0.1.0 Release

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: datasetjson
 Type: Package
 Title: Read and Write CDISC Dataset JSON Files
-Version: 0.0.1
+Version: 0.1.0
 Authors@R: c(
  person(given = "Mike",
  family = "Stackhouse",
@@ -24,8 +24,8 @@ Roxygen: list(markdown = TRUE)
 RoxygenNote: 7.2.3
 Depends: R (>= 3.5)
 Imports:
- jsonlite (>= 1.8.7),
- jsonvalidate (>= 1.3.2)
+ jsonlite (>= 1.8.0),
+ jsonvalidate (>= 1.3.1)
 Suggests:
  testthat (>= 2.1.0),
  knitr,

NEWS.md

@@ -1,3 +1,12 @@
-# datasetjson 0.0.0.9999
+# datasetjson 0.1.0
+
+- Capability to read and validate Dataset JSON files from URLs has been added (#8)
+- Remove autoset of fileOID using output path (#3)
+- Don't auto-populate optional attributes with NA (#16)
+- Push dependency versions back (#18)
+- Default `pretty` parameter on `write_dataset_json()` to false (#20)
+
+# datasetjson 0.0.1
+
+Initial development version of datasetjson, introducing core objects, readers and writers.
 
-Intial development version of datasetjson, introducing core objects, readers and writers.

R/data_metadata.R

@@ -21,7 +21,7 @@
 #' data_meta_updated <- set_metadata_version(data_meta_updated, "MDV.MSGv2.0.SDTMIG.3.3.SDTM.1.7")
 #' data_meta_updated <- set_study_oid(data_meta_updated, "SOMESTUDY")
 #'
-data_metadata <- function(study = "NA", metadata_version = "NA", metadata_ref = "NA") {
+data_metadata <- function(study = NULL, metadata_version = NULL, metadata_ref = NULL) {
 
  x <- list(
  studyOID = study,

R/file_metadata.R

@@ -7,6 +7,8 @@
 #' file."
 #' @param sys_version sourceSystemVersion, defined as "The version of the
 #' sourceSystem"
+#' @param file_oid fileOID parameter, defined as "A unique identifier for this
+#' file."
 #' @param version Dataset JSON schema version being used
 #'
 #' @return file_metadata object
@@ -26,17 +28,17 @@
 #' file_meta_updated <- set_file_oid(file_meta, "/some/path")
 #' file_meta_updated <- set_originator(file_meta_updated, "Some Org")
 #' file_meta_updated <- set_source_system(file_meta_updated, "source system", "1.0")
-file_metadata <- function(originator="NA", sys = "NA", sys_version = "NA", version = "1.0.0") {
+file_metadata <- function(originator=NULL, sys = NULL, sys_version = NULL, file_oid = NULL, version = "1.0.0") {
 
  if (!(version %in% c("1.0.0"))) {
  stop("Unsupported version specified - currently only version 1.0.0 is supported", call.=FALSE)
  }
 
  x <- list(
- "creationDateTime"= get_datetime(),
+ "creationDateTime"= character(),
  "datasetJSONVersion"= version,
- "fileOID" = character(),
- "asOfDateTime" = character(),
+ "fileOID" = file_oid,
+ "asOfDateTime" = NULL, # Not sure we want this to exist?
  "originator" = originator,
  "sourceSystem" = sys,
  "sourceSystemVersion" = sys_version
@@ -59,10 +61,20 @@ get_datetime <- function() {
  format(Sys.time(), "%Y-%m-%dT%H:%M:%S")
 }
 
-#' Set source system information
+#' File Metadata Setters
+#'
+#' Set information about the file and source system used to generate the Dataset
+#' JSON object.
+#'
+#' @details
 #'
-#' Set information about the source system used to generate the Dataset JSON
-#' object.
+#' The fileOID parameter should be structured following description outlined in
+#' the ODM V2.0 specification. "FileOIDs should be universally unique if at all
+#' possible. One way to ensure this is to prefix every FileOID with an internet
+#' domain name owned by the creator of the ODM file or database (followed by a
+#' forward slash, "/"). For example,
+#' FileOID="BestPharmaceuticals.com/Study5894/1" might be a good way to denote
+#' the first file in a series for study 5894 from Best Pharmaceuticals."
 #'
 #' @param x datasetjson object
 #' @param sys sourceSystem parameter, defined as "The computer system or

