/lib-python/2.7/mailbox.py

Large files files are truncated, but you can click here to view the full file

#! /usr/bin/env python

"""Read/write support for Maildir, mbox, MH, Babyl, and MMDF mailboxes."""

# Notes for authors of new mailbox subclasses:
#
# Remember to fsync() changes to disk before closing a modified file
# or returning from a flush() method.  See functions _sync_flush() and
# _sync_close().

import sys
import os
import time
import calendar
import socket
import errno
import copy
import email
import email.message
import email.generator
import StringIO
try:
    if sys.platform == 'os2emx':
        # OS/2 EMX fcntl() not adequate
        raise ImportError
    import fcntl
except ImportError:
    fcntl = None

import warnings
with warnings.catch_warnings():
    if sys.py3kwarning:
        warnings.filterwarnings("ignore", ".*rfc822 has been removed",
                                DeprecationWarning)
    import rfc822

__all__ = [ 'Mailbox', 'Maildir', 'mbox', 'MH', 'Babyl', 'MMDF',
            'Message', 'MaildirMessage', 'mboxMessage', 'MHMessage',
            'BabylMessage', 'MMDFMessage', 'UnixMailbox',
            'PortableUnixMailbox', 'MmdfMailbox', 'MHMailbox', 'BabylMailbox' ]

class Mailbox:
    """A group of messages in a particular place."""

    def __init__(self, path, factory=None, create=True):
        """Initialize a Mailbox instance."""
        self._path = os.path.abspath(os.path.expanduser(path))
        self._factory = factory

    def add(self, message):
        """Add message and return assigned key."""
        raise NotImplementedError('Method must be implemented by subclass')

    def remove(self, key):
        """Remove the keyed message; raise KeyError if it doesn't exist."""
        raise NotImplementedError('Method must be implemented by subclass')

    def __delitem__(self, key):
        self.remove(key)

    def discard(self, key):
        """If the keyed message exists, remove it."""
        try:
            self.remove(key)
        except KeyError:
            pass

    def __setitem__(self, key, message):
        """Replace the keyed message; raise KeyError if it doesn't exist."""
        raise NotImplementedError('Method must be implemented by subclass')

    def get(self, key, default=None):
        """Return the keyed message, or default if it doesn't exist."""
        try:
            return self.__getitem__(key)
        except KeyError:
            return default

    def __getitem__(self, key):
        """Return the keyed message; raise KeyError if it doesn't exist."""
        if not self._factory:
            return self.get_message(key)
        else:
            return self._factory(self.get_file(key))

    def get_message(self, key):
        """Return a Message representation or raise a KeyError."""
        raise NotImplementedError('Method must be implemented by subclass')

    def get_string(self, key):
        """Return a string representation or raise a KeyError."""
        raise NotImplementedError('Method must be implemented by subclass')

    def get_file(self, key):
        """Return a file-like representation or raise a KeyError."""
        raise NotImplementedError('Method must be implemented by subclass')

    def iterkeys(self):
        """Return an iterator over keys."""
        raise NotImplementedError('Method must be implemented by subclass')

    def keys(self):
        """Return a list of keys."""
        return list(self.iterkeys())

    def itervalues(self):
        """Return an iterator over all messages."""
        for key in self.iterkeys():
            try:
                value = self[key]
            except KeyError:
                continue
            yield value

    def __iter__(self):
        return self.itervalues()

    def values(self):
        """Return a list of messages. Memory intensive."""
        return list(self.itervalues())

    def iteritems(self):
        """Return an iterator over (key, message) tuples."""
        for key in self.iterkeys():
            try:
                value = self[key]
            except KeyError:
                continue
            yield (key, value)

    def items(self):
        """Return a list of (key, message) tuples. Memory intensive."""
        return list(self.iteritems())

    def has_key(self, key):
        """Return True if the keyed message exists, False otherwise."""
        raise NotImplementedError('Method must be implemented by subclass')

    def __contains__(self, key):
        return self.has_key(key)

    def __len__(self):
        """Return a count of messages in the mailbox."""
        raise NotImplementedError('Method must be implemented by subclass')

    def clear(self):
        """Delete all messages."""
        for key in self.iterkeys():
            self.discard(key)

    def pop(self, key, default=None):
        """Delete the keyed message and return it, or default."""
        try:
            result = self[key]
        except KeyError:
            return default
        self.discard(key)
        return result

    def popitem(self):
        """Delete an arbitrary (key, message) pair and return it."""
        for key in self.iterkeys():
            return (key, self.pop(key))     # This is only run once.
        else:
            raise KeyError('No messages in mailbox')

    def update(self, arg=None):
        """Change the messages that correspond to certain keys."""
        if hasattr(arg, 'iteritems'):
            source = arg.iteritems()
        elif hasattr(arg, 'items'):
            source = arg.items()
        else:
            source = arg
        bad_key = False
        for key, message in source:
            try:
                self[key] = message
            except KeyError:
                bad_key = True
        if bad_key:
            raise KeyError('No message with key(s)')

    def flush(self):
        """Write any pending changes to the disk."""
        raise NotImplementedError('Method must be implemented by subclass')

    def lock(self):
        """Lock the mailbox."""
        raise NotImplementedError('Method must be implemented by subclass')

