Werner, Loren's check of IsDebugEnabled is the accepted way to accomplish what you want. Beyond that though, if you're usage is like ours then most of the time (99.999999999999999999% :-) you don't want the logging of those large objects at all, but every once in a while you do. Well, we accepted that we'd have to restart the application to get it, and we added a static _readonly_ variable that we check in addition to IsDebugEnabled. Since it's static readonly, the JIT knows it can't be updated once it's set and will actually omit the logging from the generated machine code -- it's one performance step beyond IsDebugEnabled.
-Walden -- Walden H Leverich III Tech Software & BEC - IRBManager (516) 627-3800 x3051 wald...@techsoftinc.com http://www.TechSoftInc.com http://www.IRBManager.com Quiquid latine dictum sit altum viditur. (Whatever is said in Latin seems profound.) -----Original Message----- From: ITemplate [mailto:itempl...@hotmail.com] Sent: Thursday, April 15, 2010 3:36 AM To: log4net-user@logging.apache.org Subject: Re: Possible to test log4net filters on certain logevent? Hi Loren, Thanks for your reply! Well there are a couple of reasons but as an example, we need to log special webservices where the request and responses are VERY large objects that log4net (or any other) can't serialize correct. We have our own serializer for that job. But at the same time, these web services must perform well and object serialization here is a rather expensive and time consuming task. So I would like the ability to test if the log system would actually log the data - and if not, I would not even begin to serialize the classes to text. But on the other hand I will not begin to dig too deep into log4net to accomplish this - it would only make sense if there were some easy way like: var mustSerialize = false foreach(var appender in log4net.Appenders) { mustSerialize = appender.WouldLog(logEvent); if (mustSerialize) break; } But digging deeper would drain the goal of saving the time to serialize I guess... -- Werner Loren Keagle wrote: > > I don't think that this is possible without some dangerous coding. > Looking at the source code, there are no methods on any appenders to > test the level/filter chain in advance. Given that your log4net config > could contain dozens of appenders, each with their own set of filters, > you could not make a single call to determine if an event would be > filtered from all of them. Perhaps if you provided a little bit of > context, someone here might be able to offer an alternative approach? > > If you really must have a programmatic way of doing this, you would have > to gain access to all of your appenders, iterate through them, casting > each to AppenderSkeleton (assuming the appender derives from this helper > class!!!), test the log level of each, then iterate through all of the > filters in the FilterHead linked list property. AppenderSkeleton has a > protected method called FilterEvent that does all of this, but you > cannot access it. It would probably have solved your problem if this > method was part of the IAppender interface, but it's currently not > accessible. > > So I suppose it is possible, but you're going to end up duplicating all > of the logic that is already in AppenderSkeleton, which means you will > be executing code twice. If you have multiple appenders, you will have > to check each appender separately. > > ~Loren Keagle > > > On 4/13/2010 6:18 AM, ITemplate wrote: >> Noone knows? Perhaps I could rephrase: I need a programmatic way to >> determine >> if a certain logentry would be filtered or not. >> >> Example: I have the following logentry: >> logger.Warn("This might be logged but I dont know. I want to test it in >> code."); >> >> Now I want to find out - at runtime - if that specific line in my code >> will >> produce a log. Some programmatic way to test one or all configured >> filters >> against the data in my log-line (Logger, Level, text etc). >> >> Hope my meaning here is more clearer? >> >> Thanks. >> >> >> >> ITemplate wrote: >> >>> Hi, >>> >>> Is this "old" forum the only for log4net? Anyways - suppose I have a >>> certain logevent, is it possible to investigate up front IF that >>> particular logevent would pass all filters? I can't seem to find a >>> Filters >>> collection that I can use for this? >>> >>> -- >>> Werner >>> >>> >> >> >> >> >> No virus found in this incoming message. >> Checked by AVG - www.avg.com >> Version: 9.0.801 / Virus Database: 271.1.1/2807 - Release Date: 04/12/10 >> 11:32:00 >> >> > > > -- View this message in context: http://old.nabble.com/Possible-to-test-log4net-filters-on-certain-logevent--tp28188373p28252058.html Sent from the Log4net - Users mailing list archive at Nabble.com.
