This commit was manufactured by cvs2svn to create tag 'milter-0_8_5'.

Sprout from master 2005-12-29 22:46:07 UTC Stuart Gathman <stuart@gathman.org> 'Documentation updates.'
Cherrypick from bmsi 2005-05-31 18:23:49 UTC Stuart Gathman <stuart@gathman.org> 'Development changes since 0.7.2':
    README
    cid2spf.py
    rejects.py
    rhsbl.m4
    sample.py
    test.py
    test/amazon
    test/big5
    test/bounce
    test/bounce1
    test/bound
    test/honey
    test/missingboundary
    test/samp1
    test/spam44
    test/spam7
    test/spam8
    test/test1
    test/test8
    test/virus1
    test/virus13
    test/virus2
    test/virus3
    test/virus4
    test/virus5
    test/virus6
    test/virus7
    testsample.py

CREDITS

@@ -5,8 +5,11 @@ wrapper (Milter.py) that handles attachments, did lots of testing, packaged
 it with distutils, and generally transformed it from a quick hack to a
 real, usable Python extension.
 
-Other contributors:
+Other contributors (in random order):
 
+Dave MacQuigg 
+  for noticing that smfi_insheader wasn't supported, and creating
+  a template to help first time pymilter users create their own milter.
 Terence Way
   for providing a Python port of SPF
 Scott Kitterman

HOWTO

@@ -0,0 +1,136 @@
+	Step one.  Which DSPAM is right for you?
+
+The DSPAM project makes dspam part of the LDA (Local Delivery Agent).
+Pydspam puts dspam into the MTA (Mail Transfer Agent - sendmail with pymilter).
+
+The advantage of doing dspam in the LDA is that any aliasing has already been
+resolved.  You need only configure mailboxes.
+
+The advantage of doing dspam in the MTA is it can screen an entire 
+company as a gateway with multiple domains.  Unfortunately, this
+means you have to tell it about all the aliases that comprise each
+account.  (Also, pydspam is still uses dspam-2.6.5.2 - the Dspam API
+has changed for newer versions.)
+
+If the LDA is right for you, you'll want to use the official Dspam 
+package.  http://www.nuclearelephant.com/projects/dspam/
+
+If the MTA approach is what you want, then pydspam is what you want.
+
+In either case, you will still want pymilter to block forgeries, Windows 
+executables, etc.
+
+So, lets assume you want to install pymilter, and may or may not
+wish to install pydspam.
+
+	Step two.  Obtaining RPMS.
+
+For basic pymilter you'll need:
+
+python-2.4
+milter-0.8.2	(the RH9 rpm should work on Fedora Core - let me know)
+sendmail-8.13.x	(with milter support enabled)
+
+and for SPF you'll need:
+
+pydns-2.3.0-2.4
+
+and for SRS you'll need:
+
+pysrs-0.30.9-1.py24
+
+I'm pretty sure you will want to have SPF and SRS available.
+
+	Step three.  Activate basic milter.
+
+Activate the basic milter by editing /etc/mail/sendmail.mc and adding:
+
+INPUT_MAIL_FILTER(`pythonfilter', `S=local:/var/run/milter/pythonsock, F=T, T=C:5m;S:20s;R:5m;E:5m')
+
+You can then "make sendmail.cf" and restart sendmail.
+
+Tail /var/log/milter/milter.log while SMTP clients connect to your
+sendmail instance.  This should show you what the milter is doing.
+
+By default, milter-0.8.2 rejects on SPF fail, except for listed domains
+(that are known to be broken).  Some admins don't like that, and 0.8.3 will use
+the /etc/mail/access database to configure SPF responses.  For now,
+if you don't like SPF, you can disable spf by replacing "import spf"
+with "spf = None" around line 285 in /var/log/milter/bms.py.
+
+	Step four.  Tweaking the basic config.
+
+Most pymilter configuration is in /etc/mail/pymilter.cfg.
+
+By default, milter scans attachments for executable extensions.  You can
+turn this off by setting banned_exts to the empty list.  There are options
+to scan ZIP attachments and rfc822 attachments.  When it finds a banned
+file type, milter saves the original message in /var/log/milter/save,
+and replaces the attachment with a plain text warning message.
+
+Configure hello_blacklist with your own helo name and domains - which
+you know cannot legitimately be used by external MTAs.
+
+Configure trusted_relay with your secondary MX servers, if any.  These
+should also run pymilter with similar policies.  (But this isn't
+needed for initial testing.)
+
+Configure internal_connect with subnets of your internal SMTP clients.
+Internal connections skip SPF testing and other policies.
+
+Configure internal_domains with domains used by your internal SMTP clients.
+If they attempt to use any other domain, the attempt is blocked and the
+client is logged as a "zombie".  Conversely, any attempt by an external
+MTA to use one of your internal domains is treated as a forgery and
+blocked (a simplified form of local SPF).
+
+Adjust porn_words and spam_words - these block emails with a Subject
+containing the listed strings.  They can be empty to disable Subject
+string blocking.
+
+	Advanced SPF configuration.
+
+The sendmail access file, or another readonly database with that
+format, can be used for detail spf policy.  SPF access policy
+record are tagged with "SPF-{Result}:".  Results are
+Pass, Neutral, Softfail, Fail, PermError.  Currently supported
+policy keywords are OK, CBV, REJECT.  Currently, TempError always
+results in TEMPFAIL.
+
+The default policies are set in pymilter.cfg.  The defaults
+if none of the config options are set are as follows:
+
+SPF-Fail:	REJECT
+SPF-Softfail:	CBV
+SPF-Neutral:	OK
+SPF-PermError:	REJECT
+SPF-Pass:	OK
+
+The tag may be followed by a specific domain.  For instance, to
+require a Pass from aol.com:
+
+SPF-Neutral:aol.com	REJECT
+SPF-Softfail:aol.com	REJECT
+
+The CBV policy requires a valid HELO name.  If the EHLO name is 
+RFC2822 compliant, then a DSN is sent to the alleged sender.  The 
+template for the DSN is selected according to the SPF result:
+
+Fail:		softfail.txt
+SoftFail:	softfail.txt
+Neutral:	neutral.txt
+PermError:	permerror.txt
+None:		strike3.txt
+
+An SPF-Pass is always accepted by the milter.  Domains can be blacklisted 
+via sendmail in the access file or via a RHS DNS blacklist.
+
+	To be continued.
+
+Forthcoming topics:
+
+SRS config
+
+
+pydspam config
+wiretap config

MANIFEST.in

@@ -1,6 +1,7 @@
 include COPYING
 include TODO
 include NEWS
+include HOWTO
 include CREDITS
 include README
 include MANIFEST.in
@@ -9,6 +10,7 @@ include testmime.py
 include testbms.py
 include testdspam.py
 include rejects.py
+include report.py
 include bms.py
 include spf.py
 include cid2spf.py
@@ -16,6 +18,7 @@ include spfquery.py
 include test.py
 include sample.py
 include test/*
+include doc/*
 include Milter/*.py
 include *.spec
 include start.sh
@@ -23,5 +26,5 @@ include milter.rc
 include milter.rc7
 include milter.cfg
 include rhsbl.m4
-include softfail.txt
-include strike3.txt
+include *.txt
+include *.html

Milter.py

@@ -1,203 +0,0 @@
-
-# Author: Stuart D. Gathman <stuart@bmsi.com>
-# Copyright 2001 Business Management Systems, Inc.
-# This code is under GPL.  See COPYING for details.
-
-import os
-import milter
-import thread
-
-from milter import ACCEPT,CONTINUE,REJECT,DISCARD,TEMPFAIL,	\
-  set_flags, setdbg, setbacklog, settimeout, \
-  ADDHDRS, CHGBODY, ADDRCPT, DELRCPT, CHGHDRS,	\
-  V1_ACTS, V2_ACTS, CURR_ACTS
-
-try: from milter import QUARANTINE
-except: pass
-
-_seq_lock = thread.allocate_lock()
-_seq = 0
-
-def uniqueID():
-  """Return a sequence number unique to this process.
-  """
-  global _seq
-  _seq_lock.acquire()
-  seqno = _seq = _seq + 1
-  _seq_lock.release()
-  return seqno
-  
-class Milter:
-  """A simple class interface to the milter module.
-  """
-  def _setctx(self,ctx):
-    self.__ctx = ctx
-    if ctx:
-      ctx.setpriv(self)
-
-  # user replaceable callbacks
-  def log(self,*msg):
-    print 'Milter:',
-    for i in msg: print i,
-    print
-
-  def connect(self,hostname,unused,hostaddr):
-    "Called for each connection to sendmail."
-    self.log("connect from %s at %s" % (hostname,hostaddr))
-    return CONTINUE
-
-  def hello(self,hostname):
-    "Called after the HELO command."
-    self.log("hello from %s" % hostname)
-    return CONTINUE
-
-  def envfrom(self,f,*str):
-    """Called to begin each message.
-    f -> string		message sender
-    str -> tuple	additional ESMTP parameters
-    """
-    self.log("mail from",f,str)
-    return CONTINUE
-
-  def envrcpt(self,to,*str):
-    "Called for each message recipient."
-    self.log("rcpt to",to,str)
-    return CONTINUE
-
-  def header(self,field,value):
-    "Called for each message header."
-    self.log("%s: %s" % (field,value))
-    return CONTINUE
-
-  def eoh(self):
-    "Called after all headers are processed."
-    self.log("eoh")
-    return CONTINUE
-
-  def body(self,unused):
-    "Called to transfer the message body."
-    return CONTINUE
-
-  def eom(self):
-    "Called at the end of message."
-    self.log("eom")
-    return CONTINUE
-
-  def abort(self):
-    "Called if the connection is terminated abnormally."
-    self.log("abort")
-    return CONTINUE
-
-  def close(self):
-    "Called at the end of connection, even if aborted."
-    self.log("close")
-    return CONTINUE
-
-  # Milter methods which can be invoked from callbacks
-  def getsymval(self,sym):
-    return self.__ctx.getsymval(sym)
-
-  # If sendmail does not support setmlreply, then only the
-  # first msg line is used.
-  def setreply(self,rcode,xcode=None,msg=None,*ml):
-    return self.__ctx.setreply(rcode,xcode,msg,*ml)
-
-  # Milter methods which can only be called from eom callback.
-  def addheader(self,field,value):
-    return self.__ctx.addheader(field,value)
-
-  def chgheader(self,field,idx,value):
-    return self.__ctx.chgheader(field,idx,value)
-
-  def addrcpt(self,rcpt):
-    return self.__ctx.addrcpt(rcpt)
-
-  def delrcpt(self,rcpt):
-    return self.__ctx.delrcpt(rcpt)
-
-  def replacebody(self,body):
-    return self.__ctx.replacebody(body)
-
-  # When quarantined, a message goes into the mailq as if to be delivered,
-  # but delivery is deferred until the message is unquarantined.
-  def quarantine(self,reason):
-    return self.__ctx.quarantine(reason)
-
-  def progress(self):
-    return self.__ctx.progress()
-
-factory = Milter
-
-def connectcallback(ctx,hostname,family,hostaddr):
-  m = factory()
-  m._setctx(ctx)
-  return m.connect(hostname,family,hostaddr)
-
-def closecallback(ctx):
-  m = ctx.getpriv()
-  if not m: return CONTINUE
-  rc = m.close()
-  m._setctx(None)	# release milterContext
-  return rc
-
-def envcallback(c,args):
-  """Convert ESMTP parms to keyword parameters.
-  Can be used in the envfrom and/or envrcpt callbacks to process
-  ESMTP parameters as python keyword parameters."""
-  kw = {}
-  for s in args[1:]:
-    pos = s.find('=')
-    if pos > 0:
-      kw[s[:pos]] = s[pos+1:]
-  return apply(c,args,kw)
-
-def runmilter(name,socketname,timeout = 0):
-  # This bit is here on the assumption that you will be starting this filter
-  # before sendmail.  If sendmail is not running and the socket already exists,
-  # libmilter will throw a warning.  If sendmail is running, this is still
-  # safe if there are no messages currently being processed.  It's safer to
-  # shutdown sendmail, kill the filter process, restart the filter, and then
-  # restart sendmail.
-  pos = socketname.find(':')
-  if pos > 1:
-    s = socketname[:pos]
-    fname = socketname[pos+1:]
-  else:
-    s = "unix"
-    fname = socketname
-  if s == "unix" or s == "local":
-    print "Removing %s" % fname
-    try:
-      os.unlink(fname)
-    except:
-      pass
-
-  # The default flags set include everything
-  # milter.set_flags(milter.ADDHDRS)
-  milter.set_connect_callback(connectcallback)
-  milter.set_helo_callback(lambda ctx, host: ctx.getpriv().hello(host))
-  milter.set_envfrom_callback(lambda ctx,*str:
-  	ctx.getpriv().envfrom(*str))
-#  	envcallback(ctx.getpriv().envfrom,str))
-  milter.set_envrcpt_callback(lambda ctx,*str:
-  	ctx.getpriv().envrcpt(*str))
-#  	envcallback(ctx.getpriv().envrcpt,str))
-  milter.set_header_callback(lambda ctx,fld,val:
-  	ctx.getpriv().header(fld,val))
-  milter.set_eoh_callback(lambda ctx: ctx.getpriv().eoh())
-  milter.set_body_callback(lambda ctx,chunk: ctx.getpriv().body(chunk))
-  milter.set_eom_callback(lambda ctx: ctx.getpriv().eom())
-  milter.set_abort_callback(lambda ctx: ctx.getpriv().abort())
-  milter.set_close_callback(closecallback)
-
-  milter.setconn(socketname)
-  if timeout > 0: milter.settimeout(timeout)
-  # The name *must* match the X line in sendmail.cf (supposedly)
-  milter.register(name)
-  start_seq = _seq
-  try:
-    milter.main()
-  except milter.error:
-    if start_seq == _seq: raise	# couldn't start
-    # milter has been running for a while, but now it can't start new threads
-    raise milter.error("out of thread resources")

Milter/__init__.py

@@ -16,6 +16,8 @@ from milter import ACCEPT,CONTINUE,REJECT,DISCARD,TEMPFAIL,	\
 try: from milter import QUARANTINE
 except: pass
 