    def unlock(self):
        """Unlock the mailbox if it is locked."""
        raise NotImplementedError('Method must be implemented by subclass')

    def close(self):
        """Flush and close the mailbox."""
        raise NotImplementedError('Method must be implemented by subclass')

    def _dump_message(self, message, target, mangle_from_=False):
        # Most files are opened in binary mode to allow predictable seeking.
        # To get native line endings on disk, the user-friendly \n line endings
        # used in strings and by email.Message are translated here.
        """Dump message contents to target file."""
        if isinstance(message, email.message.Message):
            buffer = StringIO.StringIO()
            gen = email.generator.Generator(buffer, mangle_from_, 0)
            gen.flatten(message)
            buffer.seek(0)
            target.write(buffer.read().replace('\n', os.linesep))
        elif isinstance(message, str):
            if mangle_from_:
                message = message.replace('\nFrom ', '\n>From ')
            message = message.replace('\n', os.linesep)
            target.write(message)
        elif hasattr(message, 'read'):
            while True:
                line = message.readline()
                if line == '':
                    break
                if mangle_from_ and line.startswith('From '):
                    line = '>From ' + line[5:]
                line = line.replace('\n', os.linesep)
                target.write(line)
        else:
            raise TypeError('Invalid message type: %s' % type(message))


class Maildir(Mailbox):
    """A qmail-style Maildir mailbox."""

    colon = ':'

    def __init__(self, dirname, factory=rfc822.Message, create=True):
        """Initialize a Maildir instance."""
        Mailbox.__init__(self, dirname, factory, create)
        self._paths = {
            'tmp': os.path.join(self._path, 'tmp'),
            'new': os.path.join(self._path, 'new'),
            'cur': os.path.join(self._path, 'cur'),
            }
        if not os.path.exists(self._path):
            if create:
                os.mkdir(self._path, 0700)
                for path in self._paths.values():
                    os.mkdir(path, 0o700)
            else:
                raise NoSuchMailboxError(self._path)
        self._toc = {}
        self._toc_mtimes = {'cur': 0, 'new': 0}
        self._last_read = 0         # Records last time we read cur/new
        self._skewfactor = 0.1      # Adjust if os/fs clocks are skewing

    def add(self, message):
        """Add message and return assigned key."""
        tmp_file = self._create_tmp()
        try:
            self._dump_message(message, tmp_file)
        except BaseException:
            tmp_file.close()
            os.remove(tmp_file.name)
            raise
        _sync_close(tmp_file)
        if isinstance(message, MaildirMessage):
            subdir = message.get_subdir()
            suffix = self.colon + message.get_info()
            if suffix == self.colon:
                suffix = ''
        else:
            subdir = 'new'
            suffix = ''
        uniq = os.path.basename(tmp_file.name).split(self.colon)[0]
        dest = os.path.join(self._path, subdir, uniq + suffix)
        try:
            if hasattr(os, 'link'):
                os.link(tmp_file.name, dest)
                os.remove(tmp_file.name)
            else:
                os.rename(tmp_file.name, dest)
        except OSError, e:
            os.remove(tmp_file.name)
            if e.errno == errno.EEXIST:
                raise ExternalClashError('Name clash with existing message: %s'
                                         % dest)
            else:
                raise
        if isinstance(message, MaildirMessage):
            os.utime(dest, (os.path.getatime(dest), message.get_date()))
        return uniq

    def remove(self, key):
        """Remove the keyed message; raise KeyError if it doesn't exist."""
        os.remove(os.path.join(self._path, self._lookup(key)))

    def discard(self, key):
        """If the keyed message exists, remove it."""
        # This overrides an inapplicable implementation in the superclass.
        try:
            self.remove(key)
        except KeyError:
            pass
        except OSError, e:
            if e.errno != errno.ENOENT:
                raise

    def __setitem__(self, key, message):
        """Replace the keyed message; raise KeyError if it doesn't exist."""
        old_subpath = self._lookup(key)
        temp_key = self.add(message)
        temp_subpath = self._lookup(temp_key)
        if isinstance(message, MaildirMessage):
            # temp's subdir and suffix were specified by message.
            dominant_subpath = temp_subpath
        else:
            # temp's subdir and suffix were defaults from add().
            dominant_subpath = old_subpath
        subdir = os.path.dirname(dominant_subpath)
        if self.colon in dominant_subpath:
            suffix = self.colon + dominant_subpath.split(self.colon)[-1]
        else:
            suffix = ''
        self.discard(key)
        new_path = os.path.join(self._path, subdir, key + suffix)
        os.rename(os.path.join(self._path, temp_subpath), new_path)
        if isinstance(message, MaildirMessage):
            os.utime(new_path, (os.path.getatime(new_path),
                                message.get_date()))

    def get_message(self, key):
        """Return a Message representation or raise a KeyError."""
        subpath = self._lookup(key)
        f = open(os.path.join(self._path, subpath), 'r')
        try:
            if self._factory:
                msg = self._factory(f)
            else:
                msg = MaildirMessage(f)
        finally:
            f.close()
        subdir, name = os.path.split(subpath)
        msg.set_subdir(subdir)
        if self.colon in name:
            msg.set_info(name.split(self.colon)[-1])
        msg.set_date(os.path.getmtime(os.path.join(self._path, subpath)))
        return msg

