Fetching contributors…
Cannot retrieve contributors at this time
1117 lines (1023 sloc) 56.8 KB
Revision history for Log::Log4perl
1.39 (2012/10/27)
* (ms) Markus Benning reported that logcroak/confess/die stringify
their arguments, which gets in the way when throwing data
structures as exceptions. Added flag $STRINGIFY_DIE_MESSAGE.
* (ms) [ #80086] David Morel pointed out misleading
documentation in the Limiter composite appender. Fixed docs
and added C<appender_method_on_flush> parameter to support
appender flush calls by the limiter.
* (ms) [ #79960] Fabrice Gabolde asked for %X{x} to be
interpolated as NULL for the database appender if its value is
undef. The DBI appender now inits PatternLayout with the
undef_column_name parameter set to undef (defaults to "[undef]").
* (ms) Updated license/copyright/author sections in all files using
licensizer and .licensizer.yml.
* (ms) Skipped log file recreation test on Win32 as it won't remove
busy files anyway.
1.38 (2012/09/23)
* (ms) Bob Kleemann reported that logdie() in wrapper classes
printed incorrect caller lines. Fixed by adding
caller_depth_offset() utility to
* (ms) Meir Guttman reported a use case for logging messages
with dynamic levels on
Added to manual page.
* (ms) Implemented suggestion by Neil Hooey to check for and report
undefined message elements before they're assembled in
and a warning is issued without a proper code location:
* (ms) [ #78734] Added spell check on filter parameter names
1.37 (2012/05/30)
* (ms) [ #75655] Meir Guttman found the module to make
Log::Log4perl::Appender::ScreenColoredLevels work on Win32,
updated docs.
* (ms) [ #76827] UTF-8 encoded configuration files are
now supported (see Log::Log4perl::Config).
* (ms) [ #77501] Unescaped left brace in regex is deprecated
with perl 5.17. Times we live in.
1.36 (2012/02/21)
* (ms) [ #74833] Reini Urban fixed "defined @array" for
perl 5.16
* (ms) [ #74836] Cope with Carp's questionable decision to
add a trailing dot to its messages.
1.35 (2012/01/03)
* (ms) [ #73462] Changed logwarn/logcluck/logcarp/error_warn
to warn() unconditionally and send the message to log4perl which
will log it only if the log level conditions are met.
* (ms) [ #73598] Gerda Shank reported test suite problems
with DBD::CSV-0.26. Bumped up to DBD::CSV-0.33 if installed.
1.34 (2011/11/04)
* (ms) InternalDebug now replaces all instances of INTERNAL_DEBUG,
not just the first one.
* (ms) Added test case for get_logger() with a ref() on the actual
object instead of on a static category. Updated docs.
* (ms) %d{e} in PatternLayout now returns epoch seconds
* (ms) [RT 72056] Messages blocked by an appender threshold are no
longer passed on to the L4p::Appender::Buffer as undefined
1.33 (2011/05/31)
* (ms) [RT 67132] Applied patch by Darin McBride to allow for
empty syswrite messages in the file appender.
* (ms) [RT 68105] Fixed init-hash handling of subroutine references,
reported by Frew Schmidt.
* (ms) Mike Schwern noticed confusing DESTROY calls to clean up loggers
and appenders ( and, so I put on my
hazmat suit and cleaned it up. Now perl's garbage collector takes
care of disposing of logger and appender carcasses.
* (ms) Added Log::Log4perl->remove_logger($logger) to remove a logger
from the system.
1.32 (2011/02/26)
* (ms) Fixed %T caller_depth with wrapper_register(), reported
by David Christensen.
* (ms) [RT 63053] Fixed for qw() {} deprecated (Todd Rinaldo)
* (ms) [RT 62674] Fixed call to deprecated form of UNIVERSAL::can (Karen
* (ms) [RT 62896] Log::Log4perl::Appender::ScreenColoredLevels now
inherits from Log::Log4perl::Appender::Screen and therefore
supports the utf8 flag.
* (ms) [RT 64318] Andrew Sayers provided a better error message for
"threshold needs to be uppercase".
* (ms) CharleyDixon fixed LOGWARN when :no_extra_logdie_message is
in use to no longer exit().
1.31 (2010/10/27)
* (ms) Fixed the number of skipped tests for Windows for previous fix
of [RT 60665].
1.30 (2010/08/30)
* (ms) [RT 60665] HUP handlers are stacked on top of each other now,
to make sure that multiple file appenders recreate multiple
files and not just one (patch provided by Karen Etheridge).
* (ms) [RT 60197] Fixed uninitialized value warnings with
the multiline appender and provided a test case (patch provided
by Karen Etheridge)
* (ms) [ #59617] Fixed system-wide threshold without appender
thresholds. Bug reported by Dmitry Bigunyak.
* (ms) [ #24884] Using require() instead of incomplete
logic in L4p::Util::module_available(). local __DIE__
handler takes care of user-defined __DIE__ handlers
ignoring $^S (suggested by Eric Wilhelm and others).
* (ms) [ #60386] Fixed init_and_watch() which
double-bumped the caller_level and led to uninitialized
values in the pattern layout. Thanks to Mitja Bartsch for
the report.
* (ms) Applied patch by Karsten Silkenbäumer to add an optional
$log_dispatch_level to create_custom_level(). Updated
1.29 (2010/06/16)
* (ms) Added documentation on how to use Log4perl's :easy macros
with Catalyst in Log::Log4perl::Catalyst.
* (ms) wrapper_register() now deals with caller_depth automatically.
Backwards compatibility with old wrapper classes using
caller_depth directly is provided. Documentation has been
* (ms) Felix Antonius Wilhelm Ostmann reported
crashes, fixed as suggested by setting the %INC value to
the module path.
* (ms) Another caller_depth fix in Log::Log4perl::Catalyst.
* (ms) Fixed logdie() caller_depth bug reported by Rob Retter.
* (ms) [RT 56145] Saving errstr in DBI appender to survive ping()
* (ms) Added INTERNAL_DEBUG env variable to test suite triggering
all _INTERNAL_DEBUG statements to be printed for better
error diagnosis on misbehaving systems.
