Add files via upload

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: eurlex
 Type: Package
 Title: Retrieve Data on European Union Law
-Version: 0.3.5
+Version: 0.3.6
 Authors@R: c(person(given = "Michal",
              family = "Ovadek",
              role = c("aut", "cre", "cph"),
@@ -10,7 +10,6 @@ Authors@R: c(person(given = "Michal",
 Description: Access to data on European Union laws and court decisions made easy with pre-defined 'SPARQL' queries and 'GET' requests.
 License: GPL-3
 Encoding: UTF-8
-LazyData: true
 Depends:
     R (>= 3.4.0)
 Imports:
@@ -19,6 +18,7 @@ Imports:
     xml2,
     tidyr,
     httr,
+    curl,
     rvest,
     rlang,
     stringr,

NEWS.md

@@ -1,3 +1,14 @@
+# eurlex 0.3.6
+
+## Major changes
+
+- `elx_run_query()` now fails gracefully in presence of internet/server problems
+- `elx_fetch_data()` now automatically fixes urls with parentheses (e.g. "32019H1115(01)" used to fail)
+
+## Minor changes
+
+- minor fixes to vignette
+
 # eurlex 0.3.5
 
 ## Major changes

R/elx_curia_list.R

@@ -8,7 +8,8 @@
 #' @param parse If `TRUE`, references to cases and appeals are parsed out from `case_info` into separate columns
 #' @return
 #' A data frame containing case identifiers and information as character columns. Where the case id
-#' contains a hyperlink to Eur-Lex, the CELEX identifier is retrieved as well.
+#' contains a hyperlink to Eur-Lex, the CELEX identifier is retrieved as well. Hyperlinks to Eur-Lex
+#' disappeared from more recent cases.
 #' @importFrom rlang .data
 #' @export
 #' @examples

R/elx_fetch_data.R

@@ -22,6 +22,19 @@ elx_fetch_data <- function(url, type = c("title","text","ids"),
 
   language <- paste(language_1,", ",language_2,";q=0.8, ",language_3,";q=0.7", sep = "")
 
+  if (stringr::str_detect(url,"celex")){
+
+    clx <- stringr::str_extract(url, "(?<=celex\\/).*") %>%
+      stringr::str_replace_all("\\(","%28") %>%
+      stringr::str_replace_all("\\)","%29") %>%
+      stringr::str_replace_all("\\/","%2F")
+
+    url <- paste("http://publications.europa.eu/resource/celex/",
+                 clx,
+                 sep = "")
+
+  }
+
   if (type == "title"){
 
     response <- httr::GET(url=url,

R/elx_run_query.R

@@ -20,9 +20,7 @@ elx_run_query <- function(query = "", endpoint = "http://publications.europa.eu/
 
   curlready <- paste(endpoint,"?query=",gsub("\\+","%2B", utils::URLencode(query, reserved = TRUE)), sep = "")
 
-  sparql_response <- httr::GET(url = curlready,
-                               httr::add_headers('Accept' = 'application/sparql-results+xml')
-  )
+  sparql_response <- graceful_http(curlready)
 
   sparql_response_parsed <- sparql_response %>%
     elx_parse_xml()
@@ -31,3 +29,49 @@ elx_run_query <- function(query = "", endpoint = "http://publications.europa.eu/
 
 }
 
+#' Fail http call gracefully
+#'
+#' @importFrom rlang .data
+#'
+#' @noRd
+#'
+
+graceful_http <- function(remote_file) {
+
+  try_GET <- function(x, ...) {
+    tryCatch(
+      httr::GET(url = x,
+                #httr::timeout(1000000000),
+                httr::add_headers('Accept' = 'application/sparql-results+xml')),
+      error = function(e) conditionMessage(e),
+      warning = function(w) conditionMessage(w)
+    )
+  }
+
+  is_response <- function(x) {
+    class(x) == "response"
+  }
+
+  # First check internet connection
+  if (!curl::has_internet()) {
+    message("No internet connection.")
+    return(invisible(NULL))
+  }
+
+  # Then try for timeout problems
+  resp <- try_GET(remote_file)
+  if (!is_response(resp)) {
+    message(resp)
+    return(invisible(NULL))
+  }
+
+  # Then stop if status > 400
+  if (httr::http_error(resp)) {
+    httr::message_for_status(resp)
+    return(invisible(NULL))
+  }
+
+  return(resp)
+
+}
+

README.md

@@ -42,6 +42,12 @@ Please consider contributing to the maintanance and development of the package b
 
 ## Latest changes
 
+### eurlex 0.3.6
+
+- `elx_run_query()` now fails gracefully in presence of internet/server problems
+- `elx_fetch_data()` now automatically fixes urls with parentheses (e.g. "32019H1115(01)" used to fail)
+- minor fixes to vignette
+
 ### eurlex 0.3.5
 
 - it is now possible to select all resource types available with `elx_make_query(resource_type = "any")`. Since there are nearly 1 million CELEX codes, use with discretion and expect long execution times