From f379edc4ad372f96d4a5cc7cd38292f095dfdf35 Mon Sep 17 00:00:00 2001
From: John MacFarlane <jgm@berkeley.edu>
Date: Thu, 1 Nov 2018 16:08:27 -0700
Subject: HTML writer: use character entities references when possible for
 HTML5.

---
 MANUAL.txt                      |  2 +-
 src/Text/Pandoc/Writers/HTML.hs | 10 ++++++++--
 test/command/ascii.md           |  2 +-
 3 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/MANUAL.txt b/MANUAL.txt
index 5a4710474..cda70a34c 100644
--- a/MANUAL.txt
+++ b/MANUAL.txt
@@ -859,7 +859,7 @@ Options affecting specific writers {.options}
 `--ascii`
 
 :   Use only ASCII characters in output.  Currently supported for
-    XML and HTML formats (which use numerical entities instead of
+    XML and HTML formats (which use entities instead of
     UTF-8 when this option is selected), groff ms
     (which use hexadecimal escapes), and to a limited degree
     LaTeX (which uses standard commands for accented
diff --git a/src/Text/Pandoc/Writers/HTML.hs b/src/Text/Pandoc/Writers/HTML.hs
index 46f754226..4d15c827c 100644
--- a/src/Text/Pandoc/Writers/HTML.hs
+++ b/src/Text/Pandoc/Writers/HTML.hs
@@ -75,7 +75,8 @@ import Text.Pandoc.Templates
 import Text.Pandoc.Walk
 import Text.Pandoc.Writers.Math
 import Text.Pandoc.Writers.Shared
-import Text.Pandoc.XML (escapeStringForXML, fromEntities, toEntities)
+import Text.Pandoc.XML (escapeStringForXML, fromEntities,
+                        toEntities, toHtml5Entities)
 #if MIN_VERSION_blaze_markup(0,6,3)
 #else
 import Text.Blaze.Internal (preEscapedString, preEscapedText)
@@ -206,7 +207,12 @@ writeHtmlString' :: PandocMonad m
                  => WriterState -> WriterOptions -> Pandoc -> m Text
 writeHtmlString' st opts d = do
   (body, context) <- evalStateT (pandocToHtml opts d) st
-  (if writerPreferAscii opts then toEntities else id) <$>
+  (if writerPreferAscii opts
+      then
+        if stHtml5 st
+           then toHtml5Entities
+           else toEntities
+      else id) <$>
     case writerTemplate opts of
        Nothing -> return $ renderHtml' body
        Just tpl -> do
diff --git a/test/command/ascii.md b/test/command/ascii.md
index 4956ae14e..1cbb2bec4 100644
--- a/test/command/ascii.md
+++ b/test/command/ascii.md
@@ -2,7 +2,7 @@
 pandoc -t html --ascii
 äéıå
 ^D
-<p>&#228;&#233;&#305;&#229;</p>
+<p>&auml;&eacute;&inodot;&aring;</p>
 ```
 
 ```
-- 
cgit v1.2.3