1.28 (2010/02/24)
* (ms) Fixed caller stack with Buffer composite appender
* (ms) Fixed 'local caller_depth' error in various places. First
localizing a variable and then increasing it is incorrect,
as this ignores previous settings. The correct way of
increasing the caller level is: 'local depth = depth + 1'.
* (ms) Added Log::Log4perl::Catalyst for use in Catalyst applications.
1.27 (2010/02/07)
* (ms) ***WARNING: This might break backward compatibility
with some wrapper classes.
[RT 52913] Fixed category fetching in
wrapper classes (reported by Martin Evans). Wrapper classes
now need to call Log::Log4perl->wrapper_register to adapt
get_logger() category fetching. Detailed docs under
"Using Log::Log4perl with wrapper functions and classes"
* (ms) Made meta tag compatible with MakeMaker versions < 6.50
(ms) [RT 52083] Fixed manifest glitch from 1.26 (reported by
Lars Thegler).
* (ms) Added note to FAQ on 'no init happened' warnings for API
initializations, as suggested by Malcolm Nooning.
* (ms) Applied patch by Christopher Mckay which sets
Log4perl::Logger::INITIALIZED only if it's fully initialized.
* (ms) Emmanuel Rodriguez suggested changing TestBuffer's reset()
method to leave the logger population alone. Added clear()
to accomodate the need for a single buffer reset.
* (ms) Xavier Caron added %p{1} to allow abbreviated priority
strings in the pattern layout.
* (ms) Redid composite appenders to address problems with incorrect
caller() data. L4p now supports a $cache parameter to be
passed to the log() function, which stores the completely
rendered message and can be passed to log_cached() later on.
1.26 (2009/11/22)
* (ms) [RT 50495] Perl code in the config file is now evaluated/
compiled after the configuration parser has done its work,
opening up Perl subroutines to all configuration parsers, not
just PropertyConfigurator. Configuration subs for cspecs,
filter, warp_message and appender triggers are sheltered.
The previous, flawed implementation surfaced while using
a 'trigger' category, reported by Olivier Bilodeau.
* (ms) [RT 50090] Added non-portable linebreaks to PatternLayout
(requested by Zdeněk Juran).
* (ms) [RT 50094] Docfix for PatternLayout in main manpage (spotted
by Peter Rabbitson).
* (ms) [RT 28679] Added exists() to "Threshold" keyword uppercase
* (ms) Took out Class::Prototyped testcase after it got all weird
and introduced backward-incompatible changes.
1.25 (2009/09/27)
* (ms) Appender::File is now closing (or sysclosing) the file
on file_close() instead of just undef'ing the handle.
* (ms) Added l4p-tmpl helper script to help whipping up a new
log4perl configuration file.
* (ms) Fixed uninitialized warning on XML configuration files,
reported by jbkilian on the sourceforge mailing list.
* (ms) Applied patch [RT 43426] by AFF <> to have
appender_thresholds_adjust return number of appenders changed.
* (ms) [RT 34400] New :nostrict target which allows redefining a
category within a Log4perl configuration file without error
or even a warning.
* (ms) [RT 34401] Applied patch by Jae Gangemi, who fixed
code references in @INC on Win32 systems.
* (ms) [RT 32259] Patternlayout now supports %R, which returns the
number of milliseconds elapsed from last logging event to
the current logging event (thanks to Emmanuel Rodriguez for
the patch).
* (ms) [RT 30899] Color configuration and attribute support
added to ScreenColoredLevels appender by Jason Kohles.
* (ms) [RT 28987] If UNIVERSAL is available, appender existence is
now verified by checking can() on the appender's new()
method (applied modified patch by Gabriel Berriz).
1.24 (2009/07/08)
* (ms) Fixed bug with Log::Log4perl::Util::tmpfile_name which
surfaced on VMS, reported by Ben Humphreys.
* (ms) Fixed system-wide threshold to no longer lower appender
thresholds. Bug reported by Jean-Denis Muys.
* (ms) Added benchmark to determine impact of eval-free handlers
* (ms) Merged with eval_free branch. Now there are no more
eval("") statements left in the code, making it much easier
to debug. Performance on init() is about the same, performance
on init_and_watch() (noops and logged statements alike) is
25% slower but still in the range of 400,000/sec on my
1.80Ghz CPU.
1.23 (2009/05/12)
* (ms) DBI-1.608 removed a DBD::File 'feature' that allowed leaving
out parameters in a bound execute(). This caused the test
suite to fail ( Fixed the test cases to pass the correct
number of parameters every time.
* (ms) Better error message in the DBI appender on bad SQL, missing
bind parameters, or other execute() errors.
* (ms) Made DBI test suite more robust against preexisting conditions
* (ms) Added force_next_check() for init_and_watch(), cleaned up
Config::Watcher code.
* (ms) Fixed test suite to run on Strawberry Perl on Win32 (reported
by kmx on
* (ms) Added 'utf8' option to screen appender and easy mode, some of
it suggested in by
Shantanu Bhadoria.
1.22 (2009/05/02)
* (ms) is_xxx() returned true prior to L4p initialization. Fixed it
and adapted test suite.
* (ms) Added test cases on syswrite in recreate mode
* (ms) Applied patch by Jens Berthold <> to
avoid semaphore cleanup in spawned children.
* (ms) Added %m{chomp} feature, 'message_chomp_before_newline' option,
and documentation on newlines and logging messages, all
suggested by Tim Bunce (see PatternLayout).
1.21 (2009/03/16)
* (ms) Documentation typos fixed, reported by Breno G. de Oliveira
[ #42428].
* (ms) Fixed DBI appender error message, bug reported by DavidZ.
* (ms) Fixed [ #43740] reported by Martin Koehler. Now using
proper POSIX return code EEXISTS instead of error message
depending on English locale.
1.20 (2008/12/09)
* (ms) Using semctl to reset the value of the semaphore in the
Synchronized appender to prevent "Numerical result out of
range" problem caused by an unbalanced SEM_UNDO when
incrementing it. Reported by John Little.
* (ms) Added parameters in curly braces to cspecs in PatternLayout.
* (ms) As explained in
the latest LWP release (5.822) got rid of all of its internal
debugging functions, making infiltrate_lwp() and its test
case useless. Disabling it for LWP>=5.822.
