Script: (Thunberbird) mbox tree to (zimbra desktop) importable .eml tree

[fbongartz]

Our thunderbird clients have a lot of local folders that we'd like to migrate to zimbra desktop local folders. Thunderbird stores local folders in mbox files that can be structured in subdirectories. Zimbra desktop can import tgz files that contain a directory tree with .eml files where each .eml file is one message.

After playing with mb2md.pl and a bash script, I gave up and started to write my own python script which I'm sharing here:

'''
Created on 02.11.2011

@version 0.9
@author: Fabrice Bongartz (fabrice (at) fabrice d.o.t. me)
@copyright: (C) 2011 Fabrice Bongartz

This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3 of the License, or (at your option)
any later version.
This program is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
details. You should have received a copy of the GNU General Public License
along with this program; if not, see <http://www.gnu.org/licenses/>.
'''

import sys, os, os.path, time, email.utils, re, argparse, tarfile
from argparse import ArgumentTypeError

re_from = re.compile("^From ")
re_date = re.compile("^Date: +")
re_strip_from = re.compile("From ([^@]+@)*(\(.+\)|\S+) *")

def get_date_from_fromline(from_line):
    return re_strip_from.sub("", from_line)

def get_date_from_dateline(date_line):
    return re_date.sub("", date_line)

def mbox2eml(mbox_file, dest_dir, change_mtime, prefer_dateheader):
    mbox_file.seek(0) # jump back to the beginning of the file
    previous_was_newline = True
    nr = 0
    in_headers = False
    date_from = None # to store dates from the "From " header.
    date_date = None # to store dates from the "Date: " header.
    cur_path = None
    cur_file = None
    bof = True
    
    for line in mbox_file:
        if (previous_was_newline): # True for the first iteration
            if (re_from.match(line) != None):
                # we're at the beginning of a new message
                if (bof == False): # True for the first iteration
                    # finish the last message
                    if (cur_file.closed == False): cur_file.close()
                    if (change_mtime):
                        if (prefer_dateheader and date_date != None):
                            mtime = time.mktime(date_date)
                        else: mtime = time.mktime(date_from)
                        os.utime(cur_path, (mtime, mtime))
                
                # prepare for a new message
                cur_path = os.path.join(dest_dir, str(nr) + ".eml")
                cur_file = open(cur_path, "w")
                date_from = email.utils.parsedate(get_date_from_fromline(line))
                date_date = None
                in_headers = True
                bof = False
                nr += 1
            elif (in_headers): in_headers = False
        if (prefer_dateheader and in_headers and re_date.match(line) != None):
            date_date = email.utils.parsedate(get_date_from_dateline(line))
        
        # write the current line
        cur_file.write(line)
        
        # determine if we're at a newline
        if line.replace("\r\n", "\n") == "\n": previous_was_newline = True
        else: previous_was_newline = False
    
    # treat the last remaining message
    if (bof == False): # should only be True here if the file was empty
        if (cur_file.closed == False): cur_file.close()
        if (change_mtime):
            if (prefer_dateheader and date_date != None):
                mtime = time.mktime(date_date)
            else: mtime = time.mktime(date_from)
            os.utime(cur_path, (mtime, mtime))
    
    return True

def is_mbox_file(f):
    """
    Determine of the given File object is an mbox file. This simply checks
    if the file's first line has the string "From " at its start.
    """
    if (re_from.match(f.readline()) != None):
        return True
    return False

def recurse_mbox(mbox_start_dir, tmp_dir, change_mtime = True,
                 prefer_dateheader = True):
    
    for root, dirs, files in os.walk(mbox_start_dir, True, None, False):
        for f_str in files:
            if ((len(f_str) > 3 and f_str[-4:] != ".msf") or len(f_str) <= 3):
                f = open(os.path.join(root, f_str), "r")
                if (is_mbox_file(f)):
                    print "Treating mbox " + os.path.join(root, f_str)
                    rel_path = os.path.relpath(root, mbox_start_dir)
                    if (rel_path == "."): rel_path = ""
                    dest_dir = os.path.join(tmp_dir, rel_path, f_str).replace(".sbd", "")
                    if (os.path.isdir(dest_dir) == False): os.makedirs(dest_dir)
                    mbox2eml(f, dest_dir, change_mtime, prefer_dateheader)
                f.close()

def create_tgz(dir, tgz_path):
    tgz = tarfile.open(tgz_path, "w:gz")
    for root, dirnames, filenames in os.walk(dir):
        for f in filenames:
            filepath = os.path.join(root, f)
            relpath = os.path.relpath(filepath, dir)
            print "Adding to targz:", filepath
            tgz.add(filepath, relpath)
    tgz.close()

