savemessage_getnewheader was an undocmented, cryptic and overengineered
function. It generates a new unique value that can be used as a mail
header to be inserted. For this it used LOTS of randomness sources: hash
of the mail content, hash of the folder name, hash of the repository
name, the current time, a random() value, and the offlineimap version string.
All we need is something random. So reduce this to hash of content
appended by a random integer. Sufficient and somewhat faster to calculate.
Rename the function to actually describe accurately what it does or
would you have guessed that savemessage_getnewheader() did nothing more
than returning ('X-OfflineIMAP', <randomstring> )? Rename to
generate_randomheader() to make it clearer what this is all about.
Also document the function, describing what it does, and what it returns.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
There is a clumsy workaround for Darwin that chunks reads into 8kb
blocks to avoid huge memory allocations. First, this fix should not only
be required but on FreeBSD2.6 too (see
http://bugs.python.org/issue3531). Second, decent python versions (I
checked 2.6) already chunk in the SSL case anyway, so there is no need to do
that again. Remove that level of indirection.
http://evanjones.ca/python-memory.html claims that this problem has been
fixed since python 2.5, so we might consider removing the workaround
completely even for the non-SSL case.
Increase the chunk size on Mac from 8kb to 64kb. Even Macs should be
able to take that amount of memory usage nowadays.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
A typo was preventing profiling mode to really enable singlethreading
mode. Fixing the unfortunate typo of mine makes it work.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
It is currently very hard to find out what folderfilter actually does
and makes it hard to debug for a user. With this patch if the user has
enabled "-d imap" (even better would perhaps be a different debug type
for this kind of thing?), we see a message
"Filtering out folder 'foo' due to folderfilter"
in the logs.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
This enables us to make use of self.ui in all repositories without
having to import and use getglobalui() in all types of repositories and
all places.
Note that this patch only makes this available, it does not yet make use
of it.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
When e.g. specifying an invalid 'reference' value for an IMAP server to
a root folder that does not exist, we would previously have crashed with
a nonsensical and non-intuitive error message (trying to address an
element of a NoneType).
This will also raise an Exception (which should be ok, given that this
is really a misconfiguration on the user side), but it will explain to
the user WHY it exited.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
The debug output for dequote, optionsplit is very verbose, outputing
what the functions are called with and what they return. Those functions
are now very old mature and rather simple, so it suffices to output
their return value rather than cluttering out log with too much
uninteresting garbarge.
This makes log files much more readable.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
A Repository() returns the correctly instanciated dervivate of a
BaseRepository, depending on the parameters passed to it. The returned
instance is eg an ImapRepository(). This makes the code look nicer,
and we have less functions lying around outside of classes (no more
global LoadRepository() function).
This will also enable us to conveniently hand back a
LocalStatusRepository based on SQLITE rather than plain text, if the
user configures this to be the experimental and optional backend
(once it exists).
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
This seems to prevent mysterious hangs with SSL imap servers (especially
gmail?) and does not harm in any case. So let us enable keep-alive
messages for ssl connections.
Our thread pool should be made more robust against closed SSL
connections (which do not always seem to raise Exceptions), and not
deadlock while waiting for resources or data that will never arrive.
Reviewed-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Gábor Melis <mega@retes.hu>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Commit e506442996f1a4 changed getglobalui() back to UIBase.getglobalui()
although the import had changed earlier, causing a regression.
Fix this by using the correct and current way of calling the ui.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Signed-off-by: Thomas Jost <schnouki@schnouki.net>
Reviewed-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
There is no need for using the string module if all we want is to split
a string at the white space. All pythons since at least 2.4 can do that.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
We have vonverted all places in folder/* to have self.ui available,
rather than having to use UIBase.getglobalui() all the
time. Unfortunately, we did not convert the users in folder/Base.py.
This patch does it belatedly. This fixes
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=613483
Signed-off-by: Sebastian Spaeth1 <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
I accidentally added "self" as first parameter to the signal handler
method. Of course it is not called with a class instance, the signal
handler always only receives 2 parameters: the signal number and the
stack frame. So just removing the self fixes things.