    def get_string(self, key):
        """Return a string representation or raise a KeyError."""
        f = open(os.path.join(self._path, self._lookup(key)), 'r')
        try:
            return f.read()
        finally:
            f.close()

    def get_file(self, key):
        """Return a file-like representation or raise a KeyError."""
        f = open(os.path.join(self._path, self._lookup(key)), 'rb')
        return _ProxyFile(f)

    def iterkeys(self):
        """Return an iterator over keys."""
        self._refresh()
        for key in self._toc:
            try:
                self._lookup(key)
            except KeyError:
                continue
            yield key

    def has_key(self, key):
        """Return True if the keyed message exists, False otherwise."""
        self._refresh()
        return key in self._toc

    def __len__(self):
        """Return a count of messages in the mailbox."""
        self._refresh()
        return len(self._toc)

    def flush(self):
        """Write any pending changes to disk."""
        # Maildir changes are always written immediately, so there's nothing
        # to do.
        pass

    def lock(self):
        """Lock the mailbox."""
        return

    def unlock(self):
        """Unlock the mailbox if it is locked."""
        return

    def close(self):
        """Flush and close the mailbox."""
        return

    def list_folders(self):
        """Return a list of folder names."""
        result = []
        for entry in os.listdir(self._path):
            if len(entry) > 1 and entry[0] == '.' and \
               os.path.isdir(os.path.join(self._path, entry)):
                result.append(entry[1:])
        return result

    def get_folder(self, folder):
        """Return a Maildir instance for the named folder."""
        return Maildir(os.path.join(self._path, '.' + folder),
                       factory=self._factory,
                       create=False)

    def add_folder(self, folder):
        """Create a folder and return a Maildir instance representing it."""
        path = os.path.join(self._path, '.' + folder)
        result = Maildir(path, factory=self._factory)
        maildirfolder_path = os.path.join(path, 'maildirfolder')
        if not os.path.exists(maildirfolder_path):
            os.close(os.open(maildirfolder_path, os.O_CREAT | os.O_WRONLY,
                0666))
        return result

    def remove_folder(self, folder):
        """Delete the named folder, which must be empty."""
        path = os.path.join(self._path, '.' + folder)
        for entry in os.listdir(os.path.join(path, 'new')) + \
                     os.listdir(os.path.join(path, 'cur')):
            if len(entry) < 1 or entry[0] != '.':
                raise NotEmptyError('Folder contains message(s): %s' % folder)
        for entry in os.listdir(path):
            if entry != 'new' and entry != 'cur' and entry != 'tmp' and \
               os.path.isdir(os.path.join(path, entry)):
                raise NotEmptyError("Folder contains subdirectory '%s': %s" %
                                    (folder, entry))
        for root, dirs, files in os.walk(path, topdown=False):
            for entry in files:
                os.remove(os.path.join(root, entry))
            for entry in dirs:
                os.rmdir(os.path.join(root, entry))
        os.rmdir(path)

    def clean(self):
        """Delete old files in "tmp"."""
        now = time.time()
        for entry in os.listdir(os.path.join(self._path, 'tmp')):
            path = os.path.join(self._path, 'tmp', entry)
            if now - os.path.getatime(path) > 129600:   # 60 * 60 * 36
                os.remove(path)

    _count = 1  # This is used to generate unique file names.

    def _create_tmp(self):
        """Create a file in the tmp subdirectory and open and return it."""
        now = time.time()
        hostname = socket.gethostname()
        if '/' in hostname:
            hostname = hostname.replace('/', r'\057')
        if ':' in hostname:
            hostname = hostname.replace(':', r'\072')
        uniq = "%s.M%sP%sQ%s.%s" % (int(now), int(now % 1 * 1e6), os.getpid(),
                                    Maildir._count, hostname)
        path = os.path.join(self._path, 'tmp', uniq)
        try:
            os.stat(path)
        except OSError, e:
            if e.errno == errno.ENOENT:
                Maildir._count += 1
                try:
                    return _create_carefully(path)
                except OSError, e:
                    if e.errno != errno.EEXIST:
                        raise
            else:
                raise

        # Fall through to here if stat succeeded or open raised EEXIST.
        raise ExternalClashError('Name clash prevented file creation: %s' %
                                 path)

    def _refresh(self):
        """Update table of contents mapping."""
        # If it has been less than two seconds since the last _refresh() call,
        # we have to unconditionally re-read the mailbox just in case it has
        # been modified, because os.path.mtime() has a 2 sec resolution in the
        # most common worst case (FAT) and a 1 sec resolution typically.  This
        # results in a few unnecessary re-reads when _refresh() is called
        # multiple times in that interval, but once the clock ticks over, we
        # will only re-read as needed.  Because the filesystem might be being
        # served by an independent system with its own clock, we record and
        # compare with the mtimes from the filesystem.  Because the other
        # system's clock might be skewing relative to our clock, we add an
        # extra delta to our wait.  The default is one tenth second, but is an
        # instance variable and so can be adjusted if dealing with a
        # particularly skewed or irregular system.
        if time.time() - self._last_read > 2 + self._skewfactor:
            refresh = False
            for subdir in self._toc_mtimes:
                mtime = os.path.getmtime(self._paths[subdir])
                if mtime > self._toc_mtimes[subdir]:
                    refresh = True
                self._toc_mtimes[subdir] = mtime
            if not refresh:
                return
        # Refresh toc
        self._toc = {}
        for subdir in self._toc_mtimes:
            path = self._paths[subdir]
            for entry in os.listdir(path):
                p = os.path.join(path, entry)
                if os.path.isdir(p):
                    continue
                uniq = entry.split(self.colon)[0]
                self._toc[uniq] = os.path.join(subdir, entry)
        self._last_read = time.time()