def initialize_options():
        ap = argparse.ArgumentParser()
        ap.add_argument("-s", "--mbox-start-dir", required = True,
                       dest = "mbox_start_dir", help = "A source directory "
                        + "that contains a tree of mbox files.")
        ap.add_argument("-d", "--destination", required = True,
                        dest = "dest_dir", help = "Directory where the "
                        + "eml-file tree should be created.")
        ap.add_argument("-z", "--tgz", dest = "tgz", help = "Create a gzipped "
                        + "tar archive that contains the directory tree "
                        + "specified with -d/--destination at the given "
                        + "path. This is optional. Note that this might be "
                        + "faster using an optimized commandline tool like "
                        + "gnu tar.")
        ap.add_argument("-M", "--dont-change-mtimes", dest = "dont_change_mtimes",
                        action = "store_true", default = False, help = "By "
                        + "default, the mtime and atime of the created eml "
                        + "files will be changed to a date found in each email "
                        + "header. This option disabled changing mtime/atime.")
        ap.add_argument("-I", "--ignore-date-header", default = False,
                        action = "store_true", dest = "ignore_date_header",
                        help = "By default, and if -M/--dont-change-mtimes "
                        + "wasn't specified, in order to change the "
                        + "atime+mtime of each eml file, the program will look "
                        + "for a Date: line in the email headers. If no Date: "
                        + "line was found, the date from the \"From \" line at "
                        + "the beginning of the message will be used. This "
                        + "option disables looking for the Date: line so that "
                        + "the \"From \" line will always be used.")
        return ap.parse_args()

if __name__ == '__main__':
    opts = initialize_options()
    
    # check args
    if (not os.path.isdir(opts.mbox_start_dir)):
        raise ArgumentTypeError("The given mbox start dir is not a directory")
    if (not os.path.isdir(opts.dest_dir)):
        raise ArgumentTypeError("The given destination dir is not a directory")
    
    # walk the mboxes and create the destination eml file structure
    recurse_mbox(os.path.abspath(opts.mbox_start_dir),
                 os.path.abspath(opts.dest_dir),
                 (opts.dont_change_mtimes == False),
                 (opts.ignore_date_header == False))
    
    # optionally create a gzipped tar archive
    if (opts.tgz): create_tgz(opts.dest_dir, opts.tgz)

Save the script above somewhere as mbox2zeml.py. For example: /home/user/migrate/mbox2zeml.py

To get commandline help, exectute the script with the -h option like this:

python mbox2zeml.py -h

So, in order to migrate thunderbird's local folders to a zimbra desktop installation, follow these steps:

1. Make sure you read the script's commandline help. Ideally also read the source code so you understand what the script does.
2. Make sure you have python2 installed. There should be no other dependencies. (Tested with python2.7 on arch linux).
3. Copy the local folders from thunderbird to the machine where you'll be executing the script. If you're using the script on the pc that holds the thunderbird profile, you may skip this step.
4. Create a destination directory. For example: /home/user/migrate/dest. Make sure, the user that will execute the script has write privileges in that directory.
5. Launch the script. Example:
   

   python /home/user/migrate/mbox2zeml.py -s THUNDERBIRD_DIR -d /home/user/migrate/dest /home/user/migrate/importme.tgz

6. The directory tree with eml files will be created in the dest directory. After that, dest's contents will be tar/gzipped to importme.tgz
7. Open Zimbra Desktop, go to Preferences -> Local Folders -> Import / Export and import the tgz file.

Any improvements to the script are welcome. For example, I didn't implement any error handling or logging. There could be more comments and so on.

EDIT: I noticed that Zimbra Desktop doesn't seem to like tgz files over 2 GB in size on windows (can someone confirm this?), so I had to split the local folders for some big accounts.

I hope this can be useful for someone.

[olivié]

hello fbongartz,

THANKS A LOT for this script !!!!
In the past, I have change from outlook to thunderbird easily. But today, I have to go back on outlook. Since a couple of days I search and test a lot of tools that do not work.
I use your script to convert my mbox to eml keeping all the structure (folders/subfolders).
I appreciate your work, and I'm sure to not have a virus like if I test .exe files...
Now that I have the eml structure, I just have to do the rest by importing eml to outlook.
Thanks again.

[fbongartz]

hello fbongartz,
THANKS A LOT for this script !!!!
.

Welcome. I'm glad you could use it for something :-)

[ggrussenmeyer]

Thanks for this efficient script!

For future users, it is important to note that the compressed file's prefix must be "tgz" for Zimbra Desktop to accept it as a valid import file.