April 2013 - Boost - lists.preview.boost.org

[config/multiprecision/units/general] Do we have a policy for user-defined-literals?
by John Maddock 06 May '13

06 May '13

Folks, I've been experimenting with the new C++11 feature "user-defined-literals", and they're pretty impressive ;-) So far I've been able to write code such as: auto i = 0x1234567890abcdef1234567890abcdef_cppi; and generate value i as a signed 128-bit integer. It's particulary remarkable because: * It actually works ;-) * The value is generated as a constexpr - all evaluation and initialization of the multiprecision integer is done at compile time. * The actual meta-code is remarkably short and concise once you've figured out how on earth to get started! Note however my code is limited to hexadecimal constants, because it can't do compile time radix conversion - that would require a huge meta-program for each constant :-( This is obviously useful to the multiprecision library, and I'm sure for Units as well, but that leaves a couple of questions: 1) We have no config macro for this new feature (I've only tested with GCC, and suspect Clang is the only other compiler with support at present). What should it be called? Would everyone be happy with BOOST_NO_CXX11_USER_DEFINED_LITERALS ? 2) How should libraries handle these user defined suffixes? The essencial problem is that they have to be in current scope at point of use, you can never explicitly qualify them. So I suggest we use: namespace boost{ namespace mylib{ namespace literals{ mytype operator "" _mysuffix(args...); }}} Then users can import the whole namespace easily into current scope right at point of use: int main() { using namespace boost::mylib::literals; boost::mylib::mytype t = 1234_mysuffix; } 3) How should the suffixes be named? There is an obvious possibility for clashes here - for example the units lib would probably want to use _s for seconds, but no doubt other users might use it for strings and such like. We could insist that all such names added to a boost lib are suitably mangled, so "_bu_s" for boost.units.seconds, but I'm not convinced by that. Seems to make the feature much less useful? Many thanks in advance for your thoughts, John.

6 12

[operators] The Dangling Reference Polarization
by Daniel Frey 05 May '13

05 May '13

I see some good discussion going on, thanks to everyone participating! Rest assured that I value both side's contributions, it is important for me (and those agreeing with my perspective) to be able to defend this position and also to keep looking for improvements. If our side can not be defended, it's probably not good enough :) That said, here's what I hope can serve as a first summary, let's first look at it from a technical perspective: If an operator returns an rvalue reference (T&&), it is safe for all use-cases except binding the result to an rvalue- or const-lvalue-reference: T r = a + b + c; // no problem const T r = a + b + c; // no problem T& r = a + b + c; // compile-time error const T& r = a + b + c; // problem (P1) T&& r = a + b + c; // problem (P2) It seems that this is consistent behavior across at least GCC, Clang and VC++. If a compiler shows a different behavior, its likely a compiler-bug. Technically, there is no doubt that the above is a bug, but that doesn't answer the question of whether this is a bug of the user who wrote the above code or if it's a bug of the operator to return an rvalue reference instead of an rvalue. This is where we leave the technical-aspects-land and enter the minefield of opinions, expectations and decisions. I am of the opinion that the above problematic use-cases are a bug on the users side, my line-of-thoughts and my arguments are: a) The above code is explicit, you can spot it in the users code. It's not something that happens silently. b) The above code always looked suspicious to me anyways, it has a code-smell that I just don't like. c) It rely's on the fact that it extends the lifetime of a returned temporary value. The rules of the C++ standard that allow this are quite clear, and hence it's the users responsibility to make sure that there actually is a temporary value whose lifetime can be extended. It has always been an error if the expression returned a reference and not a value. Consider const std::string& oops = std::vector<std::string>(1, "hello, world" ).front(); This compiles are creates a dangling reference, as the standard says that front() returns a reference and not a value. It's not 100% comparable as in this case the standard guarantees a bug, but the main point for C++ has always been which guarantees are given and not about a certain expectation one might have. As an example: static_assert( std::is_empty< std::tuple<> >::value, "Expectation 1 broken" ); static_assert( std::is_empty< std::array< T, 0 > >::value, "Expectation 2 broken" ); // with any valid T The first case is certainly a reasonable expectation and it currently works for all implementations of std::tuple that I know of, but it is not guaranteed! If your code relies on it and it breaks in the future, you can not blame the person who modified std::tuple. The second case is also a reasonable expectation, but it breaks everywhere and AFAIK there is no standard-conforming way to implement std::array so that std::array<T,0> is empty. For the curious, here's the StackOverflow question I asked about it: <http://stackoverflow.com/q/15512827/2073257>. Another example about how to properly handle expectations vs. guarantees: <http://stackoverflow.com/q/14882588/2073257>. d) I fail to see any valid use-case for binding the result of the expression to a reference. Can someone please provide a convincing example of why (P1) or (P2) are needed/useful? Keep in mind that in the context of operator+ (or any other operator in question), we already require that the type T is copy-/moveable and that copy-elision most likely takes place if you use "T r = …" instead of "const T& r = …". e) The alternative needs to prove its claimed/theoretical efficiency in practice. Too often I have heard about "sufficiently smart compilers", but I'm more the prove-it-in-the-real-world type. So summarizing it once again: For a) and b), this is just an observation and my personal opinion. c) means it's important to talk about guarantees, not just about expectations (no matter how reasonable they are). d) poses the question if we loose something if we don't guarantee that (P1) and (P2) are allowed. I currently have several ideas/option of how we could improve Boost.Operators, some of which I already mentioned, some which only recently occurred to me. Before I will discuss these, I'd like to resolve (if possible) the Dangling Reference Polarization and gather more opinions/feedback. So far, I found the discussion both challenging and very helpful, again thanks to everyone participating! Best regards, Daniel

