docker-offlineimap/offlineimap/imapserver.py

723 lines
29 KiB
Python
Raw Normal View History

2002-06-19 06:39:00 +02:00
# IMAP server support
# Copyright (C) 2002 - 2011 John Goerzen & contributors
2002-06-19 06:39:00 +02:00
#
# 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 2 of the License, or
# (at your option) any later version.
2002-06-19 06:39:00 +02:00
#
# 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, write to the Free Software
2006-08-12 06:15:55 +02:00
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
2002-06-19 06:39:00 +02:00
from offlineimap import imaplibutil, imaputil, threadutil, OfflineImapError
from offlineimap.ui import getglobalui
from threading import Lock, BoundedSemaphore, Thread, Event, currentThread
import offlineimap.accounts
import hmac
import socket
import base64
import time
import errno
from sys import exc_info
from socket import gaierror
from ssl import SSLError, cert_time_to_seconds
try:
# do we have a recent pykerberos?
have_gss = False
import kerberos
if 'authGSSClientWrap' in dir(kerberos):
have_gss = True
except ImportError:
pass
2002-06-19 06:39:00 +02:00
class IMAPServer:
"""Initializes all variables from an IMAPRepository() instance
Various functions, such as acquireconnection() return an IMAP4
object on which we can operate.
Public instance variables are: self.:
delim The server's folder delimiter. Only valid after acquireconnection()
"""
GSS_STATE_STEP = 0
GSS_STATE_WRAP = 1
def __init__(self, repos):
self.ui = getglobalui()
self.repos = repos
self.config = repos.getconfig()
self.preauth_tunnel = repos.getpreauthtunnel()
self.transport_tunnel = repos.gettransporttunnel()
if self.preauth_tunnel and self.transport_tunnel:
raise OfflineImapError('%s: ' % repos + \
'you must enable precisely one '
'type of tunnel (preauth or transport), '
'not both', OfflineImapError.ERROR.REPO)
self.tunnel = \
self.preauth_tunnel if self.preauth_tunnel \
else self.transport_tunnel
self.username = \
None if self.preauth_tunnel else repos.getuser()
self.user_identity = repos.get_remote_identity()
self.authmechs = repos.get_auth_mechanisms()
self.password = None
self.passworderror = None
2007-07-05 06:04:14 +02:00
self.goodpassword = None
self.usessl = repos.getssl()
self.hostname = \
None if self.preauth_tunnel else repos.gethost()
self.port = repos.getport()
if self.port == None:
self.port = 993 if self.usessl else 143
self.sslclientcert = repos.getsslclientcert()
self.sslclientkey = repos.getsslclientkey()
self.sslcacertfile = repos.getsslcacertfile()
self.sslversion = repos.getsslversion()
if self.sslcacertfile is None:
self.__verifycert = None # disable cert verification
self.delim = None
self.root = None
self.maxconnections = repos.getmaxconnections()
self.availableconnections = []
self.assignedconnections = []
self.lastowner = {}
self.semaphore = BoundedSemaphore(self.maxconnections)
self.connectionlock = Lock()
self.reference = repos.getreference()
self.idlefolders = repos.getidlefolders()
self.gss_step = self.GSS_STATE_STEP
self.gss_vc = None
self.gssapi = False
2002-06-19 06:39:00 +02:00
def __getpassword(self):
"""Returns the server password or None"""
if self.goodpassword != None: # use cached good one first
2007-07-05 06:04:14 +02:00
return self.goodpassword
if self.password != None and self.passworderror == None:
return self.password # non-failed preconfigured one
# get 1) configured password first 2) fall back to asking via UI
self.password = self.repos.getpassword() or \
self.ui.getpass(self.repos.getname(), self.config,
self.passworderror)
self.passworderror = None
return self.password
# XXX: is this function used anywhere?
def getroot(self):
"""Returns this server's folder root. Can only be called after one
or more calls to acquireconnection."""
return self.root
def releaseconnection(self, connection, drop_conn=False):
"""Releases a connection, returning it to the pool.