    def _lookup(self, key):
        """Use TOC to return subpath for given key, or raise a KeyError."""
        try:
            if os.path.exists(os.path.join(self._path, self._toc[key])):
                return self._toc[key]
        except KeyError:
            pass
        self._refresh()
        try:
            return self._toc[key]
        except KeyError:
            raise KeyError('No message with key: %s' % key)

    # This method is for backward compatibility only.
    def next(self):
        """Return the next message in a one-time iteration."""
        if not hasattr(self, '_onetime_keys'):
            self._onetime_keys = self.iterkeys()
        while True:
            try:
                return self[self._onetime_keys.next()]
            except StopIteration:
                return None
            except KeyError:
                continue


class _singlefileMailbox(Mailbox):
    """A single-file mailbox."""

    def __init__(self, path, factory=None, create=True):
        """Initialize a single-file mailbox."""
        Mailbox.__init__(self, path, factory, create)
        try:
            f = open(self._path, 'rb+')
        except IOError, e:
            if e.errno == errno.ENOENT:
                if create:
                    f = open(self._path, 'wb+')
                else:
                    raise NoSuchMailboxError(self._path)
            elif e.errno in (errno.EACCES, errno.EROFS):
                f = open(self._path, 'rb')
            else:
                raise
        self._file = f
        self._toc = None
        self._next_key = 0
        self._pending = False   # No changes require rewriting the file.
        self._locked = False
        self._file_length = None        # Used to record mailbox size

    def add(self, message):
        """Add message and return assigned key."""
        self._lookup()
        self._toc[self._next_key] = self._append_message(message)
        self._next_key += 1
        self._pending = True
        return self._next_key - 1

    def remove(self, key):
        """Remove the keyed message; raise KeyError if it doesn't exist."""
        self._lookup(key)
        del self._toc[key]
        self._pending = True

    def __setitem__(self, key, message):
        """Replace the keyed message; raise KeyError if it doesn't exist."""
        self._lookup(key)
        self._toc[key] = self._append_message(message)
        self._pending = True

    def iterkeys(self):
        """Return an iterator over keys."""
        self._lookup()
        for key in self._toc.keys():
            yield key

    def has_key(self, key):
        """Return True if the keyed message exists, False otherwise."""
        self._lookup()
        return key in self._toc

    def __len__(self):
        """Return a count of messages in the mailbox."""
        self._lookup()
        return len(self._toc)

    def lock(self):
        """Lock the mailbox."""
        if not self._locked:
            _lock_file(self._file)
            self._locked = True

    def unlock(self):
        """Unlock the mailbox if it is locked."""
        if self._locked:
            _unlock_file(self._file)
            self._locked = False

    def flush(self):
        """Write any pending changes to disk."""
        if not self._pending:
            return

        # In order to be writing anything out at all, self._toc must
        # already have been generated (and presumably has been modified
        # by adding or deleting an item).
        assert self._toc is not None

        # Check length of self._file; if it's changed, some other process
        # has modified the mailbox since we scanned it.
        self._file.seek(0, 2)
        cur_len = self._file.tell()
        if cur_len != self._file_length:
            raise ExternalClashError('Size of mailbox file changed '
                                     '(expected %i, found %i)' %
                                     (self._file_length, cur_len))

        new_file = _create_temporary(self._path)
        try:
            new_toc = {}
            self._pre_mailbox_hook(new_file)
            for key in sorted(self._toc.keys()):
                start, stop = self._toc[key]
                self._file.seek(start)
                self._pre_message_hook(new_file)
                new_start = new_file.tell()
                while True:
                    buffer = self._file.read(min(4096,
                                                 stop - self._file.tell()))
                    if buffer == '':
                        break
                    new_file.write(buffer)
                new_toc[key] = (new_start, new_file.tell())
                self._post_message_hook(new_file)
        except:
            new_file.close()
            os.remove(new_file.name)
            raise
        _sync_close(new_file)
        # self._file is about to get replaced, so no need to sync.
        self._file.close()
        try:
            os.rename(new_file.name, self._path)
        except OSError, e:
            if e.errno == errno.EEXIST or \
              (os.name == 'os2' and e.errno == errno.EACCES):
                os.remove(self._path)
                os.rename(new_file.name, self._path)
            else:
                raise
        self._file = open(self._path, 'rb+')
        self._toc = new_toc
        self._pending = False
        if self._locked:
            _lock_file(self._file, dotlock=False)

    def _pre_mailbox_hook(self, f):
        """Called before writing the mailbox to file f."""
        return

    def _pre_message_hook(self, f):
        """Called before writing each message to file f."""
        return

    def _post_message_hook(self, f):
        """Called after writing each message to file f."""
        return

    def close(self):
        """Flush and close the mailbox."""
        self.flush()
        if self._locked:
            self.unlock()
        self._file.close()  # Sync has been done by self.flush() above.

