Log::Agent::Channel::Handle - I/O handle logging channel for Log::Agent
require Log::Agent::Channel::Handle; my $driver = Log::Agent::Channel::Handle->make( -prefix => "prefix", -stampfmt => "own", -showpid => 1, -handle => \*FILE, );
The handle channel performs logging to an already opened I/O handle, along with the necessary prefixing and stamping of the messages.
The creation routine make() takes the following arguments:
Specifies the I/O handle to use. It can be given as a GLOB reference, such as
\*FILE, or as an
NOTE: Auto-flushing is not enabled on the handle. Even when the channel is closed, the handle is left as-is: we simply stop sending log messages to it.
When set to true, never append any "\n" (on Unix) or "\r\n" (on Windows) to log messages.
Internally, Log::Agent relies on the channel to delimit logged lines appropriately, so this flag is not used. However, it might be useful for
Default is false, meaning newline markers are systematically appended.
When set to true, disable the prefixing logic entirely, i.e. the following options are ignored completely:
Default is false.
When set to true, don't upper-case the first letter of the log message entry when there's no prefix inserted before the logged line. When there is a prefix, a ":" character follows, and therefore the leading letter of the message should not be upper-cased anyway.
Default is false, meaning uppercasing is performed.
The application prefix string to prepend to messages.
If set to true, the PID of the process will be appended within square brackets after the prefix, to all messages.
Default is false.
-stampfmt=> (name | CODE)
Specifies the time stamp format to use. By default, my "own" format is used. See Log::Agent::Stamping for a description of the available format names.
You may also specify a CODE ref: that routine will be called every time we need to compute a time stamp. It should not expect any parameter, and should return a string.
Beware of chdir(). If your program uses chdir(), you should always specify logfiles by using absolute paths, otherwise you run the risk of having your relative paths become invalid: there is no anchoring done at the time you specify them. This is especially true when configured for rotation, since the logfiles are recreated as needed and you might end up with many logfiles scattered throughout all the directories you chdir()ed to.
Raphael Manfredi <Raphael_Manfredi@pobox.com>