November 2007 - Boost - lists.preview.boost.org

Boost Review Wizard Report for November 2007
by Ronald Garcia 21 Nov '07

21 Nov '07

============================================== Review Wizard Status Report for November 2007 ============================================== News ==== November 7, 2007 - Exception Library Accepted Announcement: http://lists.boost.org/boost-users/2007/11/31912.php We need experienced review managers. Please take a look at the list of libraries in need of managers and check out their descriptions. In general review managers are active boost participants or library contributors. If you can serve as review manager for any of them, email Ron Garcia or John Phillips, "garcia at cs dot indiana dot edu" and "jphillip at capital dot edu" respectively. A link to this report will be posted to www.boost.org. If you would like us to make any modifications or additions to this report before we do that, please email Ron or John. If you're library author and plan on submitting a library for review in the next 3-6 months, send Ron or John a short description of your library and we'll add it to the Libraries Under Construction below. We know that there are many libraries that are near completion, but we have hard time keeping track all of them. Please keep us informed about your progress. Review Queue ============ * Finite State Machines * Floating Point Utilities * Switch * Property Map (fast-track) * Graph (fast-track) * Forward (fast-track) * Singleton (fast-track) * Factory (fast-track) * Lexer * Thread-Safe Signals * Logging * Flyweight * Unordered Containers -------------------- Finite State Machines --------------------- :Author: Andrey Semashev :Review Manager: Martin Vuille :Download: `Boost Sandbox Vault <http://tinyurl.com/yjozfn>`__ :Description: The Boost.FSM library is an implementation of FSM (stands for Finite State Machine) programming concept. The main goals of the library are: * Simplicity. It should be very simple to create state machines using this library. * Performance. The state machine infrastructure should not be very time and memory-consuming in order to be applicable in more use cases. * Extensibility. A developer may want to add more states to an existing state machine. A developer should also be able to specify additional transitions and events for the machine with minimum modifications to the existing code. Floating Point Utilities ------------------------ :Author: Johan Råde :Review Manager: Need Volunteer :Download: `Boost Sandbox Vault <http://boost-consulting.com/vault/index.php? directory=Math%20-%20Numerics>`__ :Description: The Floating Point Utilities library contains the following: * Floating point number classification functions: fpclassify, isfinite, isinf, isnan, isnormal (Follows TR1) * Sign bit functions: signbit, copysign, changesign (Follows TR1) * Facets that format and parse infinity and NaN according to the C99 standard (These can be used for portable handling of infinity and NaN in text streams). Switch ------ :Author: Steven Watanabe :Review Manager: Need Volunteer :Download: `Boost Sandbox Vault <http://boost-consulting.com/vault/index.php? action=downloadfile&filename=mcs_units_v0.7.1.zip&directory=Units>`__ :Description: The built in C/C++ switch statement is very efficient. Unfortunately, unlike a chained if/else construct there is no easy way to use it when the number of cases depends on a template parameter. The Switch library addresses this issue. Property Map (fast-track) ------------------------- :Author: Andrew Sutton :Review Manager: Jeremy Siek :Download: http://svn.boost.org/svn/boost/sandbox/graph-v2 :Description: A number of additions and modifications to the Property Map Library, including: * A constant-valued property map, useful for naturally unweighted graphs. * A noop-writing property map, useful when you have to provide an argument, but just don't care about the output. * See `ChangeLog <http://svn.boost.org/trac/boost/browser/sandbox/ graph-v2/libs/property_map/ChangeLog>`__ for details. Graph (fast-track) ------------------ :Author: Andrew Sutton :Review Manager: Jeremy Siek :Download: http://svn.boost.org/svn/boost/sandbox/graph-v2 :Description: A number of additions and modifications to the Graph Library, including: * Two new graph classes (undirected and directed) which are intended to make the library more approachable for new developers * A suite of graph measures including degree and closeness centrality, mean geodesic distance, eccentricity, and clustering coefficients. * An algorithm for visiting all cycles in a directed graph (Tiernan's from 1970ish). It works for undirected graphs too, but reports cycles twice (one for each direction). * An algorithm for visiting all the cliques a graph (Bron&Kerbosch). Works for both directed and undirected. * Derived graph measures radius and diameter (from eccentricity) and girth and circumference (from Tiernan), and clique number (from Bron&Kerbosch). * An exterior_property class that helps hides some of the weirdness with exterior properties. * runtime and compile-time tests for the new algorithms. * a substantial amount of documentation * Graph cores, implemented by David Gleich (@Stanford University) * Deterministic graph generators - capable of creating or inducing specific types of graphs over a vertex set (e.g., star graph, wheel