September 2006 - Boost-users - lists.preview.boost.org

Re: [Boost-users] [iostreams] Implementing a block cipher as a filter
by Pfligersdorffer, Christian 26 Sep '06

26 Sep '06

Hello all! Seems I am the first one to reply to my own post :-) I am not so much surprised however, since this stuff is kinda weird and complicated. Nevertheless I got some news for you: after replacing the gzip- with bzip2-compression filters everything worked fine!! This means there is some (very subtle) bug or incompatibility in the gzip filters, which I of course did not expect (but rather was looking in my code of course). For the protocoll: gzip_decompressor raises a gzip_error exception with internal error_ = -3 when reaching the end of the stream. Unfortunately this behaviour could not be reproduced without actual encrypting/decrypting. Simply doing a null-transformation by replacing encrypt() and decrypt() with memcpy() does not yield the error. I did some tests: reading a file, writing it to a filtering_ostream using filter X into some file, again reading through filtering_istream using Y and comparing with the original file. X = block_crypter, Y = block_crypter ok, identity X = gzip_compressor, Y = gzip_decompressor ok, identity X = block_crypter|bzip2_compressor, Y = block_crypter|bzip2_decompressor ok, identity X = block_crypter|gzip_compressor, Y = block_crypter|gzip_decompressor error, see above and now get yourself a good seat: X = gzip_compressor|block_crypter, Y = gzip_decompressor|block_crypter ok, identity Of course compressing an encrypted file is not the same as encrypting the compressed file, which is much faster and results in smaller files. So I tried this just to see what happens the other way round. Throughout the test I used file_sink and file_source in binary-mode as devices. Have fun fixing this ;-) -- Christian Pfligersdorffer Dipl.-Ing. Bakk.rer.nat. Software Engineering EOS GmbH - Electro Optical Systems Robert-Stirling-Ring 1 D-82152 Krailling / Munich Phone +49 89 893 36-229 Fax +49 89 893 36-287 christian.pfligersdorffer(a)eos.info http://www.eos.info

1 0

[smart_ptr] comparison operators: shared_ptr == weak_ptr?
by Matt.Sullivan＠leica-geosystems.com 26 Sep '06

26 Sep '06

Hi, Probably a newbie question, but what's the best way to test if a shared_ptr and weak_ptr point to the same thing? My current code uses: if (shared_ptr != weak_ptr.lock()) but it feels wrong to create a temporary with some small locking overhead - even though I realise it will make absolutely no difference to performance. So, my question is really why is there no operator==(shared_ptr, weak_ptr)? Google provided little explanation other than this: http://std.dkuug.dk/jtc1/sc22/wg21/docs/papers/2004/n1590.html which approaches the subject but isn't exactly authoritative. thanks, Matt

3 4

boost serialization
by Anna Kloper 25 Sep '06

25 Sep '06

Hi all, I'm new to BOOST and need help with this kind of problem of pointers serialization: class A{ public: A(int *anInt):m_pi(anInt){} ~A(){} private: int *m_pi; }; class B{ public: B(int *anInt):m_a1(anInt),m_a2(anInt){} ~B(){} private: A m_a1; A m_a2; }; int main() { int i=1; B b(&i); //serialize b i=2; //serialize b //de-serialize b expected i = 1 and b.m_a1=b.m_a2=1 //de-serialize b expected i = 2 and b.m_a1=b.m_a2=2 } Now, assuming that there are serialization functions for class A and B I'd like to serialize object b swice with track never so that when de-serializing b I'd like b to contain a pointer to the location of i. The problem is that when I de-serializing b I notice that b doesn't pointer to the location of i anymore but to another location. I need this for serializing of communication, so I need to save all the data trough a pointer and read the data back step by step for preprocessing to the same location in the memory. Is there anything I can do to overcome this problem with pointers? Thank you in advance, Anna. --------------------------------- Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail.

