rev: async response bodies with DevFdResponse middleware

This new middleware should be a no-op for non-Rev concurrency
models (or by explicitly setting env['rainbows.autochunk'] to
false).

Setting env['rainbows.autochunk'] to true (the default when Rev
is used) allows (e)poll-able IO objects (sockets, pipes) to be
sent asynchronously after app.call(env) returns.

This also has a fortunate side effect of introducing a code path
which allows large, static files to be sent without slurping
them into a Rev IO::Buffer, too.  This new change works even
without the DevFdResponse middleware, so you won't have to
reconfigure your app.

This lets us epoll on response bodies that come in from a pipe
or even a socket and send them either straight through or with
chunked encoding.

1 files changed, 13 insertions, 0 deletions

diff --git a/t/large-file-response.ru b/t/large-file-response.ru
new file mode 100644
index 0000000..90dc6c5
--- /dev/null
+++ b/t/large-file-response.ru
@@ -0,0 +1,13 @@
+# lib-large-file-response will stop running if we're not on Linux here
+use Rack::ContentLength
+use Rack::ContentType
+map "/rss" do
+  run lambda { |env|
+    # on Linux, this is in kilobytes
+    ::File.read("/proc/self/status") =~ /^VmRSS:\s+(\d+)/
+    [ 200, {}, [ ($1.to_i * 1024).to_s ] ]
+  }
+end
+map "/" do
+  run Rack::File.new(Dir.pwd)
+end