Fix #20

Author: ironholds

NAMESPACE

@@ -13,6 +13,7 @@ export(port)
 export(query)
 export(scheme)
 export(suffix_extract)
+export(url_compose)
 export(url_decode)
 export(url_encode)
 export(url_parameters)

NEWS

@@ -5,6 +5,8 @@ NEW FEATURES
 * Jay Jacobs' "tldextract" functionality has been merged with urltools, and can be accessed
 with "suffix_extract"
 
+* At Nicolas Coutin's suggestion, url_compose - url_parse in reverse - has been introduced.
+
 BUG FIXES
 
 * A bug in which fragments could not be retrieved (and were incorrectly identified as parameters)

R/RcppExports.R

@@ -113,11 +113,35 @@ url_parameters <- function(urls, parameter_names) {
 #'@examples
 #'url_parse("https://en.wikipedia.org/wiki/Article")
 #'
-#'@seealso \code{url_parameters} for extracting values associated with particular keys in a URL's
-#'query string.
+#'@seealso \code{\link{url_parameters}} for extracting values associated with particular keys in a URL's
+#'query string, and \code{\link{url_compose}}, which is \code{url_parse} in reverse.
 #'
 #'@export
 url_parse <- function(urls) {
     .Call('urltools_url_parse', PACKAGE = 'urltools', urls)
 }
 
+#'@title Recompose Parsed URLs
+#'
+#'@description Sometimes you want to take a vector of URLs, parse them, perform
+#'some operations and then rebuild them. \code{url_compose} takes a data.frame produced
+#'by \code{\link{url_parse}} and rebuilds it into a vector of full URLs (or: URLs as full
+#'as the vector initially thrown into url_parse).
+#'
+#'This is currently a `beta` feature; please do report bugs if you find them.
+#'
+#'@param parsed_urls a data.frame sourced from \code{\link{url_parse}}
+#'
+#'@seealso \code{\link{scheme}} and other accessors, which you may want to
+#'run URLs through before composing them to modify individual values.
+#'
+#'@examples
+#'#Parse a URL and compose it
+#'url <- "http://en.wikipedia.org"
+#'url_compose(url_parse(url))
+#'
+#'@export
+url_compose <- function(parsed_urls) {
+    .Call('urltools_url_compose', PACKAGE = 'urltools', parsed_urls)
+}
+

man/url_compose.Rd

@@ -0,0 +1,29 @@
+% Generated by roxygen2 (4.1.1): do not edit by hand
+% Please edit documentation in R/RcppExports.R
+\name{url_compose}
+\alias{url_compose}
+\title{Recompose Parsed URLs}
+\usage{
+url_compose(parsed_urls)
+}
+\arguments{
+\item{parsed_urls}{a data.frame sourced from \code{\link{url_parse}}}
+}
+\description{
+Sometimes you want to take a vector of URLs, parse them, perform
+some operations and then rebuild them. \code{url_compose} takes a data.frame produced
+by \code{\link{url_parse}} and rebuilds it into a vector of full URLs (or: URLs as full
+as the vector initially thrown into url_parse).
+
+This is currently a `beta` feature; please do report bugs if you find them.
+}
+\examples{
+#Parse a URL and compose it
+url <- "http://en.wikipedia.org"
+url_compose(url_parse(url))
+}
+\seealso{
+\code{\link{scheme}} and other accessors, which you may want to
+run URLs through before composing them to modify individual values.
+}
+

man/url_parse.Rd

@@ -30,7 +30,7 @@ HTTP requests from R), but not for large-scale analysis.
 url_parse("https://en.wikipedia.org/wiki/Article")
 }
 \seealso{
-\code{url_parameters} for extracting values associated with particular keys in a URL's
-query string.
+\code{\link{url_parameters}} for extracting values associated with particular keys in a URL's
+query string, and \code{\link{url_compose}}, which is \code{url_parse} in reverse.
 }
 

src/RcppExports.cpp

@@ -75,3 +75,14 @@ BEGIN_RCPP
     return __result;
 END_RCPP
 }
+// url_compose
+std::vector < std::string > url_compose(DataFrame parsed_urls);
+RcppExport SEXP urltools_url_compose(SEXP parsed_urlsSEXP) {
+BEGIN_RCPP
+    Rcpp::RObject __result;
+    Rcpp::RNGScope __rngScope;
+    Rcpp::traits::input_parameter< DataFrame >::type parsed_urls(parsed_urlsSEXP);
+    __result = Rcpp::wrap(url_compose(parsed_urls));
+    return __result;
+END_RCPP
+}