:param drop_conn: If True, the connection will be released and
not be reused. This can be used to indicate broken connections."""
if connection is None: return #noop on bad connection
self.connectionlock.acquire()
self.assignedconnections.remove(connection)
# Don't reuse broken connections
if connection.Terminate or drop_conn:
connection.logout()
else:
self.availableconnections.append(connection)
self.connectionlock.release()
self.semaphore.release()
def __md5handler(self, response):
challenge = response.strip()
self.ui.debug('imap', '__md5handler: got challenge %s' % challenge)
passwd = self.__getpassword()
retval = self.username + ' ' + hmac.new(passwd, challenge).hexdigest()
self.ui.debug('imap', '__md5handler: returning %s' % retval)
return retval
def __loginauth(self, imapobj):
""" Basic authentication via LOGIN command """
self.ui.debug('imap', 'Attempting IMAP LOGIN authentication')
imapobj.login(self.username, self.__getpassword())
def __plainhandler(self, response):
"""
Implements SASL PLAIN authentication, RFC 4616,
http://tools.ietf.org/html/rfc4616
"""
authc = self.username
passwd = self.__getpassword()
authz = ''
if self.user_identity != None:
authz = self.user_identity
NULL = u'\x00'
retval = NULL.join((authz, authc, passwd)).encode('utf-8')
self.ui.debug('imap', '__plainhandler: returning %s' % retval)
return retval
# XXX: describe function
def __gssauth(self, response):
data = base64.b64encode(response)
try:
if self.gss_step == self.GSS_STATE_STEP:
if not self.gss_vc:
rc, self.gss_vc = kerberos.authGSSClientInit('imap@' +
self.hostname)
response = kerberos.authGSSClientResponse(self.gss_vc)
rc = kerberos.authGSSClientStep(self.gss_vc, data)
if rc != kerberos.AUTH_GSS_CONTINUE:
self.gss_step = self.GSS_STATE_WRAP
elif self.gss_step == self.GSS_STATE_WRAP:
rc = kerberos.authGSSClientUnwrap(self.gss_vc, data)
response = kerberos.authGSSClientResponse(self.gss_vc)
rc = kerberos.authGSSClientWrap(self.gss_vc, response,
self.username)
response = kerberos.authGSSClientResponse(self.gss_vc)
except kerberos.GSSError as err:
# Kerberos errored out on us, respond with None to cancel the
# authentication
self.ui.debug('imap', '%s: %s' % (err[0][0], err[1][0]))
return None
if not response:
response = ''
return base64.b64decode(response)
def __start_tls(self, imapobj):
if 'STARTTLS' in imapobj.capabilities and not self.usessl:
self.ui.debug('imap', 'Using STARTTLS connection')
try:
imapobj.starttls()
except imapobj.error as e:
raise OfflineImapError("Failed to start "
"TLS connection: %s" % str(e),
OfflineImapError.ERROR.REPO)
## All __authn_* procedures are helpers that do authentication.
## They are class methods that take one parameter, IMAP object.
##
## Each function should return True if authentication was
## successful and False if authentication wasn't even tried
## for some reason (but not when IMAP has no such authentication
## capability, calling code checks that).
##
## Functions can also raise exceptions; two types are special
## and will be handled by the calling code:
##
## - imapobj.error means that there was some error that
## comes from imaplib2;
##
## - OfflineImapError means that function detected some
## problem by itself.
def __authn_gssapi(self, imapobj):
if not have_gss:
return False
self.connectionlock.acquire()
try:
imapobj.authenticate('GSSAPI', self.__gssauth)
return True
except imapobj.error as e:
self.gssapi = False
raise
else:
self.gssapi = True
kerberos.authGSSClientClean(self.gss_vc)
self.gss_vc = None
self.gss_step = self.GSS_STATE_STEP
finally:
self.connectionlock.release()
def __authn_cram_md5(self, imapobj):
imapobj.authenticate('CRAM-MD5', self.__md5handler)
return True
def __authn_plain(self, imapobj):
imapobj.authenticate('PLAIN', self.__plainhandler)
return True
def __authn_login(self, imapobj):
# Use LOGIN command, unless LOGINDISABLED is advertized
# (per RFC 2595)
if 'LOGINDISABLED' in imapobj.capabilities:
raise OfflineImapError("IMAP LOGIN is "
"disabled by server. Need to use SSL?",
OfflineImapError.ERROR.REPO)
else:
self.__loginauth(imapobj)
return True
def __authn_helper(self, imapobj):
"""
Authentication machinery for self.acquireconnection().