    def _lookup(self, key=None):
        """Return (start, stop) or raise KeyError."""
        if self._toc is None:
            self._generate_toc()
        if key is not None:
            try:
                return self._toc[key]
            except KeyError:
                raise KeyError('No message with key: %s' % key)

    def _append_message(self, message):
        """Append message to mailbox and return (start, stop) offsets."""
        self._file.seek(0, 2)
        before = self._file.tell()
        try:
            self._pre_message_hook(self._file)
            offsets = self._install_message(message)
            self._post_message_hook(self._file)
        except BaseException:
            self._file.truncate(before)
            raise
        self._file.flush()
        self._file_length = self._file.tell()  # Record current length of mailbox
        return offsets



class _mboxMMDF(_singlefileMailbox):
    """An mbox or MMDF mailbox."""

    _mangle_from_ = True

    def get_message(self, key):
        """Return a Message representation or raise a KeyError."""
        start, stop = self._lookup(key)
        self._file.seek(start)
        from_line = self._file.readline().replace(os.linesep, '')
        string = self._file.read(stop - self._file.tell())
        msg = self._message_factory(string.replace(os.linesep, '\n'))
        msg.set_from(from_line[5:])
        return msg

    def get_string(self, key, from_=False):
        """Return a string representation or raise a KeyError."""
        start, stop = self._lookup(key)
        self._file.seek(start)
        if not from_:
            self._file.readline()
        string = self._file.read(stop - self._file.tell())
        return string.replace(os.linesep, '\n')

    def get_file(self, key, from_=False):
        """Return a file-like representation or raise a KeyError."""
        start, stop = self._lookup(key)
        self._file.seek(start)
        if not from_:
            self._file.readline()
        return _PartialFile(self._file, self._file.tell(), stop)

    def _install_message(self, message):
        """Format a message and blindly write to self._file."""
        from_line = None
        if isinstance(message, str) and message.startswith('From '):
            newline = message.find('\n')
            if newline != -1:
                from_line = message[:newline]
                message = message[newline + 1:]
            else:
                from_line = message
                message = ''
        elif isinstance(message, _mboxMMDFMessage):
            from_line = 'From ' + message.get_from()
        elif isinstance(message, email.message.Message):
            from_line = message.get_unixfrom()  # May be None.
        if from_line is None:
            from_line = 'From MAILER-DAEMON %s' % time.asctime(time.gmtime())
        start = self._file.tell()
        self._file.write(from_line + os.linesep)
        self._dump_message(message, self._file, self._mangle_from_)
        stop = self._file.tell()
        return (start, stop)


class mbox(_mboxMMDF):
    """A classic mbox mailbox."""

    _mangle_from_ = True

    def __init__(self, path, factory=None, create=True):
        """Initialize an mbox mailbox."""
        self._message_factory = mboxMessage
        _mboxMMDF.__init__(self, path, factory, create)

    def _pre_message_hook(self, f):
        """Called before writing each message to file f."""
        if f.tell() != 0:
            f.write(os.linesep)

    def _generate_toc(self):
        """Generate key-to-(start, stop) table of contents."""
        starts, stops = [], []
        self._file.seek(0)
        while True:
            line_pos = self._file.tell()
            line = self._file.readline()
            if line.startswith('From '):
                if len(stops) < len(starts):
                    stops.append(line_pos - len(os.linesep))
                starts.append(line_pos)
            elif line == '':
                stops.append(line_pos)
                break
        self._toc = dict(enumerate(zip(starts, stops)))
        self._next_key = len(self._toc)
        self._file_length = self._file.tell()


class MMDF(_mboxMMDF):
    """An MMDF mailbox."""

    def __init__(self, path, factory=None, create=True):
        """Initialize an MMDF mailbox."""
        self._message_factory = MMDFMessage
        _mboxMMDF.__init__(self, path, factory, create)

    def _pre_message_hook(self, f):
        """Called before writing each message to file f."""
        f.write('\001\001\001\001' + os.linesep)

    def _post_message_hook(self, f):
        """Called after writing each message to file f."""
        f.write(os.linesep + '\001\001\001\001' + os.linesep)

    def _generate_toc(self):
        """Generate key-to-(start, stop) table of contents."""
        starts, stops = [], []
        self._file.seek(0)
        next_pos = 0
        while True:
            line_pos = next_pos
            line = self._file.readline()
            next_pos = self._file.tell()
            if line.startswith('\001\001\001\001' + os.linesep):
                starts.append(next_pos)
                while True:
                    line_pos = next_pos
                    line = self._file.readline()
                    next_pos = self._file.tell()
                    if line == '\001\001\001\001' + os.linesep:
                        stops.append(line_pos - len(os.linesep))
                        break
                    elif line == '':
                        stops.append(line_pos)
                        break
            elif line == '':
                break
        self._toc = dict(enumerate(zip(starts, stops)))
        self._next_key = len(self._toc)
        self._file.seek(0, 2)
        self._file_length = self._file.tell()


class MH(Mailbox):
    """An MH mailbox."""