+__version__ = '0.8.5'
+
 _seq_lock = thread.allocate_lock()
 _seq = 0
 
@@ -42,7 +44,7 @@ class Milter:
     for i in msg: print i,
     print
 
-  def connect(self,hostname,unused,hostaddr):
+  def connect(self,hostname,family,hostaddr):
     "Called for each connection to sendmail."
     self.log("connect from %s at %s" % (hostname,hostaddr))
     return CONTINUE
@@ -104,8 +106,8 @@ class Milter:
     return self.__ctx.setreply(rcode,xcode,msg,*ml)
 
   # Milter methods which can only be called from eom callback.
-  def addheader(self,field,value):
-    return self.__ctx.addheader(field,value)
+  def addheader(self,field,value,idx=-1):
+    return self.__ctx.addheader(field,value,idx)
 
   def chgheader(self,field,idx,value):
     return self.__ctx.chgheader(field,idx,value)
@@ -215,6 +217,6 @@ def runmilter(name,socketname,timeout = 0):
     raise milter.error("out of thread resources")
 
 __all__ = globals().copy()
-for priv in ('os','milter','thread','factory','_seq','_seq_lock'):
+for priv in ('os','milter','thread','factory','_seq','_seq_lock','__version__'):
   del __all__[priv]
 __all__ = __all__.keys()

Milter/dsn.py

@@ -9,6 +9,7 @@ import smtplib
 import spf
 import socket
 from email.Message import Message
+import Milter
 
 nospf_msg = """Subject: Critical mail server configuration error
 
@@ -103,7 +104,7 @@ def send_dsn(mailfrom,receiver,msg=None):
   q = spf.query(None,None,None)
   mxlist = q.dns(domain,'MX')
   if not mxlist:
-    mxlist = (0,domain),
+    mxlist = (0,domain),	# fallback to A record when no MX
   else:
     mxlist.sort()
   smtp = smtplib.SMTP()
@@ -112,8 +113,11 @@ def send_dsn(mailfrom,receiver,msg=None):
       smtp.connect(host)
       code,resp = smtp.helo(receiver)
       # some wiley spammers have MX records that resolve to 127.0.0.1
-      if resp.split()[0] == receiver:
-        return (553,'Fraudulent MX for %s' % domain)
+      a = resp.split()
+      if not a:
+        return (553,'MX for %s has no hostname in banner: %s' % (domain,host))
+      if a[0] == receiver:
+        return (553,'Fraudulent MX for %s: %s' % (domain,host))
       if not (200 <= code <= 299):
 	raise smtplib.SMTPHeloError(code, resp)
       if msg:
@@ -151,13 +155,13 @@ def create_msg(q,rcptlist,origmsg=None,template=None):
   connectip = q.i
   receiver = q.r
   sender_domain = q.o
+  result = q.result
+  perm_error = q.perm_error
   rcpt = '\n\t'.join(rcptlist)
   try: subject = origmsg['Subject']
   except: subject = '(none)'
   try:
     spf_result = origmsg['Received-SPF']
-    if not spf_result.startswith('softfail'):
-      spf_result = None
   except: spf_result = None
 
   msg = Message()
@@ -165,11 +169,14 @@ def create_msg(q,rcptlist,origmsg=None,template=None):
   msg.add_header('To',sender)
   msg.add_header('From','postmaster@%s'%receiver)
   msg.add_header('Auto-Submitted','auto-generated (configuration error)')
+  msg.add_header('X-Mailer','PyMilter-'+Milter.__version__)
   msg.set_type('text/plain')
 
   if not template:
-    if spf_result: template = softfail_msg
-    else: template = nospf_msg
+    if spf_result and spf_result.startswith('softfail'):
+      template = softfail_msg
+    else:
+      template = nospf_msg
   hdrs,body = template.split('\n',1)
   for ln in hdrs.splitlines():
     name,val = ln.split(':',1)

NEWS

@@ -1,5 +1,28 @@
 Here is a history of user visible changes to Python milter.
 
+0.8.5	Simple trusted_forwarder implementation.
+	Fix access_file neutral policy
+	Move Received-SPF header to beginning of headers
+	Supply keyword info for all results in Received-SPF header.
+	Move guessed SPF result to separate header
+	Activate smfi_insheader only when SMFIR_INSHEADER defined
+	Handle NULL MX in spf.py
+	in-process GOSSiP server support (to be extended later)
+	Expire CBV cache and renew auto-whitelist entries
+0.8.4	Auto-whitelist recipients of outgoing email.
+	Fix SPF policy via sendmail access map (case insensitive keys).
+	Train screener on whitelisted messages
+	Optional idx parameter to addheader to invoke smfi_insheader
+	Activate progress API when SMFIR_PROGRESS defined
+0.8.3   Keep screened honeypot mail, but optionally discard honeypot only mail.
+	spf_accept_fail option for braindead SPF senders 
+	  (treats fail like softfail)
+	Option to set SPF policy via sendmail access map.
+	Option to supply Sender header from MAIL FROM when missing.
+	Consider SMTP AUTH connections internal.
+	Send DSN for SPF errors corrected by extended processing.
+	Send DSN before SCREENED mail is quarantined
+	Use logging package to keep log lines atomic.
 0.8.2	Strict processing limits per SPF RFC
 	Fixed several parsing bugs under RFC 
 	Support official IANA SPF record (type99)

TODO

@@ -1,54 +1,46 @@
+Require signed MFROM for all incoming bounces when signing all outgoing mail -
+except from trusted relays.
+
+Send DSN for permerror before processing extended result.  An additional
+DSN may be sent based on extended result.
+
+Rescind whitelist for banned extensions, in case sender is infected.
+
+Train honeypot on error only.
+
+Find rfc2822 policy for MFROM quoting.
+
+Support explicit errors for SPF policy in access file:
+SPF-Neutral:aol.com	ERROR:"550 AOL mail must get SPF PASS"
+
 Defer TEMPERROR in SPF evaluation - give precedence to security
 (only defer for PASS mechanisms).
 
 Option to add Received-SPF header, but never reject on SPF.
+  I think the above will handle this.
 
 Create null config that does nothing - except maybe add Received-SPF
 headers.  Many admins would like to turn features on one at a time.
 
-Checking in mime.py;
-/bms/cvs/milter/mime.py,v  <--  mime.py
-new revision: 1.56; previous revision: 1.55
-done
-Checking in spf.py;
-/bms/cvs/milter/spf.py,v  <--  spf.py
-new revision: 1.18; previous revision: 1.17
-done
-Checking in testmime.py;
-/bms/cvs/milter/testmime.py,v  <--  testmime.py
-new revision: 1.19; previous revision: 1.18
-
-Auto whitelist based on outgoing email - perhaps with magic subject
-or recipient prefix.
-
 Can't output messages with malformed rfc822 attachments.
 
-Example malformed SPF:
-onvunvuvvx.usafisnews.org text "v=spf1 mx ptr ip4:207.44.199.970 -all"
-
 Move milter,Milter,mime,spf modules to pymilter
 milter package will have bms.py application
 
-Support SMTP AUTH and disable SPF checks when connection is authorized.
 Web admin interface
-Check valid domains allowed by internal senders to detect PCs infected
-with spam trojans.
-Do CBV (callback verification) for mail with no published SPF record.
 message log for automated stats and blacklisting
-Skip dspam when SPF pass?
+Skip dspam when SPF pass? NO
 Report 551 with rcpt on SPF fail?
 check spam keywords with character classes, e.g.
 	{a}=[a@ãä], {i}=[i1í], {e}=[eë], {o}=[o0ö]
 
 Implement RRS - a backdoor for non-SRS forwarders.  User lists non-SRS 
 forwarder accounts, and a util provides a special local alias for the
-user to give to the forwarder.  Alias only works for mail from that
+user to give to the forwarder.  (Or user just adds arbitrary alias
+unique to that forwarder to a database.)  Alias only works for mail from that
 forwarder.  Milter gets forwarder domain from alias and uses it to
 SPF check forwarder.
 
-Another special dspam user, 'honeypot', can be listed in innoculations.
-All email to those addresses is treated as known spam.
-
 Framework for modular Python milter components within a single VM.
 Python milters can be already be composed through sendmail by running each in
 a separate process.  However, a significant amount of memory is wasted
@@ -57,8 +49,7 @@ is cumbersome (e.g., adding mail headers, writing external files).
 
 Backup copies for outgoing/incoming mail.
 
-Allow multiple wiretap groups, each with its own destination.  Perhaps 
-also copy incoming wiretap mail, even though sendmail alias works perfectly
+Copy incoming wiretap mail, even though sendmail alias works perfectly
 for the purpose, to avoid having to change two configs for a wiretap.
 
 Provide a way to reload milter.cfg without stopping/restarting milter.
@@ -72,10 +63,6 @@ Keep separate ismodified flag for headers and body.  This is important
 when rejecting outgoing mail with viruses removed (so as not to
 embarrass yourself), and also removing Received headers with hidepath.
 
-Wrap smfi_setbacklog(int) - but it is only available in sendmail >= 8.12.3,
-  so how can we detect whether to wrap it?
-
 Need a test module to feed sample messages to a milter though a live 
 sendmail and SMTP.  The mockup currently used is probably not very accurate,
 and doesn't test the threading code.
-

doc/changes.ht

@@ -0,0 +1,173 @@
+Title: Recent Changes
+
+<h2> Recent Changes </h2>
+
+Python milter has been moved to 
+<a href="http://sourceforge.net/projects/pymilter/">pymilter Sourceforge
+project</a> for development and release downloads.
+
+<h3> 0.8.5 </h3>
+
+Release 0.8.5 fixes some build bugs reported by Stephen Figgins.  It
+fixes many small things, like not auto-whitelisting recipients of
+outgoing mail when the subject contains "autoreply:".  There is a 
+simple trusted forwarder implementation.  If you have more than
+2 or so forwarders, we will need a way to "compile" SPF records into an
+IP set and TTL for it to be efficient (like libspf2 does).
+
+<h3> GOSSiP </h3>
+An alpha release of <a href="pygossip.html">pygossip</a> has been commited to
+CVS, module pygossip. A version of the bms.py milter has been commited to CVS
+which supports calling GOSSiP to track domain reputation in a local database.
+
+<h3> New website design </h3>
+
+Hey, I'm no artist, so I just used the
+<a href="http://ht2html.sourceforge.net/">ht2html</a> package
+by <a href="http://barry.wooz.org/">Barry Warsaw</a>.  The mascot
+is by <a href="http://alphard.ethz.ch/hafner/lebl.htm">Christian Hafner</a>,
+or maybe his wife.  I chose Maxwell's daemon because it tirelessly
+and invisibly sorts molecules, just as milters sort mail.
+Christian has also provided a fun 
+<a href="http://alphard.ethz.ch/hafner/PPS/PPS2002/Maxwell/simulation.htm">
+simulation</a> that lets you try your hand at sorting molecules.
+
+<h3> 0.8.4 </h3>
+
+Release 0.8.4 makes configuring SPF policy via access.db actually work.
+The honeypot idea is enhanced by auto-whitelisting recipients of
+email sent from selected domains.  Whitelisted messages are then used
+to train the honeypot.  This makes the honeypot screener entirely self
+training.  The smfi_progress() API is now automatically supported when present.
+An optional idx parameter to milter.addheader() invokes smfi_insheader().
+
+<h3> 0.8.3 </h3>
+
+Release 0.8.3 uses the standard logging module, and supports configuring
+more detailed SPF policy via the sendmail access map.  SMTP AUTH connections
+are considered INTERNAL.  Preventing forgery between internal domains is
+just a matter of specifying the user-domain map - I'll define something
+for the next version.  We now send DSNs when mail is quarantined (rejecting
+if DSN fails) and for SPF syntax errors (PermError).  There is an
+experimental option to add a Sender header when it is missing and the From
+domain doesn't match the MAIL FROM domain.  Next release, we may start
+renaming and replacing an existing Sender header when neither it nor the
+From domain matches MAIL FROM.  Since bogus MAIL FROMs are rejected
+(to varying degrees depending on the configured SPF policy), and
+both Sender and From and displayed by default in many email clients,
+this provides some phishing protection without rejecting mail based
+on headers.
+
+<h3> 0.8.2 </h3>
+
+Release 0.8.2 has changes to <a href="http://openspf.net">SPF</a> to bring it
+in line with the newly official RFC.  It adds 
+<a href="http://ses.codeshare.ca/">SES</a>
+support (the original SES without body hash) for pysrs-0.30.10, and honeypot
+support for pydspam-1.1.9.  There is a new method in the base milter module.
+milter.set_exception_policy(i) lets you choose a policy of CONTINUE, REJECT, or
+TEMPFAIL (default) for untrapped exceptions encountered in a milter callback.
+
+<h3> 0.8.0 </h3>
+
+Release 0.8.0 is the first <a href="http://sourceforge.net/">Sourceforge</a>
+release.  It supports Python-2.4, and provides an option to accept mail
+that gets an SPF softfail or fails the 3 strikes rule, provided the
+alleged sender accepts a DSN explaining the problem.  Python-2.3 is
+no longer supported by the reworked mime.py module, although API changes
+could be backported.  There are too many incompatible changes to the
+python email package.
+
+<h3> Older Releases </h3>
+
+Release 0.7.2 tightens the authentication screws with a "3 strikes and
+you're out" policy.  A sender must have a valid PTR, HELO, or SPF record
+to send email.  Specific senders can be whitelisted using the
+"delegate" option in the spf configuration section by adding a
+default SPF record for them.  The PTR and HELO are required
+by RFC anyway, so this is not an unreasonable requirement.
+There is now a coherent policy for an SPF softfail result.  A softfail
+is accepted if there is a valid PTR or HELO, or if the domain
+is listed in the "accept_softfail" option of the spf configuration section.
+A neutral result is accepted by default if there is a valid PTR or
+HELO, (and the SPF record was not guessed), unless the domain is listed in the
+"reject_neutral" option.  Common forms of PTR records for dynamic IPs are
+recognized, and do not count as a valid PTR.  This does not prevent anyone
+from sending mail from a dynamic IP - they just need to configure a
+valid HELO name or publish an SPF record.
+<p>
+As SPF adoption continues to rise, forged spam is not getting through.  So
+spammers are publishing their SPF records as predicted.  The 0.7.2 RPM
+now provides the <code>rhsbl</code> sendmail hack so that spammer domains
+can be blacklisted.  With the RPM installed, add a line like the following
+to your <code>sendmail.mc</code>.
+<pre>
+HACK(rhsbl,`blackholes.example.com',"550 Rejected: " $&{RHS} " has been spamming our customers.")dnl
+</pre>
+<p>
+Of course, spammers are now starting to register
+throwaway domains.  The next thing we need is a custom DNS server,
+in Python, that
+can recognize patterns.  For instance, one spammer registers ded304.com,
+ded305.com, ded306.com, etc.  We also need the custom DNS server to
+let SPF classic clients check SES (which will be part of pysrs).  
+The <a href="http://twistedmatrix.com/products/twisted">Twisted Python</a>
+framework provides a custom DNS server - but I
+would like a smaller implementation for our use.
+<p>
+The RPM for release 0.7.0 moves the config file and socket locations to
+/etc/mail and /var/run/milter respectively.  We now parse Microsoft CID records
+- but only hotmail.com uses them.  They seem to have applied for a patent on
+the brilliant idea of examining the mail headers to see who the message is
+from.  We aren't doing that here, so not to worry - but I am not a lawyer, so
+if you are worried, change spf.py around line 626 to return None instead of
+calling CIDParser().  There is a new option to reject mail with no PTR
+and no SPF.
+<p>
+Microsoft is pushing an anti-opensource license for their pending patent
+along with their sender-ID proposal before the IETF.
+It is royalty free - but requires anyone distributing a binary they've
+compiled from source to sign a license agreement.  The Apache Software
+Foundation <a
+href="http://www.apache.org/foundation/docs/sender-id-position.html"> explains
+the problem with sender-ID</a>, and Debian <a
+href="http://www.debian.org/News/2004/20040904">concurs</a>.  Since 
+the <a href="http://download.microsoft.com/download/4/3/9/439b024b-09fd-44ee-8ff0-10e834004c36/senderid_FAQ.PDF">Microsoft license</a> is
+<a href="http://www.circleid.com/article/732_0_1_0_C/">incompatible with free
+software in general</a> and the <a
+href="http://www.imc.org/ietf-mxcomp/mail-archive/msg03678.html">GPL in
+particular</a>, Python milter will not be able to implement sender-ID in its
+current form.  This was, no doubt, Microsoft's intent all along.
+<p>
+Sender-ID attempts to do for RFC2822 headers what SPF does for RFC2821 headers.
+Unlike SPF, it has never been tried, and is encumbered by a stupid patent.  I
+recommend ignoring it and continuing to implement and improve SPF until a
+working and unencumbered proposal for RFC2822 headers surfaces.
+
+<p>
+<a href="http://openspf.com">
+<img src="SPF.gif" align=left alt="SPF logo"></a>
+Release 0.6.6 adds support for <a href="http://openspf.com/">SPF</a>,
+a protocol to prevent forging of the envelope from address.  
+SPF support requires <a href="http://pydns.sourceforge.net/">pydns</a>.
+The included spf.py module is an updated version of the original 1.6
+version at <a href="http://www.wayforward.net/spf/">wayforward.net</a>.
+The updated version tracks the draft RFC and test suite.
+<p>
+The FAQ addresses <a href="faq.html#spf">how to get started with SPF</a>.
+<p>
+Release 0.6.1 adds a full milter based dspam application.
+<p>
+I have selected the <a href="http://www.nuclearelephant.com/projects/dspam/">
+dspam bayes filter project</a> and <a href="dspam.html">
+packaged it for python</a>.
+Release 0.6.0 offers a simple application of dspam I call "header triage",
+which rejects messages with spammy headers.  
+To use header triage, you must have <a href="dspam.html">DSPAM</a> installed,
+and select a dictionary that is well moderated by someone who gets
+lots of spam.  That dictionary can be used to block spam that is 
+obvious from the headers (e.g. X-Mailer and Subject) before it ties
+up any more resources.  I have yet to see any false positives from this
+approach (check the milter log), but if there are, the sender will
+get a REJECT with the message "Your message looks spammy."
+