1.19 (2008/10/22)
* (ms) Applied patch by Peter Rabbitson, which fixes the caller()
level when calling get_logger() on a subclass of Log4perl.
* (ms) Added documentation on is_xxx() methods and clarified that
it doesn't necessarily mean that a message gets logged if
they're returning true (requested by Conway Allen via
[ #39085].
* (ms) Applied patch by Lee Johnson to appender_by_name() to allow
for undefined appender names without issuing a warning,
which was occurring with Catalyst::Log4perl.
* (ms) Added docs on numerical levels and level strings in
* (ms) Applied patch by Anthony Foiani for support of literal
text in DateFormat format strings.
1.18 (2008/08/23)
* (ms) Added explanation that categories and loggers are the same
thing (thanks to Rabbit).
* (ms) Fixed t/053Resurrect to work with 5.005_03
* (ms) Added preinit_callback function for init_and_watch()
* (ms) Applied patch by Andy Grundman which speeds up is_LEVEL()
calls by skipping unnecessary string concatenations
* (ms) Applied patch by Jae Gangemi addding a no_warning option to the
socket appender
1.17 (2008/07/19)
* (ms) Fixed test suite to run on Strawberry Perl on Win32.
* (ms) Added 'l4p' as a valid prefix in configuration files (equal
to 'log4j' and 'log4perl' now).
1.16 (2008/05/15)
* (ms) Changed appender destruction during cleanup to show warning
messages thrown by destructors. Previously L4p ignored these
messages which caused failed DB flushes to go unnoticed with
the DB appender.
* (ms) Added explanation for Log4perl messages during global
destruction to FAQ.
* (ms) Corrected 'Trapper' listing in FAQ, thanks to Christian Reiber.
* (ms) Applied patch by Mitchell Perilstein for 5.005it and the
two-argument binmode() that 5.005 doesn't support.
* (ms) Applied patch by Emmanuel Rodriguez (POTYL) doing away with
hard-coded line numbers in 024WarnDieCarp.t to make it work
cpan2rpm for building RPM packages
* (ms) Fixed recreate_check_interval = 0 bug reported by
Bill Moseley.
* (ms) Added 'header_text' parameter to the file appender to
have it write a header every time it opens (or re-opens)
a new log file (suggested by Steven Lembark).
1.15 (2008/02/10)
* (ms) appender_thresholds_adjust() with a parameter of 0 now
does nothing (requested by Oliver Koch).
* (kg) Added 'defer_connection' to Socket appender so it's more useful
under Apache.
* (ms) [ #32738] fixed caller_depth for error_warn()
(reported by Felix Antonius Wilhelm Ostmann)
* (ms) [ #32942] fixed get_logger() for subclassed Log4perl
(reported by Felix Antonius Wilhelm Ostmann)
1.14 (2007/11/18)
* (ms) Fixed test suite bug which surfaced in Darwin because temporary
files contain '++' which freaked out the sloppy regex match.
* (ms) Better handling of empty config files (reported by Robert Raisch)
* (ms) Rewrote the Synchronized appender to use semaphores exclusivly
(got rid of IPC::Shareable).
* (ms) Added Log::Log4perl::Util::Semaphore for easy semop handling
* (ms) Fixed t/026FileApp.t to work on MSWin32.
1.13 (2007/10/11)
* (ms) Another doc fix by Craig
* (ms) Applied Fedora 7 patches
* (ms) Added create_at_logtime option to file appender
* (ms) Added trace level color (yellow) in ScreenColoredLevels
appender as suggested by Arvind Jayaprakash in
1.12 (2007/06/23)
* (ms) Added Log::Log4perl::Resurrector to resurrect commented-out
Log4perl statements in all subsequently loaded modules (allows
for deploying L4p-enabled CPAN modules without requiring L4p).
* (ms) Added ALWAYS easy mode macro (level=OFF)
* (ms) Fixed logconfess() frame level bug reported by Ali Mesdaq.
Added test case.
1.11 (2007/05/29)
* (ms) Added PatternLayout::Multiline code by Cory Bennett to
render multiline messages.
* (ms) Added log level TRACE (lets through even more messages
than DEBUG) (suggested by Craig).
* (ms) Added 'syswrite' flag to file appender to have it use
'syswrite' instead of 'print', avoiding buffered or
interleaving messages originating from different processes
(thanks to Evan Miller).
1.10 (2007/03/27)
* (kg) Nikita Dedik pointed out that Saturday is missing from
* (ms) Scott Cline noticed a potential problem with the DBI
appender reconnection logic in 'buffered' mode. Applied
a patch.
* (ms) Changed DBI reconnect logic to perform even if the DB
is pingable again.
* (ms) Applied code by Valerio Valdez Paolini with modifications
to to allow pulling values from
the property configurator by path.
1.09 (2007/02/07)
* (ms) Added $^S check to FAQ, as suggested by J. David Blackstone.
* (ms) Applied Robert Jacobson's patch for the "DDD" formatter
in L4p::DateFormats, which now formats the day-of-year values
numerically and precedes them with zeroes if necessary.
* (ms) Added %M{x} PatternLayout notation as requested by
Ankur Gupta.
* (ms) Another Win32 test suite fix, no longer deleting an open
file but moving it aside (rt.cpan:23520).
1.08 2006/11/18
* (ms) Applied test suite patch by Lars Thegler for
ancient perl 5.005_03.
* (ms) Applied patch by Jeremy Bopp to fix test suite running
under Cygwin.
* (ms) Fixed documentation bug in L4p:Appender::File,
s/recreate_signal/recreate_check_signal. Thanks to
Todd Chapman and Robert Jacobson for reporting this.
* (ms) Fixed init(), which now deletes any config file watchers
left over from previous init_and_watch() calls. Reported
by Andreas Koenig who saw sporadic errors in the test suite,
1.07 2006/10/11
* (ms) Removed checks for unlink() in t/017Watch.t since they
failed on win32.
* (ms) Fixed doc bug in Appender::File reported by Robert
* (ms) Added FAQ on why to use Log4perl and not another
logging system on CPAN.