5 13

[gsoc] Boost.Multiprecision Proposal
by Saurav 04 May '13

04 May '13

Hi, I have attached my proposal for the Boost.Multiprecision project. It has been provided in PDF format for better readability. Please have a look and give your feedback. Best regards, Saurav

3 4

Live read-only GIT mirrors of Boost trunk SVN
by Niall Douglas 04 May '13

04 May '13

The live GIT mirrors of Boost trunk SVN have become a bit stale in recent months: https://github.com/ryppl/boost-svn, long time favourite of many, now 4 months stale. https://github.com/Beman/boost-trunk-git-svn is now 2 months stale. https://github.com/alexanderkjeldaas/boost-mirror is fresh, but has no branches nor tags. They all generally use git-svn, which tends to be brittle and not translate commits perfectly unfortunately e.g. SVN eol properties and other such metadata. So I decided to try out a new Subversion <=> Git mirroring tool, SubGit v2.0 EAP (http://subgit.com/eap/) whose author very kindly gave me a free open source licence and the result is at: https://github.com/ned14/boost-trunk All branches and tags, as far as I can tell, are mirrored correctly. I've set a cron job every 24 hours. Theoretically if we had an authors mapping file it would work in the opposite direction too i.e. be read-write. My next task is to live mirror a github repo into Boost Sandbox. Currently SubGit segfaults on this due to the complex SVN properties auto-translation which requires a specially marked up .gitattributes, but the author is very keen to fix the problem. In theory then it ought to just work. Any mistakes in my Boost mirror, or any problems, or even thoughts are gratefully received. Niall

4 6

Specific-Width Floating-Point Typedefs
by Paul A. Bristow 04 May '13

04 May '13

There has recently been some discussion on improving portability of programs using floating-point. Drawing on the benefits from definition of typedefs for int32_t, int64_t, int128_t . we have drawn up a proposal to add similar typedefs for floating-point. Our draft document (in both html and pdf) is at https://svn.boost.org/svn/boost/sandbox/precision/libs/precision/doc/html/i… https://svn.boost.org/svn/boost/sandbox/precision/libs/precision/doc/precis… We hope to propose this to the SG6 Numerics subgroup of WG21 ISO C++ standards body. But would welcome comments from Boosters to help us improve this proposal. Thank you. Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

8 17

GSoC 2013 Bernoulli Numbers
by David Joy 03 May '13

03 May '13

Hi, I'm an Open University student completing my Open degree this summer (CompSci / Pure- & Applied Maths / Classical History) and am very interested in the Bernoulli numbers project for the Google Summer of Code. As well as the OU I've done quite a bit of mathematics elsewhere (Durham and KCL). Enclosed is my (naive) response to your challenge to implement the Akiyama–Tanigawa algorithm. I use the Boost framework for rational numbers, so as not to lose precision (though haven't yet adapted the solution for arbitrary precision numbers). [It would be more aesthetically pleasing if I could write the main algorithm in a STL-style so that it iterates of the vector, rather than explicitly using an index.] I have yet to get to grips with Boost.Test, but the enclosed shell script can generate a list of results, which matches the table found on the Bernoulli Number Page <http://www.bernoulli.org/> out to B_20. All code criticism gratefully accepted, David Bernoulli.cpp <http://boost.2283326.n4.nabble.com/file/n4646219/Bernoulli.cpp> Bernoulli_test.sh <http://boost.2283326.n4.nabble.com/file/n4646219/Bernoulli_test.sh> -- View this message in context: http://boost.2283326.n4.nabble.com/GSoC-2013-Bernoulli-Numbers-tp4646219.ht… Sent from the Boost - Dev mailing list archive at Nabble.com.