graph, prism graph, etc). There are several other specific types that could be added to this, but I haven't had the time just yet. Forward (fast-track) -------------------- :Author: Tobias Schwinger :Review Manager: John Torjo :Download: http://boost-consulting.com/vault/index.php? &direction=0&order=&directory=X-Files :Description: A brute-force solution to the forwarding problem. Singleton (fast-track) ---------------------- :Author: Tobias Schwinger :Review Manager: John Torjo :Download: http://boost-consulting.com/vault/index.php? &direction=0&order=&directory=X-Files :Description: Three thread-safe Singleton templates with an easy-to-use interface. Factory (fast-track) -------------------- :Author: Tobias Schwinger :Review Manager: John Torjo :Download: http://boost-consulting.com/vault/index.php? &direction=0&order=&directory=X-Files :Description: Generic factories. Lexer ----- :Author: Ben Hanson :Review Manager: Need Volunteer :Download: http://boost-consulting.com/vault/index.php? action=downloadfile&filename=boost.lexer.zip&directory=Strings%20-% 20Text%20Processing& :Description: A programmable lexical analyser generator inspired by 'flex'. Like flex, it is programmed by the use of regular expressions and outputs a state machine as a number of DFAs utilising equivalence classes for compression. Thread-Safe Signals ------------------- :Author: Frank Hess :Review Manager: Need Volunteer :Download: http://www.boost-consulting.com/vault/index.php? &direction=0&order=&directory=thread_safe_signals :Description: A thread-safe implementation of Boost.signals that has some interface changes to accommodate thread safety, mostly with respect to automatic connection management. Logging ------- :Author: John Torjo :Review Manager: Need Volunteer :Download: http://torjo.com/log2/ :Description: Used properly, logging is a very powerful tool. Besides aiding debugging/testing, it can also show you how your application is used. The Boost Logging Library allows just for that, supporting a lot of scenarios, ranging from very simple (dumping all to one destination), to very complex (multiple logs, some enabled/some not, levels, etc). It features a very simple and flexible interface, efficient filtering of messages, thread-safety, formatters and destinations, easy manipulation of logs, finding the best logger/filter classes based on your application's needs, you can define your own macros and much more! Flyweight --------- :Author: Joaquín M López Muñoz :Review Manager: Need Volunteer :Download: http://www.boost-consulting.com/vault/index.php? action=downloadfile&filename=flyweight.zip&directory=Patterns :Description: Flyweights are small-sized handle classes granting constant access to shared common data, thus allowing for the management of large amounts of entities within reasonable memory limits. Boost.Flyweight makes it easy to use this common programming idiom by providing the class template flyweight<T>, which acts as a drop-in replacement for const T. Unordered Containers -------------------- :Author: Daniel James :Review Manager: Need Volunteer :Download: http://www.boost-consulting.com/vault/index.php? action=downloadfile&filename=unordered.zip&directory=Containers :Description: An implementation of the unordered containers specified in TR1, with most of the changes from the recent draft standards. Libraries under development =========================== Dataflow -------- :Author: Stjepan Rajko :Description: The Dataflow library provides generic support for data producers, consumers, and connections between the two. It also provides layers for several specific dataflow mechanisms, namely Boost.Signals, VTK data/display pipelines, and plain pointers. The Dataflow library came out of the Signal Network GSoC project, mentored by Doug Gregor. :Status: I am polishing the Dataflow library for submission, and am expecting to add it to the review queue in the next couple of months. I am currently ironing out some faults in the design of the library, filling in missing features, and testing it on / adapting it to different dataflow mechanisms (currently VTK and soon Boost.Iostreams). As soon as I'm pretty sure that things are going the right way, I'll submit this to the review queue while I do the finishing touches. Constrained Value ----------------- :Author: Robert Kawulak :Download: http://rk.go.pl/f/constrained_value.zip http://rk.go.pl/r/constrained_value (Documentation) :Description: The Constrained Value library contains class templates useful for creating constrained objects. The simplest example of a constrained object is hour. The only valid values for an hour within a day are integers from the range [0, 23]. With this library, you can create a variable which behaves exactly like int, but does not allow for assignment of values which do not belong to the allowed range. The library doesn't focus only on constrained objects that hold a value belonging to a specified range (i.e., bounded objects). Virtually any constraint can be imposed using appropriate predicate. You can specify what happens in case of assignment of an invalid value, e.g. an exception may be thrown or the value may be adjusted to meet the constraint criterions. :Status: I'm planning to finish it in 1-2 months. Please let us know of any libraries you are currently developing that you intend to submit for review.