src/compose.cpp

@@ -0,0 +1,66 @@
+#include "compose.h"
+
+bool compose::emptycheck(std::string element){
+  if(element.size() == 0){
+    return false;
+  }
+  return true;
+}
+
+std::string compose::compose_single(std::string scheme, std::string domain, std::string port, std::string path,
+                                    std::string parameter, std::string fragment){
+  
+  std::string output;
+  
+  if(emptycheck(scheme)){
+    output += scheme + "://";
+  }
+  
+  if(emptycheck(domain)){
+    output += domain;
+    
+    if(emptycheck(port)){
+      output += ":" + port;
+    } else {
+      output += "/";
+    }
+  }
+  
+  if(emptycheck(path)){
+    output += path;
+  }
+  
+  if(emptycheck(parameter)){
+    output += "?";
+    output += parameter;
+  }
+  
+  if(emptycheck(fragment)){
+    output += "#";
+    output += fragment;
+  }
+  
+  return output;
+}
+
+std::vector < std::string > compose::compose_multiple(DataFrame parsed_urls){
+  
+  std::vector < std::string > schemes = parsed_urls["scheme"];
+  std::vector < std::string > domains = parsed_urls["domain"];
+  std::vector < std::string > ports = parsed_urls["port"];
+  std::vector < std::string > paths = parsed_urls["path"];
+  std::vector < std::string > parameters = parsed_urls["parameter"];
+  std::vector < std::string > fragments = parsed_urls["fragment"];
+  
+  std::vector < std::string > output(schemes.size());
+  
+  for(unsigned int i = 0; i < schemes.size(); i++){
+    if((i % 10000) == 0){
+      Rcpp::checkUserInterrupt();
+    }
+    output[i] = compose_single(schemes[i], domains[i], ports[i], paths[i], parameters[i],
+                               fragments[i]);
+  }
+  
+  return output;
+}

src/compose.h

@@ -0,0 +1,21 @@
+#include <Rcpp.h>
+using namespace Rcpp;
+
+#ifndef __COMPOSE_INCLUDED__
+#define __COMPOSE_INCLUDED__
+
+class compose {
+  
+private:
+  
+  bool emptycheck(std::string element);
+  
+  std::string compose_single(std::string scheme, std::string domain, std::string port, std::string path,
+                             std::string parameter, std::string fragment);
+  
+public:
+  
+  std::vector < std::string > compose_multiple(DataFrame parsed_urls);
+};
+
+#endif

src/urltools.cpp

@@ -1,6 +1,7 @@
 #include <Rcpp.h>
 #include "encoding.h"
 #include "parsing.h"
+#include "compose.h"
 
 using namespace Rcpp;
 
@@ -173,13 +174,39 @@ List url_parameters(std::vector < std::string > urls, std::vector < std::string
 //'@examples
 //'url_parse("https://en.wikipedia.org/wiki/Article")
 //'
-//'@seealso \code{url_parameters} for extracting values associated with particular keys in a URL's
-//'query string.
+//'@seealso \code{\link{url_parameters}} for extracting values associated with particular keys in a URL's
+//'query string, and \code{\link{url_compose}}, which is \code{url_parse} in reverse.
 //'
 //'@export
 //[[Rcpp::export]]
 DataFrame url_parse(std::vector < std::string > urls){
   std::vector < std::string >& urls_ptr = urls;
   parsing p_inst;
   return p_inst.parse_to_df(urls_ptr);
+}
+
+//'@title Recompose Parsed URLs
+//'
+//'@description Sometimes you want to take a vector of URLs, parse them, perform
+//'some operations and then rebuild them. \code{url_compose} takes a data.frame produced
+//'by \code{\link{url_parse}} and rebuilds it into a vector of full URLs (or: URLs as full
+//'as the vector initially thrown into url_parse).
+//'
+//'This is currently a `beta` feature; please do report bugs if you find them.
+//'
+//'@param parsed_urls a data.frame sourced from \code{\link{url_parse}}
+//'
+//'@seealso \code{\link{scheme}} and other accessors, which you may want to
+//'run URLs through before composing them to modify individual values.
+//'
+//'@examples
+//'#Parse a URL and compose it
+//'url <- "http://en.wikipedia.org"
+//'url_compose(url_parse(url))
+//'
+//'@export
+//[[Rcpp::export]]
+std::vector < std::string > url_compose(DataFrame parsed_urls){
+  compose c_inst;
+  return c_inst.compose_multiple(parsed_urls);
 }

tests/testthat/test_parsing.R

@@ -46,3 +46,8 @@ test_that("Parameter parsing works where the parameter appears earlier in the UR
   expect_that(url$to[1], equals("05/04/2015"))
 })
 
+test_that("Composing works",{
+  url <- "http://foo.bar.baz/qux/"
+  amended_url <- url_compose(url_parse(url))
+  expect_that(url, equals(amended_url))
+})