October 2012 - Boost - lists.boost.org

[fixed_point] Request for interest in a binary fixed point library
by Vicente J. Botet Escriba 02 Feb '13

02 Feb '13

Hi, the recent discussion on the MultiplePrecission Arithmetic library has show that some people has its ow fixed point library. Is there an interest in a Boost library having as base http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2012/n3352.html? I have started a prototype http://svn.boost.org/svn/boost/sandbox/fixed_point (there is no doc yet but the basic goal is quite close to the N3352 C++ proposal). This prototype should allow you to play a little bit with. If yes, what would you like to be changed, added or improved in n3352? Next follows some design decisions that IMO need to be decided before hand. * Should integers and reals be represented by separated classes? * Should signed and unsigned be represented by separated classes? * Should the library use a specific representation for signed numbers (separated sign, 2-complement? Let the user choose? * Should the library provide arbitrary range and resolution and allocators? * Should the library be open to overflow and rounding or just implement some of the possible policies? and in this case which ones? * Should fixed_point be convertible to/from integer/float/double? * Could the result of an arithmetic operation have more range and resolution than his arguments? * Is there a need for a specific I/O? * is there a need for radix other than 2 (binary)? * Should the library implement the basic functions, or should it imperatively implement the C++11 math functions? Could a first version just forward to the c++11 math functions? * Should the library support just one of the know ways to name a fixed-point, a U(a,b), nQm, ...? Provide some ways to move from one to another? * Could expect the same/better performances respect to hand written code? * What should be the size used by a fixed_point instance? _fast? _least? Should the user be able to decide which approach is better for his needs? * Which should be the namespace? boost? boost/fixed_point? boost/binary_fixed_point? boost/bfp? ... * others you can think of. Please, replay if you are interested, have some experience in the domain, good ideas, .. Best regards, Vicente

6 44

[parameter] std proposal
by Lorenzo Caminiti 07 Jan '13

07 Jan '13

Hello all, Is there a C++ standard proposal for named parameter addition? I am looking for an "N-paper" for named parameters similar to N2081 for concepts, or N1962 for contracts. Thank you. -- Sent from my mobile device Lorenzo

6 8

[config] Proposed BOOST_NOEXCEPT_NOTHROW macro
by Beman Dawes 03 Jan '13

03 Jan '13

The three current BOOST_NOEXCEPT* macro variants miss the case where C++03 code has a "throw()" throw-specifier. Classes that derive from standard library exceptions need this to avoid breaking code on C++03 compilers. I'm not enamored with the BOOST_NOEXCEPT_NOTHROW name, so feel free to offer suggestions for alternate names. Here is the full set of macros: #ifdef BOOST_NO_NOEXCEPT # define BOOST_NOEXCEPT # define BOOST_NOEXCEPT_NOTHROW throw() <---- new macro # define BOOST_NOEXCEPT_IF(Predicate) # define BOOST_NOEXCEPT_EXPR(Expression) false #else # define BOOST_NOEXCEPT noexcept # define BOOST_NOEXCEPT_NOTHROW noexcept <---- new macro # define BOOST_NOEXCEPT_IF(Predicate) noexcept((Predicate)) # define BOOST_NOEXCEPT_EXPR(Expression) noexcept((Expression)) #endif Unless there are objections, I'll commit the changes Wednesday afternoon Eastern US time. --Beman

7 11

Using c++11 features in clang
by Marshall Clow 11 Dec '12

11 Dec '12

In response to a some questions on the boost-users list (and boost bug reports), I wrote up this: http://marshall.calepin.co/llvmclang-and-standard-libraries-on-mac-os-x.html I don't know if this should be a FAQ entry for Boost, or for clang. Suggestions/comments/edits welcome. -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

3 3

Boost.Test updates in trunk: need for (mini) review?
by Gennadiy Rozenal 09 Dec '12

