sift
----

sift is a simple utility which is designed to allow you to conduct operations upon the messages located upon a remote IMAP server.



Sample Usage
------------

To get started you'll need to create a text file containing your  login details and a collection of rules.  Then you may pass that  filename as an argument to sift.

Here is one of the simplest possible rulefiles:

    server:   imap.gmail.com
    username: somebody.says
    password: testing

    folder:inbox status:new mark:read

When executed this will mark all new messages in the "inbox" as being read.



How Sift Works
--------------

Each sift rule is considered as part of a chain.  By default when you select a folder to open *all* messages are selected.

Rules in the chain have two purposes:

* Refine the list of selected messages.
* Perform an operation upon each of the currently selected messages.

For example this will select all messages:

    "folder:inbox"

This will delete all messages:

    "folder:inbox delete"

Why does that work?  Because the act of opening a folder selects all messages, and the "delete" operation deletes all currently selected  messages.

To refine thing you may reduce the messages selected by filtering on  different fields.  For example:

    "folder:inbox subject:foo delete"

This will reduce the selection of all messages to all message containing the word "foo" in their title - then delete them.



Primitives
----------

As explained there are two types of primitives, or rules, available. Those that change the selected messages on some criteria, and those that operate upon the list of selected messages.  These are known as either  "selection" or "operation" rules.

Selection Rules:

* `folder:foo`
     *  Select the specified folder.  To ease use with Googlemail all '#' characters are converted to spaces.  This allows the following rule  to work as expected:  "`folder:[Google#Mail]/Spam delete`"
* `body:foo`
     * Reduce the working set to those messages which contain '`foo`' in the body of the message.
* `!body:foo`
     * Reduce the working set to those messages which do not contain '`foo`' in the body of the message.
* `subject:foo`
     * Reduce the working set to those messages which contain '`foo`' in the subject.
* `!subject:foo`
     * Reduce the working set to those messages which do not contain '`foo`' in the subject.
* `from:bar@bar.com`
     * Reduce the working set to those messags which contain 'bar@bar.com' in the sender address.
* `!from:bar@bar.com`
     * Reduce the working set to those messags which do not contain '`bar@bar.com`' in the sender address.
* `to:bar@bar.com`
     * Reduce the working set to those messags which contain '`bar@bar.com`' in the recipient adress.
* `!to:bar@bar.com`
     * Reduce the working set to those messags which do not contain '`bar@bar.com`' in the recipient address.
* `status:[new|old]`
     * Select only messags with the given state.  "new" means unread, and "old" means messages that have been read.


Operation Rules:

* `copy:bar`
     * Copy all currently selected messages to the folder 'bar'.

* `delete`
     * Delete all currently selected messages.

* `move:bar`
     *  Move all currently selected messages into the folder '`bar`'.

* `mark:[read|unread]`
     * Mark all selected messages as either read/unread.

* `exec:/usr/local/bin/foo`
     * Run the program "`/usr/local/bin/foo`" with the body of each selected  message passed as STDIN.  This program is executed once for each selected message.

* `execonce:/usr/local/bin/foo`
     * Run the program "`/usr/local/bin/foo`".  This runs the program only once regardless of the number of selected messages.

* `dump`
     *  Provide a brief dump of the selected messages; their sender, recipient,     and subject.



Feedback?
---------

  Comments welcome to the author.


Steve
--