over-simplifying

[dcooley]

I'm simplifying one-step too much in this exmaple

test <- '{"test":[1,[2,[3]]]}'
jsonlite::fromJSON(test)
#> $test
#> $test[[1]]
#> [1] 1
#> 
#> $test[[2]]
#> $test[[2]][[1]]
#> [1] 2
#> 
#> $test[[2]][[2]]
#> [1] 3
jsonify::from_json(test)
#> $test
#> $test[[1]]
#> [1] 1
#> 
#> $test[[2]]
#>      [,1]
#> [1,]    2
#> [2,]    3

---

smaller example for testing

from_json('[1,[2]]')

[knapply]

Follow-up on eddelbuettel/rcppsimdjson#10 (comment)

Is the following what's desired?

test1 <- jsonify::from_json(
'{
    "a": [
        1,
        [
            2,
            [
                3
            ]
        ]
    ]
}', simplify = FALSE
) # except w/ the default `simplify = TRUE`

target1 <- list(
  a = list(
    1L, 
    list(
      2L, 
      list(
        3L
      )
    )
  )
)

stopifnot(identical(test1, target1))


test2 <- jsonify::from_json(
'{
    "a": [
        1,
        [
            2,
            [
                3,
                4
            ]
        ]
    ]
}'
)

target2 <- list( 
  a = list(      
    1L,          
    list(        
      2L,        
      c(3L, 
        4L)
    )
  )
)

stopifnot(identical(test2, target2))

[dcooley]

Yes.

But, now you've brought it up, I'm having a slight crisis-of-confidence on how the "simplifying" is working vs what I intended it to do back when it was written...

[dcooley]

This is seemingly a one-off edge case that needs handling, rather than something systemic

## incorrect
from_json( '[1,[2]]' )
#      [,1]
# [1,]    1
# [2,]    2

Given these exmples are correct

## right
from_json( '[[1,2],2]' )
# [[1]]
# [1] 1 2
# 
# [[2]]
# [1] 2
from_json( '[5,[6,7]]' )
# [[1]]
# [1] 5
# 
# [[2]]
# [1] 6 7

[dcooley]

Notes for me to remember in the morning:

• commit c920159 has a potential soultion
• if it works (all tests currently pass), I can remove my sequential_array_counter