I am sorry if my email points the finger at json library. That was not
intentional. I think there could be other issues:
I did a onetime read of json and then wrote out as SEXP. I noticed the
slowness persisted despite that.
So then I wrote out only 1000 rows for the time being.
Here is the conversion code, which I am sure none of you needs but
incase there is some error in what I did.
(require json)
(define data-set
(call-with-input-file "usagov_bitly_data.json"
(lambda (in)
(read-json in))))
(define (deep-ref dct keys)
(if (empty? keys)
dct
(deep-ref (dict-ref dct
(first keys))
(rest keys))))
(define cols
(for/list ([col (deep-ref data-set '(meta view columns))])
(dict-ref col 'name)))
(define rows
(for/list ([row (dict-ref data-set
'data)])
(map cons
cols
row)))
;; save a shorter list
(call-with-output-file "usagov_bitly_data.sexp"
(lambda (out)
(write (take rows 1000)
out))
#:exists 'replace)
____________________
Racket Users list:
http://lists.racket-lang.org/users
