Back Up All Your Gmail

aa_instructions.md

• Install getmail (aptitude install getmail4)
• Set Up Your Imap Server (tl;dr)
• getmail
• ruby date_based_archive.rb ~/Maildir/.Archive

date_based_archive.rb

require 'pathname'
require 'fileutils'
require 'rubygems'
require 'mail'

class Message
  attr_reader :filename

  def initialize(filename)
    @filename = filename
    @mail = Mail.read(filename)
  end

  def archive_name
    @mail.date.strftime("%Y-%m")
  rescue
    p @mail
    # Everything with an invalid date came from this month for me, ymmv
    "2004-07"
  end
end

class MessageSource
  attr_reader :dir
  def initialize(dir = Pathname.new('.'))
    @dir = dir
    @maildirs = {}
  end

  def each_message
    %w(new cur).each do |sub|
      Dir[@dir + "#{sub}/*"].each do |fn|
        yield Message.new(fn)
      end
    end
  end

  def maildir(sub_folder_name)
    @maildirs[sub_folder_name] ||= Maildir.new(self, sub_folder_name)
  end

  def destination_for(message)
    maildir(message.archive_name)
  end
end

class Maildir
  def initialize(message_source, archivename)
    @directory = "#{message_source.dir}.#{archivename}"
    if !File.directory?(@directory)
      FileUtils.mkdir(@directory)
      %w(cur new tmp).each {|d| FileUtils.mkdir(@directory + d) }
    end
  end

  def accept_message(message)
    FileUtils.mv(message.filename, @directory + 'new')
  end
end

archive = MessageSource.new(Pathname.new(ARGV[0]))
$stdout.sync=true
archive.each_message do |msg|
   maildir = archive.destination_for(msg)
   maildir.accept_message(msg)
   print '.'
end

getmailrc

[retriever]
type = SimpleIMAPSSLRetriever
server = imap.gmail.com
mailboxes = ("[Gmail]/All Mail",)
username = EMAIL
password = PASSWORD
[destination]
type = Maildir
path = ~/Maildir/.Archive/
[options]
#verbose = 2
message_log = ~/.getmail/log
read_all = false
# do not alter messages
delivered_to = false
received = false

Alternative:

1. apt-get install offlineimap
2. adjust this dotfile to taste
3. run offlineimap whenever you feel like syncing (it syncs both ways, so you can read/reply offline and then sync)

Argh, I just dig deeper into the above – OfflineIMAP will only mirror the Gmail account (as per this tweet), not do any date-based shuffling of emails. But you could adjust your date_based_archive.rb script accordingly. :)