Error on inputs greater than length 1

jimhester · jimhester · commit fbdbfe6da2f0 · 2020-04-23T08:17:57.000-04:00
diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,7 @@
 # xml2 (development version)
 
+* `read_html()` and `read_xml()` now error if passed strings of length greater than one (#121)
+
 * `read_xml.raw()` had an inadvertent regression in 1.3.0 and is now again fixed (#300)
 
 * Compilation fix on macOS 10.15.4 (@kevinushey, #296)
diff --git a/R/xml_parse.R b/R/xml_parse.R
@@ -98,6 +98,10 @@ read_xml.character <- function(x, encoding = "", ..., as_html = FALSE,
     stop("Document is empty", call. = FALSE)
   }
 
+  if (length(x) > 1) {
+    stop("`x` must be a string of length 1", call. = FALSE)
+  }
+
   options <- parse_options(options, xml_parse_options())
   if (grepl("<|>", x)) {
     read_xml.raw(charToRaw(enc2utf8(x)), "UTF-8", ..., as_html = as_html, options = options)
diff --git a/tests/testthat/test-read-xml.R b/tests/testthat/test-read-xml.R
@@ -102,3 +102,7 @@ test_that("read_html works with non-ASCII encodings", {
     "<!DOCTYPE html PUBLIC \"-//W3C//DTD HTML 4.0 Transitional//EN\" \"http://www.w3.org/TR/REC-html40/loose.dtd\">\n<html><body>\U2019</body></html>\n")
 })
 
+test_that("read_xml and read_html fail with > 1 input", {
+  expect_error(read_xml(c("foo", "bar")), "`x` must be a string of length 1")
+  expect_error(read_html(c("foo", "bar")), "`x` must be a string of length 1")
+})
