Merge pull request #13 from jeremystan/json_struct

Json struct

NAMESPACE

@@ -28,6 +28,7 @@ import(assertthat)
 import(dplyr)
 import(jsonlite)
 importFrom(purrr,map)
+importFrom(purrr,map2)
 importFrom(purrr,map_chr)
 importFrom(purrr,map_int)
 importFrom(purrr,map_lgl)

R/json_structure.r

@@ -13,24 +13,29 @@
 #'   \code{document.id} 1L if \code{x} is a single JSON string, otherwise the
 #'   index of \code{x}.
 #'
-#'   \code{parent.id} the string identifier of the parent node for this child
+#'   \code{parent.id} the string identifier of the parent node for this child.
 #'
 #'   \code{level} what level of the hierarchy this child resides at, starting
-#'   at 1L and incrementing for each level of nesting.
+#'   at \code{0L} for the root and incrementing for each level of nested
+#'   array or object.
 #'
 #'   \code{index} what index of the parent object / array this child resides
-#'   at (from \code{gather_array} for arrays)
+#'   at (from \code{gather_array} for arrays).
 #'
 #'   \code{child.id} a unique ID for this leaf in this document, represented
 #'   as <parent>.<index> where <parent> is the ID for the parent and <index>
-#'   is this index
+#'   is this index.
+#'
+#'   \code{seq} the sequence of keys / indices that led to this child
+#'   (parents that are arrays are excluded) as a list, where character strings
+#'   denote objects and integers denote array positions
 #'
 #'   \code{key} if this is the value of an object, what was the key that it
-#'   is listed under (from \code{gather_keys})
+#'   is listed under (from \code{gather_keys}).
 #'
-#'   \code{type} the type of this object (from \code{json_types})
+#'   \code{type} the type of this object (from \code{json_types}).
 #'
-#'   \code{length} the length of this object (from \code{json_lengths})
+#'   \code{length} the length of this object (from \code{json_lengths}).
 #'
 #' @export
 #' @examples
@@ -51,7 +56,7 @@ json_structure <- function(x) {
   # Create initial structure for top level
   structure <- json_structure_init(x)
 
-  this_level <- 1L
+  this_level <- 0L
   while(structure %>% should_json_structure_expand_more(this_level)) {
 
     structure <- rbind_tbl_json(
@@ -72,9 +77,10 @@ json_structure_init <- function(x) {
   x %>%
     mutate(
       parent.id = NA_character_,
-      level = 1L,
+      level = 0L,
       index = 1L,
       child.id = "1",
+      seq = replicate(n(), list()),
       key = NA_character_
     ) %>%
     json_types %>%
@@ -101,6 +107,7 @@ json_structure_empty <- function() {
       level = integer(0),
       index = integer(0),
       child.id = character(0),
+      seq = list(),
       key = character(0),
       type = factor(character(0), levels = allowed_json_types),
       length = integer(0)
@@ -141,6 +148,7 @@ json_structure_objects <- function(s) {
     transmute(
       document.id,
       parent.id = child.id,
+      seq,
       level = level + 1L
     ) %>%
     gather_keys %>%
@@ -152,9 +160,12 @@ json_structure_objects <- function(s) {
     group_by(parent.id) %>%
     mutate(index = 1L:n()) %>%
     ungroup %>%
-    mutate(child.id = paste(parent.id, index, sep = ".")) %>%
+    mutate(
+      child.id = paste(parent.id, index, sep = "."),
+      seq = map2(seq, key, c)
+    ) %>%
     select(
-      document.id, parent.id, level, index, child.id, key, type, length
+      document.id, parent.id, level, index, child.id, seq, key, type, length
     )
 
   # Reconstruct tbl_json object
@@ -169,14 +180,19 @@ json_structure_arrays <- function(s) {
     transmute(
       document.id,
       parent.id = child.id,
+      seq,
       level = level + 1L
     ) %>%
     gather_array("index") %>%
     json_types %>%
     json_lengths %>%
-    mutate(child.id = paste(parent.id, index, sep = ".")) %>%
+    mutate(
+      child.id = paste(parent.id, index, sep = "."),
+      seq = map2(seq, index, c)
+    ) %>%
     transmute(
-      document.id, parent.id, level, index, child.id, key = NA_character_, type, length
+      document.id, parent.id, level, index, child.id,
+      seq, key = NA_character_, type, length
     )
 
