views:

39

answers:

1

Adobe's JSON serialization library is very helpful, but it can lock up the UI if parsing a large string. I am wondering if it makes sense to parse JSON in a psuedo-thread (aka "chunking").

I am wondering if anyone has implemented, has suggestions for implementing, or arguments against this approach.

+1  A: 

let's just say, it's sufficiently complicated ... :)

JSON parsing is recursive, and the state of the parsing automaton is maintained in the stack ... if you want to chunk it down, you'll have to maintain the state for yourself ... i don't know of any libraries doing that ... you might want to have a look at the state pattern ...

if you have LARGE data amounts, then JSON probably isn't the best of choices, mostly because of the redundancies ... you might wanna switch to something like CSV ... it's more compact, easier to parse, and requires no recursion, thus can easily be run in a green thread ...

greetz

back2dos

back2dos