20531a097a
Added a "nation" parameter to v18_init(), in preparation for automoding.
2013-07-03 02:39:09 +08:00
ba96b391d3
fix trivial compiler error
2013-07-02 12:41:15 -05:00
134871872c
fix codec negotiation issue with the same codec at multiple rates
2013-07-02 12:16:54 -05:00
a0043b8ca2
partial rollback of ec67542628 to fix skinny breakage
2013-07-01 22:19:40 -05:00
c2c8fba14a
--resolve FS-5566
...
When you use $12345 in regex substitutions, it isn't obvious
whenever you mean $1-concatenated-2345 or $12-concatenated-345
or any other variation. In all other languages, in order to
solve this ambiguity, a braces {} are allowed to be used to
separate variable name (or a reference) from surrounding text,
like ${1}2345 or ${12}345. Use the same for freeswitch too.
While at it, fix a buffer overflow as well: the index[] variable
which is used to copy the "variable" name is 10 chars long, but
it is used in the code without bounds checking, so a reference
which is >9 chars long ($1234567890) will overflow the buffer,
crashing freeswitch.
And another overflow is in the way how size of the "substituted"
variable is handled. First, in the outer loop, we compare the
wrong variable with the size of `substituted' buffer (amount of
bytes we took from the source instead of amount of bytes we
used in `substituted'). And second, when actual regex match
is being substitured, amount of room in `substituted' variable
is not checked at all.
Patch contributed by Michael Tokarev <mjt@tls.msk.ru>
2013-07-01 17:05:07 -05:00
8bcff4ca4d
fix input event checker pre-empting return value from dtmf checker when both are present
2013-07-01 14:31:43 -05:00
e1ed386bb9
FS-5565 --resolve
2013-07-01 12:16:47 -05:00
bbdd77ec35
fix race condition on device state destroy
2013-06-28 14:21:04 -05:00
dac93d7936
FS-5527 --resolve
2013-06-28 10:42:06 -05:00
72aad68965
mod_rayo: dtmf #*A-D was broken
2013-06-28 11:16:06 -04:00
a8e0cbfcde
punt without return
2013-06-28 09:48:49 -05:00
2ea31f5d95
add preAnswer and ringReady too
2013-06-27 16:31:29 -05:00
41c1c1cb75
add answer work tag to httapi
2013-06-27 16:25:59 -05:00
cd56d774a0
parse error response for code to only change roles on 487
2013-06-27 13:00:20 -05:00
5fefae03fd
improve skinny behavior in HA deployment
2013-06-26 21:56:10 -05:00
375c102ae0
typo
2013-06-26 21:55:38 -05:00
3879a85bfb
set rtp params in skinny for rtp flush on bridge
2013-06-26 18:35:06 -05:00
7eb86bba75
fix empty +flags{} pre-empting conference_member_flags variable
2013-06-26 16:04:38 -05:00
384cf3381c
use the correct repeat key to replay messages
2013-06-26 15:17:56 -05:00
c903934628
FS-5550 --resolve
2013-06-26 10:51:44 -05:00
bf5fa172e5
FS-5547 --resolve
2013-06-26 10:47:40 -05:00
13dacdcde4
FS-5546 --resolve oddly that is the thing I was trying to fix in the first place
2013-06-26 09:17:15 -05:00
ecc4b52d4a
fix typo with some flags related to jingle
2013-06-25 14:54:34 -05:00
628b4048d9
mod_rayo: added missing modeline
2013-06-25 15:15:08 -04:00
32adc789f6
make noexpandtab explicit in all vim modelines other than xml files
2013-06-25 11:50:17 -05:00
13fdd77f7e
mod_rayo: used wrong stack
2013-06-25 11:40:38 -04:00
f750bcbcc1
mod_rayo: fixes for <prompt> and reply loops
2013-06-25 11:30:10 -04:00
7625001e49
mod_rayo: fix component JID
2013-06-25 08:08:56 -04:00
899f68759a
mod_rayo: fix cleanup of console complete functions
2013-06-25 07:54:59 -04:00
ad5ea44aeb
mod_rayo: don't send item-not-found to replies
2013-06-25 07:42:34 -04:00
cac8536170
mod_http_cache: valgrind fixes
2013-06-24 22:19:54 -04:00
a29bab368d
mod_ssml: valgrind fixes
2013-06-24 22:19:54 -04:00
8bf40819df
mod_rayo: valgrind fixes
2013-06-24 22:19:54 -04:00
e52f131460
mod_rayo: forgot to return error if unknown destination for stanza
2013-06-24 22:19:53 -04:00
c15c0924d1
mod_rayo: added message delivery threads
2013-06-24 22:19:53 -04:00
0f0f278124
mod_rayo: reworked internal messaging- all messages sent by JID only
2013-06-24 22:19:53 -04:00
911948c87a
mod_rayo: remove reply from send function - everything is send only now
2013-06-24 22:19:53 -04:00
1ed5e3e331
FS-5530 --resolve this will make it identical to the bridge behaviour even when the bridge in early media begins and now you will explicitly need continue_on_fail=true in this situation to not fail
2013-06-24 09:45:43 -05:00
6b58aea501
FS-5539 --resolve
2013-06-24 09:15:01 -05:00
f59690c02f
skinny more logging improvements and some refactoring of protocol replies
2013-06-23 21:47:58 -05:00
9279113a26
skinny - add simple expansion of text strings by id in the logging msgs
2013-06-23 16:11:54 -05:00
86e4e580b1
logging
2013-06-23 15:18:10 -05:00
4b0bae5afd
fix broken build due to running with warnings disabled, oops
2013-06-23 15:11:31 -05:00
4a267627f2
skinny logging
2013-06-23 14:58:25 -05:00
0d4967b718
fix windows build
2013-06-23 14:34:26 -05:00
e744bf0385
improvements to skinny logging, table based message text ids
2013-06-23 14:19:05 -05:00
a38ea77fa3
misc constant name adjustments to match actual text on phones
2013-06-22 13:46:26 -05:00
6faae5b178
mod_rayo: Punchblock now supports content-type on input matches
...
Fix applied to Punchblock here: cef5198e80
2013-06-22 14:27:28 -04:00
f8122e39c5
fix trivial compiler warning
2013-06-21 21:43:22 -05:00
48e1415f64
fix return values from refactor so stun timeouts can trigger error properly
2013-06-21 12:42:37 -05:00
Steve Underwood