doc/credits.ht

@@ -0,0 +1,55 @@
+Title: Credits
+
+<h1> CREDITS </h1>
+
+<a href="mailto:Jim Niemira <urmane@urmane.org>">Jim Niemira</a>
+wrote the original C module and some quick
+and dirty python to use it.  
+<a href="mailto:Stuart Gathman <stuart@bmsi.com>">Stuart D. Gathman</a>
+took that kludge and added threading and context objects to it, wrote a proper
+OO wrapper (Milter.py) that handles attachments, did lots of testing, packaged
+it with distutils, and generally transformed it from a quick hack to a
+real, usable Python extension.
+
+<h2>Other contributors (in random order):</h2>
+
+<dl>
+<dt> <a href="http://alphard.ethz.ch/hafner/lebl.htm">Christian Hafner</a>
+  <dd>for the pymilter mascot image of 
+  <a href="http://maxwelld.netfirms.com/">
+  Maxwell's daemon</a>
+<dt>Stephen Figgins
+  <dd>for reporting problems building with sendmail-8.12, and when
+  building milter.so for the first time.
+<dt>Dave MacQuigg 
+  <dd>for noticing that smfi_insheader wasn't supported, and creating
+  a template to help first time pymilter users create their own milter.
+<dt>Terence Way
+  <dd>for providing a Python port of SPF
+<dt>Scott Kitterman
+  <dd>for doing lots of testing and debugging of SPF against draft standard,
+  and for putting up a <a href="http://www.kitterman.com/spf/validate.html">
+  web page that validates SPF</a> records using spf.py
+<dt>Alexander Kourakos
+  <dd>for plugging several memory leaks
+<dt>George Graf at Vienna University of Economics and Business Administration
+  <dd>for handling None passed to setreply and chgheader.
+<dt>Deron Meranda
+  <dd>for IPv6 patches
+<dt>Jason Erikson
+  <dd>for handling NULL hostaddr in connect callback.
+<dt>John Draper
+  <dd>for porting Python milter to OpenBSD, and starting to work on tutorials
+  then pointing out that it would be easier to just write the MTA in Python.
+<dt>Eric S. Johansson
+  <dd>for helpful design discussions while working on camram
+<dt>Alex Savguira
+  <dd>for finding bugs with international headers and
+  suggesting the scan_zip option.
+<dt><a href="http://www.bmsi.com">Business Management Systems</a>
+  <dd>for hosting the website, and providing paying clients who need milter
+  service so I can work on it as part of my day job.
+</dl>
+
+If I have left anybody out, send me a reminder: 
+<a href="mailto:Stuart Gathman <stuart@bmsi.com>">stuart@bmsi.com</a>

doc/faq.ht

@@ -1,8 +1,4 @@
-<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
-<html>
-<head>
-<title>Python Milter FAQ</title>
-</head><body>
+Title: Python Milter FAQ
 
 <h1> Python Milter <a name=faq>FAQ</a> </h1>
 
@@ -22,7 +18,7 @@ shows you how to install libmilter with a separate invocation of make.
 <li> Q. Why is mfapi.h not found when I try to compile Python milter on
 RedHat 7.2?
 <p>  A. RedHat forgot to include the header in the RPM.  See the
-<a href="milter.html#rh72">RedHat 7.2 requirements</a>.
+<a href="requirements.html#rh72">RedHat 7.2 requirements</a>.
 <p>
 
 <h3> Running Python Milter </h3>
@@ -103,13 +99,14 @@ If you are running bms.py, then the block_chinese option in
 
 <li> Q. Why does sendmail coredump with milters on OpenBSD?
 <p>  A. Sendmail has a problem with unix sockets on old versions of OpenBSD.
-Use an internet domain socket instead.  For example, in
+OpenBSD users report that this problem has been fixed, so upgrading
+OpenBSD will fix this.  Otherwise, you can
+use an internet domain socket instead.  For example, in
 <code>sendmail.cf</code> use
 <pre>
 Xpythonfilter, S=inet:1234@localhost
 </pre>
 and change sample.py accordingly.
-<p> OpenBSD users report that this problem has been fixed.
 <p>
 
 <li> Q. How can I change the bounce message for an invalid recipient?
@@ -143,10 +140,15 @@ Xpythonfilter, S=local:/var/log/milter/pythonsock, F=T, T=C:5m;S:20s;R:60s;E:5m
 </pre>
 <li> Q. There is a Python traceback in the log file!  What happened to
 	my email?
-<p>  A. When the milter fails with an untrapped exception, a TEMPFAIL
-result (451) is returned to the sender.  The sender will then retry every
-hour or so for several days.  Hopefully, someone will notice the 
-traceback, and workaround or fix the problem.
+<p>  A. By default, when the milter fails with an untrapped exception, a
+TEMPFAIL result (451) is returned to the sender.  The sender will then retry
+every hour or so for several days.  Hopefully, someone will notice the 
+traceback, and workaround or fix the problem.  Beginning with milter-0.8.2,
+you can call <code>milter.set_exception_policy(milter.CONTINUE)</code>
+to cause an untrapped exception to continue processing with the
+next callback or milter instead.  For
+completeness, you can also set the exception policy to
+<code>milter.REJECT</code>.
 
 <li> Q. I read some notes such as "Check valid domains allowed by internal
 	senders to detect PCs infected with spam trojans." but could not
@@ -194,4 +196,3 @@ everything up for you.  For other systems:
 </ol>
 
 </ol>
-</html>

doc/logmsgs.ht

@@ -0,0 +1,71 @@
+Title: Python Milter Log Documentation
+<style>
+DT              { font-weight: bolder; padding-top: 1em }
+</style>
+
+<h1> Milter Log Documentation </h1>
+
+The milter log from the bms.py application has a variety of "tags" in it that
+indicate what it did.
+
+<dl>
+<dt> DSPAM: honeypot SCREENED
+<dd> message was quarantined to the honeypot quarantine
+
+<dt> REJECT: hello SPF: fail 550 access denied
+<dt> REJECT: hello SPF: softfail 550 domain in transition
+<dt> REJECT: hello SPF: neutral 550 access neither permitted nor denied
+<dd> message was rejected because there was an SPF policy for the
+HELO name, and it did not pass.
+
+<dt> CBV: sender-17-44662668-643@bluepenmagic.com
+<dd> we performed a call back verification
+
+<dt> dspam
+<dd> dspam identifier was added to the message
+
+<dt> REJECT: spam from self: jsconnor.com
+<dd> message was reject because HELO was us (jsconnor.com)
+
+<dt> INNOC: richh
+<dd> message was used to update richh's dspam dictionary
+
+<dt> HONEYPOT: michaelb@jsconnor.com
+<dd> message was sent to a honeypot address (michaelb@jsconnor.com), the 
+message was added to the honeypot dspam dictionary as spam
+
+<dt> REJECT: numeric hello name: 63.217.19.146
+<dd> message was rejected because helo name was invalid (numeric)
+
+<dt> eom
+<dd> message was successfully received
+
+<dt> TEMPFAIL: CBV: 450 No MX servers available
+<dd> we tried to do a call back verification but could not look up 
+MX record, we told the sender to try again later
+
+<dt> CBV: info@emailpizzahut.com (cached)
+<dd> call back verification was needed, we had already done it recently
+
+<dt> abort after 0 body chars
+<dd> sender hung up on us
+
+<dt> REJECT: SPF fail 550 SPF fail: see 
+  http://openspf.com/why.html?sender=m.hendersonxk@163.net&ip=213.47.161.100
+<dd> message was reject because its sender's spf policy said to
+
+<dt> REJECT: Subject: Cialis - No prescription needed!
+<dd> message was rejected because its subject contained a bad expression
+
+<dt> DSPAM: tonyc tonyc@jsconnor.com
+<dd> message was sent to tonyc@jsconnor.com and it was identified as spam 
+and placed in the tonyc dspam quarantine
+
+<dt> REJECT: CBV: 550 calvinalstonis@ix.netcom.com...User unknown
+<dt> REJECT: CBV: 553 sorry, that domain isn't in my list
+<dt> REJECT: CBV: 554 delivery error: dd This user doesn't have an account 
+<dd> message was rejected because call back verification gave us a fatal
+error
+</dl>
+
+Please add more tags to this list if you know of any. Thanks.

doc/milter.ht

@@ -1,33 +1,29 @@
-<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
-<html>
-<head>
-<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
-<title>Python Milters</title>
-</head><body>
+Title: Python Milters
 
 <P ALIGN="CENTER"><A HREF="http://www.anybrowser.org/campaign/">
-<IMG SRC="/art/brain1.gif"
+<IMG SRC="http://bmsi.com/art/brain1.gif"
 ALT="Viewable With Any Browser" BORDER="0"></A>
 
-<img src="/art/banner_4.gif" width="468" height="60" border="0"
+<img src="http://bmsi.com/art/banner_4.gif" width="468" height="60" border="0"
 	usemap="#banner_4" alt="Your vote?">
 <map name="banner_4">
   <area shape="rect" coords="330,25,426,59"
   	href="http://education-survey.org/" alt="I Disagree">
   <area shape="rect" coords="234,28,304,57" href="http://www.honestEd.com/" alt="I Agree">
 </map>
-
 </P>
+
+<img src="Maxwells.gif" alt="Maxwell's Daemon: pymilter mascot" align=left>
 <h1 align=center>Sendmail Milters in Python</h1>
 <h4 align=center>by <a href="mailto:%75%72%6D%61%6E%65%40%6E%65%75%72%61l%61%63%63%65%73%73%2E%63%6F%6D">Jim Niemira</a>
   and <a href="mailto:%73%74%75%61%72%74%40%62%6D%73%69%2E%63%6F%6D">
   Stuart D. Gathman</a><br>
 This web page is written by Stuart D. Gathman<br>and<br>sponsored by
 <a href="http://www.bmsi.com">Business Management Systems, Inc.</a> <br>
-Last updated Jun 09, 2005</h4>
+Last updated Dec 29, 2005</h4>
 
 See the <a href="faq.html">FAQ</a> | <a href="http://sourceforge.net/project/showfiles.php?group_id=139894">Download now</a> |
-<a href="/mailman/listinfo/pymilter">Subscribe to mailing list</a> |
+<a href="http://bmsi.com/mailman/listinfo/pymilter">Subscribe to mailing list</a> |
 <a href="#overview">Overview</a> |
 <a href="/python/dspam.html">pydspam</a> |
 <a href="/libdspam/dspam.html">libdspam</a>
@@ -45,151 +41,6 @@ separation features to enhance security.  Even better, sendmail 8.13
 supports socket maps, which makes <a href="pysrs.html">pysrs</a> much more
 efficient and secure.  I recommend upgrading.
 