09 Dec '12

Hi, It's been a long while since I merged any changes into boost release and by now there are whole bunch of new features there. Some are small, but some others are big ones such that they can change the way people use the library. I'd like to list them here (excluding bug fixes and removal of deprecated interfaces) and ask what should we do with these. Specifically let me know if you think these require some kind of (mini) review from community, but any other comments are welcome as well. So here we go: I. New testing tool BOOST_CHECKA This tool is based on excellent idea from Kevlin Henney. I chose the name CHECKA for this tool due to the lack of better name, but I am open to suggestions. This testing tool capable of replacing whole bunch of existing other tools like BOOST_CHECK_EQUAL, BOOST_CHECK_GT etc. Usage is the most natural you can wish for: BOOST_CHECKA( var1 - var2 >= 12 ); And the output will include as much information as we can get: error: in "foo": check var1 - var2 >= 12 failed [23-15<12] II. New "data driven test case" subsystem New data driven test case subsystem represent a generalization of parameterized test case and test case template and eventually will replace both. The idea is to allow user to specify an arbitrary (monomorphic or polymorphic) set of samples and run a test case on each of the samples. Samples can have different arity, thus you can have test cases with multiple parameters of different types. For now we support following dataset kinds (aside of generators none of dataset constructions routines performs *any* copying): a) singleton - dataset constructed out of single sample data::make(10) - singleton dataset with integer sample data::make("qwerty") - singleton dataset with string sample b) array - dataset constructed out of C array int a[] = {1,2,3,4,5,6}; data::make(a) - dataset with 6 integer values c) collection - dataset constructed out of C++ forward iterable collection std::vector<double> v{1.2, 2.3, 3.4, 5.6 }; data::make(v) - dataset with 6 double values d) join - dataset constructed by joining 2 datasets of the same type int a[] = {1,2,3}; int b[] = {7,8,9}; data::make(a) + data::make(b) - dataset with 6 integer values e) zip - dataset constructed by zipping 2 datasets of the same size, but not necessarily the same type This dataset has an arity which is sum of argument dataset arities. int a[] = {1,2,3}; char* b[] = {"qwe", "asd", "zxc"}; data::make(a) ^ data::make(b) dataset with 3 samples which are pairs of int and char*. f) grid - dataset constructed by "multiplying" 2 datasets of the same different sizes and types type This dataset has an arity which is sum of argument dataset arities. int a[] = {1,2,3}; char* b[] = {"qwe", "asd"}; double c[] = {1.1, 2.2}; data::make(a) * data::make(b) * data::make(c) dataset with 12 samples which are tuples of int and char* and double. g) xrange generator dataset which produes samples in some range data::xrange( 0., 3., 0.4 ) - dataset with 8 double samples data::xrange<int>((data::begin=9, data::end=15)) - dataset with 6 int samples data::xrange( 1., 7.5 ) - dataset with 7 double samples data::xrange( 5, 0, -1 ) - dataset with 5 int samples h) random - generator dataset with unlimited number of random samples data::random(data::distribution = std::normal_distribution<>(5.,2)) dataset with random double numbers following specified distribution data::random(( data::engine = std::minstd_rand(), data::distribution = std::discrete_distribution<>(), data::seed = 20UL )); dataset with random int numbers following specified distribution While all this interfaces can be used by itself to build complex datasets with various purposes, primary use case it was developped for is new data driven test case interface: BOOST_DATA_TEST_CASE( test_name, dataset, parameter_names... ) Here are couple examples: int samples1[] = {1,2,3}; BOOST_DATA_TEST_CASE( t1, samples1, sample ) { BOOST_CHECKA( foo(sample) > 0 ); } Above test case is going to be executed 3 times with different sample values char* strs[] = {"qwe", "asd", "zxc", "mkl" }; BOOST_DATA_TEST_CASE( t1, data::xrange(4) ^ strs ^ data::random(), intval, str, dblval ) { MyObj obj( dblval, str ); BOOST_CHECKA( obj.goo() == intval ); } Above test case will be executed 4 times with different values of parameters intval, str, and dblval. Polymorphic datasets are still being developed, but should appear soon(ish). III. Auto-registered test unit decorators. Previously is was not possible and/or convenient to assign attributes to the automatically registered test units. To alleviate this I've introduced a notion of test unit decorator. These can be "attached" to any test unit similarly to like it is done in other languages Following decorators already implemented: label - adds labels to a test unit expected_failures - set expected failures for a test