    def __init__(self, path, factory=None, create=True):
        """Initialize an MH instance."""
        Mailbox.__init__(self, path, factory, create)
        if not os.path.exists(self._path):
            if create:
                os.mkdir(self._path, 0700)
                os.close(os.open(os.path.join(self._path, '.mh_sequences'),
                                 os.O_CREAT | os.O_EXCL | os.O_WRONLY, 0600))
            else:
                raise NoSuchMailboxError(self._path)
        self._locked = False

    def add(self, message):
        """Add message and return assigned key."""
        keys = self.keys()
        if len(keys) == 0:
            new_key = 1
        else:
            new_key = max(keys) + 1
        new_path = os.path.join(self._path, str(new_key))
        f = _create_carefully(new_path)
        closed = False
        try:
            if self._locked:
                _lock_file(f)
            try:
                try:
                    self._dump_message(message, f)
                except BaseException:
                    # Unlock and close so it can be deleted on Windows
                    if self._locked:
                        _unlock_file(f)
                    _sync_close(f)
                    closed = True
                    os.remove(new_path)
                    raise
                if isinstance(message, MHMessage):
                    self._dump_sequences(message, new_key)
            finally:
                if self._locked:
                    _unlock_file(f)
        finally:
            if not closed:
                _sync_close(f)
        return new_key

    def remove(self, key):
        """Remove the keyed message; raise KeyError if it doesn't exist."""
        path = os.path.join(self._path, str(key))
        try:
            f = open(path, 'rb+')
        except IOError, e:
            if e.errno == errno.ENOENT:
                raise KeyError('No message with key: %s' % key)
            else:
                raise
        else:
            f.close()
            os.remove(path)

    def __setitem__(self, key, message):
        """Replace the keyed message; raise KeyError if it doesn't exist."""
        path = os.path.join(self._path, str(key))
        try:
            f = open(path, 'rb+')
        except IOError, e:
            if e.errno == errno.ENOENT:
                raise KeyError('No message with key: %s' % key)
            else:
                raise
        try:
            if self._locked:
                _lock_file(f)
            try:
                os.close(os.open(path, os.O_WRONLY | os.O_TRUNC))
                self._dump_message(message, f)
                if isinstance(message, MHMessage):
                    self._dump_sequences(message, key)
            finally:
                if self._locked:
                    _unlock_file(f)
        finally:
            _sync_close(f)

    def get_message(self, key):
        """Return a Message representation or raise a KeyError."""
        try:
            if self._locked:
                f = open(os.path.join(self._path, str(key)), 'r+')
            else:
                f = open(os.path.join(self._path, str(key)), 'r')
        except IOError, e:
            if e.errno == errno.ENOENT:
                raise KeyError('No message with key: %s' % key)
            else:
                raise
        try:
            if self._locked:
                _lock_file(f)
            try:
                msg = MHMessage(f)
            finally:
                if self._locked:
                    _unlock_file(f)
        finally:
            f.close()
        for name, key_list in self.get_sequences().iteritems():
            if key in key_list:
                msg.add_sequence(name)
        return msg

    def get_string(self, key):
        """Return a string representation or raise a KeyError."""
        try:
            if self._locked:
                f = open(os.path.join(self._path, str(key)), 'r+')
            else:
                f = open(os.path.join(self._path, str(key)), 'r')
        except IOError, e:
            if e.errno == errno.ENOENT:
                raise KeyError('No message with key: %s' % key)
            else:
                raise
        try:
            if self._locked:
                _lock_file(f)
            try:
                return f.read()
            finally:
                if self._locked:
                    _unlock_file(f)
        finally:
            f.close()

    def get_file(self, key):
        """Return a file-like representation or raise a KeyError."""
        try:
            f = open(os.path.join(self._path, str(key)), 'rb')
        except IOError, e:
            if e.errno == errno.ENOENT:
                raise KeyError('No message with key: %s' % key)
            else:
                raise
        return _ProxyFile(f)

    def iterkeys(self):
        """Return an iterator over keys."""
        return iter(sorted(int(entry) for entry in os.listdir(self._path)
                                      if entry.isdigit()))

    def has_key(self, key):
        """Return True if the keyed message exists, False otherwise."""
        return os.path.exists(os.path.join(self._path, str(key)))

    def __len__(self):
        """Return a count of messages in the mailbox."""
        return len(list(self.iterkeys()))

    def lock(self):
        """Lock the mailbox."""
        if not self._locked:
            self._file = open(os.path.join(self._path, '.mh_sequences'), 'rb+')
            _lock_file(self._file)
            self._locked = True

    def unlock(self):
        """Unlock the mailbox if it is locked."""
        if self._locked:
            _unlock_file(self._file)
            _sync_close(self._file)
            del self._file
            self._locked = False

    def flush(self):
        """Write any pending changes to the disk."""
        return

    def close(self):
        """Flush and close the mailbox."""
        if self._locked:
            self.unlock()

    def list_folders(self):
        """Return a list of folder names."""
        result = []
        for entry in os.listdir(self._path):
            if os.path.isdir(os.path.join(self._path, entry)):
                result.append(entry)
        return result

    def get_folder(self, folder):
        """Return an MH instance for the named folder."""
        return MH(os.path.join(self._path, folder),
                  factory=self._factory, create=False)

