From 738f4b1ab5845ba2879f3d5fff2e60767ea691e5 Mon Sep 17 00:00:00 2001
From: gagolews <egzaltowany@gmail.com>
Date: Sat, 21 Apr 2018 00:14:39 +0200
Subject: [PATCH] stri_na2empty

---
 NAMESPACE                        |  2 ++
 NEWS                             |  2 ++
 R/utils.R                        | 51 +++++++++++++++++++++++++++++++-
 man/stri_list2matrix.Rd          |  4 +++
 man/stri_na2empty.Rd             | 25 ++++++++++++++++
 man/stri_remove_empty.Rd         | 31 +++++++++++++++++++
 src/stri_encoding_conversion.cpp |  2 +-
 7 files changed, 115 insertions(+), 2 deletions(-)
 create mode 100644 man/stri_na2empty.Rd
 create mode 100644 man/stri_remove_empty.Rd

diff --git a/NAMESPACE b/NAMESPACE
index 57375cb66..2fc412760 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -147,6 +147,7 @@ export(stri_match_first)
 export(stri_match_first_regex)
 export(stri_match_last)
 export(stri_match_last_regex)
+export(stri_na2empty)
 export(stri_numbytes)
 export(stri_opts_brkiter)
 export(stri_opts_collator)
@@ -164,6 +165,7 @@ export(stri_rand_shuffle)
 export(stri_rand_strings)
 export(stri_read_lines)
 export(stri_read_raw)
+export(stri_remove_empty)
 export(stri_replace)
 export(stri_replace_all)
 export(stri_replace_all_charclass)
diff --git a/NEWS b/NEWS
index ad914caa2..7f3fb6f17 100644
--- a/NEWS
+++ b/NEWS
@@ -15,6 +15,8 @@ when input was empty.
 
 * [NEW FEATURE] #289: `stri_flatten` gained `na_empty` `omit_empty` arguments.
 
+* [NEW FEATURE] New functions: `stri_remove_empty`, `stri_na2empty`
+
 * [WARN] Removed `-Wparentheses` warnings in `icu55/common/cstring.h:38:63`
 and `icu55/i18n/windtfmt.cpp` in the ICU4C 55.1 bundle.
 
diff --git a/R/utils.R b/R/utils.R
index 5d42f1af6..2a1c942d0 100644
--- a/R/utils.R
+++ b/R/utils.R
@@ -1,5 +1,5 @@
 ## This file is part of the 'stringi' package for R.
-## Copyright (c) 2013-2017, Marek Gagolewski and other contributors.
+## Copyright (c) 2013-2018, Marek Gagolewski and other contributors.
 ## All rights reserved.
 ##
 ## Redistribution and use in source and binary forms, with or without
@@ -81,3 +81,52 @@
 stri_list2matrix <- function(x, byrow=FALSE, fill=NA_character_, n_min=0) {
    .Call(C_stri_list2matrix, x, byrow, stri_enc_toutf8(fill), n_min)
 }
+
+
+#' @title
+#' Replace NAs with empty strings
+#'
+#' @description
+#' This function replaces all missing values with empty strings
+#'
+#' @param x a character vector
+#'
+#' @return
+#' Always returns a character vector.
+#'
+#' @examples
+#' stri_na2empty(c("a", NA, "", "b"))
+#'
+#' @family utils
+#' @export
+stri_na2empty <- function(x) {
+   x <- stri_enc_toutf8(x)
+   x[is.na(x)] <- ""
+   x
+}
+
+#' @title
+#' Remove all empty strings from a character vector
+#'
+#' @description
+#' This function removes all empty strings from a character vector.
+#'
+#' @param x a character vector
+#' @param na_empty should missing values be treated as empty strings?
+#'
+#' @return
+#' Always returns a character vector.
+#'
+#' @examples
+#' stri_remove_empty(stri_na2empty(c("a", NA, "", "b")))
+#' stri_remove_empty(c("a", NA, "", "b"))
+#' stri_remove_empty(c("a", NA, "", "b"), TRUE)
+#'
+#' @family utils
+#' @export
+stri_remove_empty <- function(x, na_empty=FALSE) {
+   x <- stri_enc_toutf8(x)
+   if (identical(na_empty, TRUE))
+      x[is.na(x)] <- ""
+   x[!stri_isempty(x)]
+}
diff --git a/man/stri_list2matrix.Rd b/man/stri_list2matrix.Rd
index 6866569a8..3c7ee726a 100644
--- a/man/stri_list2matrix.Rd
+++ b/man/stri_list2matrix.Rd
@@ -55,3 +55,7 @@ stri_list2matrix(list("a", c("b", "c")), fill="")
 stri_list2matrix(list("a", c("b", "c")), fill="", n_min=5)
 
 }
+\seealso{
+Other utils: \code{\link{stri_na2empty}},
+  \code{\link{stri_remove_empty}}
+}
diff --git a/man/stri_na2empty.Rd b/man/stri_na2empty.Rd
new file mode 100644
index 000000000..e3a2b75a4
--- /dev/null
+++ b/man/stri_na2empty.Rd
@@ -0,0 +1,25 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/utils.R
+\name{stri_na2empty}
+\alias{stri_na2empty}
+\title{Replace NAs with empty strings}
+\usage{
+stri_na2empty(x)
+}
+\arguments{
+\item{x}{a character vector}
+}
+\value{
+Always returns a character vector.
+}
+\description{
+This function replaces all missing values with empty strings
+}
+\examples{
+stri_na2empty(c("a", NA, "", "b"))
+
+}
+\seealso{
+Other utils: \code{\link{stri_list2matrix}},
+  \code{\link{stri_remove_empty}}
+}
diff --git a/man/stri_remove_empty.Rd b/man/stri_remove_empty.Rd
new file mode 100644
index 000000000..ac7915277
--- /dev/null
+++ b/man/stri_remove_empty.Rd
@@ -0,0 +1,31 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/utils.R
+\name{stri_remove_empty}
+\alias{stri_remove_empty}
+\title{Remove all empty strings from a character vector}
+\usage{
+stri_remove_empty(x, na_empty = FALSE)
+}
+\arguments{
+\item{x}{a character vector}
+
+\item{na_empty}{should missing values be treated as empty strings?}
+}
+\value{
+Always returns a character vector.
+}
+\description{
+This function removes all empty strings from a character vector.
+}
+\examples{
+stri_remove_empty(stri_na2empty(c("a", NA, "", "b")))
+stri_remove_empty(c("a", NA, "", "b"))
+stri_remove_empty(c("a", NA, "", "b"), TRUE)
+
+}
+\seealso{
+stri_isempty
+
+Other utils: \code{\link{stri_list2matrix}},
+  \code{\link{stri_na2empty}}
+}
diff --git a/src/stri_encoding_conversion.cpp b/src/stri_encoding_conversion.cpp
index a2a189e03..a4f842395 100644
--- a/src/stri_encoding_conversion.cpp
+++ b/src/stri_encoding_conversion.cpp
@@ -1,5 +1,5 @@
 /* This file is part of the 'stringi' package for R.
- * Copyright (c) 2013-2017, Marek Gagolewski and other contributors.
+ * Copyright (c) 2013-2018, Marek Gagolewski and other contributors.
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without