Raises OfflineImapError() of type ERROR.REPO when
there are either fatal problems or no authentications
succeeded.
If any authentication method succeeds, routine should exit:
warnings for failed methods are to be produced in the
respective except blocks.
"""
# Authentication routines, hash keyed by method name
# with value that is a tuple with
# - authentication function,
# - tryTLS flag,
# - check IMAP capability flag.
auth_methods = {
"GSSAPI": (self.__authn_gssapi, False, True),
"CRAM-MD5": (self.__authn_cram_md5, True, True),
"PLAIN": (self.__authn_plain, True, True),
"LOGIN": (self.__authn_login, True, False),
}
# Stack stores pairs of (method name, exception)
exc_stack = []
tried_to_authn = False
tried_tls = False
mechs = self.authmechs
# GSSAPI must be tried first: we will probably go TLS after it
# and GSSAPI mustn't be tunneled over TLS.
if "GSSAPI" in mechs:
mechs.remove("GSSAPI")
mechs.insert(0, "GSSAPI")
for m in mechs:
if m not in auth_methods:
raise Exception("Bad authentication method %s, "
"please, file OfflineIMAP bug" % m)
func, tryTLS, check_cap = auth_methods[m]
# TLS must be initiated before checking capabilities:
# they could have been changed after STARTTLS.
if tryTLS and not tried_tls:
tried_tls = True
self.__start_tls(imapobj)
if check_cap:
cap = "AUTH=" + m
if cap not in imapobj.capabilities:
continue
tried_to_authn = True
self.ui.debug('imap', 'Attempting '
'%s authentication' % m)
try:
if func(imapobj):
return
except (imapobj.error, OfflineImapError) as e:
self.ui.warn('%s authentication failed: %s' % (m, e))
exc_stack.append((m, e))
if len(exc_stack):
msg = "\n\t".join(map(
lambda x: ": ".join((x[0], str(x[1]))),
exc_stack
))
raise OfflineImapError("All authentication types "
"failed:\n\t%s" % msg, OfflineImapError.ERROR.REPO)
if not tried_to_authn:
methods = ", ".join(map(
lambda x: x[5:], filter(lambda x: x[0:5] == "AUTH=",
imapobj.capabilities)
))
raise OfflineImapError("Repository %s: no supported "
"authentication mechanisms found; configured %s, "
"server advertises %s" % (self.repos,
", ".join(self.authmechs), methods),
OfflineImapError.ERROR.REPO)
# XXX: move above, closer to releaseconnection()
def acquireconnection(self):
"""Fetches a connection from the pool, making sure to create a new one
if needed, to obey the maximum connection limits, etc.
Opens a connection to the server and returns an appropriate
2002-06-19 06:39:00 +02:00
object."""
self.semaphore.acquire()
self.connectionlock.acquire()
curThread = currentThread()
2002-06-19 06:39:00 +02:00
imapobj = None
if len(self.availableconnections): # One is available.
# Try to find one that previously belonged to this thread
# as an optimization. Start from the back since that's where
# they're popped on.
imapobj = None
for i in range(len(self.availableconnections) - 1, -1, -1):
tryobj = self.availableconnections[i]
if self.lastowner[tryobj] == curThread.ident:
imapobj = tryobj
del(self.availableconnections[i])
break
if not imapobj:
imapobj = self.availableconnections[0]
del(self.availableconnections[0])
self.assignedconnections.append(imapobj)
self.lastowner[imapobj] = curThread.ident
self.connectionlock.release()
return imapobj
self.connectionlock.release() # Release until need to modify data
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
""" Must be careful here that if we fail we should bail out gracefully
and release locks / threads so that the next attempt can try...