* (ms) Fixed %M, %L, etc. level in logcarp/cluck/croak/confess
(thanks to Ateeq Altaf)
* (ms) Autocorrecting rootlogger/rootLogger typo
* (ms) Better warning on missing loggers in config sanity check
1.06 2006/07/18
* (ms) Applied patch by Robert Jacobson to fix day-of-year in
DateFormat, which was off by one.
* (ms) Added FAQ on syslog
* (ms) umask values for the file appender are now also accepted
in octal form (0xxx).
* (ms) The file appender now accepts owner/group settings of
newly created log files.
* (ms) Fixed appender cleanup, a bug caused composite appenders
to be cleaned up during global destruction, which caused an
ugly segfault with the Synchronized appender on FreeBSD.
1.05 2006/06/10
* (ms) Added recreate signal handler to L4p::Appender::File for
newsyslog support. Two new FAQ entries on dealing with
newsyslog and log files being removed by external apps.
* (ms) L4p::Config::Watch no longer sets the global $SIGNAL_CAUGHT by
default but uses an instance variable instead to prevent
clobbering L4p's config and watch mechanism.
* (ms) die() on undefined configuration (rt 18103 by
* (ms) Hugh Esco submitted a FAQ on where to put logfiles
* (ms) Applied patch provided by Chia-liang Kao to suppress an error
message and skip tests in the suite when DBI is missing.
1.04 2006/02/26
* (ms) Duplicate log4perl directives, which previously just overwrote
existing ones, are no longer permitted and cause the config
parser to throw an error.
* (ms) If a conversion pattern was specified twice in a config
file, the output was "ARRAY(0x804da00)" (bug reported by
Bill Mason). Now, gobbling up property configurator values
into an array is limited to appender properties and
excludes the conversion pattern.
* (ms) Multiple calls to import (usually happens if 'use L4p' gets
called twice within the same namespace) caused nasty warnings,
bug reported by Greg Olszewski. Fixed by ignoring subsequent
calls from the same package to import().
* (ms) Changed rendering of logdie/warn/cluck/croak/... messages
to fix a bug reported by Martin J. Evans.
* (ms) Added a L4p::Appender::String appender to handle the
rendering internally.
* (ms) Documentation patch by Matisse Enzer on increased/
decreased log levels.
* (ms) Fixed stack trace level of logcarp()
* (ms) Carl Franks reported that the test suite failed on WinXP SP2
because of a hardcoded /tmp - fixed by File::Spec->tempdir().
* (ms) Added reconnect_attempts and reconnect_sleep parameters to
DBI appender.
* (ms) Bugfix for #17886 (tmp files in test suite)
1.03 (2006/01/30)
* (ms) Some perl-5.6.1 installations have a buggy Skipping
4 test cases for these. Reported by Andy Ford and Matisse Enzer.
* (ms) The DBI appender now reconnects on stale DB connections.
* (ms) Fixed Win32 test bug as reported in by barbie.
Instead of deleting a file still in use by an appender (which
Windows doesn't like), the file gets now truncated.
1.02 (2005/12/10)
* (ms) Adapted t/006Config-Java.t to cope with Win32 path separators
* (ms) Corrected typo in Chainsaw FAQ, reported by Bernd Dirksen.
* (ms) Brian Edwards noticed that (Screen, File) were missing a
base class declaration, causing $logger->add_appender() to
fail. Fixed with test case.
* (ms) Log::Log4perl::Appender::File now handles the case where the
logfile suddenly disappears.
* (ms) Fixed section indentation in main man page
* (ms) Converted Ceki's last name to UTF-8 (a historic step!)
1.01 (09/29/2005)
* (ms) Added 'utf8' and 'binmode' flags to Log::Log4perl::Appender::File
per suggestion by Jonathan Warden.
* (ms) Made test cases 003Layout.t and 033UsrCspec.t resilient against
broken ActiveState 5.8.4 and 5.8.7.
* (ms) Skipped failing test cases for 5.005, looks like the caller() level
in carp() is wrong, but not worth fixing.
* (ms) Fixed the bug with the caller level of the first
log message sent after init_and_watch() detected a change. Added
test case to 027Watch2.t.
* (ms) Added FAQ on UTF-8.
* (ms) Applied patch by David Britton, improving performance during
the init() call.
* (ms) Fixed bug
to prevent it from modifying $_. Thanks to Steffen Winkler.
1.00 (08/13/2005)
* (ms) Added tag qw(:no_extra_logdie_message) to suppress duplicate
die() messages in scripts using simple configurations and LOGDIE().
Added logexit() as an alternative way.
* (ms) Fixed bug with logcarp/croak/cluck, which were using the
wrong Carp level.
* (kg) Fixing bug in Appender::Limit regarding $_ scope
* (ms) corrected typo in found by Rob Redmon.
* (ms) Fixed bug with Appender::File reported by Michael Smith. Checking
now if print() succeeds, catching errors with full disks and
ulimit'ed environments.
in :easy mode (suggested by Jud Dagnall).
* (ms) $INITIALIZED now gets reset during logger cleanup.
0.52 (05/08/2005)
* (ms) Jonathan Manning <> provided a patch
for to fix 3-letter month abbreviations and a
shortcut to simulate Apache's log format.
* (kg) Ola Finsbraaten provided a patch to provide a better error
message when a logger is defined twice in a config.
0.51 (01/08/2005)
* (ms) Jon Bjornstad noticed that the file appender wasn't including
$! in the die() exception thrown if open_file() fails. Added it.
* (ms) Added umask option to file appender
* (ms) Fix to L4p::Util::module::available() for Win32
compliance by Roger Yager <>
* (ms) Added check to L4p::Util::module_available() returning true
if the pm file is available in %INC, indicating that it has
already been loaded. This fixes a problem when running L4p
in a PAR binary.
* (ms) Added remove_appender() and eradicate_appender() method to, test cases and documentation on the main Log4perl
* (ms) Added a generic buffered composite appender, L4p::Appender::Buffer,
buffering messages until a trigger condition is met.
0.50 (12/08/2004)
* (ms) Added ':resurrect' source filter, which uncomments all lines
starting with "###l4p". Can be used for hidden L4p statements,
which are then activated by calling
'use Log::Log4perl qw(:resurrect)'.
