split up multiline messages sent to Log::Dispatch::Syslog

[autarch]

Migrated from rt.cpan.org #26711 (status was 'open')

Requestors:

• mark@blackmans.org

Attachments:

• Log-Dispatch-Syslog-patch.txt

From mark@blackmans.org on 2007-04-26 14:05:04:

The Text::SimpleTable module in Catalyst routinely generates multi-line
strings which Catalyst routinely sends to it's logging output which when
piped into Log::Dispatch::Syslog looks very messy in the end result.
This patch allows these outputs to show up reasonably in syslog logs.

[autarch]

From drolsky@cpan.org (@autarch) on 2007-11-02 02:37:51:

On Thu Apr 26 10:05:04 2007, mark@blackmans.org wrote:

The Text::SimpleTable module in Catalyst routinely generates multi-line
strings which Catalyst routinely sends to it's logging output which when
piped into Log::Dispatch::Syslog looks very messy in the end result.
This patch allows these outputs to show up reasonably in syslog logs.

For backwards compatibility, this would have to be an option (off by
default). Tests wouldn't hurt either ;)

[autarch]

From mark@blackmans.org on 2007-11-23 13:37:01:

On 2 Nov 2007, at 02:37, via RT wrote:

<URL: http://rt.cpan.org/Ticket/Display.html?id=26711 >

On Thu Apr 26 10:05:04 2007, mark@blackmans.org wrote:

The Text::SimpleTable module in Catalyst routinely generates multi-
line
strings which Catalyst routinely sends to it's logging output
which when
piped into Log::Dispatch::Syslog looks very messy in the end result.
This patch allows these outputs to show up reasonably in syslog logs.

For backwards compatibility, this would have to be an option (off by
default). Tests wouldn't hurt either ;)

Ok, I'm not quite sure what a straightforward mechanism for testing
syslog output is, any clues for me? Not yet checked existing tests,
so maybe that's where to start.

Cheers,
Mark

[autarch]

From autarch@urth.org (@autarch) on 2007-12-02 16:38:40:

On Fri, 23 Nov 2007, mark@blackmans.org via RT wrote:

For backwards compatibility, this would have to be an option (off by
default). Tests wouldn't hurt either ;)

Ok, I'm not quite sure what a straightforward mechanism for testing
syslog output is, any clues for me? Not yet checked existing tests,
so maybe that's where to start.

Unfortunately, I don't have any syslog tests either. I think the way to do
it would be to monkey-patch Sys::Syslog to override it's functions and
then check what gets passed to them.

-dave

/===================================================
VegGuide.Org www.BookIRead.com
Your guide to all that's veg. My book blog
===================================================/

[autarch]

From autarch@urth.org (@autarch) on 2007-12-02 17:17:47:

On Fri, 23 Nov 2007, mark@blackmans.org via RT wrote:

For backwards compatibility, this would have to be an option (off by
default). Tests wouldn't hurt either ;)

Ok, I'm not quite sure what a straightforward mechanism for testing
syslog output is, any clues for me? Not yet checked existing tests,
so maybe that's where to start.

Unfortunately, I don't have any syslog tests either. I think the way to do
it would be to monkey-patch Sys::Syslog to override it's functions and
then check what gets passed to them.

-dave

/===================================================
VegGuide.Org www.BookIRead.com
Your guide to all that's veg. My book blog
===================================================/