10 13

[numeric_traits] How to define integer difference_types
by Sebastian Theophil 21 Nov '07

21 Nov '07

Dear all, Yesterday I had run into issues compiling boost::counting_iterator under Visual Studio. The reason is integer_traits<Integer>::difference_type which is used as a default difference_type for the counting_iterator. This has already been discussed here: http://groups.google.de/group/boost-list/browse_thread/thread/79b5a5bb13 f37576/7773443825609bb1 But I don't think the discussion went far enough: - First of all, this is not just an issue concerning compiler warnings about 64 bit conversion issues (/Wp64) as Orhun suggested in the above thread. - Trying to convert __int64 to int is a level 1 compiler warning (http://msdn2.microsoft.com/en-us/library/aa984663(VS.71).aspx) and in my opinion, the compiler should warn about implicit down-casts. - David suggested in the thread, that (quote) "These 'integer size warnings' are nothing more than a nuisance in most generic code. IMO they should be turned off on the command line." To this, I strongly object. The integer size warning is correct and may be a simple programming error that a developer should at least look at. It is the supposed greater generality of defining integer_traits<int> to be __int64 that causes a problem here. - Defining integer_traits<int>::difference_type as __int64 is first of all pretty inconsistent - integer_traits<__int64>::difference_type of course also is __int64 for lack of a larger type - *Most importantly, no one defines std::ptrdiff_t as __int64 although the same reasoning would apply* Integer values may overflow. Even __int64 values may overflow. There may be cases where you need (int - int) to be a __int64 just like (int * int) should be __int64. In general however, I would like int + (int - int) to be an int again. As a developer I know this can overflow but int = int + __int64(int - int) will overflow too and if I do __int64 + (__int64 - __int64) then this may overflow too! (Besides, in the context of iterators it is extremely weird for an iterator on ints to have an operation advance(__int64).) There's no general solution preventing ints from overflowing. The numeric_traits implementation forces me to disable perfectly valid static type checks for no apparent value at all. Let me take care of overflowing problems where *I* think they may occur. Let the compiler do proper type checks. Regards and I'm looking forward to comments on this, Sebastian -- Sebastian Theophil . stheophil(a)think-cell.com Software Engineer think-cell Software GmbH . Invalidenstr. 34 . 10115 Berlin, Germany http://www.think-cell.com . phone +49-30-666473-10 . fax +49-30-666473-19 Geschaftsfuhrer: Dr. Markus Hannebauer, Arno Schoedl . Amtsgericht Charlottenburg, HRB 85229

1 0

Forcing rebuild of tests when headers change
by Anthony Williams 21 Nov '07

21 Nov '07

I've changed the headers of boost.thread several times recently, and it has become apparent that those testers doing incremental runs are not picking up the changes, and are thus not rebuilding the tests. This is especially problematic when the library files change too --- the libraries are rebuilt, but the tests are not, which leads to linker errors or crashes depending on the nature of the mismatch. Is there a way to make boost build aware of the headers belonging to a specific library, so changes to those headers will force recompilation of anything dependent on that library? Thanks, Anthony

6 7

Re: [boost] [filesytem].[aix].[1_34_0] related issue
by Gaurav Jain-SJR 21 Nov '07