* (ms) Fixed Win32 test suite bug: File::Spec->catfile() returns '/'
as a path separator on both Unix and Win32, while Log4perl's
layouts (derived from caller() info) use '\' on Win32 and '/'
on Unix. Changed tests to only verify file name, not path.
* (ms) Added 'appender_by_name()' to retrieve an appender defined
in the configuration file by name later.
* (ms) Added FAQ on "stubbing out" L4p macros in environments
that don't have L4p installed.
* (ms) Added convenience function appender_thresholds_adjust() to adjust
thresholds of chosen (or all) appenders
* (ms) Got rid of Test::Simple dependency
* (ms) Moved autoflush setting in L4p::Appender::File from log()
to file_open(), running only once, not with every message.
* (ms) Applied doc fixes suggested by Jon Bjornstad.
* (ms) Added ScreenANSIColor appender to colorize messages based on
their priority. See Log::Log4perl::Appender::ScreenANSIColor.
0.49 (11/07/2004)
* (ms) init_and_watch() no longer die()s on reloading syntactically
wrong configuration files but issues a warning and then
reloads the last working config.
* (ms) init() now also accepts an open file handle (passed in as a
glob) to a configuration file or a ref to an IO::File object.
* (ms) Jos I. Boumans <> and
Chris Winters <> reported an error thrown
by L4p in their app SPOPS: During global construction. Looks
like the Logger object's internal hash is cleared and then
the is_<level> method gets called, resulting in a runtime
exception. Added proposed remedy checking if the called
method is defined by ref.
* (ms) Added check to init_and_watch if obtaining the mod
timestamp failed.
0.48 (08/20/2004)
* (ms) fixed bug reported by Chip Salzenberg <>: logdie()
and logwarn() are now compliant with the warn() and die()
standard which suppresses the "at file line x" message if
the message ends with a "\n".
* (ms) New interface for custom config parsers.
Log::Log4perl::Config::BaseConfigurator now provides a base class
for new config parsers. Init can now be called like
Log::Log4perl->init($parser) with a parser object, which is
derived from Log::Log4perl::Config::BaseConfigurator and
provides a parse() method (no arguments). The file (or whatever)
to be parsed can be set by calling $parser->text(\@lines) or
$parser->file($name) before calling L4p->init($parser).
The Property, DOM and LDAP configurators have been
adapted, check their implementation for details.
* (ms) Added integrity check for Log4perl configurations: Log4perl
now issues a warning if a configuration doesn't define any
appenders. Should anyone not like this, it can be turned
off by setting $L4p::Config::CONFIG_INTEGRITY_CHECK = 0
before calling init().
* (ms) Fixed bug reported by Johannes Kilian <>
with __DIE__ handler and "PatternLayout" shortcut. Replaced
'eval { require ... }' by L4p::Util::module_available in
* (ms) Did away with $IS_LOADED internal variable.
* (ms) Fixed bug with L4p::INITIALIZED vs. L4P::Logger::INITIALIZED,
added t/020Easy2.t.
* (ms) Added adm/cvskwexp script to check if we're running into CVS
trouble because of <dollar>Log keyword expansion.
0.47 (07/11/2004)
* (ms) Added suggestion by Hutton Davidson <>
to make the socket appender more forgiving. New option
"silent_recovery" will silently ignore errors and recover
if possible on initiallly dead socket connections.
* (ms) Fixed bug with initialized() -- checking once caused
subsequent calls to return true.
* (ms) run t/045Composite.t only if Storable is installed -- earlier
perl versions (like 5.6.1) don't have it by default.
* (ms) fixed test case in t/020Easy.t for buggy perl 5.6.1
* (ms) added Log::Log4perl::infiltrate_lwp() to make LWP::UserAgent
play in the L4p framework upon request.
* (ms) perl 5.00503 mysteriously core dumps in t/017Watch.t, seems like
this was introduced in 0.46. Disabled these tests for now
if we're on 5.00503 to avoid installation hickups. Longer term,
need to investigate.
0.46 (06/13/2004)
* (ms) removed superfluous eval() in, reported anonymously
on the CPAN bugtracker.
* (ms) Added a cleanup() function to which is used by an
END {} block in to tear down all Loggers/Appenders
before global destruction kicks in. In addition, Kevin found
that the eval "" is the cause of an Appender memleak. Moved
assignment variable out of the eval to plug the leak.
Added $Log::Log4perl::CHATTY_DESTROY_METHODS, which shows
what L4p objects are destroyed and when.
* (ms) Kevin's idea is in now, on localizing $? in the L4p global END {}
block. It prevents logdie() et. al from exiting with unwanted
exit codes when global cleanup / global destruction modifies $?,
as seen by Tim with the Email appender.
* (ms) Dave Viner <> added isLevelEnabled() methods
as aliases to is_level().
0.45 (05/23/2004)
* (ms) fix for t/045Composite.t on perl 5.6.1 by Jeff Macdonald
<> (specify number of test cases,
getting rid of no_plan).
* (ms) Dennis Gregorovic <> provided a patch to
protect applications who are tinkering with $/. It is set
to "\n" now locally when L4p is reading the conf file. Added
a test case to t/004Config.t.
* (ms) Fixed a documentation error with initialized(), pointed
out by Victor Felix <>.
0.44 (04/25/2004)
* (ms) added filename() method to L4P::Appender::File as suggested
by Lee Carmichael <>
* (ms) added RRDs appender Log::Log4perl::Appender::RRDs and testcases
* (ms) fixed Log::Log4perl::Appender to check if a an appender package
has already been loaded and skip 'require' in this case.
Packages injected via Class::Prototyped caused an error with this.
* (ms) Extended the FAQ's "How can I write my own appender?" on
how to dynamically create new appenders via Class::Prototyped.
0.43 (03/22/2004)
* (ms) Applied patch by Markus Peter <> for 'pipe'
mode in Log::Log4perl::Appender::File
* (ms) Added composite appender Log::Log4perl::Appender::Limit to
limit message delivery to adjustable time windows.
* (ms) Fixed last 033UsrCspec.t test case to run on Win32 as well
(path fixed).
* (ms) Lars Thegler <> provided a patch to keep
compatibility with 5.005_03.