unit timeout - sets timeout for a test unit description - sets a test unit description depends_on - sets a test unit dependency enable_if/disable_if - facilitates a test unit status change fixture - assigns fixture to a test units Test unit description is new test unit attribute, which is reported by new list_content command line argument described below. Usage of labels covered below as well. enable_if/disable_if allow new, much more flexible test management. By adding enable_if/disable_if decorators to the test unit you can conditionally select which test units to run at construction time based on some compile time or run-time parameters. And we finally we have suite level fixtures which are set by attaching fixture decorator to a test suite (suite level fixtures executed once per test suite). Attachment of decorator is facilitated by using of BOOST_TEST_DECORATOR. Note that you can use any of '+', '-', '*' symbols to attach decorator (and any number of '*'): BOOST_TEST_DECORATOR( + unittest::fixture<suite_fixture>() ) BOOST_AUTO_TEST_SUITE( my_suite1 ) { } BOOST_TEST_DECORATOR( - unittest::timeout( 100 ) - unittest::expected_failures( 1 ) - unittest::enable_if( 100 < 50 ) ) BOOST_AUTO_TEST_CASE( my_test5 ) BOOST_TEST_DECORATOR( **** unittest::label( "L3" ) **** unittest::description( "suite description" ) **** unittest::depends_on( "my_suite2/my_test7" ) ) BOOST_AUTO_TEST_CASE( my_test11 ) IV Support for "run by label" plus more improvements for filtered runs Previously you only had an ability to filter test unit to execute by name. Now you can attach labels to test units and create collections to run which are located in arbitrary positions in test tree. For example you can have special label for performance test cases, which is attached to all test units which are responsible for testing performance of your various components. Or you might want to collect exception safety tests etc. To filter test units by level you still use --run_test CLA. Labels are dented by @prefix: test.exe --run=@performance You can now repeat --run_test argument to specify multiple conditions: test.exe --run=@performance,@exception_safety --run=prod/subsystem2 In addition run by name/label now recognizes dependencies. So if test unit A depends on test unit B and test unit B is disabled/is not part of current run test unit A will not run as well. Finally you now have an ability to specify "negative" conditions, by prefixing name or label with ! test.exe --run=!@performance This will run all test units which are not labeled with "performance". V. Support for failure context In many scenarios it is desirable to specify an additional information to a failure, but you do not need to see it if run is successful. Thus using regular print statements is undesirable. This becomes especially important when you develop common routine, which performs your testing and invoke it from multiple test units. In this case failure location is no help at all. To alleviate this problem two new tools are introduced: BOOST_TEST_INFO BOOST_TEST_CONTEXT BOOST_TEST_INFO attaches context to next assertion which is executed. For example: BOOST_TEST_INFO( "a=" << a ); BOOST_CHECKA( foo(a) == 0 ); BOOST_CHECK_CONTEXT attaches context to all assertions within a scope: BOOST_CHECK_CONTEXT( "Starting test foo from subset " << ss ) { BOOST_CHECKA( foo(ss, a) == 1 ); BOOST_CHECKA( foo(ss, b) == 2 ); } VI. Two new command line arguments: list_context, wait_for_debugger Using list_context you can see a full or partial test tree of your test module. wait_for_debugger allows to force test module to wait till you attach debugger. VII. Colored output. Using new CLA --color_output you can turn on colored output for error and information messages VIII New production testing tools interface introduced This feature was covered in details in my presentation in BoostCon 2010. The gist if this is an ability to use Boost.Test testing tools interfaces in production (nothing to do with testing) code. User supplied implementation is plugged in IX. Number of smaller improvements: * Notion of framework shutdown. Allows to eliminate some fake memory leaks * Added checkpoints at fixture entry points, test case entry point and test case exit point for auto registered test cases * New FPE portable interfaces introduced and FPE handling is separated from system errors handling. You can detect FPE even if catch_system_error is false * Added an ability to erase registered exception translator * execution_monitor: new interface vexecute - to be used to monitor nullary functions with no result values * test_tree_visitor interface extended to facilitate visitors applying the same action to all test units * execution_monitor use typeid to report "real" exception type if possible. * New ability to redirect leaks report into a file Thank you for your time, Gennadiy