21 Nov '07

> > > > > > "Gaurav Jain-SJR" <Gaurav.Jain <at> iflexsolutions.com> wrote on > > 11/19/2007 > > > > > > Thanks a lot this solved my problem. I am also facing another problem > > while building boost. If I build it with the following command : > > > > > > > > $bjam --v2 install toolset=vacpp --prefix=$HOME/1_34_0/32 > > --builddir=$HOME/tmp variant=release runtime-link=static,shared > > threading=single,multi --without-python Jamfile.v2 > > > > > > > > vacpp.compile.c++ > > bin.v2/libs/wave/build/vacpp/release/link-static/runtime-link-static > > > > /instantiate_cpp_grammar.o > > > > > > > > Build seems to me hanging after this. Though I am not facing a same > > problem if I specify the every target library to be built with bjam > > command. > > > > > > Frequently this type of problem is as a result of bjam's memory usage, try > "export LDR_CNTRL=MAXDATA=0x80000000" in your environment. This should > make the problem go away if the issue you are having is indeed related to > bjams memory use. I am not afraid to say its need more than this. After exporting the above mentioned environment variable I issued a fresh build, but it lead me to the same result. I am using the bjam 3.1.14 version for the build purpose. Should I upgrade to bjam 3.1.15? I come to know from boost mailing list a lot of patches and new features are gone into newest BJAM version. One such patch is I found through the mailing list discussion is Set MAXLINE in jam.h to 23k bytes for AIX. The piecemeal archive action was broken with the default MAXLINE of 102400. Because the AIX shell uses some of the 24k default buffer size for its own use, I reduced it to 23k. -- Noel B.: http://lists.boost.org/boost-build/2007/09/17386.php Is this causing the problem? BTW what's LDR_CNTRL for? :-) Regards, Gaurav Jain "There are two ways to write error-free programs; only the third works." -Alan J. Perlis DISCLAIMER: This message contains privileged and confidential information and is intended only for an individual named. If you are not the intended recipient, you should not disseminate, distribute, store, print, copy or deliver this message. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete or contain viruses. The sender, therefore, does not accept liability for any errors or omissions in the contents of this message which arise as a result of e-mail transmission. If verification is required, please request a hard-copy version.

1 0

[range_ex] Range Update Proposal now in the vault
by Neil Groves 20 Nov '07

20 Nov '07

An update to the Boost Range library that includes range adaptors has been placed in the root of the Vault. I am inviting feedback to improve the code before it is subsequently formally reviewed. It should be noted that documentation and the unit test harnesses will be submitted within the next couple of weeks. Thank you for your time and consideration. Best Wishes, Neil Groves

1 0

review process was [range][algorithm] Looking for STL adaptations
by brass goowy 20 Nov '07

20 Nov '07

> Dean Michael Berris wrote: > > Could this simple header library be put into the Fast Track review queue? > I may have missed it, but has a "Fast Track" process been approved by the moderators? The term reminds me of people who force their way into the US illegally. I think the first come, first serve approach is better as it respects those who wait their turn. Brian Wood Ebenezer Enterprises

2 1

[inplace_factory] Question about nplace_factory
by Corrado Zoccolo 20 Nov '07

20 Nov '07

Hello, I'm trying to use inplace factories on a class that performs small object optimization. If the user code pass me an inplace factory and the optimization can be done, everything is fine, but how should I handle the case in which the optimization cannot be done? Can inplace factory export an other method, that will invoke new T() with the appropriate parameters and return the pointer to it? I can't see how to do this transparently without inplace factory support, because I don't know a way to call the correct operator new if it is defined in the class ( T::operator new(sizeof(T)) works only if T declares the operator ). Corrado -- __________________________________________________________________________ dott. Corrado Zoccolo mailto:zoccolo@di.unipi.it PhD - Department of Computer Science - University of Pisa, Italy -------------------------------------------------------------------------- The self-confidence of a warrior is not the self-confidence of the average man. The average man seeks certainty in the eyes of the onlooker and calls that self-confidence. The warrior seeks impeccability in his own eyes and calls that humbleness. Tales of Power - C. Castaneda

5 11

[enable_if] how to make sure the user supplies the correct number of template arguments
by Stjepan Rajko 20 Nov '07

