ah, gmail being cleverer than me and making a link without asking me - the text is right, just the link got out of sync with reality somehow. Here it is again...
http://www.idiots.org.uk/log4net/DupeFilteringForewardingAppender_08-08-07.txt On 8/8/07, Ron Grabowski <[EMAIL PROTECTED]> wrote: > > > HTTP 404 - File not found > ----- Original Message ---- > > From: James Wilkinson <[EMAIL PROTECTED]> > > To: Log4NET User <[email protected]> > > Cc: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]> > > Sent: Wednesday, August 8, 2007 4:43:24 AM > > Subject: log4net duplicate filtering > > > > Hi, > > > > I wanted to post the code I'd written for duplicate filtering a couple of > weeks ago, but got distracted with some other work. After noticing the > recent messages about duplicate filtering, I just figured that posting it in > its current state might be of use to people. > > > > It's a modified forwarding appender that does duplicate event filtering > with a summary message inserted after the series of duplicate events has > completed. I started off trying to implement it as a filter, but hit the > problem of not being able to insert the summary message into the stream. > > > > here's the source... > > > http://www.idiots.org.uk/log4net/DupeFilteringForewardingAppender_08-08-07.txt > > > > Typical usage would be something like this... > > > > <!-- use this forwarding appender as a dupe filter which then passes on > to the console --> > > <appender name="DupeFilteringForwardingAppender" type=" > log4net.Appender.DupeFilteringForwardingAppender " > > > <appender-ref ref="ConsoleAppender" /> > > <FlushTimeout value="3" /> > > </appender> > > > > FlushTimeout (integer seconds) - see the comment in the source for the > public property for how this works. > > > > There are a whole load of things I'm not really sure I've done very well. > The name of the thing for a start. The message fingerprinting might be more > efficiently done as a checksum and from what I know, I think the reliance on > the .net 2.0 generic-based list probably isn't in the spirit of the > log4net coding standards. The hard-coded way that event fingerprints are > built up feels like it needs thought. Anyway, it's maybe a start. > > > > Also, thanks to Peter for clearing up my misconception about log4net > plugins. > > > > James > > > > > > > > > >
