The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

pastebot - the original clipboard to chat gateway

SYNOPSIS

pastebot -f configfile

DESCRIPTION

Pastebot is an IRC bot and a web server in one program. Users submit short pieces of text to the web server, and the IRC bot announces them in IRC.

A single pastebot instance can provide one or more web servers, each accepting connections on a different virtual host and/or port.

Likewise, a single pastebot may connect to zero or more IRC networks and announce pastes on zero or more channels on each. The pastebot merely acts as a web service in zero-bot mode.

CONFIGURATION SYNTAX

Nearly all of pastebot's options are set in a configuration file. The general syntax for this file follows these rules:

Sections are defined by text that begins in the first column. Section parameters are indented and consist of a parameter name and its value separated by whitespace. This example defines a web_server section named "poenet" that listens on port 8888.

  web_server
    name  poenet
    port  8888

Each section defines a component type within the pastebot. Each section requires a unique name since there may be more than one instance of a component. The names permit the program to reference specific components without confusion.

Some parameters may contain multiple values, though. Each value is defined on a separate configuration line using the same parameter name. For example, it's useful to configure multiple IRC servers for a bot to choose from. If one server is unavailable, it can try another. Here's how that's done:

  irc
    name efnet
    server irc.carrier1.net.uk  6667
    server irc.daxnet.no        6667

Configuration files may contain comments, which are text after a hash mark ("#"). Comments and blank lines are ignored.

  # Configure the paste database.
  pastes
    expire 86400

That's all.

CONFIGURING PASTEBOT

Pastebot configuration is broken down into three section types: "web_server", "irc", and "pastes". They define the web server, irc network, and paste database that the bot will use.

Optional Administrivia

These settings make it easier to administrate pastebot. Right now it only covers a PID file, but may be expanded upon in the future. This entire section is optional, but is global for the bot.

A sample:

  administrivia
    name  administrivia
    pidfile /var/run/pastebot.pid
name

The name for this section. 'administrivia' is simple and clear, but it's not limited to this.

pidfile

A path, absolute or relative, that will hold a PID for the bot. Upon startup, this file will be read and its contents interpreted as a PID. If this PID is running, pastebot will exit with a message notifying the administrator of this fact. If this PID is not running, pastebot will write its own PID to the pidfile before continuing.

Configuring a Web Server

The web_server section defines a pastebot's web server. A pastebot may have more than one web server. Each web server is defined in a separate web_server section, and each must have a different name.

A sample:

  web_server
    iface    10.0.0.80
    ifname   poe.perl.org
    iname    http://poe.perl.org:8888/
    irc      freenode_irc
    name     freenode_web
    port     8888
    proxy    127.0.0.1
    static   /path/to/static
    template Bot::Pastebot::TextTemplate
iface HOST_OR_ADDRESS

Defines the interface host or numeric address that this web server will bind to while listening for new connections. It's common for the interface address to be different than the one the outside world sees. Required.

ifname HOST_OR_ADDRESS

Defines the web server's public address, which is displayed to users in various messages and may be different than the internal address the web service has bound to. Optional. If omitted, the value of "iface" will be used instead.

iname BASE_URL

The "home page" for the pastebot's web server. It's advertised in multiple places, such as the bot's WHOIS information. The BASE_URL should agree with "ifname" and "port". Optional.

irc IRC_NAME

Identifies the IRC bot that will announce paste URLs. The IRC_NAME is the name of the bot as specified by the "name" parameter of the "irc" section. Required.

name NAME

Defines the web server's name. Each web server must have its own name so they can be referenced individually. The web server's name is also used in web page titles and IRC announcements. Required.

port PORT_NUMBER

This is the port the web server listens on. Required.

proxy ADDRESS

Optional. Used when the pastebot's web server is behind a proxy. Identifies the proxy so it may be ignored when headers are examined.

static STATIC_PATH

Optional. Defines the directory that contains static files that will be provided by the pastebot's built-in web server. Templates, graphics, and style sheets for the various served pages go here. It is essentially the Pastebot server's document root. Each web server is permitted to have its own static directory.

The Pastebot comes with a set of default static files that will be used if "static" is not specified. The files are installed in Bot::Pastebot's architecture-dependent auto directory, which may be found by running the following Perl command:

  perl -MFile::ShareDir=dist_dir -wle 'print dist_dir("Bot-Pastebot")'

If "static" is omitted, the default files will be loaded from that directory.

template CLASS

Optional. Defines the class that implements the templating for generated pages. The class must define specific methods, see Bot::Pastebot::TextTemplate for which methods and return values are expected.

