Escape Rd comments in markdown

Fixes #879

NEWS.md

@@ -1,5 +1,10 @@
 # roxygen2 (development version)
 
+* Rd comments (`%`) are automatically escaped in markdown text. Existing
+  code that uses `\%` is automatically repaired; if you're generating
+  more exotic sequences like `\\%` you'll need to put them inside 
+  backticks (#879)
+
 * As well as storing roxygen options in the `Roxygen` field of the 
   `DESCRIPTION` you can now also store them in `man/roxygen/meta.R` (#889).
 

R/markdown-link.R

@@ -137,7 +137,7 @@ parse_link <- function(destination, contents) {
     )
 
   } else {
-    contents <- gsub("%", "\\\\%", mdxml_link_text(contents))
+    contents <- mdxml_link_text(contents)
 
     list(
       paste0(

R/markdown.R

@@ -40,7 +40,7 @@ mdxml_node_to_rd <- function(xml, tag) {
 
     code_block = paste0("\\preformatted{", gsub("%", "\\\\%", xml_text(xml)), "}"),
     paragraph = paste0("\n\n", mdxml_children_to_rd(xml, tag)),
-    text = xml_text(xml),
+    text = escape_comment(xml_text(xml)),
     code = paste0("\\code{", gsub("%", "\\\\%", xml_text(xml)), "}"),
     emph = paste0("\\emph{", mdxml_children_to_rd(xml, tag), "}"),
     strong = paste0("\\strong{", mdxml_children_to_rd(xml, tag), "}"),
@@ -63,11 +63,11 @@ mdxml_node_to_rd <- function(xml, tag) {
 
 mdxml_unknown <- function(xml, tag) {
   roxy_tag_warning(tag, "Unknown xml node: ", xml_name(xml))
-  xml_text(xml)
+  escape_comment(xml_text(xml))
 }
 mdxml_unsupported <- function(xml, tag, feature) {
   roxy_tag_warning(tag, "Use of ", feature, " is not currently supported")
-  xml_text(xml)
+  escape_comment(xml_text(xml))
 }
 
 
@@ -108,7 +108,7 @@ mdxml_link <- function(xml) {
   if (!is.null(link)) {
     paste0(link, collapse = "")
   } else if (dest == "" || dest == xml_text(xml)) {
-    paste0("\\url{", xml_text(xml), "}")
+    paste0("\\url{", escape_comment(xml_text(xml)), "}")
   } else {
     paste0("\\href{", dest, "}{", mdxml_link_text(contents), "}")
   }
@@ -120,11 +120,21 @@ mdxml_link <- function(xml) {
 mdxml_link_text <- function(xml_contents) {
   text <- xml_text(xml_contents)
   text[xml_name(xml_contents) %in% c("linebreak", "softbreak")] <- " "
-  paste0(text, collapse = "")
+  escape_comment(paste0(text, collapse = ""))
 }
 
 mdxml_image = function(xml) {
   dest <- xml_attr(xml, "destination")
   title <- xml_attr(xml, "title")
   paste0("\\figure{", dest, "}{", title, "}")
 }
+
+
+escape_comment <- function(x) {
+  # Strip existing escaping - this is admittedly a hack and makes the
+  # state of % escaping more confusing, but it only affects markdown, and
+  # I think if you do want a literal \% you can do `\%`
+  x <- gsub("\\%", "%", x, fixed = TRUE)
+  x <- gsub("%", "\\%", x, fixed = TRUE)
+  x
+}

tests/testthat/test-markdown.R

@@ -344,6 +344,13 @@ So far so good. \\preformatted{ *these are not
   expect_equal(get_tag(out1, "description")[[2]], desc1)
 })
 
+test_that("% is automatically escaped", {
+  expect_equal(markdown("5%"), "5\\%")
+
+  # Even if it was escaped before
+  expect_equal(markdown("5\\%"), "5\\%")
+})
+
 test_that("% and $ and _ are not unescaped", {
   out1 <- roc_proc_text(rd_roclet(), "
     #' Title