25
1
mirror of https://github.com/processone/ejabberd.git synced 2024-10-31 15:21:38 +01:00
Commit Graph

3191 Commits

Author SHA1 Message Date
Holger Weiss
39f1005006 Fix LibYAML version in the list of dependencies
Thanks to Steve Gillespie for reporting the error.
2014-06-03 13:00:17 +02:00
Evgeniy Khramtsov
6441c284e0 Don't add 'rport' paramater to 'Via' header 2014-06-02 20:46:29 +04:00
Evgeniy Khramtsov
0aea9c74bd Rename option 'route' to 'record_route' and add new option 'routes' 2014-06-02 10:16:34 +04:00
Evgeniy Khramtsov
9a0d77571d Add global static shared_key option 2014-06-01 14:20:09 +04:00
Evgeniy Khramtsov
f446e7fc0b Sign 'Record-Route' in order to proxy unauthorized ACKs 2014-06-01 13:35:14 +04:00
Evgeniy Khramtsov
b75b5ebeb2 Fix 'via' option lookup 2014-05-31 14:22:39 +04:00
Evgeniy Khramtsov
d19903877d Add new option: route 2014-05-31 14:22:31 +04:00
Evgeny Khramtsov
f271ea6eef Merge pull request #229 from weiss/no-carbons-to-sender
XEP-0280: Don't send v1 carbon copies back to the sender
2014-05-31 12:53:18 +04:00
Evgeniy Khramtsov
c76201b6b4 Don't add 'Record-Route' header for mid-dialog requests 2014-05-31 10:00:51 +04:00
Evgeniy Khramtsov
86f2af6fdc Process bindings from multiple UACs correctly 2014-05-31 09:06:53 +04:00
Evgeniy Khramtsov
da22da23cd Don't substitute URI in ACK 2014-05-31 07:50:16 +04:00
Holger Weiss
f45654a16a Simplify mod_carboncopy:check_and_forward/4
Use the existing is_carbon_copy/1 function, and combine multiple case
clauses into a single one.
2014-05-30 23:44:19 +02:00
Holger Weiss
bb952f9ecc Let is_carbon_copy/1 recognize <received/> carbons
The mod_carboncopy:is_carbon_copy/1 function now returns true not only
for <sent/>, but also for <received/> carbon copies.
2014-05-30 23:36:02 +02:00
Holger Weiss
ad2d3964ef Don't send XEP-0280 v1 copies back to sender
An earlier version of XEP-0280 specified the <received/> and <sent/>
tags to be siblings of the <forwarded/> element, whereas the current
version mandates them to be parents of <forwarded/>.  The mod_carboncopy
module supports both variants.  However, the check that makes sure
clients won't receive a copy of the messages they sent didn't work for
the old-style schema.  This is now fixed.
2014-05-30 23:32:18 +02:00
Evgeniy Khramtsov
2cd17c7988 Fix previous commit 2014-05-30 23:49:50 +04:00
Evgeniy Khramtsov
32998f7e18 Process 'Contact' headers more accurately (as per RFC3261) 2014-05-30 23:14:52 +04:00
Evgeny Khramtsov
7261cb29ac Merge pull request #228 from weiss/turn-down-carbons-logging
Log just one [info] message on successful XEP-0280 negotiation
2014-05-29 19:19:05 +04:00
Holger Weiss
8fb1bb1f5f Log just one [info] message on Carbons negotiation
Log one instead of three [info] messages when XEP-0280 (Message Carbons)
support is enabled or disabled successfully.  On failure, log an
additional [warning].
2014-05-29 15:21:11 +02:00
Evgeny Khramtsov
5a29d56d94 Merge pull request #227 from weiss/xep-0198
XEP-0198: Cosmetic change: Reuse event handler
2014-05-28 13:43:29 +04:00
Holger Weiss
0cb9ea3643 XEP-0198: Cosmetic change: Reuse event handler
On stanza queue overflow, pass a message to self() using the exclamation
mark operator instead of send_all_state_event/2.  This allows for
reusing the existing handler for 'kick' events.
2014-05-28 11:24:38 +02:00
Evgeny Khramtsov
7d54fdea51 Merge pull request #206 from weiss/xep-0198
XEP-0198: Terminate session if stanza queue becomes too large
2014-05-28 13:18:12 +04:00
Holger Weiss
99ca8281fa XEP-0198: Terminate session on queue overflow
On queue overflow, terminate the c2s session instead of just dropping
items from the queue.  This makes sure all stanzas are either delivered
or bounced.
2014-05-27 22:56:33 +02:00
Evgeny Khramtsov
3a27b1dd0c Merge pull request #226 from weiss/simplify-state-change
XEP-0198: Cosmetic change: Simplify state change
2014-05-27 23:30:30 +04:00
Evgeny Khramtsov
f9c5e349fb Merge pull request #225 from weiss/use-fsm-next-state
Let ejabberd_c2s always use fsm_next_state/2
2014-05-27 23:27:16 +04:00
Holger Weiss
50a4c5a6ab XEP-0198: Cosmetic change: Simplify state change
When the FSM goes into the 'wait_for_resume' state, let fsm_next_state/2
take care of updating #state.mgmt_state and of writing the log line.
This doesn't change the behavior, but simplifies the code.
2014-05-27 21:14:49 +02:00
Holger Weiss
ed0c89f876 Let ejabberd_c2s always use fsm_next_state/2
Make sure any logic implemented in ejabberd_c2s:fsm_next_state/2 is
always applied.
2014-05-27 21:07:53 +02:00
Evgeny Khramtsov
702cddd4ff Merge pull request #220 from weiss/handle-send-failure
XEP-0198: Don't exit on socket send failure
2014-05-27 15:44:04 +04:00
Evgeny Khramtsov
4d1332c30f Merge pull request #222 from weiss/keep-session-on-failed-resume
XEP-0198: Don't drop session on failed resume
2014-05-27 15:38:26 +04:00
Evgeny Khramtsov
402fb9665d Merge pull request #221 from weiss/omit-redundant-guard
XEP-0198: Cosmetic change: Omit redundant guard
2014-05-27 15:34:34 +04:00
Evgeny Khramtsov
b2e84405c1 Merge pull request #218 from weiss/omit-iq-xmlns
Omit XML namespace declaration for <iq/> stanzas
2014-05-27 15:32:02 +04:00
Evgeniy Khramtsov
52221127cc Fix odbc_port option processing 2014-05-27 15:27:42 +04:00
Evgeniy Khramtsov
dceab3689d Don't forget to include 'Contact' header field in 2xx registrar responses 2014-05-26 21:34:23 +04:00
Christophe Romain
123b01aaa0 don't stop make install if epam is not compiled 2014-05-25 13:02:25 +02:00
Christophe Romain
572938aa49 install epam binary when available 2014-05-25 02:31:28 +02:00
Holger Weiss
59f6efeaf7 XEP-0198: Don't drop session on failed resume
The 'previd' value provided by the client during a session resume
request includes the client's JID and ejabberd's session ID.  If there
is a session for the requested JID but with a different session ID,
resumption should fail, but that session shouldn't be closed.  This
commit makes sure the latter won't happen.