21 66

[Numeric Conversion] Documentation Notes/Questions
by Dave Abrahams 25 Nov '12

25 Nov '12

* First thing to note: the documentation makes this library less approachable than it could be. In order to even begin, one has to bury onesself in terminology (the "Definitions" section). A super-quick "hello, world" example and a very short tutorial at the beginning would help a *lot*. numeric_cast is arguably the most generally-useful interface of the whole library and yet it's not even discussed until the end of the docs. * It's not clear what the purpose of abt(x) is. In expressions, "x" usually means its value. When you introduce a mechanism like this, it would be a good idea to explain why it's needed. Most places, the use of "abstract" just seems to complicate things: "A ulp is a quantity whose abstract magnitude" what's wrong with "A ulp is a quantity whose magnitude" ?? The end of the section on "Range and Precision" begins to explain why you want "abstract," but it takes the reader a long time to get to it. * are next() and prev() always well-defined? Seems to me that an unlimited precision rational number type might not have them. Is that type not numeric? * How do size, width, and density relate to unlimited precision rationals? I imagine: size = infinity width = infinity density = ? * This is slightly ambiguous: "This definition allows for a concise definition of subranged as given in the last section." "final section" would be clearer. A link would help too. * This is unclear: "Abstract intervals are easier to compare and overlap since only boundary values need to be considered." an example would help. How can one kind of thing be "easier to overlap" than another? * Precision has a very non-generic description, being different for types with density 1 and other types. It's hard to see what use such a concept has, and as far as Google can tell it is not used anywhere outside the Definitions section * Description of unsigned types, I think, is wrong: Notice that a numeric type, such as a C++ unsigned type, can define that any V does not overflow by always representing not V itself but the abstract value U = [ V % (abt(h)+1) ], which is always in range. Seems to me that the abstract values represented by an unsigned integer type are not the set of whole numbers, but the set of whole numbers modulo whatever-the-max+1-is. That's a well-defined mathematical abstraction, isn't it? Is there a problem with that, such as multiplication not working properly (I don't think so)? * There seems to be a contradiction here: If V >= abt(l) and V <= abt(h), V is representable (can be represented) in the type T ... Given an abstract value V represented in the type T as v, the roundoff error of the representation is the abstract difference: ( abt(v)-V). If V is not representable in type T, can you write formulae using its representation in type T? Given what I read later, maybe you should have just said "exactly representable" in the first phrase. * s/at most/less than/?: Because a correctly rounded representation is always one of adjacents of the abstract value being represented, the roundoff is guaranteed to be at most 1ulp. * s/cannot be represented/cannot be exactly represented/ ?: The integer number -1 can be exactly represented in Int, Real and Whole, but cannot be represented in Cardinal, yielding negative overflow. * Uh-oh: Floating to Floating conversions are defined only if N is representable; if it is not, the conversion has undefined behavior. in many places you've used "representable" to mean "exactly representable," but I can't believe you mean that here; it would imply most double=>float conversions have undefined behavior. What does this mean, anyway? If an abstract value can be "inexactly representable," aren't all abstract values representable if there's at least one abstract value that's exactly representable? Just convert to that and you're done. I think you need to give some serious thought to the definition of "representable." * s/then/so/: R(X) & R(Y) == R(X), then X->Y is not subranged. Thus, all values of type X are representable in the type Y. (b) Y->X: R(Y) & R(X) == R(Y), then Y->X is not subranged. Thus, all * What's the difference between a policy and a "parameter that defines the conversion" in It can optionally take some policies which can be used to customize its behavior. The Traits parameter is not a policy but the parameter that defines the conversion. ?? -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost

