How parse JSON formatted rows in a dataframe in R?

Question

I have a data frame where each row contains JSON formatted input. How can I parse these 3 rows into one table. I tired following commands, but it shows error.

a<-data.frame(v1=c('[{"ID":1,"Type":"Honda","Key":{"Service":"Destan","Name":"John"},"Fields":{"Price":23.005,"Cost":10,"DIFF":"13.005"}}]',
                   '[{"ID":2,"Type":"BMW","UpdateType":"Every hour","Key":{"Service":"Destan","Name":"Mark"},"Fields":{"Price":2.005,"Cost":1,"DIFF":"1.005"}}]',
                   '[{"ID":1,"Type":"Honda","Key":{"Service":"Destan","Name":"John"},"Fields":{"Price":13.005,"Cost":4,"DIFF":"9.005"}}]'))
a<-paste0(a$v1, collapse = ",")
l <- fromJSON(paste0(a$v1, collapse = ","))

Error is Error in a$v1 : $ operator is invalid for atomic vectors

Answer 1

If you're trying to concat the rows of json into a larger valid json string, I think you'll also need surrounding [ and ]. Like:

missing_array_ends <- paste0(a$v1, collapse=',')
json_in_r <-  glue::glue("[{missing_array_ends}]") %>% fromJSON

But, I think another complication is the nested structure of the json. Key and Field have multiple sub entries. But you can flatten (unlist) them if you work on each row (lapply). lapply returns a list of items that you want to be rows in a dataframe. dplyr::bind_rows combines them back into a single dataframe

lapply(a$v1, function(j) unlist(fromJSON(j))) %>%
  bind_rows

creates

# A tibble: 3 x 8
  ID    Type  Key.Service Key.Name Fields.Price Fields.Cost Fields.DIFF
  <chr> <chr> <chr>       <chr>    <chr>        <chr>       <chr>      
1 1     Honda Destan      John     23.005       10          13.005     
2 2     BMW   Destan      Mark     2.005        1           1.005      
3 1     Honda Destan      John     13.005       4           9.005      
# … with 1 more variable: UpdateType <chr>