20 Nov '07

In using enable_if, I came to use a simple technique to detect whether the user is supplying the correct number of (non-Enable) arguments which I think might be useful to add to the enable_if documentation. This addresses the problem which occurs when the user erroneously supplies an extra template argument (that ends up as the Enable template parameter), and it either goes undetected in the compilation or is hard to trace. I get this especially when I decide to reduce the number of template parameters to a class template which uses enable_if. E.g. // ---- WE START WITH template <typename T1, typename T2, typename Enable=void> class enabled_class { // base template implementation }; // user code enabled_class<X, Y> o; but then we change our mind and decide that T1 should not be in the template parameter list for enabled_class: // ----- WE END UP WITH template <typename T2, typename Enable=void> class enabled_class { // base template implementation }; // we forget to change the user code enabled_class<X, Y> o; // this should now be enabled_class<Y> o, but we might not even get an error! The solution I found goes like this: // ---- A PARTIAL SOLUTION template <typename T2, typename Enable=void> class enabled_class { BOOST_MPL_ASSERT((is_same<Enable, void>)); // default implementation }; // we forget to change the user code enabled_class<X, Y> o; // ERROR (except in the unfortunate case when Y is void) Is this a reasonable solution to this problem? Would it be useful to add it to the docs? I suppose this is a problem common to any time default template arguments are used, but I ran into it a lot with enable_if. Having something in the docs that addresses it would have helped me a lot. Stjepan

1 1

Why use numeric_traits<int>::difference_type in counting_iterator? Throws warnings that I'd rather treat as errors
by Sebastian Theophil 20 Nov '07

20 Nov '07

Hello everybody, 1) I tried to use a simple boost::counting_iterator<size_t> but ran into the following compiler error: error C4244: '+=' : conversion from '__int64' to 'size_t', possible loss of data void advance(typename super_t::difference_type n) { BOOST_ITERATOR_ADAPTOR_ASSERT_TRAVERSAL(random_access_traversal_tag) m_iterator += n; } Obviously, the default difference_type for size_t (and for int by the way) is __int64 as defined by template <class T> struct numeric_difference { typedef typename boost::detail::numeric_traits<T>::difference_type type; }; (counting_iterator.hpp:73 called from counting_iterator_base::ctor) I think I understand why __int64 may be a sensible choice for calculations with size_t (or int) values. You prevent some overflows this way. - Why does the counting_iterator insist on the numeric_traits difference type? Why allowing an advance operation with an argument possibly larger than m_iterator itself? 2) If I manually instantiate the counting_iterator with difference_type size_t (or int) as boost::counting_iterator<size_t, boost::use_default, size_t> it will still fail at location // A policy for wrapped numbers template <class Difference, class Incrementable1, class Incrementable2> struct number_distance { static Difference distance(Incrementable1 x, Incrementable2 y) { return numeric_distance(x, y); } }; Instantiated from template <class OtherIncrementable> difference_type distance_to(counting_iterator<OtherIncrementable, CategoryOrTraversal, Difference> const& y) const { typedef typename mpl::if_< detail::is_numeric<Incrementable> , detail::number_distance<difference_type, Incrementable, OtherIncrementable> , detail::iterator_distance<difference_type, Incrementable, OtherIncrementable> >::type d; return d::distance(this->base(), y.base()); } With error message error C4244: 'return' : conversion from '__int64' to '__w64 unsigned int', possible loss of data. Again, numeric_distance insists on using __int64 as a difference type and fails to convert the value properly. I would greatly appreciate any advice on this because using counting_iterator would save me a lot of code. Thanks for everybody's great work, Sebastian -- Sebastian Theophil . stheophil(a)think-cell.com Software Engineer think-cell Software GmbH . Invalidenstr. 34 . 10115 Berlin, Germany http://www.think-cell.com . phone +49-30-666473-10 . fax +49-30-666473-19 Geschaftsfuhrer: Dr. Markus Hannebauer, Arno Schoedl . Amtsgericht Charlottenburg, HRB 85229

1 0

[shared_ptr] explicit shared_ptr
by Sid Sacek 20 Nov '07

20 Nov '07

template<class T> class shared_ptr { template<class Y> explicit shared_ptr( Y * p ): px( p ), pn( p ) // Y must be complete { detail::sp_enable_shared_from_this( pn, p, p ); } }; I've tried really hard to figure out what it means when a constructor is both a template as well as explicit. The keywords have incompatible meanings to me and therefore I fail to make sense of it. If you template the argument, then you're saying you'll accept any type of pointer as long as it will still compile, and the explicit implies no explicit conversions to Y, which shouldn't happen anyway due to the template indicating any argument is valid as long as it complies. I googled for this and researched the C++ Templates book without success. Also, nobody I've asked this about could give me a satisfactory answer, and so I'm hoping somebody on the boost list won't mind enlightening me by providing a few concrete examples where it comes into play. And I apologize for wasting everybody else's time with this. Thanks, -Sid

5 4