Proposed-by: Christian Holme <cholme@gmx.com>
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Code was broken, as these backends import UIBase (which had been moved).
However, they don't use it, so we can just delete the import.
Sorry, I failed to find those earlier somehow.
Signed-off-by: Sebastian <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
AccountSynchronizationMixin was never used on its own and it is a very
confusing class until you understand what it is used for. (It
complemented the Account() class with a few methods to make Account()
syncable.
But we use the SyncableAccount class anyway, so merge the former Mixin'
methods directly in there.
This does away with a class that is not directly used, and was a case of
over-object-orientation which confuses more than it helps.
Touched up code documentation while going through the file.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
"mailboxes" is defined global and set to an empty list, but never used
from anywhere within offlineimap. So let us just delete it.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Signed-off-by: Edward Z. Yang <ezyang@mit.edu>
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Make --version much less verbose and only print out the version number
to facilitate easy parsing by scripts. We don't really need the verbose
copyright output, it is show with --help anyway.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Currently the Curses code is broken. Importing offlineimap.ui.Curses
will not succeed due to cyclic imports (threadutils imports ui, but ui
wants threadutils.MultiLock). So Curses cannot be chosen.
Incidentally, the only part in the code that uses "MultiLock" is the
Curses UI, to prevent concurrent access from several threads to the
ui-internal thread list and to IO resources such as the
screen. Fortunately for these purposes we don't need a MultiLock, so we
can do away with that implementation completely. A simple RLock that
allows us to have a thread "own" a lock and makes other threads wanting
access to the resource wait until the owning thread is finished.
The MultiLock implementation looked a bit weird, so simplifying code
here is a good thing, it might well be that we fix some "hangs" that
have been reported (and that would only ever occur with the Curses UI).
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
This fixes some leftover of 0b5b38d2989b7b6e6ee7846deea86daf32dc30f1.
Signed-off-by: Thomas Schwinge <thomas@schwinge.name>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
This branch is currently broken, as we were using getglobalui() but it
was not imported. This fixes the missing import and makes offlineimap
run successfully.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Improve documentation about what CustomConfigHelperMixin does,
it was not very clear without a close look to the code before.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
The current code path checked the CA cert host name, even if we did not
specify a CA cert file to use. Make the host name check dependent on a
CA cert file.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
All invocations of -d will imply the single-threaded mode unless one of
the debug options is "thread" (in which case it would be senseless to
force single-threaded mode).
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Previously, we would spawn child threads for account synchronization
even if we had single-threading enabled. This prevented us from catching
the true location of exceptions, for example. Now, in single-threaded
mode, we perform the account synchronization truely in the main thread
which will ease our debugging.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Earlier we would ignore *ALL* Exceptions that could occur during the
fsyncing of our LocalStatus database. Ignoring all Exceptions is not the
right thing here though. A recent commit improved the situation by
raising at least KeyboardInterrupt Exceptions, but that is still not
optimal.
os.fsync() is available on Unix, and Windows starting in python
2.2.3. so it should always work. If it doesn't, something is wrong.
It has been suggested to only catch EnvironmentError (ie SystemError and
OSError) here, but even those should be thrown. Something *is* wrong if
this fails and we should not ignore it.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Previously we did not catch KeyboardInterrupts explicitly as all of the
code was executed in forked child threads which would never receive
Ctrl-c exceptions. With the upcoming single threaded modus, this code
can be run in the main thread however, so we need to take care of
KeyboardInterrupts explicitly.
This was done wherever we would catch *ALL* exceptions universally and
print out an error message.
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
Previously we did not catch KeyboardInterrupts explicitly as all of the
code was executed in forked child threads which would never receive
Ctrl-c exceptions. With the upcoming single threaded modus, this code
can be run in the main thread however, so we need to take care of
KeyboardInterrupts explicitly. As this is pretty highlevel code, we also
protect against receiving a SystemExit exception which is raised e.g. in
the ui.terminate() code by calling sys.exit().
Signed-off-by: Sebastian Spaeth <Sebastian@SSpaeth.de>
Signed-off-by: Nicolas Sebrecht <nicolas.s-dev@laposte.net>
