intended? If yes, how do I best modify (part of) sequences defined by transformation algorithms? Thanks, Martin -- Dr. Martin Weiser web: www.zib.de/weiser Zuse Institute Berlin mail: weiser(a)zib.de Numerical Analysis and Modelling pgp key: www.zib.de/weiser/weiser.asc

1 0

Where are the regression reports?
by Eric Niebler 11 Jun '07

11 Jun '07

I haven't seen a regression report since June 3. Why? -- Eric Niebler Boost Consulting www.boost-consulting.com

1 1

[bind] more placeholders?
by Pavel Vozenilek 10 Jun '07

10 Jun '07

The file <boost/bind/placeholders.hpp> now contains nine of them, from _1 to _9. Would it be possible to increase the number, perhaps conditionally? I run into the problem with Luabind (Boost.Python like wrapper for Lua). In Lua it is possible to return multiple values like: result1, res2, res3 = foo(parameter1, par2, par3); Luabind internal machinery in this case uses 6 placeholders. /Pavel

1 0

Re: [boost] [multi_index] use cases for iterator_to()
by Paul Rose 10 Jun '07

10 Jun '07

>> I guess I'l finally include it, but it'd be so much nicer if we could find >> a conclusive use case. I've also thought about legacy interfaces >> requiring raw pointers, if only we could come up with a simple >> specific scenario on that line. Hi, Occasionally I want this feature, but admittedly for a fairly weak reason: Checking for NULL: if(p) ... is more convenient than checking for end: if(i == c.end()) ... If most of the time I need only to check for NULL or dereference then a pointer is more convenient. When I occasionally need to erase() then I can use iterator_to... Paul Rose

2 1

bug found in boost::intrusive::detail::ihashtable<...>::rehash
by Yoni Lavi 10 Jun '07

10 Jun '07

NOTE: I cross-posted this through gmane.comp.lib.boost.devel, I have no idea which is the correct group to post in - I rather have the gmane copy deleted if I should have posted it in here. I also made a minor correction to my post. Anyway... Position of a contained element modulo m (new_buckets_len) is known to remain equal to its previous position modulo n (old_buckets_len) ONLY IF m < n AND n %m == 0, but the latter condition isn't being checked. The code assumes it can skip rehashing of m first buckets if m < n: //If we are shrinking the bucket array, just rehash the last nodes if(same_buffer && (old_buckets_len > new_buckets_len)){ n = new_buckets_len; } //Iterate through nodes for(; n < old_buckets_len; ++n){ ... } Also, the following comment: //If this is a buffer expansion don't move if it's not necessary if(same_buffer && new_n == n){ ++before_i; ... Needs to be changed simply to "don't move if it's not necessary" - it needs to be checked/may happen when shrinking the hash table also. Here's a suggested bugfix: - check if (old_bucket_len > new_buckets_len && 0 == old_bucket_len % new_buckets_len). - if true, i'th bucket needs to store the contents of all previous buckets at positions i + j*new_buckets_len for j in [0, old_buckets_len / new_buckets_len) spliced one after another. There is no need to check on a per-element basis. I believe this optimization should be done even when the rehash isn't in-place. A special case where this optimization is useful is when using power-of-2 num_buckets. - if false, buckets have to be re-scanned, using the existing code in the body of the existing loop, but starting from n = 0 always: for(; n < old_buckets_len; ++n){ ... } Please correct me if I'm wrong on this.. but just to be sure, I made a test program that verified rehash() can break the container's invariants when shrinking the container's bucket array in place, which was indeed found to be the case. ____________________________________________________________________________________ Don't pick lemons. See all the new 2007 cars at Yahoo! Autos. http://autos.yahoo.com/new_cars.html

1 0

[bimap] relation information (was: [multi_index] User-defined indices)
by Matias Capeletto 10 Jun '07

10 Jun '07

On 6/10/07, Andrey Semashev <andysem(a)mail.ru> wrote: Hi andrey > >> No, bimap doesn't seem to suit me since the container holds not only > >> the wildcard and int. > > > You can use it anyway :) > > The last version of the lib includes a new feature called: relations > > information. If you do not need indexes over that information you can > > include it in bimap: > > > bimap< set_of<string, WildcardOrder>, unordered_set_of<int>, > > with_info< string > > > > > And then: > > > bm.left.find("ab*c")->info = "info"; // or... > > bm.left.info_at("ab*c") = "info"; > bm.right.find(3)->>info = "info"; // or... bm.right.info_at(3) = "info"; > > I guess, if I need to put more than one value as info, I have to make > a struct of them? > > struct MyInfo > { > string Data1; > int Data2; > bool Data3; > }; > > bimap< set_of<string, WildcardOrder>, unordered_set_of<int>, > with_info< MyInfo > > Exactly. > What is the value_type of the container then? Why is that important to you? Even when you do not use additional information the value type of the container is an implementation detail. But that's is why value_type exists. You do no have to worry about the implementation type. > What are the names of > the two keys and info in it? In this bimap: bm::left_map::value_type (signature-compatible to std::pair<left_type,right_type>) ---------------------------------- first : is the left element that acts as the key second : is the right element that acts as the data info : is the information, that do not participate in the map bm::right_map::value_type (signature-compatible to std::pair<right_type,left_type>) ---------------------------------- first : is the right element that acts as the key second : is the left element that acts as the data info : is the information, that do not participate in the map bm::value_type ---------------------------------- first : is the right element second : is the left element info : is the information, that do not participate in the map > As I'm not hoping it is more or less > relevant to their real semantic, I don't think it's the best choice > for me, because this value_type is exposed to other components of the > application. I do not see a problem here. > Obscure names of value_type members, They are not obscure. > an dependency on bimap are very undesirable. Ok that is your call :) If you want to give your users better names you can use tags: struct wildcard_string {}; struct id {}; struct fancy_info_name {}; typedef bimap< set_of< tagged<string, wildcard_string>, WildcardOrder>, unordered_set_of< tagged< int, id > >, with_info< tagged< MyInfo, fancy_info_name > > bm_type; bm_type bm; ... cout << bm.by<id>.find(2)->get<wilcard_string>() I am not saying that a bimap is better than a multi_index_container here. I just want to point out that you "can" use it if you want. Best Regards Matias Bimap online docs: http://tinyurl.com/22sja5

2 4