HTTP HTTP POST chunked Transfer-Encoding Content-Length # Regression test for bug where curl stops reading chunked response # when both Content-Length: 0 and Transfer-Encoding: chunked headers # are present. Per RFC 7230 Section 3.3.3, Transfer-Encoding should # take precedence and Content-Length should be ignored. # The writedelay simulates the timing issue where chunks arrive in # separate packets. # Server-side writedelay: 500 HTTP/1.1 200 OK content-type: text/json connection: keep-alive content-length: 0 transfer-encoding: chunked 1a random data in first chunk 1d another data in second chunk 15 third and last chunk 0 # Client-side http HTTP POST response with both Transfer-Encoding chunked and Content-Length 0 http://%HOSTIP:%HTTPPORT/%TESTNUMBER -d "testdata" # Verify data after the test has been "shot" POST /%TESTNUMBER HTTP/1.1 Host: %HOSTIP:%HTTPPORT User-Agent: curl/%VERSION Accept: */* Content-Length: 8 Content-Type: application/x-www-form-urlencoded testdata HTTP/1.1 200 OK content-type: text/json connection: keep-alive content-length: 0 transfer-encoding: chunked random data in first chunk another data in second chunk third and last chunk