    def add_folder(self, folder):
        """Create a folder and return an MH instance representing it."""
        return MH(os.path.join(self._path, folder),
                  factory=self._factory)

    def remove_folder(self, folder):
        """Delete the named folder, which must be empty."""
        path = os.path.join(self._path, folder)
        entries = os.listdir(path)
        if entries == ['.mh_sequences']:
            os.remove(os.path.join(path, '.mh_sequences'))
        elif entries == []:
            pass
        else:
            raise NotEmptyError('Folder not empty: %s' % self._path)
        os.rmdir(path)

    def get_sequences(self):
        """Return a name-to-key-list dictionary to define each sequence."""
        results = {}
        f = open(os.path.join(self._path, '.mh_sequences'), 'r')
        try:
            all_keys = set(self.keys())
            for line in f:
                try:
                    name, contents = line.split(':')
                    keys = set()
                    for spec in contents.split():
                        if spec.isdigit():
                            keys.add(int(spec))
                        else:
                            start, stop = (int(x) for x in spec.split('-'))
                            keys.update(range(start, stop + 1))
                    results[name] = [key for key in sorted(keys) \
                                         if key in all_keys]
                    if len(results[name]) == 0:
                        del results[name]
                except ValueError:
                    raise FormatError('Invalid sequence specification: %s' %
                                      line.rstrip())
        finally:
            f.close()
        return results

    def set_sequences(self, sequences):
        """Set sequences using the given name-to-key-list dictionary."""
        f = open(os.path.join(self._path, '.mh_sequences'), 'r+')
        try:
            os.close(os.open(f.name, os.O_WRONLY | os.O_TRUNC))
            for name, keys in sequences.iteritems():
                if len(keys) == 0:
                    continue
                f.write('%s:' % name)
                prev = None
                completing = False
                for key in sorted(set(keys)):
                    if key - 1 == prev:
                        if not completing:
                            completing = True
                            f.write('-')
                    elif completing:
                        completing = False
                        f.write('%s %s' % (prev, key))
                    else:
                        f.write(' %s' % key)
                    prev = key
                if completing:
                    f.write(str(prev) + '\n')
                else:
                    f.write('\n')
        finally:
            _sync_close(f)

    def pack(self):
        """Re-name messages to eliminate numbering gaps. Invalidates keys."""
        sequences = self.get_sequences()
        prev = 0
        changes = []
        for key in self.iterkeys():
            if key - 1 != prev:
                changes.append((key, prev + 1))
                if hasattr(os, 'link'):
                    os.link(os.path.join(self._path, str(key)),
                            os.path.join(self._path, str(prev + 1)))
                    os.unlink(os.path.join(self._path, str(key)))
                else:
                    os.rename(os.path.join(self._path, str(key)),
                              os.path.join(self._path, str(prev + 1)))
            prev += 1
        self._next_key = prev + 1
        if len(changes) == 0:
            return
        for name, key_list in sequences.items():
            for old, new in changes:
                if old in key_list:
                    key_list[key_list.index(old)] = new
        self.set_sequences(sequences)

    def _dump_sequences(self, message, key):
        """Inspect a new MHMessage and update sequences appropriately."""
        pending_sequences = message.get_sequences()
        all_sequences = self.get_sequences()
        for name, key_list in all_sequences.iteritems():
            if name in pending_sequences:
                key_list.append(key)
            elif key in key_list:
                del key_list[key_list.index(key)]
        for sequence in pending_sequences:
            if sequence not in all_sequences:
                all_sequences[sequence] = [key]
        self.set_sequences(all_sequences)


class Babyl(_singlefileMailbox):
    """An Rmail-style Babyl mailbox."""

    _special_labels = frozenset(('unseen', 'deleted', 'filed', 'answered',
                                 'forwarded', 'edited', 'resent'))

    def __init__(self, path, factory=None, create=True):
        """Initialize a Babyl mailbox."""
        _singlefileMailbox.__init__(self, path, factory, create)
        self._labels = {}

    def add(self, message):
        """Add message and return assigned key."""
        key = _singlefileMailbox.add(self, message)
        if isinstance(message, BabylMessage):
            self._labels[key] = message.get_labels()
        return key

    def remove(self, key):
        """Remove the keyed message; raise KeyError if it doesn't exist."""
        _singlefileMailbox.remove(self, key)
        if key in self._labels:
            del self._labels[key]

    def __setitem__(self, key, message):
        """Replace the keyed message; raise KeyError if it doesn't exist."""
        _singlefileMailbox.__setitem__(self, key, message)
        if isinstance(message, BabylMessage):
            self._labels[key] = message.get_labels()

    def get_message(self, key):
        """Return a Message representation or raise a KeyError."""
        start, stop = self._lookup(key)
        self._file.seek(start)
        self._file.readline()   # Skip '1,' line specifying labels.
        original_headers = StringIO.StringIO()
        while True:
            line = self._file.readline()
            if line == '*** EOOH ***' + os.linesep or line == '':
                break
            original_headers.write(line.replace(os.linesep, '\n'))
        visible_headers = StringIO.StringIO()
        while True:
            line = self._file.readline()
            if line == os.linesep or line == '':
                break
            visible_headers.write(line.replace(os.linesep, '\n'))
        body = self._file.read(stop - self._file.tell()).replace(os.linesep,
                                                                 '\n')
        msg = BabylMessage(original_headers.getvalue() + body)
        msg.set_visible(visible_headers.getvalue())
        if key in self._labels:
            msg.set_labels(self._labels[key])
        return msg

