Fix #47 (#48)

* NA aware

* Fix #47

Still not working

* Use R wrapper to filter `NULL`

- TODO Ask the Rcpp community what is the best practice to deal with
this.

```r
bench::mark(url_decode2(rep("Hello%20World", 5000)))
```

Took 1.53ms.

Author: chainsawriot

R/RcppExports.R

@@ -77,16 +77,8 @@ Rcpp_ada_get_protocol <- function(url_vec) {
     .Call(`_adaR_Rcpp_ada_get_protocol`, url_vec)
 }
 
-#' Function to percent-decode characters in URLs
-#'
-#' Similar to [utils::URLdecode]
-#'
-#' @param url a character vector
-#' @export
-#' @examples
-#' url_decode2("Hello%20World")
-url_decode2 <- function(url) {
-    .Call(`_adaR_url_decode2`, url)
+Rcpp_url_decode2 <- function(url) {
+    .Call(`_adaR_Rcpp_url_decode2`, url)
 }
 
 url_reverse <- function(urls) {

R/parse.R

@@ -30,6 +30,22 @@ ada_url_parse <- function(url, decode = TRUE) {
     df
 }
 
+#' Function to percent-decode characters in URLs
+#'
+#' Similar to [utils::URLdecode]
+#'
+#' @param url a character vector
+#' @export
+#' @examples
+#' url_decode2("Hello%20World")
+#' @export
+url_decode2 <- function(url) {
+    if (is.null(url)) {
+        return(character(0))
+    }
+    Rcpp_url_decode2(url)
+}
+
 ## NA/NULL-aware utils::URLdecode, hopefully without great performance impact
 .URLdecode <- function(URL) {
     if (is.null(URL)) {

man/url_decode2.Rd

@@ -219,14 +219,14 @@ BEGIN_RCPP
     return rcpp_result_gen;
 END_RCPP
 }
-// url_decode2
-CharacterVector url_decode2(const CharacterVector& url);
-RcppExport SEXP _adaR_url_decode2(SEXP urlSEXP) {
+// Rcpp_url_decode2
+CharacterVector Rcpp_url_decode2(CharacterVector& url);
+RcppExport SEXP _adaR_Rcpp_url_decode2(SEXP urlSEXP) {
 BEGIN_RCPP
     Rcpp::RObject rcpp_result_gen;
     Rcpp::RNGScope rcpp_rngScope_gen;
-    Rcpp::traits::input_parameter< const CharacterVector& >::type url(urlSEXP);
-    rcpp_result_gen = Rcpp::wrap(url_decode2(url));
+    Rcpp::traits::input_parameter< CharacterVector& >::type url(urlSEXP);
+    rcpp_result_gen = Rcpp::wrap(Rcpp_url_decode2(url));
     return rcpp_result_gen;
 END_RCPP
 }
@@ -262,7 +262,7 @@ static const R_CallMethodDef CallEntries[] = {
     {"_adaR_Rcpp_ada_get_pathname", (DL_FUNC) &_adaR_Rcpp_ada_get_pathname, 1},
     {"_adaR_Rcpp_ada_get_search", (DL_FUNC) &_adaR_Rcpp_ada_get_search, 1},
     {"_adaR_Rcpp_ada_get_protocol", (DL_FUNC) &_adaR_Rcpp_ada_get_protocol, 1},
-    {"_adaR_url_decode2", (DL_FUNC) &_adaR_url_decode2, 1},
+    {"_adaR_Rcpp_url_decode2", (DL_FUNC) &_adaR_Rcpp_url_decode2, 1},
     {"_adaR_url_reverse", (DL_FUNC) &_adaR_url_reverse, 1},
     {NULL, NULL, 0}
 };

src/RcppExports.cpp

@@ -2,17 +2,8 @@
 
 using namespace Rcpp;
 
-//' Function to percent-decode characters in URLs
-//'
-//' Similar to [utils::URLdecode]
-//'
-//' @param url a character vector
-//' @export
-//' @examples
-//' url_decode2("Hello%20World")
-// [[Rcpp::export]]
-CharacterVector url_decode2(const CharacterVector& url) {
-  return sapply(url, [](const String& u) {
+
+std::string decode(String u) {
     std::string input = u;
     std::string output;
     size_t i = 0;
@@ -28,9 +19,21 @@ CharacterVector url_decode2(const CharacterVector& url) {
         i += 3;
       }
     }
-
     return output;
-  });
+}
+
+// [[Rcpp::export]]
+CharacterVector Rcpp_url_decode2(CharacterVector& url) {
+  unsigned int input_size = url.size();
+  CharacterVector output(input_size);
+  for (unsigned int i = 0; i < input_size; i++) {
+    if (url[i] == NA_STRING) {
+      output[i] = NA_STRING;
+    } else {
+      output[i] = decode(url[i]);
+    }
+  }
+  return output;
 }
 
 std::string str_reverse(std::string x) {
@@ -51,4 +54,4 @@ CharacterVector url_reverse(CharacterVector& urls) {
   }
 
   return output;
-}
+}

src/urldecode.cpp

@@ -18,3 +18,9 @@ test_that("cpp implementation is correct", {
     dec <- "https://www.google.de/maps/@47.6647302,9.1389738,11z?entry=ttu"
     expect_equal(url_decode2(enc), dec)
 })
+
+test_that("corners", {
+    expect_error(url_decode2(NULL), NA)
+    expect_error(url_decode2(NA), NA)
+    expect_error(url_decode2(c("?q=%E3%83%89%E3%82%A4%E3%83%84", NA))[2], NA)
+})