* commit '0.8.9-13-g39d32e0':
Changelog for previous PR
DOC: Changelog entry fro preceeding merge from Terence
TST: Fix fail2ban.conf reader test for unreliable dictionary order
failregex when roundcube log driver is set to 'syslog'
fixed failregex line for roundcube 0.9+
TST: test all stock jails to have actions and correctly specifying blocktype
CFG: assure actions for all the jails
BF: blocktype must be defined within [Init] -- adding [Init] section. Close#232
ENH: since it seems the default is to use file based logging, $syslog is in Should-{Start|Stop} like Debian https://github.com/fail2ban/fail2ban/blob/debian/debian/fail2ban.init
ENH: opensuse script from opensuse: https://build.opensuse.org/package/view_file?expand=1&file=fail2ban.init&package=fail2ban&project=openSUSE%3AFactory
Conflicts:
ChangeLog
config/jail.conf
testcases/clientreadertestcase.py -- had to "git show XXX | patch -p2" under tests/ 2 commits: 8a57ffd7a4db4b
* origin/0.9:
TST: Fix TraceBack tests when running via "setup"
Conflicts:
fail2ban/tests/misctestcase.py -- choose my version since it also could
be ran 'nosetests' or probably even some other tester, so better just to check
if it is a straight invocation of the fail2ban-tests
* 'sshd-locked' of https://github.com/kwirk/fail2ban:
DOC: Update man page for <SKIPLINES> usage
ENH: Add new regex for locked accounts for sshd
BF: fail2ban-regex adding duplicate lines with each regex
This issue was caused by the fact that every "line" is processed for each
regex, meaning each line was duplicated for every regex. This caused
duplicate fail matches and the buffer filling too quickly and possibly
missing failures.
# Only works only if log driver: is set to 'syslog'. this is becoz fail2ban fails to 'read' the line due to the
brackets around the date timestamp on logline when log driver is set to file
Previous fix attempted shlex split which whilst worked for reading from
config file, failed when using fail2ban-client, as the input is already
effectively shelx split by the executing shell.
FilterSystemd journal match methods now handle list structures which
should be shlex split when reading from config file, and simply pass all
the relevant arguments from the shell when using fail2ban-client
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iEYEABECAAYFAlGRBZ8ACgkQjRFFY3XAJMhqzwCgvUsrv6cSjo1d8YCQUA8Na0Kk
44QAoKk7X2sqFM+wvj2vK3stsHa/80qm
=iBfR
-----END PGP SIGNATURE-----
Merge tag '0.8.9' into 0.9 (quite a bit of conflicts "resolved")
Release 0.8.9
* tag '0.8.9':
BF: add missing files to MANIFEST (I think we shoult not rely on sdist anyways -- 'git tag' tarballs are more thorough ;) )
All the (version) updates for the release of 0.8.9
BF: (travis) relax the test for needed to be presented installed directories -- allow new
BF: (travis) if tests ran under coverage -- there is a traceback parts to report (thus > would be present)
ENH: also print the failing traceback line in case of failure
ENH: include explicit list of new files which should not be there upon "install --root"
ENH: now we know that logging handlers closing was still buggy in 2.6.2
ENH: issue a warning if jail name is longer than 19 symbols (Close#222)
DOC: inline commends with ';' are in effect only if ';' follows as space
BF: Fix for filterpoll incorrectly checking for jailless state
ENH: strengthen detection of working pyinotify
ENH: use the same python executable for setup.py test
ENH: actually tune up TraceBack to determine "unittest" portions of the stack across all python releases
TST: Some primarily smoke tests for tests utils
TST: cover few more lines in fail2banreader.py
ENH: basic test for setup.py itself (when applicable, should greatly improve coverage ;) )
ENH: consistent operation of formatExceptionInfo + unittest for it
ENH: point to the status of master branch on travis
Conflicts:
ChangeLog
MANIFEST
README.md
fail2ban/version.py -- all of the above obvious version changes
below files primarily needed just a bit of help in resolution
config/jail.conf
fail2ban/server/filterpoll.py
fail2ban/server/server.py
fail2ban/tests/servertestcase.py
and following were more difficult -- git wasn't able to track renames/moves of the code
fail2ban-testcases -- needed to introduce those changes to tests/utils.py
testcases/clientreadertestcase.py -- manually applied patch from master
testcases/utils.py -- manually applied patch from master
Daniel Black