[Python-checkins] python/nondist/sandbox/spambayes split.py,NONE,1.1

[bwarsaw@users.sourceforge.net]

Update of /cvsroot/python/python/nondist/sandbox/spambayes
In directory usw-pr-cvs1:/tmp/cvs-serv28692

Added Files:
	split.py 
Log Message:
A simple mailbox splitter


--- NEW FILE: split.py ---
#! /usr/bin/env python

"""Split an mbox into two files based on a given percentage.

This script will troll through a Unix mbox file randomly assigning each
message to one of two bins.  The split is based on a given float percentage.
E.g.

    % split.py -1 mbox20 -2 mbox80 20 sourcembox

yields two mbox files, where mbox20 contains approximately 20% of the messages
and mbox80 contains 80% of the messages.  Messages are assigned to each bin
randomly.

Usage: %(programs)s -1 file -2 file [options] percent sourcembox
Options:

    -h / --help
        Print this help message and exit
    -1 file
        Names the first output file.  Approximately percent % of the messages
        from the original source file will end up in this collection.
    -2 file
        Names the second output file.  Approximately 100-percent % of the
        messages from the original source file will end up in this
        collection.

percent is a floating point number between 1 and 99.  sourcembox is a Unix
mailbox file.  All arguments except -h/--help are required.
"""

import sys
import random
import mailbox
import email
import getopt

program = sys.argv[0]



def usage(code, msg=''):
    print >> sys.stderr, __doc__
    if msg:
        print >> sys.stderr, msg
    sys.exit(code)



def _factory(fp):
    try:
        return email.message_from_file(fp)
    except email.Errors.MessageParseError:
        return None



def main():
    try:
        opts, args = getopt.getopt(sys.argv[1:], 'h1:2:', ['help'])
    except getopt.error, msg:
        usage(1, msg)

    bin1 = bin2 = percentage = mboxfile = None
    for opt, arg in opts:
        if opt in ('-h', '--help'):
            usage(0)
        elif opt == '-1':
            bin1 = arg
        elif opt == '-2':
            bin2 = arg

    if bin1 is None or bin2 is None:
        usage(1, 'Both options -1 and -2 are required')

    try:
        percent = float(args[0])
        if not (0 < percent < 100):
            raise ValueError
        percent /= 100.0
        mboxfile = args[1]
    except IndexError:
        usage(1, 'Not enough arguments')
    except ValueError:
        usage(1, 'Percent argument must be a float between 1.0 and 99.0')

    # Cruise
    bin1out = open(bin1, 'w')
    bin2out = open(bin2, 'w')
    infp = open(mboxfile)

    mbox = mailbox.PortableUnixMailbox(infp, _factory)
    for msg in mbox:
        if random.random() < percent:
            outfp = bin1out
        else:
            outfp = bin2out
        print >> outfp, msg

    outfp.close()
    bin1out.close()
    bin2out.close()



if __name__ == '__main__':
    main()