louiz’
b4a01758b0
And this one (hostname -> address)
...
ref #3273
2018-03-12 01:10:54 +01:00
louiz’
b3b90139db
Rename a few “hostname” into “address”, regarding last commit
...
ref #3273
2018-03-12 01:06:15 +01:00
louiz’
bb476f4a0d
Allow to override the addresses used to connect to an IRC network
...
fix #3273
2018-03-12 01:00:17 +01:00
louiz’
9e4a3e2b05
Fix the INSERT query for types that don’t have any Id column
2018-03-12 00:17:03 +01:00
louiz’
bb596582bd
Add a <item/> node in the presence of a leaving participant
...
fix #3339
2018-03-12 00:13:59 +01:00
louiz’
158d743bf5
Remove the virtual channel feature altogether
2018-03-12 00:13:59 +01:00
louiz’
8503d1f683
Actually display the error message from postgresql when a query failed
2018-02-16 00:58:49 +01:00
louiz’
b175084300
Log the warn/error messages coming from libpq in a correct format
2018-02-16 00:57:49 +01:00
louiz’
d3e07eee13
Sort archive messages by date (and only then by id)
...
fix #3337
Because apparently the IDs are not always incrementing. We still use them to
know the order of two messages that were received at the same second (in
this case, the ID will always be incrementing).
2018-02-16 00:56:07 +01:00
louiz’
dbb86bcc12
Fix the actual_bind versions for integrals
2018-02-15 16:54:54 +01:00
louiz’
505deda176
Fix a memory leak when psql connection fails
2018-02-14 23:25:12 +01:00
louiz’
4a2a280d76
Support the <before/> element in MAM requests
2018-02-13 03:51:03 +01:00
louiz’
50d3c4a0b2
Do not forget an early return, to return the correct item-not-found error
2018-02-13 03:39:13 +01:00
louiz’
fd7c365288
Use the same function for both history orders
2018-02-13 03:10:00 +01:00
louiz’
fd9c713938
Fix a few warnings
2018-02-11 23:39:42 +01:00
louiz’
d705541435
Send a item-not-found error when the “after” value is not in the archive
2018-02-11 23:29:58 +01:00
louiz’
0280343ced
Handle the “after” RSM value to page through results
2018-02-10 19:51:59 +01:00
louiz’
99389eefba
Always return the oldest matching messages from MAM, even if no date is set
2018-02-10 17:47:42 +01:00
louiz’
1ebd8a2321
Fix conversion warnings on 32 bits
2018-02-04 16:58:04 +01:00
louiz’
d7cf736adb
Fix argument types in declaration of send_history_message
2018-02-04 11:02:56 +01:00
louiz’
369da19d41
Remove the debug logs from previous commit
2018-01-30 09:16:12 +01:00
louiz’
58fd689166
Add a ifndef USE_DATABASE guard around send_history_message
2018-01-30 09:14:26 +01:00
louiz’
cb83178894
Add the complete='true' attribute only when appropriate
2018-01-30 09:07:21 +01:00
louiz’
17f2cb5c93
Do not forget the complete='true' attribute in MAM’s result iq
2018-01-29 21:11:20 +01:00
louiz’
7403c397ed
Correctly include the database header
2018-01-28 14:20:21 +01:00
louiz’
442f46c4ac
And an other conversion warning
2018-01-28 14:15:34 +01:00
louiz’
9bf81a2c04
This should fix the int conversion warning on 32bits arch
2018-01-28 14:10:14 +01:00
louiz’
e267512ad4
Restore the is_one_of variable template
2018-01-25 02:22:19 +01:00
louiz’
3f85137265
Merge tag '7.2'
2018-01-25 02:10:17 +01:00
louiz’
33a5f1355d
Remove a variable template usage
...
Because it’s only supported in gcc>=5.0
2018-01-25 01:22:35 +01:00
louiz’
de73a728f2
Merge tag '7.1'
2018-01-22 21:06:43 +01:00
louiz’
06271729e3
Fix a crash happening when a user cancels a non-existing ad-hoc session
2018-01-22 20:57:34 +01:00
louiz’
f371d9ca46
xep-0106 escape the JIDs listed in a disco#items server query
...
fix #3325
2018-01-18 19:34:56 +01:00
louiz’
520711023f
Use our botan gh#1276 workaround only for botan < 2.4
...
fix #3320
2018-01-17 20:59:38 +01:00
louiz’
7e64a2e361
Add a DEBUG_SQL_QUERIES to log info about the executed SQL queries
...
fix #3324
2018-01-14 21:46:49 +01:00
louiz’
3cd6490236
Remove all the ugly database debug
2018-01-14 21:44:22 +01:00
louiz’
b4b9828de5
Follow log_level even when using journald output
...
fix #3328
2018-01-12 20:08:15 +01:00
louiz’
7c8a7176d1
Only use sd_journal_* if we really are outputing to journald
...
We check that the device and inode numbers are actually the same as the
JOURNAL_STREAM value, instead of just checking that the value exists.
This fixes the logger unit tests
2018-01-12 04:03:36 +01:00
Ailin Nemui
435a63a070
Change max line length to more conservative constants
2018-01-10 13:36:53 +01:00
louiz’
8409e50c1a
Fix a subtle iid parsing error in the adhoc code
2017-12-27 19:36:36 +01:00
louiz’
2d9f516d1d
Don’t answer to some requests towards MUC participants
...
These requests are only meant to be received by the room itself. The
participant must answer with not-implemented instead.
fix #3323
2017-12-27 19:23:28 +01:00
louiz’
131ef9946f
Include the <fin><set><first/><last/></></> nodes in the MAM iq result
...
fix #3322
2017-12-26 20:21:18 +01:00
louiz’
37340e593f
Add a <x muc#user> node on outgoing private MUC messages
...
See https://xmpp.org/extensions/xep-0045.html#privatemessage
fix #3321
2017-12-26 19:52:41 +01:00
louiz’
286aa58c7b
Remove a warning (unused argument)
2017-12-17 15:21:40 +01:00
louiz’
a557718bf2
db_name also accepts postgres:// scheme for PostgreSQL connections
2017-12-16 15:34:54 +01:00
louiz’
c3313d0d41
Always free the PGresult pointer returned by PQexec
...
Fix a somewhat big memory leak
2017-12-05 01:05:45 +01:00
louiz’
24dc05dd97
Run some of the ci tests against a postgresql docker container
2017-12-03 16:46:14 +01:00
louiz’
6dc49f3284
Fix a few warnings
2017-12-02 20:08:50 +01:00
louiz’
414bbca0bc
Handle postgresql and sqlite3 libs properly
...
Do not fail to compile when one of them is missing but the other one is not.
Raise an error when trying to open a database with the missing library.
see #3237
2017-12-02 20:08:50 +01:00
Jonas Wielicki
e1c7a6518a
Actually free the database on Database::close()
...
.release() returns the pointer and releases ownership *without*
destruction.
2017-12-02 20:08:50 +01:00
louiz’