    def get_string(self, key):
        """Return a string representation or raise a KeyError."""
        start, stop = self._lookup(key)
        self._file.seek(start)
        self._file.readline()   # Skip '1,' line specifying labels.
        original_headers = StringIO.StringIO()
        while True:
            line = self._file.readline()
            if line == '*** EOOH ***' + os.linesep or line == '':
                break
            original_headers.write(line.replace(os.linesep, '\n'))
        while True:
            line = self._file.readline()
            if line == os.linesep or line == '':
                break
        return original_headers.getvalue() + \
               self._file.read(stop - self._file.tell()).replace(os.linesep,
                                                                 '\n')

    def get_file(self, key):
        """Return a file-like representation or raise a KeyError."""
        return StringIO.StringIO(self.get_string(key).replace('\n',
                                                              os.linesep))

    def get_labels(self):
        """Return a list of user-defined labels in the mailbox."""
        self._lookup()
        labels = set()
        for label_list in self._labels.values():
            labels.update(label_list)
        labels.difference_update(self._special_labels)
        return list(labels)

    def _generate_toc(self):
        """Generate key-to-(start, stop) table of contents."""
        starts, stops = [], []
        self._file.seek(0)
        next_pos = 0
        label_lists = []
        while True:
            line_pos = next_pos
            line = self._file.readline()
            next_pos = self._file.tell()
            if line == '\037\014' + os.linesep:
                if len(stops) < len(starts):
                    stops.append(line_pos - len(os.linesep))
                starts.append(next_pos)
                labels = [label.strip() for label
                                        in self._file.readline()[1:].split(',')
                                        if label.strip() != '']
                label_lists.append(labels)
            elif line == '\037' or line == '\037' + os.linesep:
                if len(stops) < len(starts):
                    stops.append(line_pos - len(os.linesep))
            elif line == '':
                stops.append(line_pos - len(os.linesep))
                break
        self._toc = dict(enumerate(zip(starts, stops)))
        self._labels = dict(enumerate(label_lists))
        self._next_key = len(self._toc)
        self._file.seek(0, 2)
        self._file_length = self._file.tell()

    def _pre_mailbox_hook(self, f):
        """Called before writing the mailbox to file f."""
        f.write('BABYL OPTIONS:%sVersion: 5%sLabels:%s%s\037' %
                (os.linesep, os.linesep, ','.join(self.get_labels()),
                 os.linesep))

    def _pre_message_hook(self, f):
        """Called before writing each message to file f."""
        f.write('\014' + os.linesep)

    def _post_message_hook(self, f):
        """Called after writing each message to file f."""
        f.write(os.linesep + '\037')

    def _install_message(self, message):
        """Write message contents and return (start, stop)."""
        start = self._file.tell()
        if isinstance(message, BabylMessage):
            special_labels = []
            labels = []
            for label in message.get_labels():
                if label in self._special_labels:
                    special_labels.append(label)
                else:
                    labels.append(label)
            self._file.write('1')
            for label in special_labels:
                self._file.write(', ' + label)
            self._file.write(',,')
            for label in labels:
                self._file.write(' ' + label + ',')
            self._file.write(os.linesep)
        else:
            self._file.write('1,,' + os.linesep)
        if isinstance(message, email.message.Message):
            orig_buffer = StringIO.StringIO()
            orig_generator = email.generator.Generator(orig_buffer, False, 0)
            orig_generator.flatten(message)
            orig_buffer.seek(0)
            while True:
                line = orig_buffer.readline()
                self._file.write(line.replace('\n', os.linesep))
                if line == '\n' or line == '':
                    break
            self._file.write('*** EOOH ***' + os.linesep)
            if isinstance(message, BabylMessage):
                vis_buffer = StringIO.StringIO()
                vis_generator = email.generator.Generator(vis_buffer, False, 0)
                vis_generator.flatten(message.get_visible())
                while True:
                    line = vis_buffer.readline()
                    self._file.write(line.replace('\n', os.linesep))
                    if line == '\n' or line == '':
                        break
            else:
                orig_buffer.seek(0)
                while True:
                    line = orig_buffer.readline()
                    self._file.write(line.replace('\n', os.linesep))
                    if line == '\n' or line == '':
                        break
            while True:
                buffer = orig_buffer.read(4096) # Buffer size is arbitrary.
                if buffer == '':
                    break
                self._file.write(buffer.replace('\n', os.linesep))
        elif isinstance(message, str):
            body_start = message.find('\n\n') + 2
            if body_start - 2 != -1:
                self._file.write(message[:body_start].replace('\n',
                                                              os.linesep))
                self._file.write('*** EOOH ***' + os.linesep)
                self._file.write(message[:body_start].replace('\n',
                                                              os.linesep))
                self._file.write(message[body_start:].replace('\n',
                                                              os.linesep))
            else:
                self._file.write('*** EOOH ***' + os.linesep + os.linesep)
         …