Module: Stomper::Support::Ruby1_9::FrameSerializer

Defined in:
lib/stomper/support/1.9/frame_serializer.rb

Overview

Implementation of FrameSerializer methods for Ruby 1.9

Instance Method Summary collapse

Instance Method Details

#determine_content_length(frame) ⇒ Fixnum

Determines the content-length of a frame being sent to a broker. For Ruby 1.9, this is the bytesize of the frame’s body.

Returns:

  • (Fixnum) —

    byte-length of frame’s body



52
53
54
 
# File 'lib/stomper/support/1.9/frame_serializer.rb', line 52

def determine_content_length(frame)
  frame.body.bytesize
end

#determine_content_type(frame) ⇒ String

Determines the content-type of a frame being sent to a broker. This version of the method is used by Ruby 1.9, and will use the encoding of the frame’s body to help determine the appropriate charset. If the body’s encoding is binary, no charset parameter will be included (even if one was manually set.) Otherwise, if the content-type is ommitted, a value of ‘text/plain’ is assumed and the encoding of the body is used as the charset parameter.

Returns:

  • (String) —

    content-type and possibly charset of frame’s body



40
41
42
43
44
45
46
47
 
# File 'lib/stomper/support/1.9/frame_serializer.rb', line 40

def determine_content_type(frame)
  ct = frame[:'content-type']
  ct &&= ct.gsub(/\;\s*charset=\"?[a-zA-Z0-9!\#$&.+\-^_]+\"?/i, '')
  enc = frame.body.encoding.name
  text = (enc != 'ASCII-8BIT') || ct =~ /^text\//
  ct = 'text/plain' if ct.nil? && text
  ct && text ? "#{ct};charset=#{enc}" : ct
end

#encode_body(body, ct_header) ⇒ String

Return the body with the specified encoding applied. An encoding is specified in the ‘content-type’ header of a frame by the charset parameter. If no encoding was explicitly specified, a UTF-8 encoding is used when the frame’s content type begins with text/, otherwise an encoding of US-ASCII is used.

Parameters:

  • body (String) —

    body of message to encode

  • ct (String) —

    content-type header of frame

Returns:

  • (String) —

    body with the appropriate encoding applied



13
14
15
16
17
18
19
20
21
 
# File 'lib/stomper/support/1.9/frame_serializer.rb', line 13

def encode_body(body, ct_header)
  body.tap do |b|
    charset = ct_header ?
      (ct_header =~ /\;\s*charset=\"?([\w\-]+)\"?/i) ? $1 :
        (ct_header =~ /^text\//) ? 'UTF-8' : 'ASCII-8BIT' :
      'ASCII-8BIT'
    b.force_encoding(charset)
  end
end

#get_body_byte ⇒ String?

Reads a single byte from the body portion of a frame. Returns nil if the character read is a frame terminator.

Returns:

  • (String, nil) —

    the byte read from the stream.



26
27
28
29
30
 
# File 'lib/stomper/support/1.9/frame_serializer.rb', line 26

def get_body_byte
  #raise "Implementation varies by Ruby version"
  c = @io.getc
  c == ::Stomper::FrameSerializer::FRAME_TERMINATOR ? nil : c
end

#get_header_line ⇒ String

Reads a line of text from the underlying IO. As per the Stomp 1.1 specification, all header strings are encoded with UTF-8.

Returns:

  • (String) —

    line of text read from IO, or ” if no text was read.



59
60
61
 
# File 'lib/stomper/support/1.9/frame_serializer.rb', line 59

def get_header_line
  (@io.gets || '').tap { |line| line.force_encoding('UTF-8') }
end