Configuring an IRC Bot

The irc section defines a pastebot's IRC bot. A pastebot may have zero, one, or more IRC bots. Each bot is defined in a separate irc section, and each must have a different name. A pastebot with no IRC bots is a web-only service.

A sample:

  irc
    away          saving humanity from evil paste
    ccinfo        ACTION VERSION CLIENTINFO USERINFO
    channel       pastebot
    channel       perl
    channel       poe
    channel       \#\#perl
    cuinfo        owned and operated by rcaputo <rcaputo@cpan.org>
    cver          pastebot 1.0 <http://thirdlobe.com/projects/pastebot/>
    flags         +i
    iname         http://poe.perl.org:8888/
    join_cfg_only 1
    localaddr     poe.perl.org
    name          freenode_irc
    nick          eatpaste
    nick          pastebot
    quit          how will you live without me?
    server        irc.freenode.net 6667
    server        irc.freenode.net 7000
    uname         pastebot
    server_pass   s3krit
    use_ssl       0
    nickserv_pass password
away AWAY_MESSAGE

Pastebots set themselves "away" immediately when they connect to IRC. They run unattended, so there is nobody to answer any messages. This parameter sets the message that will be displayed to anyone examining the bot's WHOIS information.

ccinfo CTCP_CLIENT_INFO

Sets the bot's CTCP CLIENTINFO response. Optional, but some networks require that clients respond to client information requests.

channel CHANNEL

Required. Defines the channel or channels that the pastebot will automatically join when it connects to this IRC network. A pastebot may join multiple channels, each defined as a separate "channel" entry. The channel prefix '#' may be escaped with a '\' allowing the bot to join channels such as ##perl (\#\#perl).

cuinfo CTCP_USER_INFO

Sets the bot's CTCP USERINFO response. Optional, but some networks require that clients respond to user information requests.

cver CTCP_VERSION_INFO

Sets the bot's CTCP VERSION response. Optional, but some networks require that clients respond to version requests.

flags USER_FLAGS

Optional. When specified, sets the user flags that the pastebot will set once it has connected to the IRC network. Commonly set to "+i" so the bot will be "invisible" outside the channels it's currently on. This helps reduce the amount of harassment the bot receives.

iname IRC_NAME

The pastebot's "IRC name", which is what appears in WHOIS information. By convention, pastebots include their home page URLs here. Be sure the URL's host and port match "ifname" and "port" in the appropriate web_server section. Required.

join_cfg_only BOOLEAN

Optional. Defaults to 0 (false). Normally a pastebot will join any channel it's invited into. When join_cfg_only is true, however, the pastebot will only join channels that have been specified in its configuration file.

localaddr ADDRESS

Optional. The localaddr specifies an ADDRESS that the bot will bind to before making connections. This allows the bot to appear online as coming from a particular virtual host ADDRESS, if its host machine has more than one.

name NAME

Required. This is the IRC bot's internal name. To be useful, it should match the "irc" parameter in one of the web_server sections.

nick NICKNAME

Required. Specifies the bot's nickname on the IRC network. A bot may have multiple nicknames, each specified as a separate configuration line. It will attempt to use them in order, skipping to the next in the list as it determines each is already taken. If every nickname is taken, it will append numeric digits to the names and try again.

The pastebot will periodically try to reclaim one of the original nicknames in its configuration.

quit QUIT_MESSAGE

This sets the pastebot's quit message should it need to log off for any reason.

server SERVER PORT

Sets the IRC server and port to connect to. Required. The pastebot can connect to one of several servers, each specified on a different "server" line. If one is unavailable, it will rotate to the next.

uname USER_NAME

IRC requires clients to send a short "user name" when establishing a connection. This parameter specifies that name.

use_ssl BOOLEAN

Optional. Set this flag to a true value to connect over SSL.

server_pass PASSWORD

Optional. Some IRC servers require a private password to connect. This parameter specifies that password.

nickserv_pass PASSWORD

Specifies the password to use to identify to the NickServ nickname service. Upon connecting to the irc server, the command "/msg NickServ IDENTIFY PASSWORD" will be issued to identify the bot to the NickServ service.

Configuring Pastebot's Database

Pastebot must store uploaded text somewhere.

  pastes
    check   3600
    count   0
    expire  864000
    name    pasty
    store   /home/troc/.pastebot
check SECONDS

How often to check for expired pastes. The value is the number of SECONDS between each check. 3600 is an hour, 86400 is a day.

Performing expiration checks too frequently will slow down the bot. Currently these checks are done within the main process, which is cooperatively multitasked. The entire bot pauses while the check occurs.

TODO - Revise the expiration check to work cooperatively, or fork it into a separate process altogether.

count MAX_PASTES

Set a limit on the number of pastes a pastebot's database may hold, regardless whether they're too young to expire. This limit will be disabled if MAX_PASTES is zero.

If set, only the MAX_PASTES newest pastes will remain after a periodic expiration check.

expire MAX_AGE_SECONDS

Sets the maximum limit for the age of pastes to retain. This limit may be disabled if MAX_AGE_SECONDS is zero. If set, however, pastes that are older than MAX_AGE_SECONDS will be purged during a periodic expiration check.

name NAME

Defines the database's name. Each database must have a unique name so they may be addressed individually. Database names are not currently used, but the configuration parser requires every component to be named.

store DIRECTORY

Sets the absolute path of the base directory in which pastes will be stored. This directory will be created if it doesn't exist. Parent directories will not be created, however.

REVERSE PROXYING

Pastebot is its own web server. It often cannot bind to port 80 because some other web server is already there. The recommended port 8888 may be hostile towards people behind firewalls.

Apache users can use ProxyPass to map a directory on their port-80 servers to a running pastebot. This feature requires Apache be built with both mod_proxy and mod_proxy_add_forward.c support, and that the add forward module come before the proxy module. For example:

  ./configure --enable-module=proxy \
    --activate-module=src/modules/extra/mod_proxy_add_forward.c \
    --enable-module=proxy_add_forward \
    --permute-module=proxy:END \
    ...

Once Apache is built, you can create a virtual location for the proxy directory.

  <Location /pastebot>
    Order allow,deny
    Allow from all
    Deny from none
  </Location>

Then you must tell Apache to forward any requests in the /pastebot/ directory through to the local Pastebot server.

  ProxyPass  /pastebot/  http://127.0.0.1:8888/

Now people can use your pastebot on port 80. Be sure to set your pastebot's "iname" to match your proxy's address and path. For example:

  web_server
    ...
    proxy 127.0.0.1
    iname http://example.org/pastebot/

TROUBLESHOOTING

Common problems and solutions.

Can't locate [some module] in @INC (@INC contains: ...)

Somehow Bot::Pastebot was installed without one or more prerequisites. Install the module that can't be located, and try again.

This error may show up multiple times until each required module is installed. Use a package manager, like the CPAN shell, to automate the process.

IRC client (NAME): socket error occurred: Can't assign requested address

The most likely cause is an unknown host name, either for a local machine or a remote IRC server. If a pastebot has multiple servers specified, it will rotate to the next available server in about a minute.

<something> section <name> is redefined at ##

The error is caused when two or more sections in the pastebot configuration share the same name. To fix it, configure pastebot to use different names for each section.

In practice, the error is very noisy. Fortunately you only need the first line. The following example says which section (pbtest) is redefined, and where (line 68):

  pastes section pbtest is redefined at 68
  Compilation failed in require at Server/Web.pm line 18, <MPH> line 85.
  BEGIN failed--compilation aborted at Server/Web.pm line 18, <MPH> line 85.
  Compilation failed in require at ./pastebot.perl line 9, <MPH> line 85.
  BEGIN failed--compilation aborted at ./pastebot.perl line 9, <MPH> line 85.

Renaming one of the pbtest sections will resolve the conflict.

Pastebot is not sending any pastes to IRC

In the web form, you must use a real IRC nick of someone who is on that channel.

BUG TRACKER

https://rt.cpan.org/Dist/Display.html?Queue=Bot-Pastebot

REPOSITORY

http://github.com/~rcaputo/bot-pastebot http://gitorious.org/bot-pastebot

OTHER RESOURCES

http://search.cpan.org/dist/Bot-Pastebot/ irc://irc.perl.org/poe

AUTHORS & COPYRIGHT

Pastebot is Copyright (C) 2002-2013 by Rocco Caputo. All rights are reserved. Pastebot is free software; you may redistribute and/or modify it under the same terms as Perl itself.