#!/usr/bin/env python
#
# pythonolog.py - version 1.0
#
# A log creation script similar to cronolog but also supports mass vhosting
# as described at http://httpd.apache.org/docs/2.0/vhosts/mass.html
#
# Use this script from Apache like:
#
# CustomLog "|/path/to/pythonolog.py strftime-spec" [format]
#
# See 'man strftime' for possible options. Additionally %VHOST
# can be used if the first field of the log line is the virtual host.
#
# For example:
#
# LogFormat "%V %h %l %u %t \"%r\" %>s %b \"%{Referer}i\" \"%{User-Agent}i\"" combined
# CustomLog "|/path/to/pythonolog.py /my/logs/%Y/%m-%d-%VHOST-access_log" combined
#
# The vhost data is stripped from the resulting logfile.
# Any directories that do not exist are created automatically.
#
# Report bugs to Viraj Alankar <valankar@gmail.com>

import time
import sys
import os
import getopt
import re

MAXOPEN = 10
logfile = None
format = sys.argv[1]
want_vhost = format.count('%VHOST')

class Logfiles:
	def __init__(self):
		self.logs = {}

	def close_oldest_log(self):
		oldest = ''
		for log in self.logs:
			if not oldest or self.logs[log]['lastused'] < self.logs[oldest]['lastused']:
				oldest = log

		if oldest:
			del self.logs[oldest]

	def openlog(self, filename):
		# Make directory if it doesn't exist
		if filename.count('/'):
			try:
				os.makedirs(os.path.dirname(filename))
			except OSError:
				pass

		self.logs[filename] = {}
		self.logs[filename]['fileobj'] = open(filename, 'a')
		self.logs[filename]['lastused'] = 0

	def getlog(self, filename):
		if filename not in self.logs:
			# Open a new logfile
			if len(self.logs) >= MAXOPEN:
				self.close_oldest_log()
			self.openlog(filename)
		return self.logs[filename]

	def write(self, filename, line):
		output = self.getlog(filename)
		output['fileobj'].write(line)
		output['fileobj'].flush()
		output['lastused'] = time.time()

logs = Logfiles()

while True:
	useformat = format
	line = sys.stdin.readline()
	if not line:
		break

	# Virtualhost parsing
	if want_vhost:
		vhost = line.split()[0]
		line = re.sub(r'^%s\s+' % vhost, '', line)
		useformat = re.sub(r'%VHOST', vhost, format)
		
	filename = time.strftime(useformat)

	logs.write(filename, line)