-<h2> Recent Changes </h2>
-
-Python milter is being moved to 
-<a href="http://sourceforge.net/projects/pymilter/">pymilter Sourceforge
-project</a> for development.
-<p>
-Release 0.8.0 is the first <a href="http://sourceforge.net/">Sourceforge</a>
-release.  It supports Python-2.4, and provides an option to accept mail
-that gets an SPF softfail or fails the 3 strikes rule, provided the
-alleged sender accepts a DSN explaining the problem.  Python-2.3 is
-no longer supported by the reworked mime.py module, although API changes
-could be backported.  There are too many incompatible changes to the
-python email package.
-<p>
-Release 0.7.2 tightens the authentication screws with a "3 strikes and
-you're out" policy.  A sender must have a valid PTR, HELO, or SPF record
-to send email.  Specific senders can be whitelisted using the
-"delegate" option in the spf configuration section by adding a
-default SPF record for them.  The PTR and HELO are required
-by RFC anyway, so this is not an unreasonable requirement.
-There is now a coherent policy for an SPF softfail result.  A softfail
-is accepted if there is a valid PTR or HELO, or if the domain
-is listed in the "accept_softfail" option of the spf configuration section.
-A neutral result is accepted by default if there is a valid PTR or
-HELO, (and the SPF record was not guessed), unless the domain is listed in the
-"reject_neutral" option.  Common forms of PTR records for dynamic IPs are
-recognized, and do not count as a valid PTR.  This does not prevent anyone
-from sending mail from a dynamic IP - they just need to configure a
-valid HELO name or publish an SPF record.
-<p>
-As SPF adoption continues to rise, forged spam is not getting through.  So
-spammers are publishing their SPF records as predicted.  The 0.7.2 RPM
-now provides the <code>rhsbl</code> sendmail hack so that spammer domains
-can be blacklisted.  With the RPM installed, add a line like the following
-to your <code>sendmail.mc</code>.
-<pre>
-HACK(rhsbl,`blackholes.example.com',"550 Rejected: " $&{RHS} " has been spamming our customers.")dnl
-</pre>
-<p>
-Of course, spammers are now starting to register
-throwaway domains.  The next thing we need is a custom DNS server,
-in Python, that
-can recognize patterns.  For instance, one spammer registers ded304.com,
-ded305.com, ded306.com, etc.  We also need the custom DNS server to
-let SPF classic clients check SES (which will be part of pysrs).  
-The <a href="http://twistedmatrix.com/products/twisted">Twisted Python</a>
-framework provides a custom DNS server - but I
-would like a smaller implementation for our use.
-<p>
-The RPM for release 0.7.0 moves the config file and socket locations to
-/etc/mail and /var/run/milter respectively.  We now parse Microsoft CID records
-- but only hotmail.com uses them.  They seem to have applied for a patent on
-the brilliant idea of examining the mail headers to see who the message is
-from.  We aren't doing that here, so not to worry - but I am not a lawyer, so
-if you are worried, change spf.py around line 626 to return None instead of
-calling CIDParser().  There is a new option to reject mail with no PTR
-and no SPF.
-<p>
-Microsoft is pushing an anti-opensource license for their pending patent
-along with their sender-ID proposal before the IETF.
-It is royalty free - but requires anyone distributing a binary they've
-compiled from source to sign a license agreement.  The Apache Software
-Foundation <a
-href="http://www.apache.org/foundation/docs/sender-id-position.html"> explains
-the problem with sender-ID</a>, and Debian <a
-href="http://www.debian.org/News/2004/20040904">concurs</a>.  Since 
-the <a href="http://download.microsoft.com/download/4/3/9/439b024b-09fd-44ee-8ff0-10e834004c36/senderid_FAQ.PDF">Microsoft license</a> is
-<a href="http://www.circleid.com/article/732_0_1_0_C/">incompatible with free
-software in general</a> and the <a
-href="http://www.imc.org/ietf-mxcomp/mail-archive/msg03678.html">GPL in
-particular</a>, Python milter will not be able to implement sender-ID in its
-current form.  This was, no doubt, Microsoft's intent all along.
-<p>
-Sender-ID attempts to do for RFC2822 headers what SPF does for RFC2821 headers.
-Unlike SPF, it has never been tried, and is encumbered by a stupid patent.  I
-recommend ignoring it and continuing to implement and improve SPF until a
-working and unencumbered proposal for RFC2822 headers surfaces.
-
-<p>
-<a href="http://openspf.com">
-<img src="SPF.gif" align=left alt="SPF logo"></a>
-Release 0.6.6 adds support for <a href="http://openspf.com/">SPF</a>,
-a protocol to prevent forging of the envelope from address.  
-SPF support requires <a href="http://pydns.sourceforge.net/">pydns</a>.
-The included spf.py module is an updated version of the original 1.6
-version at <a href="http://www.wayforward.net/spf/">wayforward.net</a>.
-The updated version tracks the draft RFC and test suite.
-<p>
-The FAQ addresses <a href="faq.html#spf">how to get started with SPF</a>.
-<p>
-Release 0.6.1 adds a full milter based dspam application.
-<p>
-I have selected the <a href="http://www.nuclearelephant.com/projects/dspam/">
-dspam bayes filter project</a> and <a href="dspam.html">
-packaged it for python</a>.
-Release 0.6.0 offers a simple application of dspam I call "header triage",
-which rejects messages with spammy headers.  
-To use header triage, you must have <a href="dspam.html">DSPAM</a> installed,
-and select a dictionary that is well moderated by someone who gets
-lots of spam.  That dictionary can be used to block spam that is 
-obvious from the headers (e.g. X-Mailer and Subject) before it ties
-up any more resources.  I have yet to see any false positives from this
-approach (check the milter log), but if there are, the sender will
-get a REJECT with the message "Your message looks spammy."
-
-<h2> Enough Already! </h2>
-
-Nearly a dozen people have emailed me begging for a feature to copy
-outgoing and/or incoming mail to a backup directory by user.  Ok, it
-looks like this is a most requested feature for 0.5.6.  In the meantime,
-here are some things to consider:
-<ul>
-<li> If you want to equivalent of a Bcc added to each message, this
-is very easy to do in the python code for bms.py.   See below.
-<li> If you want to copy to a file in a directory (thus avoiding having to
-set up aliases), this is slightly more involved.  The bms.py milter already
-copies the message to a temporary file for use in replacing the message body
-when banned attachments are found.  You have to open a file, and copy the
-Mesage object to it in eom().
-<li> Finally, you are probably aware that most email clients already
-keep a copy of outgoing mail?  Presumably there is a good reason for
-keeping another copy on the server.
-</ul>
-<p>
-To Bcc a message, call <code>self.add_recipient(rcpt)</code> in envfrom after
-determining whether you want to copy (e.g. whether the sender is local).  For
-example,
-<pre>
-  def envfrom(...
-    ...
-    if len(t) == 2:
-      self.rejectvirus = t[1] in reject_virus_from
-      if t[0] in wiretap_users.get(t[1],()):
-	self.add_recipient(wiretap_dest)
-      if t[1] == 'mydomain.com':
-        self.add_recipient('&lt;copy-%s&gt;' % t[0])
-      ...
-</pre>
-<p>
-To make this a generic feature requires thinking about how the configuration
-would look.  Feel free to make specific suggestions about config file
-entries.  Be sure to handle both Bcc and file copies, and designating what
-mail should be copied.  How should "outgoing" be defined?  Implementing it is
-easy once the configuration is designed.
-
 <h3><a name=overview>Overview</a></h3>
 
 This package provides a robust toolkit for Python <a
@@ -237,6 +88,8 @@ content filtering.  SPF checking
 requires <a href="http://pydns.sourceforge.net/">
 pydns</a>.  Configuration documentation is currently included as comments
 in the <a href="milter.cfg">sample config file</a> for the bms.py milter.
+See also the <a href="HOWTO">HOWTO</a> and <a href="logmsgs.html">
+Milter Log Message Tags</a>.
 <p>
 Python milter is under GPL.  The authors can probably be convinced to
 change this to LGPL if needed.
@@ -368,8 +221,14 @@ me if you successfully install milter on a system not mentioned below.
 <td>0.5.5</td><tr>
 <td>RedHat 7.3</td><td>gcc-2.96</td><td>2.3.3</td><td>8.13.1</td>
 <td>0.7.2</td><tr>
+<td>RedHat 7.3</td><td>gcc-2.96</td><td>2.4.1</td><td>8.13.5</td>
+<td>0.8.4</td><tr>
 <td>RedHat 8.0</td><td>gcc-3.2</td><td>2.2.1</td><td>8.12.6</td>
 <td>0.5.2</td><tr>
+<td>RedHat 9.0</td><td>gcc-3.2.2</td><td>2.4.1</td><td>8.13.1</td>
+<td>0.8.2</td><tr>
+<td>RedHat EL3</td><td>gcc-3.2.3</td><td>2.4.1</td><td>8.13.5</td>
+<td>0.8.4</td><tr>
 <td>Debian Linux</td><td>gcc-2.95.2</td><td>2.1.1</td><td>8.12.0</td>
 <td>0.3.7</td><tr>
 <td>Debian Linux</td><td>gcc-3.2.2</td><td>2.2.2</td><td>8.12.7</td>
@@ -380,8 +239,8 @@ me if you successfully install milter on a system not mentioned below.
 <td>0.3.4</td><tr>
 <td>AIX-4.1.5</td><td>gcc-2.95.2</td><td>2.1.3</td><td>8.12.3</td>
 <td>0.4.2</td><tr>
-<td>AIX-4.1.5</td><td>gcc-2.95.2</td><td>2.2.3</td><td>8.13.1</td>
-<td>0.7.1</td><tr>
+<td>AIX-4.1.5</td><td>gcc-2.95.2</td><td>2.4.1</td><td>8.13.1</td>
+<td>0.8.4</td><tr>
 <td>Slackware 7.1</td><td>?</td><td>?</td><td>8.12.1</td>
 <td>0.3.8</td><tr>
 <td>Slackware 9.0</td><td>gcc-3.2.2</td><td>2.2.3</td><td>8.12.9</td>
@@ -395,108 +254,54 @@ me if you successfully install milter on a system not mentioned below.
 <td>FreeBSD</td><td>gcc-2.95.3</td><td>2.2.2</td><td>?</td>
 <td>0.5.5</td><tr>
 <td>FreeBSD 4.4</td><td>gcc-2.95.3</td><td>?</td><td>8.12.10</td>
-<td>0.6.6</td><tr>
-
+<td>0.6.6</td>
 </table>
 
-<h3> Requirements </h3>
+<h2> Enough Already! </h2>
 
-<menu>
-<li> While the miltermodule will work with python 1.5, you probably
-want to use python 2.0 or better.  The python code uses a number of
-python 2 features.
-<li> Python must be configured with thread support.  This is because
-sendmail's libmilter requires thread support.
-<li> You must compile sendmail with libmilter enabled.  In versions of
-sendmail prior to 8.12 libmilter is marked FFR (For Future Release) and
-is not installed by default.  
-Sendmail 8.12 still does not enable libmilter by default.  You must 
-explicitly select the "MILTER" option when compiling.  
-<li> Python milter has been tested against sendmail-8.11 and sendmail-8.12.
-<li> Python milter must be compiled for the specific version of sendmail
-it will run with.  (Since the result is dynamically loaded, there could 
-conceivably be multiple versions available and selected at startup - but
-that will have to wait.)  This situation may only exist for sendmail
-versions prior to 8.12.  The protocol seems designed for backward 
-compatibility - and 8.12 is the first official milter release.
-<li> Mea Culpa!  After reading the Python Style guide, I realize that
-my Python code is not up to snuff.  Apparently mixed tabs and spaces
-are anathema to those using Windows editors, where tabs can be expanded using
-any arbitrary algorithm.  Other than that, my
-intuition matched Guido's pretty well - although I like to indent by 2
-rather than 4.  I will arrange to have tabs expanded to spaces when
-exporting new versions.  Until then, beware!
-</menu>
-
-<h3> <a name="aix4"> AIX 4.1.5 Requirements </a> </h3>
-To create sendmail RPMs for AIX, you can download my AIX 4.1.5 spec files 
-for <a href="/aix/sendmail.spec">sendmail-8.11.5</a> 
-or <a href="/aix/sendmail12.spec">sendmail-8.12.3</a>.  If you have
-not already set it up, I use a <a href="/aix/aix.spec">dummy RPM package</a>
-to represent the stuff that comes with AIX.  You might also want
-my <a href="/aix/python.spec">python-2.1.1</a> spec file for AIX.  It
-does not include Tk or curses modules, sorry.  If y'all trust me, you can
-download rpms for AIX 4.x from my <a href="/aix">AIX RPM directory</a>.
+Nearly a dozen people have emailed me begging for a feature to copy
+outgoing and/or incoming mail to a backup directory by user.  Ok, it
+looks like this is a most requested feature for 0.5.6.  In the meantime,
+here are some things to consider:
+<ul>
+<li> If you want to equivalent of a Bcc added to each message, this
+is very easy to do in the python code for bms.py.   See below.
+<li> If you want to copy to a file in a directory (thus avoiding having to
+set up aliases), this is slightly more involved.  The bms.py milter already
+copies the message to a temporary file for use in replacing the message body
+when banned attachments are found.  You have to open a file, and copy the
+Mesage object to it in eom().
+<li> Finally, you are probably aware that most email clients already
+keep a copy of outgoing mail?  Presumably there is a good reason for
+keeping another copy on the server.
+</ul>
 <p>
-Sendmail-8.12 renames 
-libsmutil.a to libsm.a.  Unfortunately, libsm.a is an important AIX system
-shared library.  Therefore, I rename libsm.a back to libsmutil.a for
-AIX.  This presents a problem for setup.py.
-
-<h3> <a name="rh72"> RedHat 7.2 Requirements </a> </h3>
-
-If you are running Redhat 7.2, the distributed version of sendmail
-now enables libmilter by default.  RedHat 7.2 bundles
-the development libraries with the main sendmail package, so
-there is no sendmail-devel package.  However, they forgot to include the
-headers!  So you'll have to get the SRPM and modify it.  I suggest
-moving the static libs to a devel package and adding the headers.  If
-this is too much trouble, you can get the <a href="mfapi.h">mfapi.h</a>
-header for sendmail-8.6.11 from here and manually install it as
-<code>/usr/include/libmilter/mfapi.h</code>.
+To Bcc a message, call <code>self.add_recipient(rcpt)</code> in envfrom after
+determining whether you want to copy (e.g. whether the sender is local).  For
+example,
+<pre>
+  def envfrom(...
+    ...
+    if len(t) == 2:
+      self.rejectvirus = t[1] in reject_virus_from
+      if t[0] in wiretap_users.get(t[1],()):
+	self.add_recipient(wiretap_dest)
+      if t[1] == 'mydomain.com':
+        self.add_recipient('&lt;copy-%s&gt;' % t[0])
+      ...
+</pre>
 <p>
-If you do modify the SRPM, I suggest renaming libsmutil.a
-to libsm.a - just like sendmail-8.12 will.  If you manually install
-mfapi.h or don't rename libsmutil.a, you'll
-need to force <code>libs = ["milter", "smutil"]</code> in setup.py.
-<p>
-If you have installed python2, and want
-python-milter to use python2, add <code>python=python2</code> to setup.cfg
-and build with <code>python2 setup.py bdist_rpm</code>.   
+To make this a generic feature requires thinking about how the configuration
+would look.  Feel free to make specific suggestions about config file
+entries.  Be sure to handle both Bcc and file copies, and designating what
+mail should be copied.  How should "outgoing" be defined?  Implementing it is
+easy once the configuration is designed.
 