2 1

newbie [boost::test] linking in MSVC6
by Brian Hall 25 Sep '06

25 Sep '06

I have been using boost and STLPort for sometime and I have built up unit tests for my projects that have worked fine up until now. However, I've been lazy and used the #include unit_test_framework.hpp route instead linking to libraries. Now I'm stuck. My tests won't compile with STLPort 5.0.x, apparently because the complier needs more memory (I'm using MS VC++6(SP5) and I've set the complier option /Zm to extend to its maximum). So I decided to build the boost libraries (boost 1_33_1). Unfortunately, now I can't get my testing code to link with these libraries! Could anyone tell me which of the following I should be linking to (I am building in 'release' mode with multithreading)? libboost_unit_test_framework-vc6-mt-gdp-1_33_1.lib libboost_unit_test_framework-vc6-mt-gdp.lib libboost_unit_test_framework-vc6-mt-p-1_33_1.lib libboost_unit_test_framework-vc6-mt-p.lib libboost_unit_test_framework-vc6-mt-sgdp-1_33_1.lib libboost_unit_test_framework-vc6-mt-sgdp.lib libboost_unit_test_framework-vc6-mt-sp-1_33_1.lib libboost_unit_test_framework-vc6-mt-sp.lib In fact, I've tried them all without success. Here is an example error message: libboost_unit_test_framework-vc6-mt-sp-1_33_1.lib(framework.obj) : error LNK2001: unresolved external symbol "class boost::unit_test::test_suite * __cdecl init_unit_test_suite(int,char * * const)" (?init_unit_test_suite@@YAPAVtest_suite@unit_test@boost@@HQAPAD@Z) Release/mst_gum_tree_real_dev.exe : fatal error LNK1120: 1 unresolved externals Error executing link.exe. Also, I simply removed the #include of unit_test_framework.hpp in my original source code. Is that correct? The remaining includes are: unit_test.hpp and floating_point_comparison.hpp. As an alternative to linking to the libraries, I have tried to include the source files in my project, as suggested in the unit test documentation, but that doesn't work either! The documentation does not appear to be up to date. What is a poor boy to do?!! Please give me some ideas

3 3

reference in new- and old-style iterators
by Jens Theisen 25 Sep '06

25 Sep '06

Hello, I'm wondering what the reference member in the iterator_traits of an iterator is required to be. I couldn't find any requirement at all in the C++98 standard, and more surprisingly, I can't see any requirement in the new iterator proposal either. In particular, operator* does not seem to be required to return the iterator_traits< Iterator >::reference for lvalue iterators. At one point, however, the proposal suggests that this was the intention: "...would have a return type of reference; the same as operator*." where reference has a font style indicating that it's the iterator_traits's notion of reference. I'm looking at the proposal version currently published along in the documentation of the boost iterator library. So my questions: - Am I right in my suspicion that in the STL, the reference iterator_trait is entirely useless if taken strictly? - Should the proposal not better require the reference to match with operator*'s return type for lvalue iterators? If not, what's its use? - Is this off-topic here? :) Regards, Jens

2 1

bug in boost::random?
by boost＠larroy.com 24 Sep '06

24 Sep '06

Hi, First of all boost::random is very clever and complete, thanks! Using boost::random I found uniform_int distribution not working in my box. I have version 1.33 on a Debian system. This is an example of a program that triggers a non-uniform integer distribution. I'm I doing something wrong? Regards. #include <iostream> #include <string> #include <cstdlib> #include <cassert> #include<vector> #include<stdexcept> #include <boost/random.hpp> #include <limits> using namespace std; int main(int argc, char *argv[]) { boost::mt19937 rng; boost::uniform_int<int64_t> ui(numeric_limits<int64_t>::min(),numeric_limits<int64_t>::max()); boost::variate_generator<boost::mt19937&, boost::uniform_int<int64_t> > rgen(rng, ui); while(true) { cout << rgen() << endl; } } -- Pedro Larroy Tovar, pedro at larroy dot com | http://pedro.larroy.com/ * Las patentes de programación son nocivas para la innovación * http://proinnova.hispalinux.es/

1 0

VS2003 Linking issuse
by Anthony Mason 22 Sep '06

22 Sep '06

Hi guys, I know this is probably a common problem, but I haven't found any help on any google searches so far... I've got my app compiling in debug mode fine using the boost_date_time and boost_filesystem libraries, compiling using Multithreaded Debug, but when I set up my Release mode compile using just plain Multithreaded I get a ton of LNK2005 messages similar to : error LNK2005: "public: static unsigned int const std::ctype<char>::table_size"(?table_size@$ctype@D@std@@2IB)already defined in libboost_date)time-vc71-mt-s-1_33_1.lib(greg_month.obj) I get about 70 of these errors between date_time and filesystem libraries. I'm linking against the same libraries (mt-s) both debug and release mode. I was hoping that someone could give me some advice. Thanks, Anthony

1 0

[iostreams] Implementing a block cipher as a filter
by Pfligersdorffer, Christian 22 Sep '06