* (ms) Added a patch to avoid warnings on undefined MDC values referenced
via %X in PatternLayout. Now, the string "[undef]" is used. Bug
was reported by Ritu Kohli <>
0.42 (02/14/2004)
* (kg) added filters to XML DOMConfig and DTD
* (ms) Fixed caller level to cspecs by adding one
* (ms) Added init_once() and documentation
* (ms) Worked around the perl bug that triggers __DIE__ handlers
even if die() occurs within an eval(). So if you did
BEGIN { $SIG{__DIE__} = sub { print "ouch!"; die }; }
use Log::Log4perl;
and Time::HiRes wasn't available, the
eval { require Time::HiRes }
in triggered the __DIE__ handler. Now there's
a function module_available() in L4p::Util to check if a
module is installed.
* (ms) Fixed %M cspec in PatternLayout in case a logging
method is called within one (or more) eval {} block(s).
caller(n+m) will be called repeatedly if necessary
to get the next real subroutine. Anonymous subroutines will
still be called __ANON__, but this can be overridden by
local *__ANON__ = "subroutine_name";
in them explicitely (thanks, Perlmonks :).
0.41 (12/12/2003)
* (ms) Applied documentation update for Synchronized appender, suggested
by David Viner E<lt>dviner@yahoo-inc.comE<gt>
* (ms) Added option to Log::Log4perl::Layout::PatternLayout to
enable people to provide their own timer functions.
0.40 (11/11/2003)
* (ms) perl 5.005_03 fix for l4p::Appender::Synchronized
* (ms) Fixed a bug in 0.39 (thanks to James King for finding) which
caused composite appenders like Synchronized to just use
SimpleLayout. With the fix, composite appenders are now relaying
messages unmodified to their delegates, which can then apply
any layout they desire.
* (ms) Added file_open(), file_close() and file_switch() to
0.39 (10/23/2003)
* (kg) fixed bug in interaction between Logger::Level and Level::is_valid
so that now you can do $logger->level('INFO') instead of just $INFO.
* (ms) Added logic for 'composite appenders'. Appenders can now be
configured to relay messages to other appenders. Added
Log::Log4perl::Appender::Synchronized, an appender guaranteeing
atomic logging of messages via semaphores.
* (ms) Added basic substitution to PropertyConfigurator. Now you can
define variables (like in "name=value") and subsequent patterns
of "${name}" will be replaced by "value" in the configuration file.
* (kg) Followed Mike's lead and added variable substitution to the
* (ms) Added Log::Log4perl::Appender::Socket as a simple Socket
appender featuring connection recovery.
0.38 (09/29/2003)
* (kg) fixed bug where custom_levels beneath DEBUG didn't work
* (ms) fixed 5.00305 incompatibility reported by
Brett Rann <> (constants with leading _).
* (ms) Log::Log4perl->easy_init() now calls ->reset() first to make sure
it's not duplicating the existing logging environment. Thanks
to William McKee <> for bringing this up.
* (ms) fixed bug with error_die() - printed the wrong function/line/file.
Reported by Brett Rann <>.
* (ms) added %T to PatternLayout as a stack traced as suggested by
Brett Rann <>.
0.37 (09/14/2003)
* (kg) adjusting tests for XML::Parser 2.32 having broken
XML::DOM 1.42 and lower
* (ms) Added signal handling to init_and_watch
* (ms) renamed l4p-internal DEBUG constant to avoid confusion with
DEBUG() and $DEBUG as suggested by Jim Cromie <>.
* (ms) Applied patch by Mac Yang <> for
Log::Log4perl::DateFormat to calculate the timezone for the 'Z'
conversion specifier.
0.36 (07/22/2003)
* (ms) Matthew Keene <> suggested to have
an accessor for all appenders currently defined -- added
appenders() method
* (ms) Test case 041SafeEval.t didn't share $0 explicitely and
created some warnings, fixed that with (jf)'s help.
* (ms) Added performance improvements suggested by
Kyle R. Burton <>. is_debug/is_info/etc.
are now precompiled, similar to the debug/info/etc. methods.
* (ms) Added a fix to have is_debug()/is_info()/etc. pay
attention to on-the-fly config file changes via init_and_watch().
* (ms) Fixed bug that reloaded the config under init_and_watch()
every time the check period expired, regardless if the config
file itself had changed. Added test case.
0.35 06/21/2003
* (kg) got rid of warnings during make test in 014ConfErrs.t
added user-defined hooks to JavaMap
* Jim Cromie <> provided a patch to get
rid of deprecated our-if syntax in
* (ms) removed test case for RollingFileAppender because of recent
instability. Added dependency for Log::Dispatch::RollingFile 1.10
in Log/Log4perl/JavaMap/
0.34 06/08/2003
* (ms) James FitzGibbon <> noticed a major
bug in Log::Log4perl::Appender::File and provided a patch. Problem
was that 0.33 was reusing the same file handle for every opened file,
causing all messages to end up in the same file.
0.33 05/30/2003
* (kg) CPAN rt#2636, coordinating XML::DOM version required across modules
and unit tests
* (ms) Removed Log::Dispatch dependency, added standard
Log::Log4perl::Appender appenders File and Screen.
Log::Dispatch is still supported for backwards compatibility
and special purpose appenders implemented within this hierarchy.
0.32 05/17/2003
* (ms) Added fix to Makefile.PL to compensate for MakeMaker bug
in perl < 5.8.0, causing man pages below Log::Log4perl::Config
not to be installed. Thanks to Mathieu Arnold <>
for bringing this up.
* (ms) 0.31 had a Win32 test suite glitch, replaced getpwuid()
(not implemented) by stat() for Safe test.
0.31 05/08/2003
* (kg) fixed bug Appender::DBI where it was consuming the message
array before other appenders could get to it
* (ms) changed config_and_watch to ignore clock differences between
system time and file system time (helpful with skewed NFS
systems). Added Log::Log4perl::Config::Watch.
* James FitzGibbon <>: Added support for
optionally restricting eval'd code to Safe compartments.
* (ms) allow/deny code in configuration files should now be controlled
via the accessor Log::Log4perl::Config->allow_code(0/1).
$Log::Log4perl::ALLOW_CODE_IN_CONFIG_FILE is still supported
for backwards compatibility.