"""
success = 0
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
try:
while not success:
# Generate a new connection.
if self.tunnel:
self.ui.connecting('tunnel', self.tunnel)
imapobj = imaplibutil.IMAP4_Tunnel(self.tunnel,
timeout=socket.getdefaulttimeout())
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
success = 1
elif self.usessl:
self.ui.connecting(self.hostname, self.port)
fingerprint = self.repos.get_ssl_fingerprint()
imapobj = imaplibutil.WrappedIMAP4_SSL(self.hostname,
self.port,
self.sslclientkey,
self.sslclientcert,
self.sslcacertfile,
self.__verifycert,
self.sslversion,
timeout=socket.getdefaulttimeout(),
fingerprint=fingerprint
)
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
else:
self.ui.connecting(self.hostname, self.port)
imapobj = imaplibutil.WrappedIMAP4(self.hostname, self.port,
timeout=socket.getdefaulttimeout())
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
if not self.preauth_tunnel:
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
try:
self.__authn_helper(imapobj)
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
self.goodpassword = self.password
success = 1
except OfflineImapError as e:
self.passworderror = str(e)
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
raise
# update capabilities after login, e.g. gmail serves different ones
typ, dat = imapobj.capability()
if dat != [None]:
imapobj.capabilities = tuple(dat[-1].upper().split())
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
if self.delim == None:
listres = imapobj.list(self.reference, '""')[1]
if listres == [None] or listres == None:
# Some buggy IMAP servers do not respond well to LIST "" ""
# Work around them.
listres = imapobj.list(self.reference, '"*"')[1]
if listres == [None] or listres == None:
# No Folders were returned. This occurs, e.g. if the
# 'reference' prefix does not exist on the mail
# server. Raise exception.
err = "Server '%s' returned no folders in '%s'" % \
(self.repos.getname(), self.reference)
self.ui.warn(err)
raise Exception(err)
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
self.delim, self.root = \
imaputil.imapsplit(listres[0])[1:]
self.delim = imaputil.dequote(self.delim)
self.root = imaputil.dequote(self.root)
with self.connectionlock:
self.assignedconnections.append(imapobj)
self.lastowner[imapobj] = curThread.ident
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
return imapobj
except Exception as e:
"""If we are here then we did not succeed in getting a
connection - we should clean up and then re-raise the
error..."""
Patch for error handling / separation of accounts etc. Dear All, I have made the attached patch to try and make offlineimap a bit more stable in challenging situations. It's extremely useful in slow connection environments - but sometimes if one account had the wrong password or the connection went down then unfortunately the whole program would crash. I have tested this on our connection and tried throwing at it just about every situation - connection, up down, up, down again, change password, error whilst copying one message, etc. I have been running this patch for the last 5 days or so syncing 6 accounts at the moment... It seems to work and stay alive nicely (even if your connection does not)... Hope that this can go in for the next release... Please let me know if anyone notices any problems with this... Regards, -Mike -- Attached file included as plaintext by Ecartis -- -- File: submit From 1d6777cab23637eb830031c7cab0ae9b8589afd6 Mon Sep 17 00:00:00 2001 From: mike <mike@mikelaptop.(none)> Date: Mon, 24 Aug 2009 19:37:59 +0430 Subject: [PATCH] This patch attempts to introduce a little more error handling - e.g. if one account has an error because of a changed password or something that should not affect the other accounts. Specifically: If one sync run has an issue this is in a try-except clause - if it has an auto refresh period the thread will sleep and try again - this could be quite useful in the event of the connection going down for a little while, changed password etc. If one folder cannot be created an error message will be displayed through the UI and the program will continue (e.g. permission denied to create a folder) If one message does not want to copy for whatever resaon an error message will be displayed through the UI and at least the other messages will be copied If one folder run has an exception then the others will still run
2009-08-24 17:17:57 +02:00
self.semaphore.release()
severity = OfflineImapError.ERROR.REPO
if type(e) == gaierror:
#DNS related errors. Abort Repo sync
#TODO: special error msg for e.errno == 2 "Name or service not known"?
