[SPARK-25821][SQL] Remove SQLContext methods deprecated in 1.4

## What changes were proposed in this pull request?

Remove SQLContext methods deprecated in 1.4

## How was this patch tested?

Existing tests.

Closes #22815 from srowen/SPARK-25821.

Authored-by: Sean Owen <[email protected]>
Signed-off-by: Sean Owen <[email protected]>

Author: srowen

R/pkg/NAMESPACE

@@ -420,13 +420,11 @@ export("as.DataFrame",
        "currentDatabase",
        "dropTempTable",
        "dropTempView",
-       "jsonFile",
        "listColumns",
        "listDatabases",
        "listFunctions",
        "listTables",
        "loadDF",
-       "parquetFile",
        "read.df",
        "read.jdbc",
        "read.json",

R/pkg/R/SQLContext.R

@@ -343,7 +343,6 @@ setMethod("toDF", signature(x = "RDD"),
 #' path <- "path/to/file.json"
 #' df <- read.json(path)
 #' df <- read.json(path, multiLine = TRUE)
-#' df <- jsonFile(path)
 #' }
 #' @name read.json
 #' @method read.json default
@@ -363,51 +362,6 @@ read.json <- function(x, ...) {
   dispatchFunc("read.json(path)", x, ...)
 }
 
-#' @rdname read.json
-#' @name jsonFile
-#' @method jsonFile default
-#' @note jsonFile since 1.4.0
-jsonFile.default <- function(path) {
-  .Deprecated("read.json")
-  read.json(path)
-}
-
-jsonFile <- function(x, ...) {
-  dispatchFunc("jsonFile(path)", x, ...)
-}
-
-#' JSON RDD
-#'
-#' Loads an RDD storing one JSON object per string as a SparkDataFrame.
-#'
-#' @param sqlContext SQLContext to use
-#' @param rdd An RDD of JSON string
-#' @param schema A StructType object to use as schema
-#' @param samplingRatio The ratio of simpling used to infer the schema
-#' @return A SparkDataFrame
-#' @noRd
-#' @examples
-#'\dontrun{
-#' sparkR.session()
-#' rdd <- texFile(sc, "path/to/json")
-#' df <- jsonRDD(sqlContext, rdd)
-#'}
-
-# TODO: remove - this method is no longer exported
-# TODO: support schema
-jsonRDD <- function(sqlContext, rdd, schema = NULL, samplingRatio = 1.0) {
-  .Deprecated("read.json")
-  rdd <- serializeToString(rdd)
-  if (is.null(schema)) {
-    read <- callJMethod(sqlContext, "read")
-    # samplingRatio is deprecated
-    sdf <- callJMethod(read, "json", callJMethod(getJRDD(rdd), "rdd"))
-    dataFrame(sdf)
-  } else {
-    stop("not implemented")
-  }
-}
-
 #' Create a SparkDataFrame from an ORC file.
 #'
 #' Loads an ORC file, returning the result as a SparkDataFrame.
@@ -434,6 +388,7 @@ read.orc <- function(path, ...) {
 #' Loads a Parquet file, returning the result as a SparkDataFrame.
 #'
 #' @param path path of file to read. A vector of multiple paths is allowed.
+#' @param ... additional external data source specific named properties.
 #' @return SparkDataFrame
 #' @rdname read.parquet
 #' @name read.parquet
@@ -454,20 +409,6 @@ read.parquet <- function(x, ...) {
   dispatchFunc("read.parquet(...)", x, ...)
 }
 
-#' @param ... argument(s) passed to the method.
-#' @rdname read.parquet
-#' @name parquetFile
-#' @method parquetFile default
-#' @note parquetFile since 1.4.0
-parquetFile.default <- function(...) {
-  .Deprecated("read.parquet")
-  read.parquet(unlist(list(...)))
-}
-
-parquetFile <- function(x, ...) {
-  dispatchFunc("parquetFile(...)", x, ...)
-}
-
 #' Create a SparkDataFrame from a text file.
 #'
 #' Loads text files and returns a SparkDataFrame whose schema starts with

R/pkg/tests/fulltests/test_sparkSQL.R

@@ -628,14 +628,10 @@ test_that("read/write json files", {
     jsonPath3 <- tempfile(pattern = "jsonPath3", fileext = ".json")
     write.json(df, jsonPath3)
 
