CREATE UNIQUE INDEX tag_s ON tag USING btree (tag);

-- table to hold tag - site links
CREATE TABLE urls (
	date_added timestamp without time zone DEFAULT ('now'::text)::timestamp(0) without time zone NOT NULL,
	id_site smallint NOT NULL,
	id_tag smallint NOT NULL
	id_tag smallint NOT NULL,
	regex text
);

ALTER TABLE ONLY urls
	ADD CONSTRAINT urls_pkey PRIMARY KEY (date_added);

CREATE UNIQUE INDEX urls_id_site ON urls USING btree (id_site);

-- rule to join all tables into one to simplify access
-- automaticall uses current day and time data
CREATE VIEW site_rule AS
SELECT a.redirect_url, a.netmask, b.site
SELECT a.redirect_url, a.netmask, b.site, b.regexp
FROM ((
	SELECT rules.redirect_url, tag.tag AS rule_tag, rules.netmask
	FROM (rules NATURAL JOIN tag)
	WHERE ((((('now'::text)::time without time zone >= rules.from_time)
		AND (('now'::text)::time without time zone <= rules.to_time))
		AND (date_part('dow'::text, now()) >= (rules.from_weekday)::double precision))
		AND (date_part('dow'::text, now()) <= (rules.to_weekday)::double precision))
	FROM rules NATURAL JOIN tag
	WHERE ('now'::text)::time without time zone >= rules.from_time
		AND ('now'::text)::time without time zone <= rules.to_time
		AND date_part('dow'::text, now()) >= (rules.from_weekday)::double precision
		AND date_part('dow'::text, now()) <= (rules.to_weekday)::double precision
) a JOIN (
	SELECT site.site, tag.tag AS url_tag
	FROM ((urls NATURAL JOIN tag) NATURAL JOIN site)
) b ON ((b.url_tag && a.rule_tag)));
	SELECT site.site, tag.tag AS url_tag, regexp
	FROM urls NATURAL JOIN tag NATURAL JOIN site
) b ON (b.url_tag && a.rule_tag));

#!/usr/bin/env python3.1

import configparser, optparse, os, postgresql.api, re, sys, _thread

class Logger:
	__slots__ = frozenset(['_silent', '_syslog'])

	def __init__(self, silent = True):
		if silent:
	def __init__(self):
		config.section('log')
		if config['silent'] = 'yes':
			self._silent = True
		else:
			import syslog
			self._syslog = syslog
			self._syslog.openlog('squidTag')
			self._silent = False

class tagDB:
	__slots__ = frozenset(['_prepared', '_check_stmt', '_db'])

	def __init__(self):
		self._prepared = set()
		self._db = False
		self._check_stmt = self._curs().prepare("select redirect_url from site_rule where site <@ tripdomain($1) and netmask >> $2::text::inet limit 1")
		self._check_stmt = self._curs().prepare("select redirect_url, regexp from site_rule where site <@ tripdomain($1) and netmask >> $2::text::inet order by array_length(site, 1) desc limit 1")

	def _curs(self):
		if not self._db:
			config.section('database')
			self._db = postgresql.open(
				'pq://{0}:{1}@{2}/{3}'.format(
					config['user'],
					config['password'],
					config['host'],
					config['database'],
			) )
		return(self._db)

	def check(self, ip_address, site):
		result = self._check_stmt(site, ip_address)
		if len(result) > 0:
			if result[1]:
				if re.compile(result[1]).match(url_path):
					return result[0]
				else:
					return None
			return result[0]
		else:
			return None

class CheckerThread:
	__slots__ = frozenset(['_db', '_lock', '_lock_queue', '_log', '_queue'])

		parser.add_option('-c', '--config', dest = 'config',
			help = 'config file location', metavar = 'FILE',
			default = '/usr/local/etc/squid-tagger.conf')

		(options, args) = parser.parse_args()

		if not os.access(options.config, os.R_OK):
			print("Can't read {0}: exitting".format(options.config))
			print("Can't read {}: exitting".format(options.config))
			sys.exit(2)

		self._config = configparser.ConfigParser()
		self._config.readfp(open(options.config))

	def section(self, section):
		self._section = section

	def __getitem__(self, name):
		return self._config.get(self._section, name)

config = Config()

log = Logger(False)
log = Logger()
db = tagDB()
checker = CheckerThread(db,log)

while True:
	line = sys.stdin.readline()
	if len(line) == 0:
		break
	checker.check(line)