R/read_dataset_json.R

@@ -1,9 +1,10 @@
 #' Read a Dataset JSON to datasetjson object
 #'
-#' This function validated a dataset JSON file on disk against the Dataset JSON schema, and if valid
-#' returns a datasetjson object
+#' This function validates a dataset JSON file against the Dataset JSON schema,
+#' and if valid returns a datasetjson object. The Dataset JSON file can be
+#' either a file path on disk of a URL which contains the Dataset JSON file.
 #'
-#' @param file File path on disk, or a pre-loaded Dataset JSON file in a single element character string
+#' @param file File path or URL of a Dataset JSON file
 #'
 #' @return datasetjson object
 #' @export
@@ -12,15 +13,24 @@
 #' # Read from disk
 #' \dontrun{
 #' dat <- read_dataset_json("path/to/file.json")
+#' # Read file from URL
+#' dat <- dataset_json('https://www.somesite.com/file.json')
 #' }
 #'
 #' # Read from an already imported character vector
 #' ds_json <- dataset_json(iris, "IG.IRIS", "IRIS", "Iris", iris_items)
 #' js <- write_dataset_json(ds_json)
 #' dat <- read_dataset_json(js)
 read_dataset_json <- function(file) {
+
+ if (path_is_url(file)) {
+ file_contents <- read_from_url(file)
+ } else {
+ file_contents <- readLines(file)
+ }
+
  # Validate the input file against the schema
- valid <- jsonvalidate::json_validate(file, schema_1_0_0, engine="ajv")
+ valid <- jsonvalidate::json_validate(file_contents, schema = schema_1_0_0, engine="ajv")
 
  if (!valid) {
  stop(paste0(c("Dataset JSON file is invalid per the JSON schema. ",
@@ -29,7 +39,7 @@ read_dataset_json <- function(file) {
  }
 
  # Read the file and convert to datasetjson object
- ds_json <- jsonlite::fromJSON(file)
+ ds_json <- jsonlite::fromJSON(file_contents)
 
  # Pull the object out with a lot of assumptions because the format has already
  # been validated

R/utils.R

@@ -54,3 +54,64 @@ set_col_attr <- function(nm, d, attr, items) {
  attr(x, attr) <- items[items$name == nm,][[attr]]
  x
 }
+
+#' Get the index of nulls in a list
+#'
+#' @param x A list
+#'
+#' @return Integer vector of indices
+#' @noRd
+get_null_inds <- function(x) {
+ which(vapply(x, is.null, FUN.VALUE = TRUE))
+}
+
+#' Remove nulls from a Dataset JSON object
+#'
+#' Only targets the file and data metadata to pull off optional elements
+#'
+#' @param x A Dataset JSON object
+#'
+#' @return A Dataset JSON object
+#' @noRd
+remove_nulls <- function(x) {
+
+ # Specifically target the data metadata
+ dm_nulls <- get_null_inds(x[[get_data_type(x)]])
+ if (length(dm_nulls) > 0) {
+ x[[get_data_type(x)]] <- x[[get_data_type(x)]][-dm_nulls]
+ }
+
+ # Top level
+ fm_nulls <- get_null_inds(x)
+ if (length(fm_nulls) > 0) {
+ x <- x[-fm_nulls]
+ }
+
+ x
+}
+
+#' Check if given path is a URL
+#'
+#' @param path character string
+#'
+#' @return Boolean
+#' @noRd
+path_is_url <- function(path) {
+ grepl("^((http|ftp)s?|sftp|file)://", path)
+}
+
+#' Read data from a URL
+#'
+#' This function will let you pull data that's provided from a simple curl of a
+#' URL
+#'
+#' @param path valid URL string
+#'
+#' @return Contents of URL
+#' @noRd
+read_from_url <- function(path) {
+ con <- url(path, method = "libcurl")
+ x <- readLines(con, warn=FALSE) # the EOL warning shouldn't be a problem for readers
+ close(con)
+ x
+}

R/validate_dataset_json.R

@@ -1,10 +1,11 @@
 #' Validate a Dataset JSON file
 #'
-#' This function calls `jsonvalidate::json_validate()` directly, with the parameters
-#' necessary to retrieve the error information of an invalid JSON file per the
-#' Dataset JSON schema.
+#' This function calls `jsonvalidate::json_validate()` directly, with the
+#' parameters necessary to retrieve the error information of an invalid JSON
+#' file per the Dataset JSON schema.
 #'
-#' @param x Path to a Dataset JSON file or a character vector holding JSON text
+#' @param x File path or URL of a Dataset JSON file, or a character vector
+#' holding JSON text
 #'
 #' @return A data frame
 #' @export
@@ -13,14 +14,22 @@
 #'
 #' \dontrun{
 #' validate_dataset_json('path/to/file.json')
+#' validate_dataset_json('https://www.somesite.com/file.json')
 #' }
 #'
 #' ds_json <- dataset_json(iris, "IG.IRIS", "IRIS", "Iris", iris_items)
 #' js <- write_dataset_json(ds_json)
 #'
 #' validate_dataset_json(js)
 validate_dataset_json <- function(x) {
- v <- jsonvalidate::json_validate(x, schema_1_0_0, engine="ajv", verbose=TRUE)
+ # If contents are a URL then pull out the content
+ if (path_is_url(x)) {
+ js <- read_from_url(x)
+ } else {
+ js <- x
+ }
+
+ v <- jsonvalidate::json_validate(js, schema_1_0_0, engine="ajv", verbose=TRUE)
  if (!v) {
  warning("File contains errors!")
  return(attr(v, 'errors'))

