May 2024 - Boost-users - lists.boost.org

Fwd: [beast] Trying to Integrate Chunked Encoding into server-sync Example
by Max D 29 May '24

29 May '24

Hi everyone, I'm trying to integrate the manual chunked encoding example [0] into the sync-server example [1] and have difficulties understanding a compiler error. I thought I might find some help in this mailing list. Here's my diff to the original example: @@ -151,6 +151,22 @@ handle_request( req.method() != http::verb::head) return send(bad_request("Unknown HTTP-method")); + if (req.target().find("/chunked") == 0) { + std::vector<uint8_t> data; + for (int i = 0; i < 10; i++) { + data.push_back(0x00); + } + net::const_buffer buf {data.data(), 10}; + + send.send_chunk_headers(); + + for (int i = 0; i < 3; i++) { + send.send_chunk(buf); + } + + send.send_last_chunk(); + } + // Request path must be absolute and not contain "..". if( req.target().empty() || req.target()[0] != '/' || @@ -243,6 +259,29 @@ struct send_lambda http::serializer<isRequest, Body, Fields> sr{msg}; http::write(stream_, sr, ec_); } + + void + send_chunk_headers() + { + http::response<http::empty_body> res {http::status::ok, 11}; + res.set(http::field::server, "Beast"); + res.chunked(true); + + http::response_serializer<http::empty_body> sr{res}; + http::write_header(stream_, sr); + } + + void + send_last_chunk() + { + http::write(stream_, http::make_chunk_last()); + } + + void + send_chunk(net::const_buffer buf) const + { + http::write(stream_, http::make_chunk(buf)); + } }; // Handles an HTTP server connection I'm passing a TCP socket as well as the result of http::make_chunk() with a net::const_buffer to http::write() which, if I understand correctly, is the same thing that is done in the example. However, I'm getting the following error when compiling: .../server-sync/src/beast.cpp: In instantiation of ‘void send_lambda<Stream>::send_chunk(boost::asio::const_buffer) const [with Stream = boost::asio::basic_stream_socket<boost::asio::ip::tcp>]’: .../server-sync/src/beast.cpp:164:28: required from ‘void handle_request(boost::beast::string_view, boost::beast::http::request<Body, boost::beast::http::basic_fields<Allocator> >&&, Send&&) [with B ody = boost::beast::http::basic_string_body<char>; Allocator = std::allocator<char>; Send = send_lambda<boost::asio::basic_stream_socket<boost::asio::ip::tcp> >&; boost::beast::string_view = boost::basic_string_view<char, std::char_traits< char> >; boost::beast::http::request<Body, boost::beast::http::basic_fields<Allocator> > = boost::beast::http::message<true, boost::beast::http::basic_string_body<char>, boost::beast::http::basic_fields<std::allocator<char> > >]’ .../server-sync/src/beast.cpp:313:57: required from here .../server-sync/src/beast.cpp:283:20: error: no matching function for call to ‘write(boost::asio::basic_stream_socket<boost::asio::ip::tcp>&, boost::beast::http::chunk_body<boost::asio::const_buffer>) ’ 283 | http::write(stream_, http::make_chunk(buf)); | ~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The same thing happens for the send_last_chunk() call. Why isn't the call to write() working? Did I misunderstand something in the example? I'm using boost 1.71 and am compiling with GCC 10.3.0. Let me know if you need any more details. Thanks in advance for any help Max --- [0] https://www.boost.org/doc/libs/1_71_0/libs/beast/doc/html/beast/using_http/… [1] https://www.boost.org/doc/libs/1_71_0/libs/beast/example/http/server/sync/h…

2 1

boost::flyweights::intermodule_holder
by Foelsche, Peter 10 May '24

10 May '24

Dear All, Thanks for the boost library! Apparently the boost::flyweight::intermodule_holder on LINUX is creating shared memory objects, which by accident (crash?) can persist after the process having created them is already gone. As a result I ran into the following situation: I could not successfully run some code, as it attempted to create an already existing shared memory object created by another user. Would the following be a solution to this problem: Use a copy of the code of the static_holder inside another namespace as attached and compile one module with __INCLUDE_HOLDER__ defined and all the rest without this define? Peter

4 4

Boost Graph - adjacency_list in_edge_iterator are only input_iterator_tag?
by Nic 10 May '24

10 May '24

Hi, Hopefully this is the right place for this. This with Boost 1.81.0 It looks like boost::adjacency_list's in_edge_iterator, out_edge_iterator, and for bidirectional graphs also edge_iterator all are only input iterators. This seems a bit restrictive, and is regardless of the type of EdgeList template parameter. For example, this code (which seems a relatively reasonable thing to want to do) doesn't work: https://godbolt.org/z/n3PdEYo49 It doesn't work because transform_view doesn't define iterator_category if the iterator is not at least a forward iterator (I'm a bit confused as to why that is, but that's not a boost-related thing). I did some digging and found out why the iterator is always an input iterator: The iterators are defined as following: template < class BaseIter, class VertexDescriptor, class EdgeDescriptor, class Difference > struct in_edge_iter : iterator_adaptor< in_edge_iter< BaseIter, VertexDescriptor, EdgeDescriptor, Difference >, BaseIter, EdgeDescriptor, use_default, EdgeDescriptor, Difference > { // .... }; EdgeDescriptor is passed in as the reference type, but it is not a reference so the iterator adaptor always chooses input_iterator_tag for the iterator_category. Does anyone know of a workaround, or know if this is the desired behavior for edge iterators? Am I missing something obvious here? Thank you very much, Nicolas Morales

2 1