2 4

Voronoi benchmark update
by Sebastien Loriot 03 May '13

03 May '13

Hello, I'm involved in the CGAL Project, and saw the Voronoi benchmark. It would be nice if you could change the CGAL code slightly, as it is not as bad as it currently seems. Attached is the diff to the trunk of boost. Concerning the Voronoi diagram of points, a CGAL user, would use the Delaunay_triangulation_2 class. Concerning the Voronoi diagram of segments (that do not intersect) one would make a better choice for number types and traits classes. Also one better first inserts the endpoints and then the segments (I agree, this should definitely go into a member function). Admittedly our examples could be better. Cheers, Sebastien

3 4

[GSOC 2013]Boost.Math Proposal
by Nikhar Agrawal 03 May '13

03 May '13

Hi everyone, I have been in touch with mentors for the project "Boost.Math Bernoulli Numbers and Applications to Special Functions" for the past three weeks. And after regular interactions and assignments and suggestions from them, I have finally posted my proposal. I was a little late because in submitting the proposal because of my end semester examinations. But that now they are over, I have all the free time to completely concentrate on the project. Here's the link to the proposal. http://www.google-melange.com/gsoc/proposal/review/google/gsoc2013/nikhar/1 I'd be really grateful if you guys could review it and suggest improvements. Thanks a lot. :) Nikhar P.S.:- If anyone wants I could also forward the conversations I've had with the mentors until now and I could upload the assignments I've done if anyone's interested.