-<h3> <a name="rh62"> Redhat 6.2 Requirements </a> </h3>
-
-If you are running Redhat 6.2, the distributed version of sendmail
-does not enable libmilter.  You can download the Redhat 7.2 sendmail.spec
-modified to compile on RedHat 6.2:
-<a href="http://www.bmsi.com/linux/rh62/sendmail-rhmilter.spec">
-sendmail-rhmilter.spec</a>.  The <a
-href="ftp://updates.redhat.com/7.0/en/os/SRPMS/sendmail-8.11.6-1.7.0.src.rpm">
-SRPM for sendmail-8.11.6</a> is available from
-<a href="http://www.redhat.com">Redhat</a> under 
-<a href="http://www.redhat.com/support/errata/RHSA-2001-106.html">
-Errata for RH6.2</a>.  But that doesn't include the latest security
-patches since RH6.2 is no longer supported.
-<p>
-If y'all trust me, you can pick up source and binary sendmail RPMs for RH6.2
-from my <a href="http://www.bmsi.com/linux/rh62">linux downloads</a> directory.
-The lastest RPMs were built by taking a RH7.2 SRPMS and removing some
-RPM features from the spec file that RH6.2 doesn't support, then
-recompiling on RH6.2.  You can check this by installing the RH7.2 SRPM,
-then diffing my sendmail.spec with theirs.  Then run
-"rpm -bb sendmail-rhmilter.spec" when you are satisfied.
-<p>
-If you have installed python2, and want
-python-milter to use python2, add <code>python=python2</code> to setup.cfg
-and build with <code>python2 setup.py bdist_rpm</code>.
-You'll need to install the sendmail-devel package to compile milter.
 
 <hr>
 <p>
 <a href="http://validator.w3.org/check/referer">
-<img border=0 src="/vh32.png" alt=" [ Valid HTML 3.2! ] " height=31 width=88></a>
+<img border=0 src="http://bmsi.com/vh32.png" alt=" [ Valid HTML 3.2! ] " height=31 width=88></a>
 <a href="http://www.redhat.com">
-<img src="/art/powered_by.gif" width="88" height="31" alt=" [ Powered By Red Hat Linux ] " border="0"></a>
+<img src="http://bmsi.com/art/powered_by.gif" width="88" height="31" alt=" [ Powered By Red Hat Linux ] " border="0"></a>
 </p>
-
-</body></html>

doc/policy.ht

@@ -0,0 +1,237 @@
+Title: Python Milter Mail Policy
+
+<h1> Python Milter Mail Policy </h1>
+
+These are the policies implemented by the <code>bms.py</code> milter
+application.  The milter and Milter modules do not implement any policies
+by themselves.  Eventually, I'll get the bms.py milter moved to its
+own package.
+
+<h3> Classify connection </h3>
+
+When the SMTP client connects, the connection IP address is
+saved for later verification, and the connection
+is classified as INTERNAL or EXTERNAL by matching the ip
+address against the <code>internal_connect</code> configuration.
+IP addresses with no PTR, and PTR names that look like
+the kind assigned to dynamic IPs (as determined by a heuristic
+algorithm) are flagged as DYNAMIC.  IPs that match the
+<code>trusted_relay</code> configuration are flagged as TRUSTED.
+<p>
+Examples from the log file (<i>not</i> the SMTP error message returned):
+<pre>
+2005Jul29 13:56:53 [71207] connect from p50863492.dip0.t-ipconnect.de at ('80.134.52.146', 1858) EXTERNAL DYN
+2005Jul29 18:10:15 [74511] connect from foopub at ('1.2.3.4', 46513) EXTERNAL TRUSTED
+2005Jul29 14:41:00 [71805] connect from foobar at ('192.168.0.1', 41205) INTERNAL
+2005Jul29 14:41:15 [71806] connect from cncln.online.ln.cn at ('218.25.240.137', 35992) EXTERNAL
+</pre>
+<p>
+Certain obviously evil PTR names are blocked at this point:
+"localhost" (when IP is not 127.*) and ".".
+<pre>
+2005Jul29 14:49:50 [71918] connect from localhost at ('221.132.0.6', 50507) EXTERNAL
+2005Jul29 14:49:50 [71918] REJECT: PTR is localhost
+</pre>
+
+<h3> HELO Check </h3>
+
+The HELO name provided by the client is saved for later verification
+(for example by SPF).  We could validate the HELO at this point
+by verifying that an A record for the HELO name matches the connect ip.
+However, currently we only block certain obvious problems.  
+HELO names that look like an IP4 address 
+and ones that match the <code>hello_blacklist</code> configuration
+are immediately rejected.  The hello_blacklist typically contains
+the current MTAs own HELO name or email domains.
+Clients that attempt to skip HELO are immediately rejected.
+<pre>
+2005Jul29 18:10:15 [74512] hello from example.com
+2005Jul29 18:10:15 [74512] REJECT: spam from self: example.com
+2005Jul29 18:17:09 [74581] hello from 80.191.244.69
+2005Jul29 18:17:09 [74581] REJECT: numeric hello name: 80.191.244.69
+</pre>
+
+<h3> MAIL FROM Check </h3>
+
+Before calling our milter, sendmail checks a DNS blacklist to 
+block banned sender domains.  We never see a blocked domain.
+<p>
+The MAIL FROM address is saved for possible use by the smart-alias
+feature.  First, the <code>internal_domains</code> is used for
+a simple screening if defined.  If the MAIL FROM for an INTERNAL connection 
+is NOT in <code>internal_domains</code>, then it is rejected (the
+PC is most likely infected and attempting to send out spam).
+If the MAIL FROM for an EXTERNAL connection IS in
+<code>internal_domains</code>, then the message is immediately rejected.
+This is quick and effective for most small company MTAs.  For more
+complex mail networks, it is too simplistic, and should not be defined.
+SPF will handle the complex cases.
+
+<h4> wiretap </h4>
+
+The wiretap feature can screen and/or monitor mail to/from certain
+users.  If the MAIL FROM is being wiretapped, the recipients are
+altered accordingly.
+
+<!--table-stop-->
+
+<h2> SPF check </h2>
+
+Finally, the MAIL FROM, connect IP, and HELO name are checked against
+any SPF records published via DNS for the alleged sender (MAIL FROM).
+If there is no SPF record, we check for a local substitute under the
+domain defined in the <code>[spf]delegate</code> configuration.
+Further checks depend on the result.
+
+<table border=1>
+<tr><th>NONE</th><td>
+If there is no SPF record (official or delegated), then we 
+initiate a "three strikes and your out" regime, which looks for
+<b>some</b> form of validated identification.
+<ol>
+<li>We try a "best guess" SPF record of "v=spf1 a/24 mx/24 ptr".  If this
+    passes, good.  
+<li> We try to validate the HELO name. First check for an SPF record.
+   Otherwise, check whether the connect IP matches any A record for
+   the HELO name, or any A record for any MX name for the HELO name,
+   or is at least in the same /24 subnet as any of the above.
+   (In other words, a HELO SPF "best guess" of "v=spf1 a/24 mx/24".)
+   If so, good.  We consider the HELO validated.  If the HELO SPF
+   check fails, we reject the email.
+</ol>
+<pre>
+2005Jul30 19:45:16 [93991] connect from [221.200.41.54] at ('221.200.41.54', 3581) EXTERNAL DYN
+2005Jul30 19:45:18 [93991] hello from adelphia.net
+2005Jul30 19:45:19 [93991] mail from <wendy.stubbsua@link-it.com> ()
+2005Jul30 19:45:19 [93991] REJECT: hello SPF: fail 550 access denied
+</pre>
+<ol>
+<li> If there is a validated PTR name, and it doesn't look
+   like a dynamic name, good.  We consider the connection validated.
+</ol>
+If any of the above can be validated, we continue on.
+If none of the above can be validated, and the <code>[SPF]reject_noptr</code>
+option is true, we reject the message immediately with the explanation
+that we need some form of valid identification before we accept an email.
+If <code>[SPF]reject_noptr</code> is false, we flag the message as
+needing Call Back Validation.  
+The Call Back Valildation sends a DSN to the purported sender informing
+them of the lack of identification.  If the message is legitimate, the
+sender needs to know that their email setup is broken and should be corrected.
+If the message is forged, the sender is informed of the forgery, 
+and their need to publish an SPF record or at least use a valid HELO name.
+If the purported sender does not accept the DSN,
+then the message is rejected.  The CBV status is cached to avoid
+annoying the purported sender with too many DSNs.  Currently, the DSN
+is repeated to the same sender once per month.
+<p>
+In this example, although 3com.com has no SPF record, we assume that 
+any legitimate mail from them will at least have a valid HELO or PTR.
+<pre>
+2005Jul30 23:52:03 [96777] connect from [222.252.233.200] at ('222.252.233.200', 29934) EXTERNAL DYN
+2005Jul30 23:52:03 [96777] hello from 3mail.3com.com
+2005Jul30 23:52:04 [96777] mail from <etec_nic_family@3mail.3com.com> ()
+2005Jul30 23:52:04 [96777] REJECT: no PTR, HELO or SPF
+</pre>
+</td></tr>
+
+<tr><th>PASS</th><td>
+A pass result normally lets the email continue on, but the domain is 
+tracked for reputation (and may be blocked), and may skip content scanning if
+it matches a whitelist.
+<pre>
+2005Jul24 17:44:26 [2104] mail from <gnucash-devel-bounces@gnucash.org> ('SIZE=4410',)
+2005Jul24 17:44:26 [2104] Received-SPF: pass (mail.bmsi.com: domain of gnucash.org
+	designates 204.107.200.65 as permitted sender)
+	client-ip=204.107.200.65; envelope-from=gnucash-devel-bounces@gnucash.org; helo=cvs.gnucash.org;
+</pre>
+</td></tr>
+
+<tr><th>NEUTRAL</th><td>
+A neutral result normally lets the email continue on, but the domain is not
+tracked for reputation or matched against any whitelists. 
+Highly forged domains listed in <code>[SPF]reject_neutral</code> are
+rejected.
+<pre>
+2005Jul24 17:41:37 [2070] connect from cp500627-a.dbsch1.nb.home.nl at ('84.27.225.3', 3465) EXTERNAL
+2005Jul24 17:41:37 [2070] hello from cp500627-a.dbsch1.nb.home.nl
+2005Jul24 17:41:38 [2070] mail from <nwarjejkw@yahoo.com> ()
+2005Jul24 17:41:38 [2070] REJECT: SPF neutral for nwarjejkw@yahoo.com
+</pre>
+</td></tr>
+
+<tr><th>SOFTFAIL</th><td>
+A softfail result normally lets the email continue on, but the domain is not
+tracked for reputation or matched against any whitelists.  Furthermore,
+the message is flagged as needing Call Back Validation,
+and the highly forged domains listed in <code>[SPF]reject_neutral</code> are
+rejected as well.  
+<p>
+At present, we also require a valid HELO or PTR to avoid rejecting 
+a softfail.  But this should probably change to only require a
+successful CBV.
+<p>
+The Call Back Valildation sends a DSN to the purported sender informing
+them of the softfail.  If the message is legitimate, the sender needs
+to know about the softfail so that their email setup can be corrected.
+If the message is forged, the sender is informed of the forgery, confirming
+that SPF is protecting their reputation and encouraging a rapid transition
+to a strict policy.  If the purported sender does not accept the DSN,
+then the message is rejected.  The CBV status is cached to avoid
+annoying the purported sender with too many DSNs.  Currently, the DSN
+is repeated to the same sender once per month.
+<pre>
+2005Jul24 15:41:33 [801] mail from <Aitp@horafeliz.com> ()
+2005Jul24 15:41:33 [801] Received-SPF: softfail (mail.bmsi.com: transitioning domain of horafeliz.com
+	does not designate 221.184.83.185 as permitted sender)
+	client-ip=221.184.83.185; envelope-from=Aitp@horafeliz.com;
+	helo=p8185-ipad30funabasi.chiba.ocn.ne.jp;
+2005Jul24 15:41:33 [801] rcpt to <david@example.com> ()
+2005Jul24 15:41:35 [801] Subject: Microsoft, Adobe, Macromedia, Corel software. Up to 80% discount.
+2005Jul24 15:41:35 [801] X-Mailer: Microsoft Outlook, Build 10.0.2605
+2005Jul24 15:41:35 [801] CBV: Aitp@horafeliz.com
+2005Jul24 15:41:38 [801] REJECT: CBV: 550 <Aitp@horafeliz.com>: User unknown
+</pre>
+</td></tr>
+
+<tr><th>FAIL</th><td>
+The message is rejected with a reference the SPF why page.
+<pre>
+2005Jul30 19:53:27 [94070] connect from [212.70.52.16] at ('212.70.52.16', 3192) EXTERNAL DYN
+2005Jul30 19:53:27 [94070] hello from winzip.com
+2005Jul30 19:53:27 [94070] mail from <dan@winzip.com> ()
+2005Jul30 19:53:27 [94070] REJECT: SPF fail 550 SPF fail:
+	see http://openspf.com/why.html?sender=dan@winzip.com&ip=212.70.52.16
+</pre>
+</td></tr>
+
+<tr><th>PERMERROR</th><td>
+Permanent errors were called "unknown", and are still show that way
+in the log.  The message is rejected.  Previously, we enabled "lax" parsing
+of the SPF record, but rejecting is better because it informs the
+sender about their problem.  The next milter version will 
+look for a local substitute SPF record (as for a missing SPF record)
+before rejecting.  This will inform the sender of their problem, but
+also let the receiver install a temporary workaround.
+<pre>
+2005Jul24 18:05:37 [2312] mail from <b-mihdbcgaacaa-becibijh-000-@msg.euxiphipops.com> () 
+2005Jul24 18:05:37 [2312] REJECT: SPF unknown 550 SPF Permanent Error:
+	include mechanism missing domain: include
+</pre>
+The SPF record for msg.euxiphipops.com looked like this at the time of the
+above error:
+<pre>
+msg.euxiphipops.com	TXT "v=spf1 mx ptr a include"
+</pre>
+</td></tr>
+
+<tr><th>TEMPERROR</th><td>
+Temporary errors result in a 451 "Try again later" response.  The sender
+should retry the message at a later time.
+<pre>
+2005Jul24 07:33:13 [29846] mail from <quickenloans@rate.quicken.com> ('SIZE=73775', 'BODY=8BITMIME')
+2005Jul24 07:33:43 [29846] TEMPFAIL: SPF error 450 SPF Temporary Error: DNS Timeout
+</pre>
+</td></tr>
+
+</table>

