Hi Gerhard, thanks for your answers. I checked out the trunk and it's working as i expect it. I forgot to mention that i'm using facelets and trinidad, but the component support handles this great.
Regards, Markus -----Ursprüngliche Nachricht----- Von: Gerhard Petracek [mailto:[email protected]] Gesendet: Dienstag, 1. Dezember 2009 21:45 An: MyFaces Discussion Betreff: Re: [EXTVal] Validators wihtout implicit required validations hi markus, i introduced marker annotations [1] so that you can easily mark validation strategies which are aware of empty- and null-values. i committed the changes as well as test-cases. there will be a new (milestone) version quite soon. regards, gerhard [1] http://issues.apache.org/jira/browse/EXTVAL-74 http://www.irian.at Your JSF powerhouse - JSF Consulting, Development and Courses in English and German Professional Support for Apache MyFaces 2009/11/30 Gerhard Petracek <[email protected]> > short addition: > > i see your point and i'll change that. > for now: my previous post contains a list of simple solutions you can use. > > regards, > gerhard > > > http://www.irian.at > > Your JSF powerhouse - > JSF Consulting, Development and > Courses in English and German > > Professional Support for Apache MyFaces > > > 2009/11/30 Gerhard Petracek <[email protected]> > > hi, >> >> it's a bit different. >> the parameter mentioned by jakob is a jsf 2.0 parameter especially for the >> jsr 303 integration. >> if you set it to true, the not-null constraint will be interpreted like >> required (= true). >> extval uses it in an abstract class in a similar way. that means: you just >> have to use it to explicitly deactivate the behavior described above. >> it's automatically available for custom validation interceptors. an >> internal example is the validation interceptor of the bean-validation module >> (which allows to use jsr 303 with all jsf versions). >> >> the property validation module of the next release (as well as the >> available milestone) overrides this new default behavior so that an empty >> field doesn't lead to a null value for validation (due to backward >> compatibility reasons). however, you can easily customize this behavior. >> >> back to the original question: >> >> extval just delegates to the (configured) jsf length validator. so >> basically the behavior depends on the implementation you are using in your >> application. >> >> in this case the default implementation throws an exception because it >> isn't aware of empty values. to change this behavior you can: >> - register your custom jsf-length-validator implementation (see: >> javax.faces.Length) which handles it as you expect it (that's plain jsf) >> or >> - replace the default extval validation strategy for @Length to ignore >> empty values >> or >> - implement your custom extval annotation >> >> all 3 possibilities are quite easy. >> >> regards, >> gerhard >> >> http://www.irian.at >> >> Your JSF powerhouse - >> JSF Consulting, Development and >> Courses in English and German >> >> Professional Support for Apache MyFaces >> >> >> 2009/11/30 Jakob Korherr <[email protected]> >> >> Hi Markus, >>> >>> Take a look at the >>> javax.faces.INTERPRET_EMPTY_STRING_SUBMITTED_VALUES_AS_NULL >>> config parameter in web.xml. >>> >>> Setting this to true will cause that your property "input" will be null, >>> if >>> the user submits an empty field and thus @Length will not be processed. >>> >>> Also take a look at http://jsfatwork.irian.at/semistatic/jsf.html, >>> chapter >>> 2.12 Validierung fur further information. It's written in german, but I >>> assume you're from Germany... >>> >>> Regards >>> >>> Jakob Korherr >>> >>> >>> 2009/11/30 Dreher, Markus <[email protected]> >>> >>> > Hi all, >>> > >>> > i want to validate a field only when the user filled it out. >>> > >>> > It's not a reuqired field, but when it is filled out, the input should >>> > be at least for example 3 digits. >>> > >>> > @Length(minimum=3, maximum = 60) >>> > private String input; >>> > >>> > With @Length this leads to a required field. >>> > >>> > Do i miss something? >>> > >>> > Regards, >>> > >>> > Markus >>> > >>> > >>> > >>> >> >> >
