From: Simon Marlow <marlowsd@gmail.com>
Date: Thu, 2 Jul 2009 08:14:15 +0000 (+0000)
Subject: a byte between 0x80 and 0xBF is illegal immediately (#3341)
X-Git-Tag: ghc-darcs-git-switchover~376
X-Git-Url: http://git.megacz.com/?a=commitdiff_plain;h=97080d8293694fbf4e51556352e3fd196a43bdde;p=ghc-base.git

a byte between 0x80 and 0xBF is illegal immediately (#3341)
---

diff --git a/GHC/IO/Encoding/UTF8.hs b/GHC/IO/Encoding/UTF8.hs
index 0efb187..99928e0 100644
--- a/GHC/IO/Encoding/UTF8.hs
+++ b/GHC/IO/Encoding/UTF8.hs
@@ -83,7 +83,7 @@ utf8_decode
                            if not (validate3 c0 c1 c2) then invalid else do
                            writeCharBuf oraw ow (chr3 c0 c1 c2)
                            loop (ir+3) (ow+1)
-                  | otherwise ->
+                  | c0 >= 0xf0 ->
                            if iw - ir < 4 then done ir ow else do
                            c1 <- readWord8Buf iraw (ir+1)
                            c2 <- readWord8Buf iraw (ir+2)
@@ -91,6 +91,8 @@ utf8_decode
                            if not (validate4 c0 c1 c2 c3) then invalid else do
                            writeCharBuf oraw ow (chr4 c0 c1 c2 c3)
                            loop (ir+4) (ow+1)
+                  | otherwise ->
+                           invalid
          where
            invalid = if ir > ir0 then done ir ow else ioe_decodingError