--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/spamc-milter.py	Wed Nov 02 12:03:09 2011 +0100
@@ -0,0 +1,103 @@
+# Simple milter deamon that handles email messages off to spamd for
+# classification.  Messages classified as spam are rejected.  Other
+# are augmented with Spamassassin additional headers.
+
+import email.parser
+from cStringIO import StringIO
+
+import Milter as milter
+import spamc
+
+
+# headers to add to scanned messages
+HEADERS = ('X-Spam-Checker-Version', 'X-Spam-Level', 'X-Spam-Status', )
+
+
+class Milter(milter.Base):
+
+    def __init__(self):
+        # integer incremented with each call
+        self.id = milter.uniqueID()
+        self.fd = StringIO()
+
+    @milter.noreply
+    def connect(self, hostname, family, hostaddr):
+        self.hostname = hostname
+        self.hostaddr = hostaddr
+        return milter.CONTINUE
+
+    def hello(self, heloname):
+        self.helo = heloname
+        return milter.CONTINUE
+
+    def envfrom(self, mailfrom, *str):
+        self.mailfrom = mailfrom
+        return milter.CONTINUE
+
+    @milter.noreply
+    def envrcpt(self, recipient, *str):
+        self.recipient = recipient
+        return milter.CONTINUE
+
+    @milter.noreply
+    def header(self, name, val):
+        # add header to buffer
+        self.fd.write('%s: %s\n' % (name, val))
+        return milter.CONTINUE
+
+    @milter.noreply
+    def eoh(self):
+        # terminate headers
+        self.fd.write('\n')
+        return milter.CONTINUE
+
+    @milter.noreply
+    def body(self, chunk):
+        # add body to buffer
+        self.fd.write(chunk)
+        return milter.CONTINUE
+
+    def eom(self):
+        # ask spamd about the received message
+        self.conn = spamc.Client('localhost')
+        resp = self.conn.request('HEADERS', self.fd.getvalue())
+        # reject spam messages
+        if resp.result:
+            return milter.REJECT
+        # parse returned message headers
+        data = resp.read()
+        print data
+        headers = email.parser.Parser().parsestr(data)
+        # add headers
+        for name in HEADERS:
+            # transform multiple line headers to single line ones
+            value = ' '.join(headers.get(name, '').split())
+            self.addheader(name, value)
+        # accept message
+        return milter.ACCEPT
+
+    def close(self):
+        # clean up
+        self.fd.close()
+        self.conn.close()
+        return milter.CONTINUE
+
+    def abort(self):
+        # client disconnected prematurely
+        return milter.CONTINUE
+
+
+def main():
+    socketname = "/var/spool/postfix/private/spamc"
+    timeout = 600
+
+    # register the milter factory
+    milter.factory = Milter
+    # ask milter features we use
+    milter.set_flags(milter.CHGHDRS + milter.ADDHDRS)
+    # run
+    milter.runmilter("spamc-milter", socketname, timeout)
+
+
+if __name__ == "__main__":
+    main()

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/spamc.py	Wed Nov 02 12:03:09 2011 +0100
@@ -0,0 +1,235 @@
+# implements http://svn.apache.org/repos/asf/spamassassin/trunk/spamd/PROTOCOL
+
+import re
+import socket
+import email.parser
+
+
+# default spamd port
+PORT = 783
+
+# maximal line length when calling readline()
+MAXLINE = 65536
+
+# maximal amount of data to read at one time in
+MAXAMOUNT = 1048576
+
+# marker
+UNKNOWN = 'UNKNOWN'
+
+
+class ProtocolError(Exception):
+    pass
+
+
+class Response(object):
+    def __init__(self, sock, method):
+        self.fd = sock.makefile('rb', 0)
+        self.method = method
+        
+        self.version  = UNKNOWN
+        self.status   = UNKNOWN
+        self.reason   = UNKNOWN
+        self.result   = UNKNOWN
+        self.score    = UNKNOWN
+        self.required = UNKNOWN
+
+    def _read_status_line(self):
+        line = self.fd.readline()
+        if not line:
+            raise ProtocolError('empty status line')
+
+        try:
+            version, status, reason = line.split()
+        except ValueError:
+            raise ProtocolError('bad status line: %s' % line)
+
+        if not version.startswith('SPAMD/'):
+            raise ProtocolError('bad status line: %s' % line)
+
+        try:
+            version = float(version[6:])
+            status = int(status)
+            reason = reason.strip()
+        except ValueError:
+            raise ProtocolError('bad status line: %s' % line)
+
+        if not self.version > 1.0:
+            raise ProtocolError('unknown protocol: %s' % self.version)
+
+        return version, status, reason
+
+    def _read_headers(self):
+        headers = []
+        while True:
+            line = self.fd.readline(MAXLINE + 1)
+            if len(line) > MAXLINE:
+                raise ProtocolError('header line too long')
+            headers.append(line)
+            if line in ('\r\n', '\n', ''):
+                break
+        hstring = ''.join(headers)