 }
@@ -190,4 +206,3 @@ rbind_tbl_json <- function(x, y) {
   )
 
 }
-

R/tidyjson-package.r

@@ -4,6 +4,6 @@
 #' @docType package
 #' @import assertthat
 #' @import jsonlite
-#' @importFrom purrr map map_lgl map_int map_chr
+#' @importFrom purrr map map_lgl map_int map_chr map2
 #' @import dplyr
 NULL

tests/testthat/test-json_structure.r

@@ -8,9 +8,10 @@ test_that("simple string works", {
       data_frame(
         document.id = 1L,
         parent.id = NA_character_,
-        level = 1L,
+        level = 0L,
         index = 1L,
         child.id = "1",
+        seq = list(list()),
         key = NA_character_,
         type = "string" %>% factor(levels = allowed_json_types),
         length = 1L
@@ -29,9 +30,10 @@ test_that("simple object works", {
       data_frame(
         document.id = c(1L, 1L),
         parent.id = c(NA_character_, "1"),
-        level = c(1L, 2L),
+        level = c(0L, 1L),
         index = c(1L, 1L),
         child.id = c("1", "1.1"),
+        seq = list(list(), list("key")),
         key = c(NA_character_, "key"),
         type = c("object", "string") %>% factor(levels = allowed_json_types),
         length = c(1L, 1L)
@@ -50,9 +52,10 @@ test_that("simple array works", {
       data_frame(
         document.id = c(1L, 1L, 1L),
         parent.id = c(NA_character_, "1", "1"),
-        level = c(1L, 2L, 2L),
+        level = c(0L, 1L, 1L),
         index = c(1L, 1L, 2L),
         child.id = c("1", "1.1", "1.2"),
+        seq = list(list(), list(1L), list(2L)),
         key = rep(NA_character_, 3),
         type = c("array", "number", "number") %>% factor(levels = allowed_json_types),
         length = c(2L, 1L, 1L)
@@ -71,9 +74,10 @@ test_that("nested object works", {
       data_frame(
         document.id = c(1L, 1L, 1L),
         parent.id = c(NA_character_, "1", "1.1"),
-        level = c(1L, 2L, 3L),
+        level = c(0L, 1L, 2L),
         index = c(1L, 1L, 1L),
         child.id = c("1", "1.1", "1.1.1"),
+        seq = list(list(), list("k1"), list("k1", "k2")),
         key = c(NA_character_, "k1", "k2"),
         type = c("object", "object", "string") %>% factor(levels = allowed_json_types),
         length = c(1L, 1L, 1L)
@@ -94,9 +98,10 @@ test_that("works with empty values appropriately", {
       data_frame(
         document.id = 1L,
         parent.id = NA_character_,
-        level = 1L,
+        level = 0L,
         index = 1L,
         child.id = "1",
+        seq = list(list()),
         key = NA_character_,
         type = "null" %>% factor(levels = allowed_json_types),
         length = 0L
@@ -115,9 +120,10 @@ test_that("works with tbl_json already", {
       data_frame(
         document.id = c(1L, 2L),
         parent.id = rep(NA_character_, 2),
-        level = rep(1L, 2),
+        level = rep(0L, 2),
         index = rep(1L, 2),
         child.id = rep("1", 2),
+        seq = list(list(), list()),
         key = rep(NA_character_, 2),
         type = rep("string", 2) %>% factor(levels = allowed_json_types),
         length = rep(1L, 2)
@@ -127,3 +133,15 @@ test_that("works with tbl_json already", {
   )
 
 })
+
+test_that("seq works for a deeply nested sequence", {
+
+  expect_identical(
+    '{"a": {"2": [1, {"3": "value"}] } }' %>%
+      json_structure %>%
+      `[[`("seq") %>%
+      `[[`(6),
+    list("a", "2", 2L, "3")
+  )
+
+})