1 1

[gsoc-2013] proposal for approximate string matching
by Yu Jiang 03 May '13

03 May '13

Hi, I'm a student from Tsinghua University in china and interested in the idea about "approximate string matching". Below is my proposal for this. Could you please view it and give me any kinds of suggestions on it? Thanks! (1) Personal Details * Name: Yu Jiang * College/University: Tsinghua University, China * Course/Major: Computer Science and Technology * Degree Program: M.Sc. * Email: sunlightt07(a)gmail.com * Homepage: do not have one * Availability: - How much time do you plan to spend on your GSoC? Since I have a long summer holiday, I can spend 30 hours a week. - What are you intended start and end dates? I will follow the calendar of GSoC. I have plenty of time from June to September. - What other factors affect your availability (exams, courses, moving, work, etc.)? I have some courses this term which will end before June. 15th. However, these courses don't have heavy workload, I still have enough time to work on GSoC. (2) Background Information * Please summarize your educational background (degrees earned, courses taken, etc.). I majored in Computer Science and got my bachelor degree in Tsinghua University from 2008 to 2012. I learned some basic courses such as algorithm, data structure, algebra and logic. I also learned many advanced courses such as database, operating system, network, formal language, compiler, computer architecture. Currently I'm working for my master degree. My research interest is database. * Please summarize your programming background (OSS projects, internships, jobs, etc.). This is the first time I plan to take part in an OSS projects. I have worked as an intern in Hulu, an online video website, for one year, mainly focus on frontend webpages. I have used C++ to finish many course projects, such as a simple ftp server and client, completing a tiny os and so on. Recently I am learning the new features in C++11 and get to know much more about the C++ language. * Please tell us a little about your programming interests. I really love to optimize on the implantation of an algorithm by making full use of the compiler and the architecture(e.g. cache, SIMD). Besides, I like to write webpages using different kinds of excellent frameworks. * Please tell us why you are interested in contributing to the Boost C++ libraries. Boost is the most famous C++ libraries. It is powerful and can be used in many different kinds of programs. It's my honor to add some new features into it. * What is your interest in the project you are proposing? It's related to my research interest. Nowadays data are produced rapidly. To clean or integrate them so that we can get a better view, "Approximate string matching" is rather a powerful tool. I'd like to see such functions in boost library. * Have you done any previous work in this area before or on similar projects? Yes. My lab has focused on researching on such algorithms for 3-5 years. The "best" algorithm for general string similarity search and join is called "passjoin", which is invented by our lab. I have implemented it and done a lot of optimizations. Early this year we attended a string similarity search & join competition[1] and achieved champion. The code is mainly written by me. [1] http://www2.informatik.hu-berlin.de/~wandelt/searchjoincompetition2013/Resu… * What are your plans beyond this Summer of Code time frame for your proposed work? I will make efforts to make it accepted by the boost library. If it can be released someday, I will maintain it. * Please rate, from 0 to 5(0 being no experience, 5 being expert), your knowledge of the following languages, technologies, or tools: -C++: 3 -C++ Standard Library: 3.5 -Boost C++ Libraries: 2 -Subversion: 3 * What software development environments are you most familiar with? Eclipse. I use it in both windows and linux. * What software documentation tool are you most familiar with? Doxygen. (3) Project Proposal Approximate string matching is a "hotspot" in recent years. It's widely used in big data processing such as search engine, DNA matching. It also can be used in simple tasks like spell checking. However, there isn't a high performance and well-designed library in such area. I wish to see such algorithms in boost so that many people can benefit from this! My proposal contains two parts: A. To implement basic similarity metrics. Edit distance and Jaccard may be the most popular ones. Well, I'll implement more similarity metrics, such as Hamming distance, dice and cosine. B. To implement the fastest algorithm for string similarity search and join. Similarity metrics will be edit distance only since it is most commonly used. Similarity search is to find out all the strings in a given set S, which are "similar" to a query string. Similarity join is to find out all the string pairs<r,s>, where r is in set R and s is in set S and r is similar to s. They are the most useful operators in large data processing. However, the brute-force method for such problem is too slow. And it's not easy to write a correct and high performance algorithm without reading the details in many research papers and debugging them for a long time. To avoid reinventing the wheel, I strongly propose to implement it in boost library. Here is a prototype for the interface: Part A. basic similarity metrics (use edit distance as an example) (1) int edit_distance(const Range1T&, const Range2T&); // return edit distance. (2) int edit_distance(const Range1T&, const Range2T&, int threshold); // return edit distance if it is no larger than threshold, else return -1 or threshold + 1. (3) int edit_distance(const Range1T&, const Range2T&, equal_func); // use uqual_func to determine whether two elements are equal or not. (4) int edit_distance(const Range1T&, const Range2T&, insert_cost_func, delete_cost_func, substitute_cost_func); // use three cost functions to calculate the cost for one insert/delete/substitute operation. (5) some functions to return the sequence of operations? Version (1) may be the most common case. Version (2) can use the given threshold to do optimization for the speed (O(m*n) -> O(threshold * min(m, n))). It's useful since sometimes we only care about the exact edit distance if it is small. Version (3) and (4) extend the flexibility of this utility algorithm. I'm not sure whether to support (5) or not. Part B. string similarity search and join algorithms, focus on edit distance class searcher { void insert(IDType id, const string &word); // insert a word to the index, id is used as an identifier of this word. void erase(IDType id); // erase a word from the index, id should be the same one used in insertion process. vector<Result> search(const string &query, int threshold); // performing similarity search, which will return all the strings which are similar to the query string. Result may be a pair of the id and real edit distance. }; class joiner { vector<PairResult> join(const string_and_id_container &R, const string_and_id_container &S, int threshold); // performing similarity join. Usually we are doing on-line join, which means we cannot pre-build the index. So we only need this one function. It will return all the <r,s> pair where r is in R and s is in S and r is similar to s. PairResult may be a tuple of r's id, s's id and real edit distance. }; (4) Proposed Milestones and Schedule Now – June 15 To be more familiar with the Boost library. Discuss the details and interfaces with my mentor and on the develop mail list. June 15 – July 15 Develop a beta version of the approximate matching library. Most functions will be done at this time. July 15 – August 1 Review my code and improve its performance and quality. Discuss with my mentor for mid-term evaluation and future works. August 1 – September Write documentation, test code for my library. Make it "boost" like.