0.30 03/14/2003
* (ms) Added Log4perl custom filter logic and standard filter set
* (kg) Added url support to init(), finally documenting it
* (kg) Finished implementation of DOMConfigurator allowing xml configs.
* (ms) Corrected DateFormat inconsistencies as reported by
Roger Perttu <>
0.29 01/30/2003
* (kg) Removing debugging from 0.28, big woops
* (kg) Fixing 036JSyslog.t, Syslog implementations are too often broken
to base any results on.
* (kg) Fixing XML-DOM tests, Data::Dumper doesn't return data exactly the
same way.
0.28 (01/28/2003)
* (ms) '#' in the conf file are now interpreted as comment starters only
if they're at the start of a line with optional whitespace.
The previous setting (comments starting anywhere) had problems
with code containing '#''s, like in layout.cref = sub { $#_ = 1 }
* (ms) warp_message accepts code refs or function names
* (kg) Split config bits into PropertyConfigurator and implemented
DOMConfigurator for XML configs.
* (kg) Adding appender.warp_message parameter as a help to DBI
* (kg) Added NoopLayout to help DBI appender
* (ms) Added message output filters:
log({filter => \&filter, value => $value})
* (kg) t/024WarnDieCarp was assuming / as directory separator, failed
on Win32
* (kg) implemented JavaMaps for NTEventLogAppender, SyslogAppender
* (kg) found and addressed circular ref problem in Logger->reset
* (kg) moved TestBuffer under Appender/ directory along with DBI
* (kg) fixed docs, Pattern layout, %f not supported, s/b %F
* (kg) added Log::Log4perl::Appender::DBI to implement JDBCAppender
* (ms) Every value in the config file can now be a perl function,
dynamically replaced by its return value at configuration
parse time
* (ms) NDC now prints entire stack, not just
top element (as mandated by Log4j)
* (ms) Allow trailing spaces after a line-breaking '\' in the
config file to be fault-tolerant on cut-and-pasted code
0.27 12/06/2002
* (ms) Updated FAQ with "Recipes of the Week"
* (ms) Added Log::Log4perl::NDC (Nested Diagnostic Contexts) and
Log::Log4perl::MDC (Mapped Diagnostic Contexts)
* (ms) LOGDIE and LOGWARN added to stealth loggers
* (ms) Logging methods ($lo->debug(), $lo->info() ...) now return
a value, indicating the number of appenders that the message
was propagated to. If the message was suppressed due to level
constraints, undef is returned. Updated manpage (new section
"return values").
* (ms) Fixed bug reported by Francisco Olarte Sanz.
<>: ISO date format and documentation mixed
up MM with mm in the simple date format
* (kg) User-defined conversion specifiers for PatternLayout in
configuration file and as C API
* (kg) implementing map to log4j.RollingFileAppender
* (kg) trying out oneMessagePerAppender parameter
* (kg) changed unit tests to use File::Spec
0.26 11/11/2002
* (kg) enabled %l (was missing from PatternLayout::define)
* (kg) got rid of "Use of uninitialized value in join or string" message
when some of $logger->debug(@array) when some of @array are undef
* (ms) Stealth loggers and documentation
* (kg) Better error message for case reported by Hai Wu
* (ms) Added Log/Log4perl/, which the homepage links to
* (ms) Took dependency on Test::More and Test::Simple out of the PPD file
because of a problem with Activestate 5.6.1 reported
by James Hahn <>
* (ms) Added Log::Dispatch equivalent levels to the Log4perl loggers,
which are passed on the Log::Dispatch appenders now according
to the priority of the message instead of the default "DEBUG"
* (ms) Added %P process ID to PatternLayout as suggested by
Paul Harrington <>. Also added
%H as hostname
* (kg) Added %min.max formatter to PatternLayout
* (ms) Updated docs for Log::Log4perl::DateFormat
0.25 10/06/2002
* (ms) backwards-compatibility with perl 5.00503
* (ms) added system-wide threshold, fixed java-app thresholds
* (kg) Nested configuration structures for appenders like L::D::Jabber
* (ms) ::Log4perl::Appender::threshold() accepts strings or integer
levels (as submitted by Aaron Straup Cope <>)
* (ms) Fixed logdie/logwarn caller(x) offset bug reported by
Brian Duffy <Brian.Duffy@DFA.STATE.NY.US>
* (ms) dies now on PatternLayout without ConversionPattern (helps detecting
typos in conf files)
0.24 09/26/2002
* (kg) Fix for init_and_watch and test cases
* (ms) Added documentation for Log::Log4perl::Config
* (ms) Added log4perl.additivity.loggerName conf file syntax
* (ms) Assume Log::Log4perl::Layout prefix of 'relative'
layout class names in conf file (say 'SimpleLayout'
instead of 'Log::Log4perl::Layout::SimpleLayout').
* (ms) accidently appending a ';' at the end of an appender
class in a conf file now spits out a reasonable error message
* (ms) added a by_name() method to TestBuffer to retrieve an
instance of the TestBuffer population by name instead of
relying on the order of creation via POPULATION[x]
(for testing only).
* (kg) Win32 compatibility fixes
0.23 09/14/2002
* Both Log4perl/log4perl is now accepted in conf file
* Added documentation to Log::Log4perl::Appender
* Made Time::HiRes optional. If it's missing, PatternLayout will
just use full seconds as %r.
* SimpleDateFormat "%d{HH:SS}", including predefined formats (DATE etc.)
* Added another cut-and-paste example to the docs (EXAMPLE)
* Added new logdie/logwarn/error_warn/error_die/logcarp/
logcluck/logcroak/logconfess functions written by
Erik Selberg <>
* Added PatternLayout documentation
* Changed suppression of duplicate newline in log message algorithm
* Custom levels and inc_level/dec_level/more_logging/less_logging
added by Erik Selberg <>
* Append to logfile by default if Log::Dispatch::File is used
(previously clobbered by default)
* Kevin's init_and_watch fix
0.22 8/17/2002
* Threshold settings of appenders:
log4j.appender.A.Threshold = ERROR
* Chris R. Donnelly <>
submitted two patches:
- extended init() to take obj references (added, also added a test
case and documentation)
- fixed %F and %L if Log4perl is used by a wrapper class (accepted,
but changed variable name to Log::Log4perl::caller_depth as
a tribute to Log::Dispatch::Config, added test case 022Wrap
and documentation
0.21 8/08/2002
* Synopsis shows code samples in
* Slight Log4j incompatibility but useful: %F{n} lets you
limit the number of entries the source file path is logged
* Erik W. Selberg ( suggested having
suppress another \n if the messages already contains a \n and the
format requires a %n. Done.