+        return email.parser.Parser().parsestr(hstring)
+
+    def begin(self):
+        # parse response status line
+        self.version, self.status, self.reason = self._read_status_line()            
+
+        # parse headers
+        self.headers = self._read_headers()
+
+        # response length
+        try:
+            self.length = int(self.headers.get('content-length', ''))
+        except ValueError:
+            self.length = None
+
+        # if the used methods returns a Spam header
+        if self.method in ('CHECK', 'SYMBOLS', 'REPORT',
+                           'REPORT_IFSPAM', 'PROCESS', 'HEADERS'):
+            # parse Spam header
+            value = self.headers.get("spam")
+            try:
+                m = re.match(r'^(\w+)\s;\s(-?\d+.?\d?)\s/\s(\d+.?\d?)$', value)
+                result, score, required = m.groups()
+                self.result = result == 'True'
+                self.score = float(score)
+                self.required = float(required)
+            except (AttributeError, ValueError):
+                raise ProtocolError('bad results header: %s' % value)
+            
+    def _read_data(self, size):
+        s = []
+        while size > 0:
+            chunk = self.fd.read(min(size, MAXAMOUNT))
+            if not chunk:
+                raise ProtocolError('incomplete read')
+            s.append(chunk)
+            size = size - len(chunk)
+        return ''.join(s)
+
+    def read(self, size=None):
+        if self.fd is None:
+            return ''
+
+        if self.length is not None:
+            if size is None or size > self.length:
+                # clip read to response size
+                size = self.length
+            s = self._read_data(size)
+            self.length = self.length - size
+            if not self.length:
+                # we read everything
+                self.close()            
+            return s
+                
+        s = self.fd.read(size)
+        return s
+
+    def close(self):
+        if self.fd:
+            self.fd.close()
+            self.fd = None
+
+    def __str__(self):
+        return '%s %s' % (self.status, self.reason)
+
+
+class Client(object):
+    def __init__(self, host, port=PORT, timeout=socket._GLOBAL_DEFAULT_TIMEOUT, source=None):
+        self.host = host
+        self.port = port
+        self.timeout = timeout
+        self.source = source
+        self.sock = None
+        self.method = None
+        self.response = None
+
+    def connect(self):
+        self.sock = socket.create_connection((self.host,self.port),
+                                             self.timeout, self.source)
+
+    def close(self):
+        if self.sock:
+            self.sock.close()
+            self.sock = None
+        if self.response:
+            self.response.close()
+            self.response = None
+
+    def request(self, method, body=None, user=None, headers={}):
+        # save the used method
+        self.method = method
+
+        # buffer output
+        output = []
+
+        # construct request
+        request = '%s SPAMC/1.5' % method
+        output.append(request)
+        if body is not None:
+            output.append('Content-length: %d' % len(body))
+        if user is not None:
+            output.append('User: %s' % user)
+            # remove user from provided headers
+            headers.pop('User', None)
+        for header, value in headers.iteritems():
+            output.append('%s: %s' % (header, value))
+        output.append('')
+        if body is not None:
+            output.append(body)
+        output.append('')
+
+        # send
+        data = '\r\n'.join(output)
+        del output[:]
+        self.connect()
+        self.sock.sendall(data)
+
+        # parse response
+        self.response = Response(self.sock, self.method)
+        self.response.begin()
+        return self.response
+
+
+
+MESSAGE = """\
+Message-ID: <4E9D52C3.7060305@example.org>
+Date: Tue, 18 Oct 2011 12:19:47 +0200
+From: Example <example@example.org>
+MIME-Version: 1.0
+To: example@example.org
+Subject: Test
+Content-Type: text/plain; charset=ISO-8859-1
+Content-Transfer-Encoding: 7bit
+
+Hello
+
+"""
+
+
+def main():
+
+    conn = Client(HOST, PORT)
+    resp = conn.request('PING')
+    print resp
+    print resp.read()
+
+    for method in ('CHECK', 'SYMBOLS', 'REPORT', 'PROCESS', 'HEADERS', ):
+        print method
+        conn = Client(HOST, PORT)
+        resp = conn.request(method, MESSAGE, user='daniele')
+        print resp
+        print resp.read(1024)
+
+    method = 'HEADERS'
+    conn = Client(HOST, PORT)
+    resp = conn.request(method, MESSAGE, user='daniele')
+
+    headers = email.parser.Parser().parsestr(resp.read())
+    for name in ('X-Spam-Checker-Version', 'X-Spam-Level', 'X-Spam-Status', ):
+        print name, headers.get(name, '')
+    
+
+if __name__ == '__main__':
+    main()