|
From: David G. <dg...@co...> - 2003-08-15 19:45:53
|
Thanks, Bruce. Throwing an exception in a Script stage gives me exactly what I need. FYI, it seems that the logic for failOnError is exactly the opposite as described in the Users Guide, in that my testing seems to indicate that it defaults to true unless explicitly set to false. Speaking of the Users Guide, I'm having an interesting time finding something that will allow me to edit the docbook file in a WYSIWYG format. Being new to docbook, can I ask what you use (preferably Java or Linux-based)? Thanks again, David On Friday 15 August 2003 11:12 am, McDonald, Bruce wrote: > All seems good - In in specialized processing like this - The Scripting > stage is your friend. > > > See comments inline: > > 1. I understand that a stage that fails will force a FAIL status. However, > what I'm trying to do is force a FAIL status when a stage completes > successfully, but the application detects an error. For example, in my > case I post SOAP messages, and if the response is a SOAPFAULT, I want to > force an error into the journal and stop the pipeline processing. My > question is: can I do this via a script stage (and if so, how), or must I > create a new stage type to support it? > > > If you throw an exception from the pipelinestage - that stage will write a > failed to the journal. You might need to expand the SoapWriter stage and > add a config option like 'failOnFault' (default false) to do this. <<<< > > > 2. I also understand about setting the 'tracked' attribute for the stages. > The question I have regards the use of the journal tool. Using it to view > entries, is there any indication that a particular entry in the journal > contains a tracked document? The reason is that, in the case of a failure, > I expect the operator to do the following: > > a. View the journal entries to determine which stage failed. > b. Correct the cause of the failure. > c. Assuming the problem was not due to the document, replay the journal > from the stage that failed. If the particular stage that failed is not > tracking the document, I assume he will need to find the closest previous > stage that has a document associated with it, and replay it from that > point. If this is correct, he'll need some way to determine which stage to > use. If this is not correct, how should I be using the replay > functionality? > > > YOu might need to write a specialized journal view to do this. I am not > sure if the default tools support this. <<<< > > Thanks again for your help, > > David > > On Friday 15 August 2003 12:05 am, Dejan Krsmanovic wrote: > > Hi David, > > > > > Hello to the Great and Powerful List, > > > > > > I would like to use the journal to track all documents through the > > > system. > > > > In > > > > > order to do this, I'll need to do the following: > > > > > > 1. If I encounter an application error, I'd like to force a FAIL status > > > > into > > > > > the Journal, with an appropriate error message. Is there a way to do > > > > this? > > > > All pipeline stages by default forces FAIL status in case of error. Also, > > processing > > is stopped in case of error. This can be changed using errorStage and > > failOnError > > config options. I think there is a section about this is User Guide. > > > > > 2. I'll be using pretty large documents for some of my operations, so I > > > > don't > > > > > want to track it in every stage; only those stages where it will be > > > transformed in some way. Is there anyway to know from the journal > > > which stages track document? This will simplify things once I turn > > > over > > > > monitoring > > > > > of the system to the client. > > > > By default documents are NOT tracked. If you want to track documents you > > should > > use tracked config option which is false by default > > > > > 3. Also, is the document tracked on entry to the stage or on exit from > > > it? > > > > In > > > > > other words, is tracking saving the document before any transformation > > > is applied by the stage, or after it's transformed? > > > > Before. > > > > > 4. I would like to organize pipeline fragments to operate as > > > subroutines > > > > that > > > > > can be called by multiple stages, which I believe that I can now do > > > with > > > > the > > > > > updated CallStage functionality. I would also like to place these > > > > fragments > > > > > in a separate file so that I can use them in multiple pipeline > > > configurations. Can this be done? > > > > This should be possible now, right Bruce? Bruce and I have discussion > > about that few > > months ago and I think he has implemented it. I havent used it so I am > > not sure if it is working. > > > > > > Dejan > > > > > Thanks, > > > > > > David > > > > > > -- > > > David Glick > > > Transmit Consulting, Inc > > > 619-475-4052 > > > dg...@tr... > > > > > > > > > > > > ------------------------------------------------------- > > > This SF.Net email sponsored by: Free pre-built ASP.NET sites including > > > Data Reports, E-commerce, Portals, and Forums are available now. > > > Download today and enter to win an XBOX or Visual Studio .NET. > > > > http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/ > >01 > > > > > _______________________________________________ > > > Babeldoc-devel mailing list > > > Bab...@li... > > > https://lists.sourceforge.net/lists/listinfo/babeldoc-devel > > > > ------------------------------------------------------- > > This SF.Net email sponsored by: Free pre-built ASP.NET sites including > > Data Reports, E-commerce, Portals, and Forums are available now. > > Download today and enter to win an XBOX or Visual Studio .NET. > > http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/ > >01 _______________________________________________ > > Babeldoc-devel mailing list > > Bab...@li... > > https://lists.sourceforge.net/lists/listinfo/babeldoc-devel -- David Glick Transmit Consulting, Inc 619-475-4052 dg...@tr... |