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

boost::asio::ssl::context::use_certificate_chain_file and boost::asio::ssl::context::use_certificate_file Methods Give "bad method" Error
by Osman Zakir 21 Jul '22

21 Jul '22

Hello. Message for update to previously-reported issue. It turns out that the "bad method" error was because of the boost::asio::ssl::context object methods I mentioned in the subject of this email. Reading the certs in from the files into std::strings manually and using use_certificate and use_certificate_chain instead is better, at least from what I saw. I still have an error in the browser (code: ERR_SSL_PROTOCOL_ERROR) when I try to visit my running app saying that the connection for the site isn't secure and that it gave back an invalid response. And running the command openssl s_client -connect <ip_address>:5501 -servername dragonosman.dynu.net gave this result: CONNECTED(00000148) 11784:error:1409442E:SSL routines:ssl3_read_bytes:tlsv1 alert protocol version:ssl\record\rec_layer_s3.c:1544:SSL alert number 70 --- no peer certificate available --- No client certificate CA names sent --- SSL handshake has read 7 bytes and written 322 bytes Verification: OK --- New, (NONE), Cipher is (NONE) Secure Renegotiation IS NOT supported Compression: NONE Expansion: NONE No ALPN negotiated Early data was not sent Verify return code: 0 (ok) and when command exits, I see this error in my server console window: Lines 625 and 626: handshake: unsupported protocol (SSL routines, tls_early_post_process_client_hello) Lines 646 and 647: read: unspecified system error I'm assuming this is a HTTP/2 upgrade request inside a TLS ClientHello message, but is that enough reason for the SSL error I'm getting in my browser? Anyway, I've sent an email on the OpenSSL users mailing list to ask how to parse the ClientHello message, so hopefully I'll get something from there.

1 0

[container] flat_set using a reversed small_vector
by Phil Endecott 20 Jul '22

20 Jul '22

Dear Experts, I currently have: template <typename KEY, typename VALUE, size_t SZ> using small_flat_map = boost::container::flat_map< KEY, VALUE, std::less<KEY>, boost::container::small_vector< std::pair<KEY,VALUE>, SZ > >; This works OK, but one thing that I do with it quite often is auto i = m.upper_bound(...); m.erase(m.begin(),i); Erasing items at the start of the vector is of course the worst thing you can do for performance with a flat_map. So I was wondering if I can have a flat_map that stores its elements in the reverse order in the underlying vector. Question: can anyone think of a way to do this that doesn't require changing all of the downstream code? For a few milliseconds I thought I could just change std::less for std::greater but that reverses the externally-visible behaviour. Is there some Boost utility that can wrap the implementation vector to work in reverse? Thanks, Phil.

1 0

Server App using Boost.Beast, Sending Only One Certificate out of Chain
by Osman Zakir 16 Jul '22

16 Jul '22

Hi, everyone. I upgraded my server app to Boost 1.79.0 and tried running it, but there are issues: 1. I renewed my Let's Encrypt certs, but when I try to check them, I see it's only sending one cert out of a chain of three. And I'm loading the certs and the RSA private key using the SSL context methods to load the cert chain and key from files, too. 2. The read operation on line 640 in the do_session function in currency_converter.cpp gave a "bad method" error before, but mostly gives "the connection was forcibly closed by the remote host" error. I have the source code on GitHub here: https://github.com/DragonOsman/currency_converter.

1 0

[release] Boost 1.80.0 beta 1 released
by Glen Fernandes 14 Jul '22

14 Jul '22

Boost release 1.80.0 beta 1 is now available at: https://boostorg.jfrog.io/artifactory/main/beta/1.80.0.beta1/source/ The SHA256 checksums are: boost_1_80_0_b1.tar.bz2 - b29191c1bd4504511f9b1155c6f79f06cc7044444780392ed7358f5778ac7d16 boost_1_80_0_b1.tar.gz - ad3df9f63acd667dcf13375fa0e90e7303f851251e91d1505974bdbbf1f94907 boost_1_80_0_b1.7z - 7bfdf705b1d058c3525be90ed5cb47aacf919cd74aa7d5752d55e78aa60880ba boost_1_80_0_b1.zip - 1b509ed848f153086984658d4d4a57af6230f496a53bf88aba805bb49039012d Release notes (still need updates): https://www.boost.org/users/history/version_1_80_0.html Please try the beta and report any problems you encounter. Thanks, -- Boost Release Team

1 0

Higher than expected memory usage by boost::iostreams::gzip_compressor
by Alex Petty 13 Jul '22

13 Jul '22

Hello, I'm looking at adopting boost::iostreams for reading/writing compressed streams. I'm testing with textual files commonly on the order of a couple of GB uncompressed, and a couple hundred MB compressed. I noticed during some initial testing that using a gzip_compressor to write a compressed stream through boost::iostreams results in significant memory usage, I'm seeing hundreds of MB per compressed stream. It looks like internally the gzip_compressor must be doing some GC-type behavior, because I see its memory usage increase to just under 1GB per stream with a gzip_compressor, and then decrease again. Is there any way to control the memory that gzip_compressor is using? I've tried both increasing and decreasing the buffer size when gzip_compressor is constructed, but that buffer is tiny and cannot be what's driving the increased memory usage. Thanks, Alex

