Michael, Here's how the JAXB gang prescribes avoiding this problem.
https://jaxb.dev.java.net/guide/Mapping_cyclic_references_to_XML.html What of this seems plausible to you (or other readers) for Aegis? --benson On Thu, Apr 2, 2009 at 1:07 AM, Michael Berkowitz <[email protected]>wrote: > Quite right. I haven’t forgotten! In fact, I just brought it up with my > staff at yesterday’s meeting, but we’ve all been swamped since I last > wrote. I hope to get to it either over the weekend or early next week. > > > > *From:* Benson Margulies [mailto:[email protected]] > *Sent:* Thursday, April 02, 2009 3:31 AM > > *To:* Michael Berkowitz > *Subject:* Re: boolean webmethod with Aegis throwing NPE > > > > You never did post me a test case, did you? > > On Mon, Mar 23, 2009 at 3:57 AM, Michael Berkowitz < > [email protected]> wrote: > > I don’t think I ever mentioned examples of where Aegis is anal about > JavaBeans conformance: I’ve found that if I have a method that starts with > *get* or *is* and there’s no corresponding *set* (perhaps because the > thing being returned isn’t an actual field but rather some sort of > calculation), it complains. Ultimately I found that I could use an > annotation to tell Aegis to ignore it. > > I believe it also complained if I had a data member for which there was no > getter and setter. > > > > *From:* Benson Margulies [mailto:[email protected]] > > *Sent:* Sunday, March 22, 2009 4:05 AM > > *To:* Michael Berkowitz > *Subject:* Re: boolean webmethod with Aegis throwing NPE > > > > Aegis isn't necessarily 'strict', but it's defintely unimaginative. Have > you played with .aegis.xml files to give it instructions? What precisely is > the case at hand? Is there something you want it to turn into an element > that it's oblivious to? > > Thats bug's pretty humorous. > > I'm about the only Aegis maintainer. If you could perchance create a little > Maven project that has the necessary dependencies so that I can run it and > watch it explode, I'll see what I can do. It doesn't look likely to succumb > to just reading the code. > > Anyway, I'd rather be reconnected to someone with some embarassment than > not, so feel free to rat me out. > > On Sat, Mar 21, 2009 at 4:50 PM, Michael Berkowitz < > [email protected]> wrote: > > I’ll do that, but it may have to wait ‘til tomorrow. Your graduation date > is certainly the right general timeframe; I won’t tell him you’ve forgotten > J > > > > I have a general question about Aegis, while I’ve got your attention: I’ve > noticed that it’s very, very strict about Javabean conformance, whereas > another library I’ve used for serializing/unserializing to/from XML, XStream > from ThoughtWorks, seems to do the whole business using reflection > regardless of whether the classes are Javabeans. Is there something > specific about the SOAP dialect that makes this necessary? > > > > And one more thing, if I may: there’s another Aegis bug on the books, > https://issues.apache.org/jira/browse/CXF-1670, that nobody seems to be > working on, and I was wondering what the chances are of that happening. I’d > do it myself if I could find the time to learn the code, as it’s forcing me > to create a whole set of “mock” objects to pass back-and-forth because > passing the actual Hibernate objects is prevented by this bug. > > > > Anyway, thanks for your attention, and I’ll try that bug with a new version > soon. > > > > *From:* Benson Margulies [mailto:[email protected]] > *Sent:* Saturday, March 21, 2009 10:24 PM > *To:* Michael Berkowitz > > > *Subject:* Re: boolean webmethod with Aegis throwing NPE > > > > I have a horrible head for names. I graduated in '82 after entering with > the class of '81. It's entirely possible that I knew him then. > > I'd be grateful if you give the newer version a whirl, since there is > without a doubt a bug in there somewhere. > > On Sat, Mar 21, 2009 at 4:16 PM, Michael Berkowitz < > [email protected]> wrote: > > I've worked around it by changing the webmethod signature, so I'm okay for > now, but if it would help you for me to try it with those versions I'll be > glad to do so; please just let me know. > > BTW, did you go to MIT with my friend Jeremy Barkan? I seem to recall his > mentioning your name. > > > -----Original Message----- > From: Benson Margulies [mailto:[email protected]] > Sent: Saturday, March 21, 2009 10:00 PM > To: [email protected] > Subject: Re: boolean webmethod with Aegis throwing NPE > > Aegis has bugs. Can I ask you to try 2.1.5-SNAPSHOT? Or even 2.2? If I have > to fix this, I'd rather be fixing it in the current code. > > On Sat, Mar 21, 2009 at 3:28 PM, Michael Berkowitz > <[email protected]>wrote: > > > I've got a web method defined thus: > > > > @WebMethod > > > > boolean haveNumber(String number); > > > > I'm using Aegis as my databinding. When I call the method from my > > client, the response looks like this: > > > > <soap:Envelope xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/";> > > > > <soap:Body> > > > > <ns1:haveNumberResponse > > xmlns:ns1="http://voim.xconnect.com/";> > > > > <return>false</return> > > > > </ns1:haveNumberResponse> > > > > </soap:Body> > > > > </soap:Envelope> > > > > which seems fine, but my client throws a NullPointerException because > > the proxy code has trouble somehow parsing the response. Specifically, > > in org.apache.cxf.endpoint.ClientImpl the line > > > > resList = inMsg.getContent(List.class); > > > > sets resList to an array with one null member. > > > > I've tried both Boolean and boolean, with the same results. > > > > I'm using CXF 2.1.3 on both ends. Any ideas? > > > > Michael > > > > > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ > > > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ > > > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ > > > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ >
