INSTALL

Compiling
=========

If you downloaded the sources using Mercurial, you will need to execute
./autogen.sh first to build the automake structure in your source tree. This
process requires autotools and libtool to be installed.

If you installed Dovecot from sources, Pigeonhole's configure script should be
able to find the installed dovecot-config automatically:

./configure
make
sudo make install

If your system uses a $prefix different than the default /usr/local, the
configure script can still find the installed dovecot-config automatically when
supplied with the proper --prefix argument:

./configure --prefix=/usr
make
sudo make install

If this doesn't work, you can use --with-dovecot=<path> configure option, where
the path points to a directory containing dovecot-config file. This can point to
an installed file:

./configure --with-dovecot=/usr/local/lib/dovecot
make
sudo make install

or to a Dovecot source directory that is already compiled:

./configure --with-dovecot=../dovecot-2.1.0
make
sudo make install

The following additional parameters may be of interest for the configuration of
the Pigeonhole build:

 --with-managesieve=yes
   Controls whether Pigeonhole ManageSieve is compiled and installed, which is
   the default.

 --with-unfinished-features=no
   Controls whether unfinished features and extensions are built. Enabling this
   will enable the compilation of code that is considered unfinished and highly
   experimental and may therefore introduce bugs and unexpected behavior.
   In fact, it may not compile at all. Enable this only when you are eager to
   test some of the new development functionality.

Configuration
=============

The Pigeonhole package provides the following items:

  - The Sieve interpreter plugin for Dovecot's Local Delivery Agent (LDA): This
    facilitates the actual Sieve filtering upon delivery.

  - The ManageSieve Service: This implements the ManageSieve protocol through
    which users can remotely manage Sieve scripts on the server.

The functionality of these items is described in more detail in the README file.
In this file and in this section their configuration is described. Example
configuration files are provided in the doc/example-config directory of this
package.

Sieve Interpreter - Basic Configuration
---------------------------------------

To use Sieve, you will first need to make sure you are using Dovecot LDA
or Dovecot LMTP for delivering incoming mail to users' mailboxes. Then, you need
to enable the Sieve interpreter plugin for LDA/LMTP in your dovecot.conf:

protocol lda {
..
  mail_plugins = sieve # ... other plugins like quota
}

protocol lmtp {
..
  mail_plugins = sieve # ... other plugins like quota
}

The Sieve interpreter recognizes the following configuration options in the
plugin section of the config file (default values are shown if applicable):

 sieve = ~/.dovecot.sieve
   The location of the user's main active script.

 sieve_default =
   The location of the default personal sieve script file, which gets executed
   ONLY if user's private Sieve script does no exist, e.g.
   /var/lib/dovecot/default.sieve. This is usually a global script, so be sure
   to pre-compile this script manually using the sievec command line tool, as
   explained in the README file. This setting used to be called
   `sieve_global_path', but that name is now deprecated.

 sieve_global_dir =
   Location for :global include scripts for the Sieve include extension.

 sieve_dir = ~/sieve
   Location for :personal include scripts for the Sieve include extension.

 sieve_extensions =
   Which Sieve language extensions are available to users. By default, all
   supported extensions are available, except for deprecated extensions or those
   that are still under development. Some system administrators may want to
   disable certain Sieve extensions or enable those that are not available by
   default. This setting can use '+' and '-' to specify differences relative to
   the default. For example `sieve_extensions = +imapflags' will enable the
   deprecated imapflags extension in addition to all extensions were already
   enabled by default.

 sieve_global_extensions =
   Which Sieve language extensions are ONLY avalable in global scripts. This can
   be used to restrict the use of certain Sieve extensions to administrator
   control, for instance when these extensions can cause security concerns. This
   setting has higher precedence than the `sieve_extensions' setting (above),
   meaning that the extensions enabled with this setting are never available to
   the user's personal script no matter what is specified for the
   `sieve_extensions' setting. The syntax of this setting is similar to
   the `sieve_extensions' setting, with the difference that extensions are
   enabled or disabled for exclusive use in global scripts. Currently, no
   extensions are marked as such by default.

 sieve_plugins =
   The Pigeonhole Sieve interpreter can have plugins of its own. Using this
   setting, the used plugins can be specified. Check the Dovecot wiki
   (wiki2.dovecot.org) or the pigeonhole website (http://pigeonhole.dovecot.org)
   for available plugins. The sieve_extprograms plugin is included in this
   release.

 sieve_user_log =
   The path to the file where the user log file is written. If not configured, a
   default location is used. If the main user's personal Sieve (as configured
   with sieve=) is a file, the logfile is set to <filename>.log by default. If
   it is not a file, the default user log file is ~/.dovecot.sieve.log.

 recipient_delimiter = +
   The separator that is expected between the :user and :detail address parts
   introduced by the subaddress extension. This may also be a sequence of
   characters (e.g. '--'). The current implementation looks for the separator
   from the left of the localpart and uses the first one encountered. The :user
   part is left of the separator and the :detail part is right. This setting is
   also used by Dovecot's LMTP service.

For example:

plugin {
...
  # The location of the user's active script:
  sieve = ~/.dovecot.sieve

  # If the user has no personal active script (i.e. if the file
  # indicated in sieve= does not exist), use this one:
  sieve_default = /var/lib/dovecot/sieve/default.sieve

  # The include extension fetches the :personal scripts from this
  # directory. When ManageSieve is used, this is also where scripts
  # are uploaded.
  sieve_dir = ~/sieve

  # The include extension fetches the :global scripts from this
  # directory.
  sieve_global_dir = /var/lib/dovecot/sieve/global/
}

Sieve Interpreter - Configurable Limits
---------------------------------------

 sieve_max_script_size = 1M
   The maximum size of a Sieve script. The compiler will refuse to compile any
   script larger than this limit. If set to 0, no limit on the script size is
   enforced.

 sieve_max_actions = 32
   The maximum number of actions that can be performed during a single script
   execution. If set to 0, no limit on the total number of actions is enforced.

 sieve_max_redirects = 4
   The maximum number of redirect actions that can be performed during a single
   script execution. If set to 0, no redirect actions are allowed.

Sieve Interpreter - Script Locations
------------------------------------

The location of Sieve scripts is not limited to the file system. The Sieve
interpreter can be extended to retrieve Sieve scripts from other sources as
well, such as a database. Currently, all settings that are used to obtain the
location of a single Sieve script, such as sieve=, sieve_default=, sieve_dir=
and sieve_global_dir= accept the following extended syntax:

location = [<type>:]path[;<option>[=<value>][;...]]

The following script location types are implemented by default:

  file    - The location path is a file system path pointing to the script file
            or a directory containing script files with names structured as
            `<script-name>.sieve'.
	dict    - Dovecot dict lookup. The location path is a dict uri. Read
            doc/scipt-location-dict.txt for more information and examples.

If the type prefix is omitted, the script location type is 'file'.

The following options are defined for all location types:

  name=<script-name>
    Set the name of the Sieve script that this location points to. If the name
    of the Sieve script is not contained in the location path, this option is
    required (e.g. for dict locations that must point to a particular script).
    If the name of the script is contained in the location, the value of the
    name option overrides the name retrieved from the location. If the Sieve
    interpreter explicitly queries for a specific name (e.g. to include a script
    from the sieve_dir= location), this option has no effect.

  bindir=<dirpath>
		Points to the directory where the compiled binaries for this script location
    are stored. If this option is omitted, the behavior depends on the location
    type. For `file' type locations, the binary is then stored in the same
    directory as where the script file was found if possible. For `dict' type
    locations, the binary is not stored at all in that case. Don't specify the
    same directory for different script locations, as this will result in
    undefined behavior.

Sieve Interpreter - Per-user Sieve Script Location
--------------------------------------------------

By default, the Pigeonhole LDA Sieve plugin looks for the user's Sieve script
file in the user's home directory (~/.dovecot.sieve). This requires that the
home directory is set for the user.

If you want to store the script elsewhere, you can override the default using
the sieve= setting, which specifies the location of the user's script file. This
can be done in two ways:

 1. Define the sieve setting in the plugin section of dovecot.conf.
 2. Return sieve extra field from userdb extra fields.

For example, to use a Sieve script file named <username>.sieve in
/var/sieve-scripts, use:

plugin {
...

 sieve = /var/sieve-scripts/%u.sieve
}

You may use templates like %u, as shown in the example. Refer to
http://wiki.dovecot.org/Variables for more information.

A relative path (or just a filename) will be interpreted to point under the
user's home directory.

Sieve Interpreter - Executing Multiple Scripts Sequentially
-----------------------------------------------------------

Pigeonhole's Sieve interpreter allows executing multiple Sieve scripts
sequentially. The extra scripts can be executed before and after the user's
private script. For example, this allows executing global Sieve policies before
the user's script. The following settings in the plugin section of the Dovecot
config file control the execution sequence:

 sieve_before =
 sieve_before2 =
 sieve_before3 = (etc..)
   Path to a script file or a directory containing script files that need to be
   executed before the user's personal script. If the path points to a
   directory, all the Sieve scripts contained therein (with the proper .sieve
   extension) are executed. The order of execution within that directory is
   determined by the file names, using a normal 8bit per-character comparison.

   Multiple script file or directory paths can be specified by appending an
   increasing number. The Sieve scripts found from these paths are added to the
   script execution sequence in the specified order. Reading the numbered
   sieve_before settings stops at the first missing setting, so no numbers may
   be skipped.

 sieve_after =
 sieve_after2 =
 sieve_after3 = (etc..)
   Identical to sieve_before, but the specified scripts are executed after the
   user's script (only when keep is still in effect, as explained below).

The script execution ends when the currently executing script in the sequence
does not yield a "keep" result: when the script terminates, the next script is
only executed if an implicit or explicit "keep" is in effect. Thus, to end all
script execution, a script must not execute keep and it must cancel the implicit
keep, e.g. by executing "discard; stop;". This means that the command "keep;"
has different semantics when used in a sequence of scripts. For normal Sieve
execution, "keep;" is equivalent to "fileinto "INBOX";", because both cause the
message to be stored in INBOX. However, in sequential script execution, it only
controls whether the next script is executed. Storing the message into INBOX
(the default folder) is not done until the last script in the sequence executes
(implicit) keep. To force storing the message into INBOX earlier in the
sequence, the fileinto command can be used (with ":copy" or together with
"keep;").

Apart from the keep action, all actions triggered in a script in the sequence
are executed before continuing to the next script. This means that when a script
in the sequence encounters an error, actions from earlier executed scripts are
not affected. The sequence is broken however, meaning that the script execution
of the offending script is aborted and no further scripts are executed. An
implicit keep is executed in stead if necessary, meaning that the interpreter
makes sure that the message is at least stored in the default folder (INBOX).

Just as for executing a single script the normal way, the Pigeonhole LDA Sieve
plugin takes care never to duplicate deliveries, forwards or responses. When
vacation actions are executed multiple times in different scripts, the usual
error is not triggered: the subsequent duplicate vacation actions are simply
discarded.

For example:

plugin {
...
   # Global scripts executed before the user's personal script.
   #   E.g. handling messages marked as dangerous
   sieve_before = /var/lib/dovecot/sieve/discard-virusses.sieve

   # User-specific scripts executed before the user's personal script.
   #   E.g. a vacation script managed through a non-ManageSieve GUI.
   sieve_before2 = /var/vmail/%d/%n/sieve-before

   # User-specific scripts executed after the user's personal script.
   # (if keep is still in effect)
   #   E.g. user-specific default mail filing rules
   sieve_after = /var/vmail/%d/%n/sieve-after

   # Global scripts executed after the user's personal script
   # (if keep is still in effect)
   #   E.g. default mail filing rules.
   sieve_after2 = /var/lib/dovecot/sieve/after.d/
}

IMPORTANT: The scripts specified by sieve_before and sieve_after are often
located in global locations to which the Sieve interpreter has no write access.
In that case be sure to manually pre-compile those scripts using the sievec
tool, as explained in the README file.

Sieve Interpreter - Extension Configuration
-------------------------------------------

- Editheader extension:

  The editheader extension [RFC5293] enables sieve scripts to interact with
  other components that consume or produce header fields by allowing the script
  to delete and add header fields.

  The editheader extension requires explicit configuration and is not enabled
  for use by default. Refer to doc/extensions/editheader.txt for configuration
  information.

- Vacation extension:

  The Sieve vacation extension [RFC5230] defines a mechanism to generate
  automatic replies to incoming email messages.

  The vacation extension is available by default, but it has its own specific
  configuration options. Refer to doc/extensions/vacation.txt for settings
  specific to the vacation extension.

- Include extension:

  The Sieve include extension (draft) permits users to include one Sieve script
  into another.

  The include extension is available by default, but it has its own specific
  configuration options. Refer to doc/extensions/include.txt for settings
  specific to the include extension.

- Spamtest and virustest extensions:

  Using the spamtest and virustest extensions (RFC 5235), the Sieve language
  provides a uniform and standardized command interface for evaluating spam and
  virus tests performed on the message. Users no longer need to know what headers
  need to be checked and how the scanner's verdict is represented in the header
  field value. They only need to know how to use the spamtest (spamtestplus) and
  virustest extensions. This also gives GUI-based Sieve editors the means to
  provide a portable and easy to install interface for spam and virus filter
  configuration.

  The spamtest, spamtestplus and virustest extensions require explicit
  configuration and are not enabled for use by default. Refer to
  doc/extensions/spamtest-virustest.txt for configuration information.

- Vnd.dovecot.duplicate extension:

  The vnd.dovecot.duplicate extension augments the Sieve filtering
  implementation with a test that allows detecting and handling duplicate
  message deliveries, e.g. as caused by mailinglists when people reply both to
  the mailinglist and the user directly.

  The vnd.dovecot.duplicate extension requires explicit configuration and is not
  enabled for use by default. Refer to doc/extensions/vnd.dovecot.duplicate.txt
  for configuration information.

- Extprograms plugin;
    vnd.dovovecot.pipe, vnd.dovecot.filter, vnd.dovecot.execute extensions:

  The "sieve_extprograms" plugin provides extensions to the Sieve filtering
  language adding new action commands for invoking a predefined set of external
  programs. Messages can be piped to or filtered through those programs	and
  string data can be input to and retrieved from those programs.

  This plugin and the extensions it provides require explicit configuration and
  are not enabled for use by default. Refer to doc/plugins/sieve_extprograms.txt
	for more information.

Sieve Interpreter - Migration from CMUSieve (Dovecot v1.0/v1.1)
---------------------------------------------------------------

For the most part, migration from CMUSieve to the Pigeonhole LDA Sieve plugin is
just a matter of changing the used plugin name from 'cmusieve' to 'sieve' in the
mail_plugins option in the protocol lda section of the config file (as explained
above). However, there are a few important differences:

 * The imapflags extension is now called imap4flags. The CMUSieve implementation
   is based on an old draft specification that is not completely compatible.
   Particularly, the mark and unmark commands were removed from the new
   specification. For backwards compatibility, support for the old imapflags
   extension can be enabled using the sieve_extensions setting (as explained
   above). This is disabled by default.

 * The notify extension is now called enotify. The CMUSieve implementation is
   based on an old draft specification that is not completely compatible.
   Particularly, the denotify command and $text$ substitutions were removed from
   the new specification. For backwards compatibility, support for the old
   imapflags extension can be enabled using the sieve_extensions setting (as
   explained above). This is disabled by default.

 * The include extension now requires your script file names to end with
   ".sieve". This means that ` include :personal "myscript"; ' won't work unless
   you rename "myscript" to "myscript.sieve"

Sieve Interpreter - Migration from Dovecot Sieve v0.1.x (Dovecot v1.2)
----------------------------------------------------------------------

 * Dovecot v2.0 adds support for LMTP. Much like the Dovecot LDA, it can make
   use of the Pigeonhole Sieve plugin. Since the LMTP service has its own
   prototocol lmtp section in the config file, you need to add the Sieve plugin
   to the mail_plugins setting there too when you decide to use LMTP.
 * The 'sieve_subaddress_sep' setting for the Sieve subaddress extension is now
   known as 'recipient_delimiter'. Although sieve_subaddress_sep is still
   recognized for backwards compatibility, it is recommended to update the
   setting to the new name, since the LMTP service also uses the
   recipient_delimiter setting.

ManageSieve Service - Basic Configuration
-----------------------------------------

IMPORTANT:

  If you have used the LDA Sieve plugin without ManageSieve before and you have
  .dovecot.sieve files in user directories, you are advised to make a backup
  before installing ManageSieve. Although the ManageSieve service takes care to
  move these files to the Sieve directory before it is substituted with a
  symbolic link, this is not a very well tested operation, meaning that there is
  a slim possibility that existing Sieve scripts get lost.

Just like all other binaries that Dovecot uses, the managesieve and
managesieve-login binaries are installed during make install.

There are two things that have be done to activate the ManageSieve support in
Dovecot:

 * The first step is to add `sieve' to the protocols= configuration line in
   your dovecot.conf.

 * The next step is specifying an additional service type for the ManageSieve
   service. This could be done in Dovecot's conf.d/master.conf or you can use
   the 20-managesieve.conf file from the doc/example-config/conf.d directory of
   this package.

   For example:

    service managesieve-login {
      inet_listener sieve {
        port = 4190
      }
    }

Because the implementation of the ManageSieve daemon is largely based on the
original IMAP implementation, it is very similar in terms of configuration. The
following settings can be configured in the 'protocol sieve' section:

 managesieve_max_line_length = 65536
   The maximum ManageSieve command line length in bytes. This setting is
   directly borrowed from IMAP. But, since long command lines are very unlikely
   with ManageSieve, changing this will not be very useful.

 managesieve_logout_format = bytes=%i/%o
   Specifies the string pattern used to compose the logout message of an
   authenticated session. The following substitutions are available:

     %i - total number of bytes read from client
     %o - total number of bytes sent to client

 managesieve_implementation_string = Dovecot Pigeonhole
   To fool ManageSieve clients that are focused on CMU's timesieved you can
   specify the IMPLEMENTATION capability that the Dovecot reports to clients
   (e.g. 'Cyrus timsieved v2.2.13').

 managesieve_max_compile_errors = 5
   The maximum number of compile errors that are returned to the client upon
   script upload or script verification.

 managesieve_sieve_capability =
 managesieve_notify_capability =
   Respectively the SIEVE and NOTIFY capabilities reported by the ManageSieve
   service before authentication. If left unassigned these will be assigned
   dynamically according to what the Sieve interpreter supports by default
   (after login this may differ depending on the authenticated user).

Additionally, the ManageSieve service uses the following settings from the
plugin section of the config file. These settings are the same ones used by
the LDA Sieve plugin.

 sieve_dir = ~/sieve
   This specifies the path to the directory where the uploaded scripts are
   stored. Scripts are stored as separate files with extension '.sieve'. All
   other files are ignored when scripts are listed by a ManageSieve client. The
   Sieve interpreter also uses this setting to locate the user's personal
   scripts for use with the Sieve include extension. A storage location
   specified by sieve_dir is always generated automatically if it does not exist
   (as far as the system permits the user to do so; no root privileges are
   used). This is similar to the behavior of the mail daemons regarding the
   mail_location configuration.

 sieve = ~/.dovecot.sieve
   This specifies the location of the symbolic link pointing to the active
   script in the Sieve storage directory. The Sieve interpreter uses this
   setting to locate the main script file that needs to be executed upon
   delivery. When using ManageSieve, this is a symbolic link managed by the
   ManageSieve service. ManageSieve thereby determines which script (if any) in
   the sieve_dir directory is executed for incoming messages. If a regular file
   already exists at the location specified by the sieve setting, it is moved to
   the sieve_dir location before the symbolic link is installed. It is renamed
   to dovecot.orig.sieve and therefore listed as `dovecot.orig' by a ManageSieve
   client.

The following provides an example configuration for Sieve and ManageSieve. Only
sections and settings relevant to ManageSieve are shown. Refer to
20-managesieve.conf in the doc/example-config/conf.d directory for a full
example with comments, but don't forget to configure Sieve and add sieve to the
'protocols = ...' setting if you use it.

# *** conf.d/20-managesieve.conf ***

service managesieve-login {
  inet_listener sieve {
    # Bind the daemon only to the specified address(es)
    # (default: *, ::)
    address = 127.0.0.1
    # Specify an alternative port the daemon must listen on
    # (default: 4190)
    port = 2000
  }
}

protocol sieve {
  managesieve_max_compile_errors = 10
}

# ***  conf.d/90-sieve.conf ***

plugin {
  sieve=~/.dovecot.sieve
  sieve_dir=~/sieve
}

# *** dovecot.conf ***

# .... (other config items)

# Start imap, pop3, lmtp and managesieve services
protocols = imap pop3 lmtp sieve

# .... (other config items)

ManageSieve Service - Quota Support
-----------------------------------

By default, users can manage an unlimited number of Sieve scripts on the server
through ManageSieve. However, ManageSieve can be configured to enforce limits on
the number of personal Sieve scripts per user and/or the amount of disk storage
used by these scripts. The maximum size of individual uploaded scripts is
dictated by the configuration of the Sieve plugin. The limits are configured in
the plugin section of the Dovecot configuration as follows:

 sieve_max_script_size = 1M
   The maximum size of a Sieve script. If set to 0, no limit on the script size
   is enforced.

 sieve_quota_max_scripts = 0
   The maximum number of personal Sieve scripts a single user can have. If set
   to 0, no limit on the number of scripts is enforced.

 sieve_quota_max_storage = 0
   The maximum amount of disk storage a single user's scripts may occupy. If set
   to 0, no limit on the used amount of disk storage is enforced.

ManageSieve Service - Proxying
------------------------------

Like Dovecot's imapd, the ManageSieve login daemon supports proxying to multiple
backend servers. Although the underlying code is copied from the imapd sources
for the most part, it has some ManageSieve-specifics that have not seen much
testing.

The proxy configuration wiki page for POP3 and IMAP should apply to ManageSieve
as well:

http://wiki.dovecot.org/PasswordDatabase/ExtraFields/Proxy

ManageSieve Service - Migration
-------------------------------

The following has changed since the ManageSieve releases for Dovecot v1.x:

 * For Dovecot v1.0 and v1.1, the sieve_dir setting used by ManageSieve was
   called sieve_storage. Also, the sieve and sieve_storage settings were located
   inside the 'protocol managesieve' section of the configuration. As per
   Dovecot v1.2, these settings are shared with the Sieve plugin and located in
   the 'plugin' section of the configuration. Make sure you have updated the
   name of the sieve_dir setting and the location of both these settings if you
   are upgrading from ManageSieve for Dovecot v1.0/v1.1.
 * Pigeonhole ManageSieve does not use the mail_location configuration as a
   fall-back anymore to determine a default location for storing Sieve scripts.
   It always uses the sieve_dir setting, with default value ~/sieve.
 * The Pigeonhole ManageSieve service now binds to TCP port 4190 by default due
   to the IANA port assignment for the ManageSieve service. When upgrading from
   v1.x, this should be taken into account. For a smooth transition, the service
   can be configured manually to listen on both port 2000 and port 4190, as
   demonstrated in the example above.
 * The Dovecot configuration now calls the ManageSieve protocol 'sieve' instead
   of 'managesieve' because it is registered as such with IANA. The binaries and
   the services are still called 'managesieve' and 'managesieve-login'. The
   example above demonstrates how this affects the configuration.

Test Suite
==========

This package includes a test suite to verify the basic processing of the Sieve
interpreter on your particular platform. Note that the test suite is not
available when this package is compiled against the Dovecot headers only. The
test suite executes a list of test cases and halts when one of them fails. If it
executes all test cases successfully, the test suite finishes. You can execute
the basic test suite using `make test`, which does not include the plugins. You
can execute the full test suite using `make check`.

A failing test case is always a bug and a report is greatly appreciated.