Fix #18

Author: ironholds

NEWS

@@ -2,9 +2,14 @@ Version 1.2.0 [WIP]
 -------------------------------------------------------------------------
 
 NEW FEATURES
-*Jay Jacobs' "tldextract" functionality has been merged with urltools, and can be accessed
+* Jay Jacobs' "tldextract" functionality has been merged with urltools, and can be accessed
 with "suffix_extract"
 
+BUG FIXES
+
+* A bug in which fragments could not be retrieved (and were incorrectly identified as parameters)
+has been fixed. Thanks to Nicolas Coutin for reporting it and providing a reproducible example.
+
 Version 1.1.1
 -------------------------------------------------------------------------
 BUG FIXES

src/parsing.cpp

@@ -49,7 +49,7 @@ std::string parsing::path(std::string& url){
       return output;
     }
     output = url.substr(0,fragment);
-    url = url.substr(fragment+1);
+    url = url.substr(fragment);
     return output;
   }
 

tests/testthat/test_parsing.R

@@ -26,6 +26,11 @@ test_that("Check parsing can handle missing elements", {
   expect_that(data$fragment[1], equals("ending"))
 })
 
+test_that("Parsing does not up and die and misplace the fragment",{
+  data <- url_parse("http://www.yeastgenome.org/locus/S000005366/overview#protein")
+  expect_that(data$fragment[1], equals("protein"))
+})
+
 test_that("Parameter parsing can handle multiple, non-existent and pre-trailing parameters",{
   urls <- c("https://www.google.com:80/foo.php?api_params=parsable&this_parameter=selfreferencing&hiphop=awesome",
             "https://www.google.com:80/foo.php?api_params=parsable&this_parameter=selfreferencing&hiphop=awesome#foob",
@@ -39,4 +44,5 @@ test_that("Parameter parsing works where the parameter appears earlier in the UR
   url <- url_parameters("www.housetrip.es/tos-de-vacaciones/geo?from=01/04/2015&guests=4&to=05/04/2015","to")
   expect_that(ncol(url), equals(1))
   expect_that(url$to[1], equals("05/04/2015"))
-})
+})
+