INSTALLATION
  ============

  compile binary:

      tar xfj squclac-x.y.z.tar.bz
      cd squclac-x.y.z
      make
      cp squclac /usr/local/squid/bin/   # or any other place where squid can execute it


  create config file, place in /etc or and make sure it is readable by squid
  user.

  copy squclac binary in some place accessible by squid and verify that
  squid user have +rx persmission.

  edit squid.conf, so first place is redirect_programm

      redirect_program /usr/local/squid/bin/squclac
      redirect_children 10

  now find line:

      acl localhost src 127.0.0.1/255.255.255.255

  and somewhere after this line add:

      redirector_access deny localhost

  make sure localhost can use proxy server, because redirector will use your
  proxy server as 127.0.0.1:8080. If your proxy server is running on diferent
  port use configuration directive proxy.

  to allow access from localhost, put something like that on squid.conf:

      http_access allow localhost


  restart or reconfigure squid, and check the squclac log file.



  USAGE
  =====

  If you change configuration file, you do not need to reconfigure squid,
  because it will impact performace. It is not necesary to kill all
  redirecors. Just send SIGHUP to all of them. On SIGHUP redirectors reloads
  configuration file and checks for new clamav db version.

  normaly there are muliple instances of squclac redirector, and it is easy to
  send SIGHUP to all of them with pkill:

      pkill -HUP squclac



  If you start squclac from shell, hopefuly you do it for testing purposes.
  To get out of it, press ctrl-d, so pressing ctrl-c will not exit.



  minsize, maxsize and timeout explanation:
  -----------------------------------------


  maxsize - if file is biger than this, it is not scaned, but pased or
  blocked. if it is possible to determine size before download, file will not
  be downloaded. There are many situatiations when size can't be determined
  before download, so when this situation is meet, squclac will try to
  download it and when maxsize is reached, download is stoped and file is
  passed. if file size is under maxsize, then size and timeout parameters are
  used.

  minsize - file size to download, only after this size timeout timer will
  start to count. usualy virus is small file. if it is served from slow
  connection using only timout can pass viruss, thats why minsize is used.

  timeout - after size bytes are downloaded, timer start to count, and if its
  value is reached, download is stoped and file is passed or blocked. So the
  maximal scaned sizes depends on your connection speed and load.


  [  size  | timout --> | ................. | maxsize ]

  now you can define what to do with files bigger than timeout. They can
  be passed or blocked for some of users or ips:

  example:

      # set default timeout action to redirect
      timeout action redirect;

      # add some src/dst
      timeout exeption whitelist;
      timeout exeption good_clients;



  troubleshooting
  ---------------


  strange behaviors:

  * cpu load getting high, pattern mathed urls does not load at all, squid
  reloads redirectors all the time

    possibly some of squid logfiles readhed size 2Gb, do ./squid -k rotate

  * all redirector processes get busy

    possibly you have slow internet connection, then get faster one.
  
    users are using slow wep pages, add pages to whitelists, change pattern.