--- !ditz.rubyforge.org,2008-03-06/issue 
title: optimize json encode/decode performance
desc: |-
  Performance of encoding/decoding plain ASCII data (which is often 99+% of
  JSON encoded data) should be improved.
  
  Decoding can take advantage of the fact that a NUL cannot occur inside a
  valid JSON document.  So, as long as the input is NUL terminated, and all
  syntax checking is correct, an attempt to read beyond the input buffer will
  cause a syntax error when the NUL is encountered.  Explicit bounds checking
  is thus not necessary at all.
  
  Encoding should have a fast path which first ensures that there is enough
  space to process N input characters safely (taking into account the maximum
  possible expansion caused by e.g. escaping).  Inside the safe loop there is
  no need to check for output space for each output character.
type: :task
component: duk
release: 
reporter: sva <sami.vaarala@iki.fi>
status: :unstarted
disposition: 
creation_time: 2013-09-10 19:39:08.860112 Z
references: []

id: 42b394dbd86284c1b8e15f3f35733ac1efdc7f88
log_events: 
- - 2013-09-10 19:39:09.081188 Z
  - sva <sami.vaarala@iki.fi>
  - created
  - ""