Commit Graph

954 Commits

Author SHA1 Message Date
Oswald Buddenhagen
b1842617f7 make MaxMessages work for new mails as well
this helps enormously on the first sync of a 100k message box with a
limit of 1k messages. it also happens to make the syncing idempotent.

in a few conditionals we now explicitly test for max_messages being
enabled, not smaxxuid != 0, as after the initial fetch with no important
messages smaxxuid is zero, but we still have to skip over 99k messages
in the above case.
2013-12-01 13:36:28 +01:00
Oswald Buddenhagen
d3f6347021 delay propagation of new messages
previous sequence:
  examine & propagate new => examine old => propagate old
new sequence:
  examine new => examine old => propagate new => propagate old

this alone does not buy us much ...
2013-12-01 13:36:28 +01:00
Oswald Buddenhagen
391ec01f28 make message propagation recording less magic
assign the sync record to the source message asap, and later on rely
on a more explicit condition than not doing so.
2013-12-01 13:36:28 +01:00
Oswald Buddenhagen
7f784fd235 log maxuid bumping less aggressively
we can bump the internal variable whereever convenient, but we cannot
log it until we know that all messages were copied, as otherwise we
could miss some new messages after an interruption. with the new
approach, interruption would merely cause some additonal traffic.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
8b76412b0d document message expiration transactions 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
ecb4c7ab07 propagate deletions with other flag changes
less code duplication, more logical order of issued driver commands
(especially after the next commit), and the "side effect" of letting the
message expiration code see those deletions if they are asynchronous.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
273ac899f3 don't delay loading master even if messages were expired
the delay optimized the corner case of previously important but now
expired messages on the slave disappearing, either through an external
expunge or after a journal replay. no point in pessimizing the common
case.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
12676f28da remove cleanup of expired entries during setup of master load
the removed code would only ever trigger if a) we were after a journal
replay or b) something external expunged the expired messages - both are
corner cases not worth the extra code.
however, this means that the syncing code further down now needs to take
care of these zombies.
in the end, the normal cleanup will take care of all expired entries,
new and old.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
9a62521cff micro-optimization/-clarification: swap condition order 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
014d9b9081 make message counting in expiration code less confusing 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
83b834cdfd count unread messages like flagged messages when expiring
that is, don't count them towards the total only below the cut-off
point. making them extend the working set even though they are inside it
is counterintuitive.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
9e186ae88b use post-sync "seen" flag to determine expirability
otherwise it wouldn't be idempotent.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
15216947fb don't protect recent messages from MaxMessages
while maildir has a clearly defined meaning of "recent" and for example
mutt handles it graciously, IMAP's definition is fubared to the point
that some servers (for example gmail) simply refuse to support it.
for symmetry reasons it is best to pretend that it doesn't exist at all.
it doesn't seem too useful anyway (the user can simply mark the messages
as read to allow pruning).
and last but not least, the man page of mbsync says nothing about
"recent", only "unread". unlike the isync man page, though.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
6b7b2b1106 always get slave flags when we are expiring
even if we are not propagating new messages, the appearance of new
messages on the slave can lead to expiring older messages. for that, we
need to know their importance, and thus flags.

the alternative would be not doing an expiration run when not fetching
new messages, but that would mean more conditionals all over the place.
as the decision is somewhat arbitrary, just do the simpler thing.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
f1eea7d9a5 do not trash expired messages
we are not actually deleting them, so there is no point in saving them
in the trash.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
48754ecc74 make sync state header format less obscure
the header is not space-critical, so use proper name-value pairs.
this has the additional advantage that subsequent format changes can be
done much easier.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
83bb1cf716 make state loading in showstate() similar to ckstate() 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
f044adbfa4 take configs out of target state defs
cleaner and less duplication
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
5297425918 more precise failure reporting 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
3d81ccbf21 make it possible to run only selected tests 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
3dcb393de2 set srec->msg[] when finding messages by tuid
otherwise we would propagate phantom deletions.