doc/requirements.ht

@@ -0,0 +1,99 @@
+Title: Requirements
+
+<h2> Requirements </h2>
+
+<menu>
+<li> While the miltermodule will work with python 1.5, you probably
+want to use python 2.0 or better.  The python code uses a number of
+python 2 features.  The email support requires python 2.4.
+<li> Python must be configured with thread support.  This is because
+pymilter uses sendmail's libmilter which requires thread support.
+<li> You must compile sendmail with libmilter enabled.  In versions of
+sendmail prior to 8.12 libmilter is marked FFR (For Future Release) and
+is not installed by default.  
+Sendmail 8.12 still does not enable libmilter by default.  You must 
+explicitly select the "MILTER" option when compiling.  
+<li> When compiling Python milter against sendmail versions earlier than
+8.13, you must set MAX_ML_REPLY to 1 in setup.py.  There is no way to tell from
+the libmilter includes that smfi_setmlreply is not supported.
+<li> You probably want to use sendmail-8.13, since that supports multi-line
+SMTP error descriptions and SOCKETMAP.  You want SOCKETMAP for use with
+pysrs.
+<li> Python milter has been tested against sendmail-8.11 through sendmail-8.13.
+<li> Python milter must be compiled for the specific version of sendmail
+it will run with.  (Since the result is dynamically loaded, there could 
+conceivably be multiple versions available and selected at startup - but
+that will have to wait.)  This situation may only exist for sendmail
+versions prior to 8.12.  The protocol seems designed for backward 
+compatibility - and 8.12 is the first official milter release.
+<li> Mea Culpa!  After reading the Python Style guide, I realize that
+my Python code is not up to snuff.  Apparently mixed tabs and spaces
+are anathema to those using Windows editors, where tabs can be expanded using
+any arbitrary algorithm.  Other than that, my
+intuition matched Guido's pretty well - although I like to indent by 2
+rather than 4.  I will arrange to have tabs expanded to spaces when
+exporting new versions.  Until then, beware!
+</menu>
+
+<h3> <a name="aix4"> AIX 4.1.5 Requirements </a> </h3>
+To create sendmail RPMs for AIX, you can download my AIX 4.1.5 spec files 
+for <a href="/aix/sendmail.spec">sendmail-8.11.5</a> 
+or <a href="/aix/sendmail12.spec">sendmail-8.12.3</a>.  If you have
+not already set it up, I use a <a href="/aix/aix.spec">dummy RPM package</a>
+to represent the stuff that comes with AIX.  You might also want
+my <a href="/aix/python.spec">python-2.1.1</a> spec file for AIX.  It
+does not include Tk or curses modules, sorry.  If y'all trust me, you can
+download rpms for AIX 4.x from my <a href="/aix">AIX RPM directory</a>.
+<p>
+Sendmail-8.12 renames 
+libsmutil.a to libsm.a.  Unfortunately, libsm.a is an important AIX system
+shared library.  Therefore, I rename libsm.a back to libsmutil.a for
+AIX.  This presents a problem for setup.py.
+
+<h3> <a name="rh72"> RedHat 7.2 Requirements </a> </h3>
+
+If you are running Redhat 7.2, the distributed version of sendmail
+now enables libmilter by default.  RedHat 7.2 bundles
+the development libraries with the main sendmail package, so
+there is no sendmail-devel package.  However, they forgot to include the
+headers!  So you'll have to get the SRPM and modify it.  I suggest
+moving the static libs to a devel package and adding the headers.  If
+this is too much trouble, you can get the <a href="mfapi.h">mfapi.h</a>
+header for sendmail-8.6.11 from here and manually install it as
+<code>/usr/include/libmilter/mfapi.h</code>.
+<p>
+If you do modify the SRPM, I suggest renaming libsmutil.a
+to libsm.a - just like sendmail-8.12 will.  If you manually install
+mfapi.h or don't rename libsmutil.a, you'll
+need to force <code>libs = ["milter", "smutil"]</code> in setup.py.
+<p>
+If you have installed python2, and want
+python-milter to use python2, add <code>python=python2</code> to setup.cfg
+and build with <code>python2 setup.py bdist_rpm</code>.   
+
+<h3> <a name="rh62"> Redhat 6.2 Requirements </a> </h3>
+
+If you are running Redhat 6.2, the distributed version of sendmail
+does not enable libmilter.  You can download the Redhat 7.2 sendmail.spec
+modified to compile on RedHat 6.2:
+<a href="http://www.bmsi.com/linux/rh62/sendmail-rhmilter.spec">
+sendmail-rhmilter.spec</a>.  The <a
+href="ftp://updates.redhat.com/7.0/en/os/SRPMS/sendmail-8.11.6-1.7.0.src.rpm">
+SRPM for sendmail-8.11.6</a> is available from
+<a href="http://www.redhat.com">Redhat</a> under 
+<a href="http://www.redhat.com/support/errata/RHSA-2001-106.html">
+Errata for RH6.2</a>.  But that doesn't include the latest security
+patches since RH6.2 is no longer supported.
+<p>
+If y'all trust me, you can pick up source and binary sendmail RPMs for RH6.2
+from my <a href="http://www.bmsi.com/linux/rh62">linux downloads</a> directory.
+The lastest RPMs were built by taking a RH7.2 SRPMS and removing some
+RPM features from the spec file that RH6.2 doesn't support, then
+recompiling on RH6.2.  You can check this by installing the RH7.2 SRPM,
+then diffing my sendmail.spec with theirs.  Then run
+"rpm -bb sendmail-rhmilter.spec" when you are satisfied.
+<p>
+If you have installed python2, and want
+python-milter to use python2, add <code>python=python2</code> to setup.cfg
+and build with <code>python2 setup.py bdist_rpm</code>.
+You'll need to install the sendmail-devel package to compile milter.

milter.cfg

@@ -8,7 +8,7 @@ tempdir = /var/log/milter/save
 log_headers = 0
 # connection ips and hostnames are matched against this glob style list
 # to recognize internal senders.
-;internal_connect = 192.168.*.*
+;internal_connect = 192.168.*.*,127.*
 
 # mail that is not an internal_connect and claims to be from an
 # internal domain is rejected.  Furthermore, internal mail that
@@ -17,7 +17,7 @@ log_headers = 0
 # flexible.  However, SPF is not currently checked for outgoing
 # (internal_connect) mail because it doesn't yet handle authorizing 
 # internal IPs locally.
-;internal_domains = mycorp.com
+;internal_domains = mycorp.com,localhost.localdomain
 
 # connections from a trusted relay can trust the first Received header
 # SPF checks are bypassed for internal connections and trusted relays.
@@ -70,6 +70,10 @@ config=/etc/mail/pysrs.cfg
 ;fwdomain = mydomain.com
 # turn this on after a grace period to reject spoofed DSNs
 reject_spoofed = 0
+# Many braindead MTAs send DSNs with a non-DSN MFROM (e.g. to report that
+# some virus claiming to be sent by you).  This heuristic
+# refuses mail from user names commonly abused in that way.
+;banned_users = postmaster, mailer-daemon, clamav
 
 # See http://spf.pobox.com for more info on SPF.
 [spf]
@@ -85,6 +89,21 @@ reject_spoofed = 0
 ;reject_noptr = 0
 # always accept softfail from these domains, or send DSN otherwise
 ;accept_softfail = bounces.amazon.com
+# Treat fail from these domains like softfail: because their SPF record
+# or an important sender is screwed up.  Must have valid HELO, however.
+;accept_fail = custhelp.com
+# Use sendmail access map or similar format for detailed spf policy.
+# SPF entries in the access map will override any defaults set above.
+;access_file = /etc/mail/access.db
+# Add MAIL FROM as Sender when Sender is missing and From domain
+# doesn't match MAIL FROM.  Outlook and other email clients will then display
+# something like: "Sent by sender@domain.com on behalf of from@example.com"
+;supply_sender = 0
+# Connections that get an SPF pass for a pretend MAIL FROM of 
+# postmaster@sometrustedforwarder.com skip SPF checks for the real MAIL FROM.
+# This is for non-SRS forwarders.  It is a simple implementation that
+# is inefficient for more than a few entries.
+;trusted_forwarder = careerbuilder.com
 
 # features intended to clean up outgoing mail
 [scrub]
@@ -101,6 +120,7 @@ blind = 1
 # (sendmail aliases let you monitor incoming mail)
 #
 ;users = disloyal@bigcorp.com, bigmouth@bigcorp.com
+# multiple destinations can use smart_alias
 ;dest = spy@bigcorp.com
 # discard outgoing mail without alerting sender
 # can be used in conjunction with wiretap to censor outgoing mail
@@ -108,7 +128,10 @@ blind = 1
 #
 # smart aliases trigger on both sender and recipient
 #
-;smart_alias = copycust,walter
+;smart_alias = copycust,walter,spy1,spy2
+# multiple wiretap monitors
+;spy1 = disloyal@bigcorp.com,spy@bigcorp.com
+;spy2 = bigmouth@bigcorp.com,spy@bigcorp.com
 # mail from client@clientcorp.com to sue@bigcorp.com is redirected to 
 # local alias copycust
 ;copycust = client@clientcorp.com,sue@bigcorp.com
@@ -128,7 +151,13 @@ blind = 1
 # only EXTERNAL messages are dspam filtered
 ;dspam_dict=/var/lib/dspam/moderator.dict
 
-# Opt-opt recipients from dspam screening and header triage
+# Recipients of mail sent from these senders are added to the auto_whitelist.
+# Auto_whitelisted senders with an SPF PASS are never rejected by dspam, and
+# messages from auto_whitelisted senders will be used to train screener
+# dictionaries as innocent mail.
+;whitelist_senders = @mycorp.com
+
+# Opt-out recipients entirely from dspam screening and header triage
 ;dspam_exempt=getitall@mycorp.com
 # Do not scan mail (ostensibly) from these senders
 ;dspam_whitelist=getitall@sender.com

milter.rc

@@ -5,7 +5,7 @@
 # chkconfig: 2345 80 30
 # description: Milter is a process that filters messages sent through sendmail.
 # processname: milter
-# config: /var/log/milter/bms.py
+# config: /etc/mail/pymilter.cfg
 # pidfile: /var/run/milter/milter.pid
 
 python="python2.3"

milter.rc7

@@ -5,7 +5,7 @@
 # chkconfig: 2345 80 30
 # description: Milter is a process that filters messages sent through sendmail.
 # processname: milter
-# config: /var/log/milter/bms.py
+# config: /etc/mail/pymilter.cfg
 # pidfile: /var/run/milter/milter.pid
 
 python="python2.3"

milter.spec

@@ -1,6 +1,6 @@
 %define name milter
-%define version 0.8.2
-%define release 2.RH7
+%define version 0.8.5
+%define release 1.RH7
 # what version of RH are we building for?
 %define redhat9 0
 %define redhat7 1
@@ -39,11 +39,11 @@ Prefix: %{_prefix}
 Vendor: Stuart D. Gathman <stuart@bmsi.com>
 Packager: Stuart D. Gathman <stuart@bmsi.com>
 Url: http://www.bmsi.com/python/milter.html
-Requires: %{python} >= 2.4, sendmail >= 8.12.10
+Requires: %{python} >= 2.4, sendmail >= 8.13
 %ifos Linux
 Requires: chkconfig
 %endif
-BuildRequires: %{python}-devel , sendmail-devel >= 8.12.10
+BuildRequires: %{python}-devel , sendmail-devel >= 8.13
 
 %description
 This is a python extension module to enable python scripts to
@@ -52,10 +52,15 @@ modules provide for navigating and modifying MIME parts.
 