22 Sep '06

Dear community, I'm on my way to learn writing filters for the iostreams library and find it very hard. I want to compress and encrypt data, so I use the gzip-filters (that are already provided) and some self-tailored crypto-filter. Using a symmetric block cipher should be just fine with the framework, since data from a stream is usually fed in portions to the filter (128 Bytes by default). Of course not when you use an aggregate_filter, which was my first attempt because it is so simple. Implementing a multichar_dual_use_filter seemed simple too but I am experiencing weird behaviour. Such as having runtime errors when reading an encrypted compressed file but not when reading compressed encrypted files. Also I am not entirely sure that I understood the documentation... Here is what I do, perhaps you gurus instantly see what I'm doing wrong: _________________________________________________________________________ class block_crypter : public boost::iostreams::multichar_dual_use_filter { const std::string pw; char* buffer; int length; void reallocate_buffer(std::size_t n) { if (length < n) { if (buffer) free(buffer); buffer = (char*) malloc(n); length = n; } } public: block_crypter(const char* password) : pw(password), buffer(NULL), length(0) {} ~block_crypter() { if (buffer) free(buffer); } template<typename Source> std::streamsize read(Source& src, char* s, std::streamsize n) { // Read up to n filtered characters into the buffer s, // returning the number of characters read or -1 for EOF. // Use src to access the unfiltered character sequence reallocate_buffer(n); std::streamsize count = boost::iostreams::read(src, buffer, n); assert(count <= n); if (count <= 0) return -1; decrypt(pw, buffer, count, s); return count; } template<typename Sink> std::streamsize write(Sink& dest, const char* s, std::streamsize n) { // Consume up to n filtered characters from the buffer s, // writing filtered characters to dest. Return the number // of characters consumed. reallocate_buffer(n); encrypt(pw, s, n, buffer); std::streamsize count = boost::iostreams::write(dest, buffer, n); assert(count <= n); return count; } }; BOOST_IOSTREAMS_PIPABLE(block_crypter, 0) _________________________________________________________________________ The encrypt() and decrypt() functions produce a character sequence of the exact same length as the input sequence. All char values can occur which might be a problem but I don't know. Runtime errors always occur at the end of the stream, i.e. on the last run of the filter. Perhaps I did not understand the EOF mechanics or am mixing up something. But then, it compiles and seems to behave properly if used alone. Only in read-mode preceding gzip-deflation there is an error. I hope someone can help me there. BTW: I did not even try to implement my filter as symmetric_filter. I cannot see how anyone would be able to do this given the present documentation or existing "examples". Another thing that clearly arises is the block size issue n which has to be the same for writing and reading in order to produce the correct plain text again. Maybe the optimally_buffered_tag should be used instead of trusting in always the same default filter buffer size? Thanks for your suggestions, -- Christian Pfligersdorffer Dipl.-Ing. Bakk.rer.nat. Software Engineering EOS GmbH - Electro Optical Systems Robert-Stirling-Ring 1 D-82152 Krailling / Munich Phone +49 89 893 36-229 Fax +49 89 893 36-287 christian.pfligersdorffer(a)eos.info http://www.eos.info

1 0

boost::interprocess::managed_external_buffer class restrictions
by alexmark 22 Sep '06

22 Sep '06

The problem is the class managed_external_buffer currently does not provide public default c-tor and interface functions create(), open(), close(). I'm writing DLL, where I need a global object of type managed_external_buffer. This object is "connected" to mapped view of paging file within processing of DLL_PROCESS_ATTACH and disconnected at DLL_PROCESS_DETACH like that: HANDLE g_hfm; LPVOID g_buf_base_addr; managed_external_buffer g_buf; BOOL APIENTRY DllMain(,,,. ULONG reason, ...) { switch (reason) { case DLL_PROCESS_ATTACH: if (!Attach()) return FALSE; break; case DLL_PROCESS_DETACH: Detach(); break; } return TRUE; } bool Attach () { ... g_hfm = CreateFileMapping(0xFFFFFFFF, ...); // use paging file g_buf_base_addr = MapViewOfFile(g_hfm, ...); g_buf.create(g_buf_base_addr, ...); ... MyObj* p = g_buf.construct(...); ... } void Detach () { ... g_buf.destroy_ptr(p); ... g_buf.close(); UnmapViewOfFile(g_buf_base_addr); CloseHandle(g_hfm); ... } As a workaround, I wrote my own class "managed_user_buffer" derived from boost::interprocess::detail::basic_managed_memory_impl and providing public default c-tor and create()/open()/close() methods. I just made a copy of managed_external_buffer.hpp and slightly modified the code. But it's not so clear, why managed_external_buffer class itself exposes the restricted interface? I suppose the class would be more usable w/o the restrictions like that. Or not? Moreover, in boost::shmem::named_shared_object class this functionality was available. Thanks, Alex.

2 1

BGL and maximum spanning trees
by aitor 22 Sep '06

22 Sep '06

Hello, in my project i need to compute the maximum spanning tree of an undirected weighted graph. One possible solution is to could iteratively convert each weight wij to 1/wij and apply prim_minimum_spanning_tree algorithm over it. However, i wonder if i could slightly change the prim_minimum_spanning_tree algorithm to a prim_maximum_spanning_tree function by changing only the compare predicate (in line 39 of prim_minimum_spanning_tree.hpp) from std::less<W> compare; to std::greater<W> compare; Wold then the function compute a maximum spanning tree ? Or am i missing something important ?

2 1