4 4

[#pragma once]
by Sid Sacek 18 Nov '12

18 Nov '12

I've worked with the MSVC, Sun and GCC compilers. Each one of them supports #pragma once. Is there any reason to wrap this pragma within #ifdefs? Are there any compilers out there today that don't support this #pragma? #if defined( _MSC_VER ) && ( _MSC_VER >= 1020 ) # pragma once #endif Could we change this coding pattern so that it no longer looks like MSVC is being selfish? -Sid Sacek

17 20

[iterator] [property_map] Enhancement proposals
by Louis Dionne 12 Nov '12

12 Nov '12

Dear boost, First of all, let me quickly introduce myself as this is my first post here: I am a 20 years old programming enthusiast with interest for metaprogramming, functional programming and library design. I also have a strong will to contribute to open source projects since I love the philosophy behind it and it is a great way to develop new skills and to work with talented people. The object of this message is to collect feedback on several enhancement proposals I have prepared for two existing libraries: [boost.iterator] [boost.iterator.accessor_iterator] The accessor_iterator is an iterator adaptor that returns a reference to a member of its pointee when dereferenced. I see it as another step in the direction of easier iterator composition, which has helped me writting less code in the past. [boost.iterator.chained_output_iterator] The chained_output_iterator is an iterator adaptor allowing to apply a functional transformation to a value before outputing it into another output iterator. It is much like a function_output_iterator that would forward its result to another output iterator. It allows for cool stuff like creating a pipeline of operations without having to create temporary storage. [boost.property_map] [boost.property_map.accessor_property_map] The accessor_property_map is a property map returning a reference to a member of the key it is given. It is much like the accessor_iterator for property maps. [boost.property_map.chained_property_map] The chained_property_map is a tool for property map composition. It takes two property maps and uses a value in the first one as a key in the second one to perform its put, get, and operator[] operations. [boost.property_map.property_traits] I feel like the boost/property_map/property_map.hpp header is a little bit bloated and not quite as modular as it could be. For now, I propose that we move the code dealing with property_traits into the boost/property_map/property_traits.hpp header and that we provide metafunctions to check for adherence to property map types. By providing an include in the boost/property_map/property_map.hpp header, backward compatibility is maintained and we simply increase modularity. All the code is available on github: https://github.com/ldionne/boost-submissions To see how each class can be used, you may look at the unit tests for the corresponding class. If anything is accepted, I will be pleased to write documentation and examples for the concerned feature and to integrate it into the existing documentation. Looking forward to your feedback, Louis Dionne

2 11

filesystem::path passed to GTest's assertion leads to stack overflow because of infinite recursion
by Andrey Upadyshev 10 Nov '12

10 Nov '12

Hi all, I found an interesting bug with 'filesystem::path' used with GTest: https://svn.boost.org/trac/boost/ticket/7564 I just want to discuss the design of 'path' class which iterator contains the 'path' itself. Don't you think it's nice to fix thing? Best regards, Andrey Upadyshev P.S. Sorry if this is wrong mailing list. Just let me know and I'll repost the message to boost-users <boost-users(a)lists.boost.org> mailing list.

2 5

Release branch closed
by Marshall Clow 08 Nov '12

08 Nov '12

Boosters, The release branch is now closed for all commits, except by explicit permission of release managers. -- The Release Managers, Beman Dawes Daniel James Eric Niebler Rene Rivera Vladimir Prus Marshall Clow

4 5