* Erik W. Selberg ( suggested loggers should take
any number of messages and concatenate them. Done.
* Fixed double-init problem and added a test case. Now the entire
configuration is cleared before the second init(). However, this
surfaced a problem with init_and_watch: If a program obtains
references to one or more loggers, rewriting the configuration
file during program execution and re-initing makes these reference
point to loggers which hold obsolete configurations. Fixed that by
code in debug(), info(), etc. which *replaces* (shudder) the
logger reference the program hands in to them with a new one of
the same category. This happens every time if 'init_and_watch' has
been enabled. However, this introduces a small runtime penalty.
This is different from the original log4j, which does some
half-assed re-initialization, because Java isn't expressive enough
to allow for it. Making this thread-safe might be tough, though.
* Added DEBUG statements to and to trace execution
(debugging won't work because of "eval"s). Both files define a
constant named DEBUG towards the top of the file, which will
have perl optimize away the debug statements in case it's set to 0.
* A warning is issued now (once) if init() hasn't been called or
no appenders have been defined.
* Added ':levels' target to Log::Log4perl to import $DEBUG, $ERROR,
etc. levels (just like 'use Log::Log4perl::Level' works).
* Added ':easy' target to allow for simple setup
* Code references can be passed in as log messages to avoid parameter
passing penalty
0.20 7/23/2002
* Strip trailing spaces in config file
* Accept line continuations in properties file
* Refactored for speed, defined the logging behavior when
the logger is created, not when a message is logged
* Fixing test suites so that SimpleFormat newline is accounted for
* Fixed a bug with root inheritance where the category name wasn't
coming through
* added init_and_watch
0.19 07/16/2002
* Added Log::Log4perl::Appender::TestBuffer back in the distribution, otherwise
regression test suite would fail.
0.18 07/16/2002
* Failed attempt to fix the Log::Dispatch::Buffer problem.
0.17 07/11/2002
* Updated documentation according to Dave Rolsky's suggestions
* Lots of other documentation fixes
* Fixed bug in renderer, %M was displayed as the logger function
bumped up the level by 1
* Fixed %% bug
0.16 07/10/2002
* Updated documentation for CPAN release
* Applied Kevin's patch to limit it to one Log::Dispatcher
0.15 07/10/2002
* There were name conflicts in Log::Dispatch, because we used *one*
Log::Dispatch object for the *all* loggers in the Log::Log4perl
universe (it still worked because we were using log_to() for
Log::Dispatch to send messages to specific appenders only). Now
every logger has its own Log::Dispatch object. doesn't
call Kevin's anti-dupe logic anymore -- is this ok? Maybe there's
some leftovers which need to be cleaned up.
* Kevin fixed t/014ConfErrs.t after last night's change
0.14 07/09/2002
* (!) Added new class Log::Log4perl::Appender as a wrapper around
Log::Dispatch::*. Layouts are no longer attached to the loggers,
but to the appenders instead. $app->layout($layout) sets the
layout. $logger->add_appender($app) is the new syntax to add
an appender to a logger. The $logger->layout method is gone
for that reason.
* Added documentation on categories
* Added documentation on Log::Log4perl::Appender,
0.13 07/09/2002
* in the config files, 'debug' is not a level, 'DEBUG' is
* expanded the layouts so that we can add subclassess, added
SimpleLayout, note that api usage changes
-$logger->layout('buf',"The message is here: %m");
Log::Log4perl::Layout::PatternLayout('buf',"The message is
here: %m"));
* did benchmarks, see doc/benchmark*, t/013Bench.t
* further tweaked errors for bad configuration, added a test for those
0.12 07/08/2002
* Log::Log4perl::Logger->get_logger now accessible via
* Log::Log4perl::Config->init now accessible via
* Adapted test cases to new shortcuts
* Constrained some files to 80 chars width
* Added test case t/009Deuce.t for two appenders in one category
via the config file
* Changed default layout in case there's none defined (SimpleLayout)
* Implemented dictatory date format for %d: yyyy/MM/dd hh:mm:ss
0.11 07/07/2002
* added documentation to Log/
* added is_debug/is_error/is_info etc. functions to,
test cases to t/002Logger.t
0.10 07/05/2002
* %p should return level name of the calling function, so
$logger->warn('bad thing!!') should print 'WARN - bad thing'
even if the category is set to debug, so took level_str out of (kg)
0.09 07/03/2002
* %p should return level name, not number, adding level_str to (kg)
* - discriminating: priorities are 1-4, levels are
'info','debug',etc (kg)
0.08 07/03/2002
* Non-root loggers are working now off the config file
0.07 07/02/2002
* Updated documentation
* removed "diagnostics"
0.06 07/01/2002
* Bug discovered by Kevin Goess <>, revealed
in 004-Config.t: Wrong layout used if Appender is inherited.
* Changed Log::Log4perl::Appender::TestBuffer to keep track of the
object population -- so we can easily reference them
in the Log::Log4perl test cases. Got rid of get_buffer().
* Added a reset() method to Log::Log4perl and Log::Log4perl::Logger
for easier testing. It resets all persistent loggers to
the inital state.
* Added documentation
0.05 06/30/2002
* Fixed bug with mapped priorities between java/Log::Dispatch
* Java/Perl integration with conf file
0.04 06/30/2002
* Layout tests
* %r to layout
* Added lib4j configuration file stuff and tests
0.03 06/30/2002
* Layout
* Curly braces in Layout first ops
0.02 06/30/2002
* Created Logger and test cases
0.01 06/22/2002
* Where it all began
TODO List:
* '%t'
* Wild idea: Could we possibly utilize the compiler
frontend to eliminate log statements that are not going to be
triggered? This would be a HUGE performance increase!
* get_logger() thread safety (two try to create it at the same time)
* Thread safety with re-reading the conf file (watch)
* log4j.logger.blah = INHERITED, app