R/write_dataset_json.R

@@ -16,22 +16,19 @@
 #' \dontrun{
 #' write_dataset_json(ds_json, "path/to/file.json")
 #' }
-write_dataset_json <- function(x, file, pretty=TRUE) {
+write_dataset_json <- function(x, file, pretty=FALSE) {
  stopifnot_datasetjson(x)
 
- # Populate the as-of datetime
- x[['asOfDateTime']] <- get_datetime()
+ # Populate the creation datetime
+ x[['creationDateTime']] <- get_datetime()
+
+ x <- remove_nulls(x)
 
  if (!missing(file)) {
  # Make sure the output path exists
  if(!dir.exists(dirname(file))) {
  stop("Folder supplied to `file` does not exist", call.=FALSE)
  }
-
- # Attach the file OID
- x <- set_file_oid(x, tools::file_path_sans_ext(file))
- } else{
- x <- set_file_oid(x, "NA")
  }
 
  # Create the JSON text

README.Rmd

@@ -7,11 +7,7 @@ knitr::opts_chunk$set(echo = TRUE)
 library(datasetjson)
 ```
 
-# datasetjson
-
-<!-- Please do not edit the README.md file as it is auto-generated after PR merges. Only edit the README.Rmd file -->
-<!-- The purpose of this is to enable dynamic links using dyn_link function above to access devel/main admiral homepage respectively -->
-<!-- To test this in your feature branch use code: rmarkdown::render("README.Rmd", output_format ="md_document") -->
+# **datasetjson** <img src="man/figures/logo.svg" align="right" alt="" width="120" />
 
 <!-- badges: start -->
 [<img src="https://img.shields.io/codecov/c/github/atorus-research/datasetjson">](https://app.codecov.io/gh/atorus-research/datasetjson)
@@ -67,7 +63,7 @@ write_dataset_json(ds_updated, file = "./iris.json")
 Or if you don't provide a file path, the JSON text will return directly. 
 
 ```{r write_print}
-js_text <- write_dataset_json(ds_updated)
+js_text <- write_dataset_json(ds_updated, pretty=TRUE)
 cat(js_text)
 ```
 
@@ -87,6 +83,8 @@ print(attr(dat$Sepal.Width, "type"))
 ```
 Note that Dataset JSON is an early CDISC standard and is still subject to change, as as such this package will be updated. Backwards compatibility will be enforced once the standard itself is more stable. Until then, it is not recommended to use this package within production activities. 
 
+# [<img src="man/figures/cdisc.png" alt="" width="120" />](https://www.cdisc.org/)
+
 ## Acknowledgements
 
 Thank you to Ben Straub and Eric Simms (GSK) for help and input during the original CDISC Dataset JSON hackathon that motivated this work.

README.md

@@ -1,9 +1,6 @@
 
-# datasetjson
+# **datasetjson** <img src="man/figures/logo.svg" align="right" alt="" width="120" />
 
-<!-- Please do not edit the README.md file as it is auto-generated after PR merges. Only edit the README.Rmd file -->
-<!-- The purpose of this is to enable dynamic links using dyn_link function above to access devel/main admiral homepage respectively -->
-<!-- To test this in your feature branch use code: rmarkdown::render("README.Rmd", output_format ="md_document") -->
 <!-- badges: start -->
 
 [<img src="https://img.shields.io/codecov/c/github/atorus-research/datasetjson">](https://app.codecov.io/gh/atorus-research/datasetjson)
@@ -73,15 +70,14 @@ write_dataset_json(ds_updated, file = "./iris.json")
 Or if you don’t provide a file path, the JSON text will return directly.
 
 ``` r
-js_text <- write_dataset_json(ds_updated)
+js_text <- write_dataset_json(ds_updated, pretty=TRUE)
 cat(js_text)
 ```
 
  ## {
- ## "creationDateTime": "2023-09-15T17:57:31",
+ ## "creationDateTime": "2023-09-25T12:20:23",
  ## "datasetJSONVersion": "1.0.0",
- ## "fileOID": "NA",
- ## "asOfDateTime": "2023-09-15T17:57:31",
+ ## "fileOID": "/some/path",
  ## "originator": "Some Org",
  ## "sourceSystem": "source system",
  ## "sourceSystemVersion": "1.0",
@@ -174,7 +170,7 @@ attached as attributes on the data frame itself:
 print(attr(dat, "creationDateTime"))
 ```
 
- ## [1] "2023-09-15T17:57:31"
+ ## [1] "2023-09-25T12:20:23"
 
 ``` r
 print(attr(dat$Sepal.Length, "OID"))
@@ -194,6 +190,8 @@ compatibility will be enforced once the standard itself is more stable.
 Until then, it is not recommended to use this package within production
 activities.
 
+# [<img src="man/figures/cdisc.png" alt="" width="120" />](https://www.cdisc.org/)
+
 ## Acknowledgements
 
 Thank you to Ben Straub and Eric Simms (GSK) for help and input during