 %prep
 %setup
-#%patch -p1
+#%patch -p0 -b .bms
 
 %build
-env CFLAGS="$RPM_OPT_FLAGS" %{python} setup.py build
+if %{redhat9}; then
+  LDFLAGS="-g"
+else
+  LDFLAGS="-s"
+fi
+env CFLAGS="$RPM_OPT_FLAGS" LDFLAGS="$LDFLAGS" %{python} setup.py build
 
 %install
 rm -rf $RPM_BUILD_ROOT
@@ -63,7 +68,7 @@ rm -rf $RPM_BUILD_ROOT
 mkdir -p $RPM_BUILD_ROOT/var/log/milter
 mkdir -p $RPM_BUILD_ROOT/etc/mail
 mkdir $RPM_BUILD_ROOT/var/log/milter/save
-cp bms.py strike3.txt softfail.txt $RPM_BUILD_ROOT/var/log/milter
+cp bms.py *.txt $RPM_BUILD_ROOT/var/log/milter
 cp milter.cfg $RPM_BUILD_ROOT/etc/mail/pymilter.cfg
 
 # logfile rotation
@@ -146,7 +151,7 @@ rm -rf $RPM_BUILD_ROOT
 
 %files -f INSTALLED_FILES
 %defattr(-,root,root)
-%doc README NEWS TODO CREDITS sample.py
+%doc README HOWTO NEWS TODO CREDITS sample.py
 /etc/logrotate.d/milter
 /etc/cron.daily/milter
 %ifos aix4.1
@@ -160,12 +165,49 @@ rm -rf $RPM_BUILD_ROOT
 %dir /var/log/milter/save
 %config /var/log/milter/start.sh
 %config /var/log/milter/bms.py
-%config /var/log/milter/strike3.txt
-%config /var/log/milter/softfail.txt
+%config(noreplace) /var/log/milter/strike3.txt
+%config(noreplace) /var/log/milter/softfail.txt
+%config(noreplace) /var/log/milter/neutral.txt
+%config(noreplace) /var/log/milter/quarantine.txt
+%config(noreplace) /var/log/milter/permerror.txt
 %config(noreplace) /etc/mail/pymilter.cfg
 /usr/share/sendmail-cf/hack/rhsbl.m4
 
 %changelog
+* Thu Dec 29 2005 Stuart Gathman <stuart@bmsi.com> 0.8.5-1
+- Simple trusted_forwarder implementation.
+- Fix access_file neutral policy
+- Move Received-SPF header to beginning of headers
+- Supply keyword info for all results in Received-SPF header.
+- Move guessed SPF result to separate header
+- Activate smfi_insheader only when SMFIR_INSHEADER defined
+- Handle NULL MX in spf.py
+- in-process GOSSiP server support (to be extended later)
+- Expire CBV cache and renew auto-whitelist entries
+* Fri Oct 21 2005 Stuart Gathman <stuart@bmsi.com> 0.8.4-2
+- Don't supply sender when MFROM is subdomain of header from/sender.
+- Don't send quarantine DSN for DSNs
+- Skip dspam for replies/DSNs to signed MFROM
+* Thu Oct 20 2005 Stuart Gathman <stuart@bmsi.com> 0.8.4-1
+- Fix SPF policy via sendmail access map (case insensitive keys).
+- Auto whitelist senders, train screener on whitelisted messages
+- Optional idx parameter to addheader to invoke smfi_insheader
+- Activate progress when SMFIR_PROGRESS defined
+* Wed Oct 12 2005 Stuart Gathman <stuart@bmsi.com> 0.8.3-1
+- Keep screened honeypot mail, but optionally discard honeypot only mail.
+- spf_accept_fail option for braindead SPF senders (treats fail like softfail)
+- Consider SMTP AUTH connections internal.
+- Send DSN for SPF errors corrected by extended processing.
+- Send DSN before SCREENED mail is quarantined
+- Option to set SPF policy via sendmail access map.
+- Option to supply Sender header from MAIL FROM when missing.
+- Use logging package to keep log lines atomic.
+* Fri Jul 15 2005 Stuart Gathman <stuart@bmsi.com> 0.8.2-4
+- Limit each CNAME chain independently like PTR and MX
+* Fri Jul 15 2005 Stuart Gathman <stuart@bmsi.com> 0.8.2-3
+- Limit CNAME lookups (regression)
+* Fri Jul 15 2005 Stuart Gathman <stuart@bmsi.com> 0.8.2-2
+- Handle corrupt ZIP attachments
 * Fri Jul 15 2005 Stuart Gathman <stuart@bmsi.com> 0.8.2-1
 - Strict processing limits per SPF RFC
 - Fixed several parsing bugs under RFC 
@@ -174,7 +216,6 @@ rm -rf $RPM_BUILD_ROOT
 - Extended SPF processing results beyond strict RFC limits
 - Support original SES for local bounce protection (requires pysrs-0.30.10)
 - Callback exception processing option in milter module
-- Handle corrupt ZIP attachments
 * Thu Jun 16 2005 Stuart Gathman <stuart@bmsi.com> 0.8.1-1
 - Fix zip in zip loop in mime.py
 - Fix HeaderParseError in bms.py header callback

miltermodule.c

@@ -34,6 +34,15 @@ $ python setup.py help
      libraries=["milter","smutil","resolv"]
 
  * $Log$
+ * Revision 1.8  2005/10/20 23:23:36  customdesigned
+ * Include smfi_progress is SMFIR_PROGRESS defined
+ *
+ * Revision 1.7  2005/10/20 23:04:46  customdesigned
+ * Add optional idx for position of added header.
+ *
+ * Revision 1.6  2005/07/15 22:18:17  customdesigned
+ * Support callback exception policy
+ *
  * Revision 1.5  2005/06/24 04:20:07  customdesigned
  * Report context allocation error.
  *
@@ -967,28 +976,39 @@ milter_setreply(PyObject *self, PyObject *args) {
 }
 
 static char milter_addheader__doc__[] =
-"addheader(field, value) -> None\n\
+"addheader(field, value, idx=-1) -> None\n\
 Add a header to the message. This header is not passed to other\n\
 filters. It is not checked for standards compliance;\n\
 the mail filter must ensure that no protocols are violated\n\
 as a result of adding this header.\n\
 field - header field name\n\
 value - header field value\n\
+idx - optional position in internal header list to insert new header\n\
 Both are strings.  This function can only be called from the EOM callback.";
 
 static PyObject *
 milter_addheader(PyObject *self, PyObject *args) {
   char *headerf;
   char *headerv;
+  int idx = -1;
   SMFICTX *ctx;
   PyThreadState *t;
 
-  if (!PyArg_ParseTuple(args, "ss:addheader", &headerf, &headerv)) return NULL;
+  if (!PyArg_ParseTuple(args, "ss|i:addheader", &headerf, &headerv, &idx))
+    return NULL;
   ctx = _find_context(self);
   if (ctx == NULL) return NULL;
   t = PyEval_SaveThread();
-  return _thread_return(t,smfi_addheader(ctx, headerf, headerv),
-			 "cannot add header");
+#ifdef SMFIR_INSHEADER
+  return _thread_return(t, (idx < 0) ? smfi_addheader(ctx, headerf, headerv) :
+      smfi_insheader(ctx, idx, headerf, headerv), "cannot add header");
+#else
+  if (idx < 0)
+    return _thread_return(t, smfi_addheader(ctx, headerf, headerv),
+	"cannot add header");
+  PyErr_SetString(MilterError, "insheader not supported");
+  return NULL;
+#endif
 }
 
 static char milter_chgheader__doc__[] =
@@ -1143,7 +1163,7 @@ milter_quarantine(PyObject *self, PyObject *args) {
 }
 #endif
 
-#if _FFR_SMFI_PROGRESS
+#ifdef SMFIR_PROGRESS
 static char milter_progress__doc__[] =
 "progress() -> None\n\
 Notify the MTA that we are working on a message so it will reset timeouts.";
@@ -1174,7 +1194,7 @@ static PyMethodDef context_methods[] = {
 #ifdef SMFIF_QUARANTINE
   { "quarantine",  milter_quarantine,  METH_VARARGS, milter_quarantine__doc__},
 #endif
-#if _FFR_SMFI_PROGRESS
+#ifdef SMFIR_PROGRESS
   { "progress",  milter_progress,  METH_VARARGS, milter_progress__doc__},
 #endif
   { NULL, NULL }

neutral.txt

@@ -0,0 +1,34 @@
+Subject: SPF %(result)s (POSSIBLE FORGERY)
+
+This is an automatically generated Delivery Status Notification.
+
+THIS IS A WARNING MESSAGE ONLY.
+
+YOU DO *NOT* NEED TO RESEND YOUR MESSAGE.
+
+Delivery to the following recipients has been delayed.
+
+       %(rcpt)s
+
+Subject: %(subject)s
+Received-SPF: %(spf_result)s
+
+Your sender policy (or lack thereof) indicated that the above email was not
+sent via an authorized SMTP server, but may still be legitimate.  Since there
+is no positive confirmation that the message is really from you, we have
+to give it extra scrutiny - including verifying that the sender really
+exists by sending you this DSN.  We will remember this sender and not
+bother you again for a while.  You can avoid this message entirely for
+legitimate mail by using an authorized SMTP server.  Contact your mail
+administrator and ask how to configure your email client to use an
+authorized server.  
+
+If you never sent the above message, then your domain has been forged.
+Your mail admin needs to publish a strict SPF record so that I can reject
+those forgeries instead of bugging you about them.
+
+If you need further assistance, please do not hesitate to contact me.
+
+Kind regards,
+
+postmaster@%(receiver)s

permerror.txt

@@ -0,0 +1,31 @@
+Subject: Critical SPF configuration error
+
+This is an automatically generated Delivery Status Notification.
+
+THIS IS A WARNING MESSAGE ONLY.
+
+YOU DO *NOT* NEED TO RESEND YOUR MESSAGE.
+
+Delivery to the following recipients has been delayed.
+
+	%(rcpt)s
+
+Subject: %(subject)s 
+
+Your spf record has a permanent error.  The error was:
+
+	%(perm_error)s
+
+We will reinterpret your record using "lax" processing heuristics
+which may result in your mail being accepted anyway.  But you or your
+mail administrator need to fix your SPF record as soon as possible.
+
+We are sending you this message to alert you to the fact that
+you have problems with your email configuration.
+
+If you need further assistance, please do not hesitate to
+contact me again.
+
+Kind regards,
+
+postmaster@%(receiver)s

quarantine.txt

@@ -0,0 +1,26 @@
+Subject: DELIVERY STATUS (POSSIBLE SPAM)
+
+This is an automatically generated Delivery Status Notification.
+
+THIS IS A WARNING MESSAGE ONLY.
+
+YOU DO *NOT* NEED TO RESEND YOUR MESSAGE.
+
+Delivery to the following recipients has been delayed.
+
+       %(rcpt)s
+
+Subject: %(subject)s
+Received-SPF: %(spf_result)s
+
+A statistical analysis of your message has classified it as junk mail,
+and it has been quarantined.  Eventually, the recipients will review
+their quarantined mail and may notice your message.  If your message is
+important, please contact them via other means.  You may also try sending
+them a simple plain text message.
+
+If you need further assistance, please do not hesitate to contact me.
+
+Kind regards,
+
+postmaster@%(receiver)s

report.py

@@ -0,0 +1,138 @@
+# Analyze milter log to find abusers
+import traceback
+import sys
+
+def parse_addr(a):
+  beg = a.find('<')
+  end = a.find('>')
+  if beg >= 0:
+    if end > beg: return a[beg+1:end]
+  return a
+
+class Connection(object):
+  def __init__(self,dt,tm,id,ip=None,conn=None):
+    self.dt = dt
+    self.tm = tm
+    self.id = id
+    if ip:
+      _,self.host,self.ip = ip.split(None,2)
+    elif conn:
+      self.ip = conn.ip
+      self.host = conn.host
+      self.helo = conn.helo
+    self.subject = None
+    self.rcpt = []
+    self.mfrom = None
+    self.helo = None
+    self.innoc = []
+    self.whitelist = False
+
+def connections(fp):
+  conndict = {}
+  termdict = {}
+  for line in fp:
+    if line.startswith('{'): continue
+    a = line.split(None,4)
+    if len(a) < 4: continue
+    dt,tm,id,op = a[:4]
+    if (id,op) == ('bms','milter'):
+      # FIXME: optionally yield all partial connections in conndict
+      conndict = {}
+      termdict = {}
+      continue
+    if id[0] == '[' and id[-1] == ']':
+      try:
+	key = int(id[1:-1])
+      except:
+        print >>sys.stderr,'bad id:',line.rstrip()
+	continue
+    else: continue
+    if op == 'connect':
+      ip = a[4].rstrip()
+      conn = Connection(dt,tm,id,ip=ip)
+      conndict[key] = conn
+    elif op in (
+	'DISCARD:','TAG:','CBV:','Large','No',
+	'NOTE:','From:','Sender:','TRAIN:'):
+      continue
+    else:
+      op = op.lower()
+      try:
+	conn = conndict[key]
+      except KeyError:
+        try:
+	  conn = termdict[key]
+	  del termdict[key]
+	  conndict[key] = conn
+	except KeyError:
+	  print >>sys.stderr,'key error:',line.rstrip()
+	  continue
+      try:
+	if op == 'subject:':
+	  if len(a) > 4:
+	    conn.subject = a[4].rstrip()
+	elif op == 'innoc:':
+	  conn.innoc.append(a[4].rstrip())
+	elif op == 'whitelist':
+	  conn.whitelist = True
+	elif op == 'x-mailer:':
+	  if len(a) > 4:
+	    conn.mailer = a[4].rstrip()
+        elif op == 'x-guessed-spf:':
+          conn.spfguess = a[4]
+	elif op == 'received-spf:':
+	  conn.spfres,conn.spfmsg = a[4].rstrip().split(None,1)
+	elif op == 'received:':
+	  conn.received = a[4].rstrip()
+	elif op == 'temp':
+	  _,conn.tempfile = a[4].rstrip().split(None,1)
+	elif op == 'srs':
+	  _,conn.srsrcpt = a[4].rstrip().split(None,1)
+	elif op == 'mail':
+	  _,conn.mfrom = a[4].rstrip().split(None,1)
+	elif op == 'rcpt':
+	  _,rcpt = a[4].rstrip().split(None,1)
+	  conn.rcpt.append(rcpt)
+	elif op == 'hello':
+	  _,conn.helo = a[4].rstrip().split(None,1)
+	elif op in ('eom','dspam','abort'):
+	  del conndict[key]
+	  conn.enddt = dt
+	  conn.endtm = tm
+	  conn.result = op
+	  yield conn
+	  termdict[key] = Connection(conn.dt,conn.tm,conn.id,conn=conn)
+	elif op in ('reject:','dspam:','tempfail:','reject','fail:','honeypot:'):
+	  del conndict[key]
+	  conn.enddt = dt
+	  conn.endtm = tm
+	  conn.result = op
+	  conn.resmsg = a[4].rstrip()
+	  yield conn
+	  termdict[key] = Connection(conn.dt,conn.tm,conn.id,conn=conn)
+	elif op in ('fp:','spam:'):
+	  del conndict[key]
+	  termdict[key] = Connection(conn.dt,conn.tm,conn.id,conn=conn)
+	else:
+	  print >>sys.stderr,'unknown op:',line.rstrip()
+      except Exception:
+	print >>sys.stderr,'error:',line.rstrip()
+        traceback.print_exc()
+
+if __name__ == '__main__':
+  import gzip
+  for fn in sys.argv[1:]:
+    if fn.endswith('.gz'):
+      fp = gzip.open(fn)
+    else:
+      fp = open(fn)
+    for conn in connections(fp):
+      if conn.rcpt and conn.mfrom:
+        for r in conn.rcpt:
+	  if r.lower().find('iancarter') > 0: break
+	else:
+	  if conn.mfrom.lower().find('iancarter') < 0: continue
+	print >>sys.stderr,conn.result,conn.dt,conn.tm,conn.id,conn.subject,parse_addr(conn.mfrom),
+	for a in conn.rcpt:
+	  print parse_addr(a),
+	print

setup.cfg

@@ -1,5 +1,5 @@
 [bdist_rpm]
-python=python2
+python=python2.4
 doc_files=README NEWS TODO
 packager=Stuart D. Gathman <stuart@bmsi.com>
-release=2.4
+release=1

setup.py

@@ -12,7 +12,8 @@ if sys.version < '2.2.3':
   DistributionMetadata.classifiers = None
   DistributionMetadata.download_url = None
 
-setup(name = "milter", version = "0.8.2",
+# NOTE: importing Milter to obtain version fails when milter.so not built
+setup(name = "milter", version = '0.8.5',
 	description="Python interface to sendmail milter API",
 	long_description="""\
 This is a python extension module to enable python scripts to
@@ -31,6 +32,7 @@ querying SPF records.
 	ext_modules=[
 	  Extension("milter", ["miltermodule.c"],
 	    libraries=libs,
+	    # set MAX_ML_REPLY to 1 for sendmail < 8.13
 	    define_macros = [ ('MAX_ML_REPLY',32) ]
 	  ),
 	],

softfail.txt

@@ -1,4 +1,4 @@
-Subject: SPF softfail (POSSIBLE FORGERY)
+Subject: SPF %(result)s (POSSIBLE FORGERY)
 
 This is an automatically generated Delivery Status Notification.
 