-    # Test read.json()/jsonFile() works with multiple input paths
+    # Test read.json() works with multiple input paths
     jsonDF1 <- read.json(c(jsonPath2, jsonPath3))
     expect_is(jsonDF1, "SparkDataFrame")
     expect_equal(count(jsonDF1), 6)
-    # Suppress warnings because jsonFile is deprecated
-    jsonDF2 <- suppressWarnings(jsonFile(c(jsonPath2, jsonPath3)))
-    expect_is(jsonDF2, "SparkDataFrame")
-    expect_equal(count(jsonDF2), 6)
 
     unlink(jsonPath2)
     unlink(jsonPath3)
@@ -655,20 +651,6 @@ test_that("read/write json files - compression option", {
   unlink(jsonPath)
 })
 
-test_that("jsonRDD() on a RDD with json string", {
-  sqlContext <- suppressWarnings(sparkRSQL.init(sc))
-  rdd <- parallelize(sc, mockLines)
-  expect_equal(countRDD(rdd), 3)
-  df <- suppressWarnings(jsonRDD(sqlContext, rdd))
-  expect_is(df, "SparkDataFrame")
-  expect_equal(count(df), 3)
-
-  rdd2 <- flatMap(rdd, function(x) c(x, x))
-  df <- suppressWarnings(jsonRDD(sqlContext, rdd2))
-  expect_is(df, "SparkDataFrame")
-  expect_equal(count(df), 6)
-})
-
 test_that("test tableNames and tables", {
   count <- count(listTables())
 
@@ -2658,17 +2640,14 @@ test_that("read/write Parquet files", {
     expect_is(df2, "SparkDataFrame")
     expect_equal(count(df2), 3)
 
-    # Test write.parquet/saveAsParquetFile and read.parquet/parquetFile
+    # Test write.parquet/saveAsParquetFile and read.parquet
     parquetPath2 <- tempfile(pattern = "parquetPath2", fileext = ".parquet")
     write.parquet(df, parquetPath2)
     parquetPath3 <- tempfile(pattern = "parquetPath3", fileext = ".parquet")
     suppressWarnings(saveAsParquetFile(df, parquetPath3))
     parquetDF <- read.parquet(c(parquetPath2, parquetPath3))
     expect_is(parquetDF, "SparkDataFrame")
     expect_equal(count(parquetDF), count(df) * 2)
-    parquetDF2 <- suppressWarnings(parquetFile(parquetPath2, parquetPath3))
-    expect_is(parquetDF2, "SparkDataFrame")
-    expect_equal(count(parquetDF2), count(df) * 2)
 
     # Test if varargs works with variables
     saveMode <- "overwrite"

docs/sparkr.md

@@ -709,8 +709,12 @@ You can inspect the search path in R with [`search()`](https://stat.ethz.ch/R-ma
 
 ## Upgrading to SparkR 2.3.1 and above
 
- - In SparkR 2.3.0 and earlier, the `start` parameter of `substr` method was wrongly subtracted by one and considered as 0-based. This can lead to inconsistent substring results and also does not match with the behaviour with `substr` in R. In version 2.3.1 and later, it has been fixed so the `start` parameter of `substr` method is now 1-base. As an example, `substr(lit('abcdef'), 2, 4))` would result to `abc` in SparkR 2.3.0, and the result would be `bcd` in SparkR 2.3.1.
+ - In SparkR 2.3.0 and earlier, the `start` parameter of `substr` method was wrongly subtracted by one and considered as 0-based. This can lead to inconsistent substring results and also does not match with the behaviour with `substr` in R. In version 2.3.1 and later, it has been fixed so the `start` parameter of `substr` method is now 1-based. As an example, `substr(lit('abcdef'), 2, 4))` would result to `abc` in SparkR 2.3.0, and the result would be `bcd` in SparkR 2.3.1.
 
 ## Upgrading to SparkR 2.4.0
 
  - Previously, we don't check the validity of the size of the last layer in `spark.mlp`. For example, if the training data only has two labels, a `layers` param like `c(1, 3)` doesn't cause an error previously, now it does.
+
+## Upgrading to SparkR 3.0.0
+
+ - The deprecated methods `parquetFile`, `jsonRDD` and `jsonFile` in `SQLContext` have been removed. Use `read.parquet` and `read.json`.