reason = "Could not resolve name '%s' for repository "\
"'%s'. Make sure you have configured the ser"\
"ver name correctly and that you are online."%\
(self.hostname, self.repos)
raise OfflineImapError(reason, severity)
elif isinstance(e, SSLError) and e.errno == 1:
# SSL unknown protocol error
# happens e.g. when connecting via SSL to a non-SSL service
if self.port != 993:
reason = "Could not connect via SSL to host '%s' and non-s"\
"tandard ssl port %d configured. Make sure you connect"\
" to the correct port." % (self.hostname, self.port)
else:
reason = "Unknown SSL protocol connecting to host '%s' for"\
"repository '%s'. OpenSSL responded:\n%s"\
% (self.hostname, self.repos, e)
raise OfflineImapError(reason, severity)
elif isinstance(e, socket.error) and e.args[0] == errno.ECONNREFUSED:
# "Connection refused", can be a non-existing port, or an unauthorized
# webproxy (open WLAN?)
reason = "Connection to host '%s:%d' for repository '%s' was "\
"refused. Make sure you have the right host and port "\
"configured and that you are actually able to access the "\
"network." % (self.hostname, self.port, self.reposname)
raise OfflineImapError(reason, severity)
# Could not acquire connection to the remote;
# socket.error(last_error) raised
if str(e)[:24] == "can't open socket; error":
raise OfflineImapError("Could not connect to remote server '%s' "\
"for repository '%s'. Remote does not answer."
% (self.hostname, self.repos),
OfflineImapError.ERROR.REPO)
else:
# re-raise all other errors
raise
def connectionwait(self):
"""Waits until there is a connection available. Note that between
the time that a connection becomes available and the time it is
requested, another thread may have grabbed it. This function is
mainly present as a way to avoid spawning thousands of threads
to copy messages, then have them all wait for 3 available connections.
It's OK if we have maxconnections + 1 or 2 threads, which is what
this will help us do."""
self.semaphore.acquire()
self.semaphore.release()
2002-06-21 09:25:24 +02:00
def close(self):
# Make sure I own all the semaphores. Let the threads finish
# their stuff. This is a blocking method.
with self.connectionlock:
# first, wait till all connections had been released.
# TODO: won't work IMHO, as releaseconnection() also
# requires the connectionlock, leading to a potential
# deadlock! Audit & check!
threadutil.semaphorereset(self.semaphore, self.maxconnections)
for imapobj in self.assignedconnections + self.availableconnections:
imapobj.logout()
self.assignedconnections = []
self.availableconnections = []
self.lastowner = {}
# reset kerberos state
self.gss_step = self.GSS_STATE_STEP
self.gss_vc = None
self.gssapi = False
def keepalive(self, timeout, event):
"""Sends a NOOP to each connection recorded. It will wait a maximum
of timeout seconds between doing this, and will continue to do so
until the Event object as passed is true. This method is expected
to be invoked in a separate thread, which should be join()'d after
the event is set."""
self.ui.debug('imap', 'keepalive thread started')
while not event.isSet():
self.connectionlock.acquire()
numconnections = len(self.assignedconnections) + \
len(self.availableconnections)
self.connectionlock.release()
threads = []
for i in range(numconnections):
self.ui.debug('imap', 'keepalive: processing connection %d of %d' % (i, numconnections))
if len(self.idlefolders) > i:
# IDLE thread
idler = IdleThread(self, self.idlefolders[i])
else:
# NOOP thread
idler = IdleThread(self)
idler.start()
threads.append(idler)
self.ui.debug('imap', 'keepalive: waiting for timeout')
event.wait(timeout)
self.ui.debug('imap', 'keepalive: after wait')
for idler in threads:
# Make sure all the commands have completed.
idler.stop()
idler.join()
self.ui.debug('imap', 'keepalive: all threads joined')
self.ui.debug('imap', 'keepalive: event is set; exiting')
return
def __verifycert(self, cert, hostname):
'''Verify that cert (in socket.getpeercert() format) matches hostname.
CRLs are not handled.
Returns error message if any problems are found and None on success.
