encoding: use getutf8char in toutf8b

This correctly avoids the ambiguity of U+FFFD already present in the
input and similar confusion by working a character at a time.

mercurial/encoding.py

@@ -470,16 +470,19 def toutf8b(s):
         s.decode('utf-8')
         return s
     except UnicodeDecodeError:
-        # surrogate-encode any characters that don't round-trip
-        s2 = s.decode('utf-8', 'ignore').encode('utf-8')
+        pass
+
         r = ""
         pos = 0
-        for c in s:
-            if s2[pos:pos + 1] == c:
+    l = len(s)
+    while pos < l:
+        try:
+            c = getutf8char(s, pos)
+            pos += len(c)
+        except UnicodeDecodeError:
+            c = unichr(0xdc00 + ord(s[pos])).encode('utf-8')
+            pos += 1
                 r += c
-                pos += 1
-            else:
-                r += unichr(0xdc00 + ord(c)).encode('utf-8')
         return r
 
 def fromutf8b(s):