Consistently use surrogate escape by default.

Milter/__init__.py

@@ -328,7 +328,7 @@ class Base(object):
   # or trap utf-8 conversion exception, etc.
   def envfrom_bytes(self,*b):
     try:
-      s = (v.decode(encoding='utf-8') for v in b)
+      s = (v.decode(encoding='utf-8',errors='surrogateescape') for v in b)
     except UnicodeDecodeError: s = b
     return self.envfrom(fld,*s)
   ## Called when the SMTP client says MAIL FROM. Called by the
@@ -349,7 +349,7 @@ class Base(object):
   # or trap utf-8 conversion exception, etc.
   def envrcpt_bytes(self,*b):
     try:
-      s = (v.decode(encoding='utf-8') for v in b)
+      s = (v.decode(encoding='utf-8',errors='surrogateescape') for v in b)
     except UnicodeDecodeError: s = b
     return self.envrcpt(fld,*s)
   ## Called when the SMTP client says RCPT TO. Called by the
@@ -377,7 +377,7 @@ class Base(object):
   # to pass bytes to @link #header the header callback @endlink instead.
   def header_bytes(self,fld,val):
     try:
-      s = val.decode(encoding='utf-8')
+      s = val.decode(encoding='utf-8',errors='surrogateescape')
     except UnicodeDecodeError: s = val
     return self.header(fld,s)
   ## Called for each header field in the message body.

mime.py

@@ -548,7 +548,7 @@ if __name__ == '__main__':
     return Milter.CONTINUE
 
   for fname in sys.argv[1:]:
-    fp = open(fname,'rb')
+    with open(fname,'rb') as fp:
       msg = message_from_file(fp)
     email.iterators._structure(msg)
     check_attachments(msg,_list_attach)

sample.py

@@ -73,7 +73,7 @@ class sampleMilter(Milter.Milter):
 
       # even if we wanted the Taiwanese spam, we can't read Chinese
       # (delete if you read chinese mail)
-      print('val=',val)
+      print('val=',val.encode(errors='surrogateescape'))
       if val.startswith('=?big5') or val.startswith('=?ISO-2022-JP'):
         self.log('REJECT: %s: %s' % (name,val))
 	#self.setreply('550','','Go away spammer')