In practice, this will only make a difference in odd corner cases.
2014-05-23 23:38:04 +02:00
Holger Weiss
737b0ae5dc XEP-0198: Cosmetic change: Omit redundant guard
The stream management state is never 'pending' when the c2s FSM is in
the 'session_established' state.
2014-05-23 20:46:17 +02:00
Evgeniy Khramtsov
8925975c86 Fix proxying of ACK requests for 2xx responses 2014-05-23 20:14:53 +04:00
Holger Weiss
ab9667f917 XEP-0198: Don't exit on socket send failure
If stream management is enabled, don't exit the c2s process when
ejabberd_socket:send/2 fails, but close the socket instead.  This gives
the client a chance to resume the session.

Thanks go to Matthias Rieber for reporting the issue, providing detailed
logs, and testing the fix.
2014-05-23 11:38:54 +02:00
Badlop
6baf3a24de Merge branch 'weiss-check-packet-type' into 3 2014-05-21 18:45:43 +02:00
Badlop
735bd95659 Merge branch 'check-packet-type' of git://github.com/weiss/ejabberd into weiss-check-packet-type
Conflicts:
	src/ejabberd_c2s.erl
2014-05-21 18:45:28 +02:00
badlop
69abb48c90 Merge pull request #217 from weiss/fix-extauth-cache-usage
Don't use cached passwords if "extauth_cache: 0"
2014-05-21 17:35:54 +02:00
badlop
419a98d45a Merge pull request #216 from lavrin/p1-c2s
Cleanup some pieces of ejabberd_c2s
2014-05-21 17:35:24 +02:00
badlop
0dc8429d16 Merge pull request #215 from weiss/fix-url-markup
Fix URL markup in the guide
2014-05-21 17:33:56 +02:00
badlop
b9210d491a Merge pull request #208 from weiss/update-travis-config
Travis CI: Remove unnecessary configure flag
2014-05-21 17:31:49 +02:00
badlop
5d855f3723 Merge pull request #207 from weiss/xep-0334
Honor XEP-0334: Message Processing Hints
2014-05-21 17:31:22 +02:00
Holger Weiss
6b996061a2 Omit XML namespace declaration for <iq/> stanzas
Only the child elements of <iq/> stanzas are qualified by the namespaces
in question, not the <iq/> stanzas themselves.

This change just clarifies the code.  It doesn't alter the behaviour, as
those <iq/> stanzas are handed over to jlib:iq_to_xml/1, and that
function ignores the 'xmlns' attribute anyway.
2014-05-21 00:07:57 +02:00
Holger Weiss
fca640f50f Don't use cached passwords if "extauth_cache: 0"
Regarding "extauth_cache", the guide says: "The integer 0 (zero) enables
caching for statistics, but doesn't use that cached information to
authenticate users."  Make sure the cached password isn't used even if
the user is currently logged in with another resource.
2014-05-20 23:00:28 +02:00
Badlop
5010cea1a4 If log uses file:write, no need to double escape ~ in messages (EJAB-1696) 2014-05-20 14:49:52 +02:00
Radosław Szymczyszyn
5726636053 Fix check_from/2 formatting 2014-05-20 12:31:28 +01:00
Radosław Szymczyszyn
b7a542e074 Sanitize copy-pasted get_statustag/1 2014-05-20 12:28:14 +01:00