@@ -14,7 +14,9 @@ Subject: %(subject)s
 Received-SPF: %(spf_result)s
 
 Your sender policy indicated that the above email was likely forged and that
-feedback was desired.
+feedback was desired.  If you are sending from a foreign ISP,
+then you may need to follow your home ISPs instructions for configuring
+your outgoing mail server.
 
 If you need further assistance, please do not hesitate to contact me.
 

spf.py

@@ -19,7 +19,7 @@ AND THERE IS NO OBLIGATION WHATSOEVER TO PROVIDE MAINTENANCE,
 SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
 
 For more information about SPF, a tool against email forgery, see
-	http://spf.pobox.com/
+	http://openspf.org/
 
 For news, bugfixes, etc. visit the home page for this implementation at
 	http://www.wayforward.net/spf/
@@ -47,6 +47,42 @@ For news, bugfixes, etc. visit the home page for this implementation at
 # Terrence is not responding to email.
 #
 # $Log$
+# Revision 1.17  2005/12/23 21:44:15  customdesigned
+# Always include keyword data in Received-SPF header.
+#
+# Revision 1.16  2005/12/01 22:42:32  customdesigned
+# improve gossip support.
+# Initialize srs_domain from srs.srs config property.  Should probably
+# always block unsigned DSN when signing all.
+#
+# Revision 1.15  2005/10/30 01:08:14  customdesigned
+# Ignore records missing spaces.
+#
+# Revision 1.14  2005/08/12 17:36:51  customdesigned
+# Trean non-existant include as no match in "lax" mode.
+#
+# Revision 1.13  2005/07/22 16:00:23  customdesigned
+# Limit CNAME chains independently of DNS lookup limit
+#
+# Revision 1.31  2005/07/22 02:11:50  customdesigned
+# Use dictionary to check for CNAME loops.  Check limit independently for
+# each top level name, just like for PTR.
+#
+# Revision 1.30  2005/07/21 20:07:31  customdesigned
+# Translate DNS error in DNSLookup.  This completely isolates DNS
+# dependencies to the DNSLookup method.
+#
+# Revision 1.29  2005/07/21 17:49:39  customdesigned
+# My best guess at what RFC intended for limiting CNAME loops.
+#
+# Revision 1.28  2005/07/21 17:37:08  customdesigned
+# Break out external DNSLookup method so that test suite can
+# duplicate CNAME loop bug.  Test zone data dictionary now
+# mirrors structure of real DNS.
+#
+# Revision 1.27  2005/07/21 15:26:06  customdesigned
+# First cut at updating docs.  Test suite is obsolete.
+#
 # Revision 1.26  2005/07/20 03:12:40  customdesigned
 # When not in strict mode, don't give PermErr for bad mechanism until
 # encountered during evaluation.
@@ -253,6 +289,20 @@ if not hasattr(DNS.Type,'SPF'):
   DNS.Type.typemap[99] = 'SPF'
   DNS.Lib.RRunpacker.getSPFdata = DNS.Lib.RRunpacker.getTXTdata
 
+def DNSLookup(name,qtype):
+  try:
+    req = DNS.DnsRequest(name, qtype=qtype)
+    resp = req.req()
+    #resp.show()
+    # key k: ('wayforward.net', 'A'), value v
+    return [((a['name'], a['typename']), a['data']) for a in resp.answers]
+  except DNS.DNSError,x:
+    raise TempError,'DNS ' + str(x)
+
+def isSPF(txt):
+  "Return True if txt has SPF record signature."
+  return txt.startswith('v=spf1 ') or txt == 'v=spf1'
+
 # 32-bit IPv4 address mask
 MASK = 0xFFFFFFFFL
 
@@ -305,6 +355,7 @@ DEFAULT_SPF = 'v=spf1 a/24 mx/24 ptr'
 MAX_LOOKUP = 10 #draft-schlitt-spf-classic-02 Para 10.1
 MAX_MX = 10 #draft-schlitt-spf-classic-02 Para 10.1
 MAX_PTR = 10 #draft-schlitt-spf-classic-02 Para 10.1
+MAX_CNAME = 10 # analogous interpretation to MAX_PTR
 MAX_RECURSION = 20
 ALL_MECHANISMS = ('a', 'mx', 'ptr', 'exists', 'include', 'ip4', 'ip6', 'all')
 COMMON_MISTAKES = { 'prt': 'ptr', 'ip': 'ip4', 'ipv4': 'ip4', 'ipv6': 'ip6' }
@@ -409,6 +460,9 @@ class query(object):
 	>>> q.check(spf='v=spf1 ip4:192.0.0.0/8 ?all moo')
 	('unknown', 550, 'SPF Permanent Error: Unknown mechanism found: moo')
 
+	>>> q.check(spf='v=spf1 =a ?all moo')
+	('unknown', 550, 'SPF Permanent Error: Unknown qualifier, IETF draft para 4.6.1, found in: =a')
+
 	>>> q.check(spf='v=spf1 ip4:192.0.0.0/8 ~all')
 	('pass', 250, 'sender SPF verified')
 
@@ -450,8 +504,6 @@ class query(object):
 			  self.perm_error.ext = rc
 			  raise self.perm_error
 			return rc
-		except DNS.DNSError,x:
-			return ('error', 450, 'SPF DNS Error: ' + str(x))
 		except TempError,x:
 			return ('error', 450, 'SPF Temporary Error: ' + str(x))
 		except PermError,x:
@@ -555,8 +607,15 @@ class query(object):
 			return ('none', 250, EXPLANATIONS['none'])
 
 		# split string by whitespace, drop the 'v=spf1'
-		#
-		spf = spf.split()[1:]
+		spf = spf.split()
+		# Catch case where SPF record has no spaces
+		# Can never happen with conforming dns_spf(), however
+		# in the future we might want to give permerror
+		# for common mistakes like IN TXT "v=spf1" "mx" "-all"
+		# in relaxed mode.
+		if spf[0] != 'v=spf1':   
+		    raise PermError('Invalid SPF record in', self.d)
+		spf = spf[1:]
 
 		# copy of explanations to be modified by exp=
 		exps = self.exps
@@ -601,9 +660,15 @@ class query(object):
 		      if res == 'pass':
 			break
 		      if res == 'none':
-			raise PermError(
-			  'No valid SPF record for included domain: %s'%arg,
-			  mech)
+			try:
+			  if self.strict or not self.perm_error:
+			    raise PermError(
+			      'No valid SPF record for included domain: %s'%arg,
+			      mech)
+			except PermError,x:
+			  if self.strict:
+			    raise x
+			  self.perm_error = x
 		      continue
 		    elif m == 'all':
 			    break
@@ -775,11 +840,11 @@ class query(object):
 		is found.
 		"""
 		# for performance, check for most common case of TXT first
-		a = [t for t in self.dns_txt(domain) if t.startswith('v=spf1')]
+		a = [t for t in self.dns_txt(domain) if isSPF(t)]
 		if len(a) == 1 and self.strict < 2:
 		    return a[0]   			
 		# check official SPF type first when it becomes more popular
-		b = [t for t in self.dns_99(domain) if t.startswith('v=spf1')]
+		b = [t for t in self.dns_99(domain) if isSPF(t)]
 		if len(b) == 1:
 		    # FIXME: really must fully parse each record
 		    # and compare with appropriate parts case insensitive.
@@ -792,7 +857,7 @@ class query(object):
 		if DELEGATE:	# use local record if neither found
 		    a = [t
 		      for t in self.dns_txt(domain+'._spf.'+DELEGATE)
-			if t.startswith('v=spf1')
+			if isSPF(t)
 		    ]
 		    if len(a) == 1: return a[0]
 		return None
@@ -823,7 +888,9 @@ class query(object):
 
 	def dns_a(self, domainname):
 		"""Get a list of IP addresses for a domainname."""
-		return self.dns(domainname, 'A')
+		if domainname:
+		  return self.dns(domainname, 'A')
+		return []
 
 	def dns_aaaa(self, domainname):
 		"""Get a list of IPv6 addresses for a domainname."""
@@ -844,7 +911,7 @@ class query(object):
 		"""Get a list of domain names for an IP address."""
 		return self.dns(reverse_dots(i) + ".in-addr.arpa", 'PTR')
 
-	def dns(self, name, qtype):
+	def dns(self, name, qtype, cnames=None):
 		"""DNS query.
 
 		If the result is in cache, return that.  Otherwise pull the
@@ -861,31 +928,33 @@ class query(object):
 		result = self.cache.get( (name, qtype) )
 		cname = None
 		if not result:
-			req = DNS.DnsRequest(name, qtype=qtype)
-			resp = req.req()
-			#resp.show()
-			for a in resp.answers:
-			    # key k: ('wayforward.net', 'A'), value v
-			    k, v = (a['name'], a['typename']), a['data']
+			for k,v in DNSLookup(name,qtype):
 			    if k == (name, 'CNAME'):
-				    cname = v
+				cname = v
 			    self.cache.setdefault(k, []).append(v)
 			result = self.cache.get( (name, qtype), [])
 		if not result and cname:
-			result = self.dns(cname, qtype)
+			if not cnames:
+			  cnames = {}
+			elif len(cnames) >= MAX_CNAME:
+			  raise PermError(
+			    'Length of CNAME chain exceeds %d' % MAX_CNAME)
+			cnames[name] = cname
+			if cname in cnames:
+			  raise PermError,'CNAME loop'
+			result = self.dns(cname, qtype, cnames=cnames)
 		return result
 
 	def get_header(self,res,receiver=None):
 	  if not receiver:
 	    receiver = self.r
-	  if res in ('pass','fail','softfail'):
-	    return '%s (%s: %s) client-ip=%s; envelope-from=%s; helo=%s;' % (
-	  	res,receiver,self.get_header_comment(res),self.i,
+	  if res in ('unknown','permerror'):
+	    txt = ' '.join([res] + self.mech)
+	  else:
+	    txt = res
+	  return '%s (%s: %s) client-ip=%s; envelope-from=%s; helo=%s;' % (
+	  	txt,receiver,self.get_header_comment(res),self.i,
 	        self.l + '@' + self.o, self.h)
-	  if res == 'unknown':
-	    return '%s (%s: %s)' % (' '.join([res] + self.mech),
-	      receiver,self.get_header_comment(res))
-	  return '%s (%s: %s)' % (res,receiver,self.get_header_comment(res))
 
 	def get_header_comment(self,res):
 		"""Return comment for Received-SPF header.
@@ -907,10 +976,10 @@ class query(object):
 		    "%s is neither permitted nor denied by domain of %s" \
 		    	% (self.i,sender)
 		    #"%s does not designate permitted sender hosts" % sender
-		elif res == 'unknown': return \
+		elif res in ('unknown','permerror'): return \
 		    "permanent error in processing domain of %s: %s" \
 		    	% (sender, self.prob)
-		elif res == 'error': return \
+		elif res in ('error','temperror'): return \
 		    "temporary error in processing during lookup of %s" % sender
 		elif res == 'fail': return \
 		    "domain of %s does not designate %s as permitted sender" \

strike3.txt

@@ -23,7 +23,7 @@ SMTP (email) servers to prevent criminals from forging your
 domain.  The simplest step is usually to publish an SPF record
 with your Sender Policy.  
 
-For more information, see: http://openspf.com
+For more information, see: http://openspf.org
 
 I hate to annoy you with a DSN (Delivery Status
 Notification) from a possibly forged email, but since you

testbms.py

@@ -1,4 +1,5 @@
 import unittest
+import doctest
 import Milter
 import bms
 import mime
@@ -22,7 +23,7 @@ class TestMilter(bms.bmsMilter):
 
   def getsymval(self,name):
     if name == 'j': return 'test.milter.org'
-    return bms.bmsMilter.getsymval(self,name)
+    return ''
 
   def replacebody(self,chunk):
     if self._body:
@@ -284,7 +285,10 @@ class BMSMilterTestCase(unittest.TestCase):
 #    self.failUnless(rc == Milter.REJECT)
 #    milter.close();
 
-def suite(): return unittest.makeSuite(BMSMilterTestCase,'test')
+def suite(): 
+  s = unittest.makeSuite(BMSMilterTestCase,'test')
+  s.addTest(doctest.DocTestSuite(bms))
+  return s
 
 if __name__ == '__main__':
   if len(sys.argv) > 1:
@@ -296,4 +300,5 @@ if __name__ == '__main__':
       fp = milter._body
       sys.stdout.write(fp.getvalue())
   else:
-    unittest.main()
+    #unittest.main()
+    unittest.TextTestRunner().run(suite())