this affected only sync runs after an interruption while storing
messages, so it went (mostly?) unnoticed.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
3814f19661 remove pointless assignment
we already know that tmsg->srec is null at this point.
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
e63e16ab45 assert no stray TUIDs 2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
8e49300cf7 initialize struct tm
strptime() does not initialize at least tm_isdst, which leads to an
uninited value reference in mktime().
2013-12-01 13:36:27 +01:00
Oswald Buddenhagen
62a60997c3 make use of strptime() portable
it does not (officially) support the %z conversion, so re-implement that
part by hand.
2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
55e65147df fix compilation with older gcc versions
the warning suppression pragma within function scope is apparently a new
thing.
as i don't want to disable the check for the entire function (even if
this currently would make no difference), just use a wrapper function
to suppress the format string check.
2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
a49893f32e fix strftime() %z conversion specifier check
only glibc does something sane with gmtime()+strftime(). on bsd (incl.
mac os), strftime() can be used only with localtime().
2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
2b27216b86 ignore automake's "compile" script 2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
945e05cfdd use autoreconf instead of calling separate tools
this has been the correct way since a long time.

Pointed-out-by: Felipe Contreras <felipe.contreras@gmail.com>
2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
32def5dc0a add/fix comments and improve debug messages 2013-12-01 13:36:26 +01:00
Oswald Buddenhagen
a9a331c98a simplify condition
... and document the cases.
2013-12-01 13:35:02 +01:00
Oswald Buddenhagen
03f8bfdfb2 micro-optimization/-clarification 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
00076a6971 move initializations for clarity 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
61ef099cd5 MaxMessages: make condition exactly symmetrical to condition below 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
080740f867 rewrite condition for readability and consistency 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
b10fd0c21c remove assumption about value of M constant 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
a893cba483 fix enum abuse
amends 9c86ec344.

S_FIND was for the sync record status field. it has no business in the
sync vars status fields. its value coincided with ST_SELECTED, which
luckily only means that we always tried to match up TUIDs even if there
was nothing to do.

the need for TUID matching arises in two mostly independent
circumstances, so add two separate flags ST_FIND_{OLD,NEW}.
2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
9a0e65f899 create unseen messages in /new/
seen messages still go to /cur/.
this is consistent with the actual maildir driver.
2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
0a684bd933 be a bit more verbose 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
394aca03a2 properly handle unexpected exit while replaying journal 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
07377cb753 ensure that the journal replay pass really does nothing 2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
100f9487f4 sort messages by serial number instead of UID in box dumper
the input data is sorted that way, so it's easier to compare.
2013-12-01 13:35:01 +01:00
Oswald Buddenhagen
2568459a7b fix error message in sync state dumper 2013-12-01 13:35:00 +01:00
Oswald Buddenhagen
b570c17766 deal with messages disappearing between being listed and fetched 2013-09-26 09:17:08 +02:00
Oswald Buddenhagen
17c4748dfa support backslash-escaping in the config file
note that no attempt is made at making this work in the compat wrapper.
2013-09-26 09:17:08 +02:00
Oswald Buddenhagen
2213d6976c support backslashes and quotes in quoted IMAP strings
the RFCs require it - well hidden in the BNF at the bottom.

patch somewhat inspired by "guns" <self@sungpae.com>.
2013-09-26 09:17:08 +02:00
Oswald Buddenhagen
c0bf867669 make next_arg() more readable & efficient 2013-09-26 09:17:08 +02:00
Oswald Buddenhagen
3ceb553102 IPv6 support
inspired by a patch by "Todd T. Fries" <todd@fries.net>.
2013-09-01 17:39:07 +02:00
Oswald Buddenhagen
4a39cae8c4 support multi-homed servers 2013-09-01 17:39:07 +02:00
Oswald Buddenhagen
0b59ee0df3 support multi-character path separators
this applies to both the IMAP PathDelimiter (which is needed by Lotus
Domino), as well as the Flatten-ed separators.
2013-08-11 10:20:02 +02:00