nixspamsum
author Pascal Volk <user@localhost.localdomain.org>
Fri, 03 Jul 2009 21:12:03 +0000
changeset 4 ae58d9881be5
parent 3 6b0d09cdfbdb
child 5 65590f05bb97
permissions -rwxr-xr-x
showResult(): count matches; break if no matches were found sort by domain name if the output format is CSV

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# Copyright 2009 Pascal Volk

__author__ = 'Pascal Volk'
__version__ = '0.1.2'
__date__ = '2009-07-03'

import os
import re

class NiXSapmSum:
    """Do sth ..."""

    """Regular expression pattern for mail logs from Postfix"""
    RE_PF = '''^[\w\s:-]{17,80}\spostfix\/smtpd\[[\d]{3,5}\]: NOQUEUE: reject:.*blocked using ix.dnsbl.manitu.net; Spam sent to the mailhost ((?:[a-z0-9-]{1,63}\.){1,}[a-z]{2,6}) was detected by NiX Spam.*$'''

    def __init__(self):
        self._doms = {}
        self._mxs  = {}
        self._repo = None

    def setLogFormat(self, format='postfix'):
        if format == 'postfix':
            self._repo = re.compile(NiXSapmSum.RE_PF)
        else:
            raise Exception('MTA/Logformat not supported yet.')

    def parseLog(self, filehandle):
        for l in filehandle:
            mo = self._repo.match(l)
            if mo:
                mx = mo.group(1)
                try:
                    self._mxs[mx] += 1
                except KeyError:
                    self._mxs[mx]  = 1

    def countByDom(self):
        for mx in self._mxs.keys():
            dom = '.'.join(mx.split('.')[-2:])
            try:
                self._doms[dom] += self._mxs[mx]
            except KeyError:
                self._doms[dom]  = self._mxs[mx]

    def getDomains(self):
        return self._doms

    def getMXs(self):
        return self._mxs

def getOptionParser():
    from optparse import OptionParser
    description = 'do something ...'
    usage  = 'usage: %prog [options] maillog [maillog [...]]'
    version = '%prog '+__version__
    parser = OptionParser(description=description,usage=usage,version=version)
    parser.add_option('-d', action='store_true', dest='countByDom',
            default=False, help='summarize all MX by domain')
    parser.add_option('-m', action='store_false', dest='countByDom',
            help='count per MX host [default]')
    parser.add_option('-o', dest='oFormat', default='table',metavar='FORMAT',
            help='the output format: table or csv [default: %default]')
    parser.add_option('-p', action='store_true', dest='percent', default=False,
            help='show also percentages in table output [default: %default]')
    parser.add_option('-s', dest='order', default='name', metavar='SORTBY',
            help='arrange output by: name or count [default: %default]')
    parser.add_option('-t', dest='format', default='postfix',metavar='MTA',
            help='MTA that generated the maillog [default: %default]')
    return parser

def openLogFile(fname):
    try:
        fh = open(fname)
        return fh
    except IOError, e:
        os.sys.stderr.write('Warning: %s\nskipped file: %s\n' % (e.strerror,
            fname))

def getDomLen(domainnames):
    dlen = 0
    for d in domainnames:
        l = len(d)
        if l > dlen:
            dlen = l
    return dlen

def buildTable(output, domains, percent, orderBy):
    k = 0 if orderBy == 'name' else 1
    doms = sorted(domains.items(), lambda d,c: cmp(d[k],c[k]), reverse=k)
    dlen = getDomLen(domains.keys())+1
    clen = len(str(max(domains.values())))
    total = sum(domains.values())
    if percent:
        format = ' %%%ds  %%%dd  %%6.2f %%%%\n' % (-dlen, clen)
        for d, c in doms:
            dfrac = 100./total*c
            output.write(format % (d, c, dfrac))
        output.write('%s\n' % ((clen+dlen+14)*'-'))
        output.write(format % ('total', total, 100))
    else:
        format = ' %%%ds  %%%dd\n' % (-dlen, clen)
        for d in doms:
            output.write(format % d)
        output.write('%s\n' % ((clen+dlen+4)*'-'))
        output.write(format % ('total', total))

def showResult(nixspamsum, options):
    if options.countByDom:
        nixspamsum.countByDom()
        domains = nixspamsum.getDomains()
    else:
        domains = nixspamsum.getMXs()
    if not len(domains):
        print "No Nix Spam RBL rejects found"
        return

    from cStringIO import StringIO
    output = StringIO()
    # build the table
    if options.oFormat == 'table':
        buildTable(output, domains, options.percent, options.order)
    # generate comma separated values
    elif options.oFormat == 'csv':
        order = domains.keys()
        order.sort()
        for d in order:
            output.write("'%s',%d\n" % (d, domains[d]))
    # should never be reached
    else:
        print "Oops, error in function showResult() happend"
    # show the result
    print output.getvalue()

def main():
    parser = getOptionParser()
    opts, args = parser.parse_args()
    if opts.oFormat not in ('csv', 'table'):
        parser.error("Output format '%s' is not supported" % opts.oFormat)
    if len(args) < 1:
        parser.error('No logfiles specified')
    nixss = NiXSapmSum()
    nixss.setLogFormat(opts.format)
    for fn in args:
        fh = openLogFile(fn)
        if fh is not None:
            nixss.parseLog(fh)
            fh.close()
    showResult(nixss, opts)

if __name__ == '__main__':
    main()