_split_header(): If we have a header which is a byte string containing

8-bit data, we cannot split it safely, so return the original string
unchanged.

_is8bitstring(): Helper function which returns True when we have a
byte string that contains non-ascii characters (i.e. mysterious 8-bit
data).

diff --git a/Lib/email/Generator.py b/Lib/email/Generator.py
index 7f05218d48599493f01d278556ef442dafdc8811..58e2f91d622a4c1ef07baecfa50fcbc822abe599 100644 (file)
--- a/Lib/email/Generator.py
+++ b/Lib/email/Generator.py
@@ -8,7 +8,7 @@ import time
 import re
 import random
 
-from types import ListType
+from types import ListType, StringType
 from cStringIO import StringIO
 
 from email.Header import Header
@@ -35,6 +35,14 @@ SPACE8 = ' ' * 8
 
 fcre = re.compile(r'^From ', re.MULTILINE)
 
+def _is8bitstring(s):
+    if isinstance(s, StringType):
+        try:
+            unicode(s, 'us-ascii')
+        except UnicodeError:
+            return True
+    return False
+
 
 \f
 class Generator:
@@ -174,6 +182,14 @@ class Generator:
             # No line was actually longer than maxheaderlen characters, so
             # just return the original unchanged.
             return text
+        # If we have raw 8bit data in a byte string, we have no idea what the
+        # encoding is.  I think there is no safe way to split this string.  If
+        # it's ascii-subset, then we could do a normal ascii split, but if
+        # it's multibyte then we could break the string.  There's no way to
+        # know so the least harm seems to be to not split the string and risk
+        # it being too long.
+        if _is8bitstring(text):
+            return text
         # The `text' argument already has the field name prepended, so don't
         # provide it here or the first line will get folded too short.
         h = Header(text, maxlinelen=maxheaderlen,