1 0

[Graph] Can users on Xcode 7-10 help out?
by Jeremy Murphy 12 Jul '22

12 Jul '22

Dear Boost.Graph users, Is anyone that uses old Xcode 7-10 able to help out with some curious compilation errors? https://github.com/boostorg/graph/issues/298 We'll do our best to fix them but I assume it might be easier for someone that actually uses these particular platforms. Thanks, cheers. Jeremy

1 0

xpressive: How to combine vector of regex to one?
by Anton Sergunov 08 Jul '22

08 Jul '22

Hello Boost Users! I'm seeking some help from the community because can't see answer in documentation or in source code. I need to find which regex of N my string matches. I don't like to compare them one by one, to speed up the process. It's easy to implement if I have them compile time: ```cpp sregex re = (s1 = re1)|(s2 = re2)|(s3=re3); ... std::set<int> foo(const std::string& str) { std::set<int> result; smatch match; if(regex_match(str, match, re)) { for(int i = 0; i<3; ++i) if(match[i+1]) result.insert(i); } return result; } ``` But I have a vector: ```cpp std::set<int> foo(const std::string& str, const std::vector<sregex>& reVector) { sregex re; int index = 1; for(auto&& item : reVector) { re |= (mark_tag(index++) = item); // (1) } std::set<int> result; smatch match; if(regex_match(str, match, re)) { for(int i = 0; i<reVector; ++i) if(match[i+1]) result.insert(i); } } return result; } ``` There at (1) `|=` doesn't work as I expect for some reson, It didn't catch some values. If I replace it with ```cpp re = re | (mark_tag(index++) = item); ``` it just crashes. If I replace with: ```cpp auto t = re; re = t | (mark_tag(index++) = item); ``` it making captures of previous tags to be subexpresions matches. So code doesn't work as I want. Is it any way to prevent sub-expression capture isolation? As I understand it should be some intermediate type before sregex to express expression under construction. So it will not isolate matches of existing part. I'm using v1.65.1 https://github.com/boostorg/xpressive/issues/24

1 0

Embedded install paths in CMake files for Boost 1.79
by Alex 05 Jul '22

05 Jul '22

Apologies if this is the wrong list. I tried boost-build but I'm not sure if my message was verified (I don't see it in the archive). I'm trying to make a build of Boost that includes the generated CMake files. For my case, I need to install the files to a staging area before packaging them up for installation in the target environment. I've tried "b2 --stagedir <dir> stage" but the generated CMake files still include the stagedir path in them, instead of the install prefix specified during bootstrap, which triggers errors in the packaging system I need to use. I'm also not sure if it's then possible to just install the headers to the same staging location. Is there a way to install to a staging location while the CMake files have the paths to the final install location? CMake uses CMAKE_STAGING_PREFIX for this but I can't find an equivalent for b2. I guess it must be required for cross-compiling though, if you want to support CMake on the target? Regards, Alex

1 0

Excessive dynamic allocation when using boost::split(..., boost::is_any_of(", "))?
by Dominique Pellé 05 Jul '22

05 Jul '22

Hi This following program shows many allocations coming from boost::split(..., boost::is_any_of(",")): ``` $ cat split.cpp #include <string> #include <cstdio> #include <boost/algorithm/string.hpp> #include <boost/container/small_vector.hpp> int main() { std::string str = "foo,bar,foobar,quux"; std::size_t s = 0; for (int i = 0; i < 100; ++i) { boost::container::small_vector<std::string, 4> v; #ifdef USE_IS_ANY_OF boost::split(v, str, boost::is_any_of(",")); #else boost::split(v, str, [](const char c) { return c == ',';}); #endif s += v.size(); } fprintf(stderr, "%zu\n", s); } ``` When using boost::is_any_of(","), there are 7 dynamic allocations per loop iteration whereas when using the lambda predicate, there are 0 dynamic allocation per loop as shown below: ``` $ clang++ -std=c++14 -DUSE_IS_ANY_OF split.cpp $ heaptrack ./a.out ... allocations: 702 $ clang++ -std=c++14 split.cpp $ heaptrack ./a.out ... allocations: 2 ``` I suspect that there are potential optimizations in Boost here to avoid so many allocations when boost::is_any_of(",") is used. Regards Dominique

3 2

Boost.URL review planned for 13-22 August 2022
by Mateusz Loskot 29 Jun '22

29 Jun '22

Folks, Klemens Morgenstern kindly volunteered [1] as the Review Manager (thanks Klemens!) for the Boost.URL [2], so we have planned the review shortly after the upcoming release of the Boost 1.80. Check the Review Schedule [3] for details! [1] https://lists.boost.org/Archives/boost/2022/06/253225.php [2] https://lists.boost.org/Archives/boost/2022/04/252814.php [3] https://www.boost.org/community/review_schedule.html Best regards, -- The Boost Review Wizards Mateusz Loskot

2 2