4 12

[gsoc-2013] Boost.Expected
by Pierre T. 02 May '13

02 May '13

Hello, My name is Pierre Talbot, I participated to GSoC 2011 and began the implementation of the Boost.Check library (for checking the validity of any number having a check digit — credit card number, ISBN, …). It was two years ago and the design is still evolving due to many variation points in the code. One of the function looks like : template <typename check_algo, typename range> boost::optional<typename check_algo::checkdigit_type> compute_checkdigit(const range &x); Using boost::optional tell us that: "A check digit should be computed if 'x' is a valid sequence". Since 'x' has many reasons to be incorrect, many errors could be raised. Based on a policy template class, it launches exception or returns with an empty optional . Then I though a lot about a better way to do it, allowing the user to get an exception or an error code. But it was quite complex for a so little part of my library… I decided that the policy "throw exception or nothing" should be enough. Yesterday, I watched the video of Mr. Alexandrescu on Boost.Expected and I think it would be a very useful library. It could mainly be useful in system programming (where many error codes can arise from a single call), but in any code where many exception errors can be thrown (like in Boost.Check). As you suspected, I'm interested in coding Boost.Expected during the summer as a GSoC student. Firstly, it could be very useful to list the resources on the subject (aside the talk), I have several articles that I will talk about later. Secondly, and hoping you'll debate, I would like to ask your opinion about several ideas and facts: 1) In the Boost project description, we can read: "adding a class expected-or-error_code". Obviously, the main design decision made by Alexandrescu is to consider that error code are exception. Do we need an abstraction of exception/error code ? Why do you think a "expected-or-error_code" class is interesting ? 2) Consider this code (from Alexandrescu slides) : // Caller string s = readline(); auto x = parseInt(s).get(); // throw on error auto y = parseInt(s); // won’t throw if (!y.valid()) { // handle locally if (y.hasException<std::invalid_argument>()) { // ------------------ < The flagged line. // no digits ... } y.get(); // just "re"throw } The flagged line has some tastes of "return to the past" flavor. Back to the C procedural language, the basic error code handling system was a lot criticized because: * Readability of the code decrease ; * Error handling occurs in the middle of the execution flow ; * <Add your favourite reason here>. Several links on the subjects (if you have others, I'm interested) http://www.joelonsoftware.com/articles/Wrong.html http://www.joelonsoftware.com/items/2003/10/13.html http://blogs.msdn.com/b/oldnewthing/archive/2005/01/14/352949.aspx http://nedbatchelder.com/text/exceptions-vs-status.html Basically, only the last one is clearly for exception. The main argument against the procedural approach is the readability. I would say that the Expected approach just differ by allowing to rethrow exception. But if you want to handle it, you must code multiple if-than-else statements. So I considered a complementary approach working with Expected to handle multiple error cases: string visa_number = readline(); expected<char> expected_checkdigit = compute_checkdigit<visa>(visa_number); if(expected_checkdigit.valid(visa_error_resolver)) { visa_number += expected_checkdigit.get(); std::cout << visa_number << std::endl; } With this code, there is only a if statement, and no more multiple error cases handles. But what is this error_resolver ? It may be declared as : // Somewhere in visa.hpp. A type list. typedef error_list<size_error_exception, unknown_character_exception, checkdigit_encoding_exception, …> visa_errors; // Somewhere in the user code. error_resolver<visa_errors, expected_type> visa_error_resolver; // in this case, expected_type is a char. // Initialize error handler on specific exception/errors. visa_error_resolver.on<size_error_exception>(size_error_handler) .on<unknown_character_exception>(unknown_character_exception) ... Now we are agree that visa_error_resolver can be reused everywhere we want to resolve an error on a visa number. What are the handlers ? There are [Function|Functor|Lambda] (pick up your favourite) with this form : expected<ExpectedType> size_error_handler(const size_error_exception&) expected<ExpectedType> unknown_character_exception(const unknown_character_exception&) Now you can understand for what the type list "error_list" stands for, we can store these handlers into the error_resolver and call them without any virtual cost. Why the return type of error handler is expected<ExpectedType> ? Consider this size_error_handler code : expected<ReturnType> size_error_handler(const size_error_exception& e) { std::cout << "The number you gave has a bad size." << std::endl; std::cout << "Enter it again : " << std::endl; return read_visa_checkdigit(); } read_visa_checkdigit can call recursively valid() until it's valid. Though there are some ways to make this treatment iterative. A basic treatment could be to print an warning message and just returns (in this case, valid returns false): expected<ReturnType> size_error_handler(const size_error_exception& e) { std::cout << "Warning: the VISA field is incorrect." << std::endl; return expected<ReturnType>::fromException(e); } Results: * The error code handling is delegated to specific functions ; * The readability is still excellent ; * You can easily re-use your error handler function ; * If you don't like it, you can still throw exception on failure with "get()". I can code a "proof of concept" if you think this is a good idea. Do not hesitate to comment it, point out programming pitfalls, request further clarification, or anything you judge useful. Thank you for reading it ! Pierre Talbot

10 63