class GzipOutputFilter
        def initialize(app)
                @app = app
        end

        def call(env)
                out = @app.call(env)

                gzip = IO.popen('gzip', 'r+')
                gzip.write out[2]    # handles String only, would be nice to check
for IO object too
                gzip.close_write

                out[1]['Content-encoding'] = 'gzip'
                out[2] = gzip

                out
        end
end

class HelloHandler
        def call(env)
                [ 200, { 'Content-type' => 'text/plain' }, 'hello, world' ]
        end
end

use Rack::Lint
use GzipOutputFilter
run HelloHandler.new

On Jun 11, 2008, at 8:30 AM, spicyj wrote:

require 'zlib'

class Deflater
   include Zlib

   def initialize(app)
     @app = app
   end

   def call(env)
     accept = env['HTTP_ACCEPT_ENCODING']
     status, headers, body = @app.call(env)
     if accept and accept.include? 'deflate'
       headers["Content-Encoding"] = "deflate"
       body = deflate(body)
     end
     [status, headers, body]
   end

   def deflate(body)
     deflater = Deflate.new(
       DEFAULT_COMPRESSION,
       # drop the zlib header which causes both Safari and IE to choke
       -MAX_WBITS,
       DEF_MEM_LEVEL,
       DEFAULT_STRATEGY)

     case body
     when IO
       body.rewind if body.respond_to? :rewind
       gzout = StringIO.new(deflater.deflate(body.read, FINISH))
       body.close
     when String
       gzout = StringIO.new(deflater.deflate(body, FINISH))
     end
     gzout.rewind
     gzout
   end

end

On Jun 12, 2008, at 7:33 AM, Christian Neukirchen wrote:

Cheers-

Content-Encoding: gzip

( headers["Content-Encoding"] = "gzip" )

Which is better/right - deflate of gzip?

On Jun 12, 5:12 am, Ezra Zygmuntowicz <e...@engineyard.com> wrote:

On Jun 12, 2008, at 6:40 PM, Stoyan Zhekov wrote:

  def call(env)
    accept = env['HTTP_ACCEPT_ENCODING']
    status, headers, body = @app.call(env)
    if accept and [ 'deflate', 'gzip'].include? accept
      headers["Content-Encoding"] = "deflate"
      body = deflate(body)
    end
    [status, headers, body]
  end

On Jun 12, 2008, at 8:56 PM, Ezra Zygmuntowicz wrote:

On Jun 13, 2008, at 5:36 AM, Christian Neukirchen wrote:

Accessing the value of Accept-Encoding seems to be more complex to me?

http://ruby-doc.org/stdlib/libdoc/webrick/rdoc/classes/WEBrick/HTTPUt...

It returns an array sorted in order of browser preference. (Though it
appears not to remove q=0.0 encodings.)

The accepted_encodings methods definitely needs a second look and a new name.
Provide a little feedback and I might finish it for inclusion. :-)

I don't think it's worth generalizing the current mimeparse.rb code to
cover both Accept and Accept-Encoding; handling Accept-Encoding is
much simpler than Accept (mostly because there's no need to deal with
type specificity).

Once I extend my Accept-Encoding code to handle * and drop the complex
regexp I'll ask the list for a review.

Comments are very welcome - especially on the select_best_encoding tests.

Comments are very welcome, as usual.

I can't get a full test case, because gzip needs a header for the
mtime of the data, changing the output. Is there a good way to test
the output (maybe gunzipping the output and comparing it to the
original)? I should probably also change it to use the Last-Modified
header as the mtime.

Also, should this be in a separate file? Apache's mod_deflate handles
both deflate and gzip encodings.