'''
errstr = "CA Cert verifying failed: "
if not cert:
return ('%s no certificate received' % errstr)
dnsname = hostname.lower()
certnames = []
# cert expired?
notafter = cert.get('notAfter')
if notafter:
if time.time() >= cert_time_to_seconds(notafter):
return '%s certificate expired %s' % (errstr, notafter)
# First read commonName
for s in cert.get('subject', []):
key, value = s[0]
if key == 'commonName':
certnames.append(value.lower())
if len(certnames) == 0:
return ('%s no commonName found in certificate' % errstr)
# Then read subjectAltName
for key, value in cert.get('subjectAltName', []):
if key == 'DNS':
certnames.append(value.lower())
# And finally try to match hostname with one of these names
for certname in certnames:
if (certname == dnsname or
'.' in dnsname and certname == '*.' + dnsname.split('.', 1)[1]):
return None
return ('%s no matching domain name found in certificate' % errstr)
class IdleThread(object):
def __init__(self, parent, folder=None):
"""If invoked without 'folder', perform a NOOP and wait for
self.stop() to be called. If invoked with folder, switch to IDLE
mode and synchronize once we have a new message"""
self.parent = parent
self.folder = folder
self.stop_sig = Event()
self.ui = getglobalui()
if folder is None:
self.thread = Thread(target=self.noop)
else:
self.thread = Thread(target=self.idle)
self.thread.setDaemon(1)
def start(self):
self.thread.start()
def stop(self):
self.stop_sig.set()
def join(self):
self.thread.join()
def noop(self):
#TODO: AFAIK this is not optimal, we will send a NOOP on one
#random connection (ie not enough to keep all connections
#open). In case we do the noop multiple times, we can well use
#the same connection every time, as we get a random one. This
#function should IMHO send a noop on ALL available connections
#to the server.
imapobj = self.parent.acquireconnection()
try:
imapobj.noop()
except imapobj.abort:
self.ui.warn('Attempting NOOP on dropped connection %s' % \
imapobj.identifier)
self.parent.releaseconnection(imapobj, True)
imapobj = None
finally:
if imapobj:
self.parent.releaseconnection(imapobj)
self.stop_sig.wait() # wait until we are supposed to quit
def __dosync(self):
remoterepos = self.parent.repos
account = remoterepos.account
localrepos = account.localrepos
remoterepos = account.remoterepos
statusrepos = account.statusrepos
remotefolder = remoterepos.getfolder(self.folder)
hook = account.getconf('presynchook', '')
account.callhook(hook)
offlineimap.accounts.syncfolder(account, remotefolder, quick=False)
hook = account.getconf('postsynchook', '')
account.callhook(hook)
ui = getglobalui()
ui.unregisterthread(currentThread()) #syncfolder registered the thread
def __idle(self):
"""Invoke IDLE mode until timeout or self.stop() is invoked"""
def callback(args):
"""IDLE callback function invoked by imaplib2
This is invoked when a) The IMAP server tells us something
while in IDLE mode, b) we get an Exception (e.g. on dropped
connections, or c) the standard imaplib IDLE timeout of 29
minutes kicks in."""
result, cb_arg, exc_data = args
if exc_data is None and not self.stop_sig.isSet():
# No Exception, and we are not supposed to stop:
self.needsync = True
self.stop_sig.set() # continue to sync
while not self.stop_sig.isSet():
self.needsync = False
success = False # successfully selected FOLDER?
while not success:
imapobj = self.parent.acquireconnection()
try:
imapobj.select(self.folder)
except OfflineImapError as e:
if e.severity == OfflineImapError.ERROR.FOLDER_RETRY:
# Connection closed, release connection and retry
self.ui.error(e, exc_info()[2])
self.parent.releaseconnection(imapobj, True)
else:
raise e
else:
success = True
if "IDLE" in imapobj.capabilities:
imapobj.__idle(callback=callback)
else:
self.ui.warn("IMAP IDLE not supported on server '%s'."
"Sleep until next refresh cycle." % imapobj.identifier)
imapobj.noop()
self.stop_sig.wait() # self.stop() or IDLE callback are invoked
try:
# End IDLE mode with noop, imapobj can point to a dropped conn.
imapobj.noop()
except imapobj.abort:
self.ui.warn('Attempting NOOP on dropped connection %s' % \
imapobj.identifier)
self.parent.releaseconnection(imapobj, True)
else:
self.parent.releaseconnection(imapobj)
if self.needsync:
# here not via self.stop, but because IDLE responded. Do
# another round and invoke actual syncing.
self.stop_sig.clear()
self.__dosync()