From d30791a38166538be60a134196f1d2675275017d Mon Sep 17 00:00:00 2001
From: John MacFarlane <jgm@berkeley.edu>
Date: Mon, 22 Feb 2021 14:17:22 -0800
Subject: Fall back to latin1 if UTF-8 decoding fails...

...when handling URL argument served with no charset in the mime type.
The assumption is that most pages that don't specify a charset
in the mime type are either UTF-8 or latin1.  I think that's a good
assumption, though I'm not sure.
---
 src/Text/Pandoc/App.hs | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/src/Text/Pandoc/App.hs b/src/Text/Pandoc/App.hs
index 59af029b5..40fb34834 100644
--- a/src/Text/Pandoc/App.hs
+++ b/src/Text/Pandoc/App.hs
@@ -1,3 +1,4 @@
+{-# LANGUAGE LambdaCase          #-}
 {-# LANGUAGE OverloadedStrings   #-}
 {-# LANGUAGE CPP                 #-}
 {-# LANGUAGE ScopedTypeVariables #-}
@@ -352,7 +353,12 @@ readURI src = do
     Just "UTF-8"      -> return $ UTF8.toText bs
     Just "ISO-8859-1" -> return $ T.pack $ B8.unpack bs
     Just charset      -> throwError $ PandocUnsupportedCharsetError charset
-    Nothing           -> return $ UTF8.toText bs
+    Nothing           -> liftIO $ -- try first as UTF-8, then as latin1
+                          E.catch (return $! UTF8.toText bs)
+                                  (\case
+                                      TSE.DecodeError{} ->
+                                        return $ T.pack $ B8.unpack bs
+                                      e -> E.throwIO e)
 
 readFile' :: MonadIO m => FilePath -> m BL.ByteString
 readFile' "-" = liftIO BL.getContents
-- 
cgit v1.2.3