ant-ivy-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Xavier Hanin" <xavier.ha...@gmail.com>
Subject Re: Ivy VersionRangeMatcher lower range bug [4.2.0.3,4.2.+]
Date Sun, 11 Nov 2007 12:40:17 GMT
On Nov 11, 2007 12:58 PM, Eranb <eranb@personeta.com> wrote:

> Hi Xavier,
>
> Thank you for your prompt reply.
>
> However, it seems that I was not very clear.
> I actually ment the lower bound.
> To be more specific - in case the version in the release repository is:
> 4.2.0.3 (just as the lower bound) it is rejected.
>
> It seems to me like it is a critical bug. For first release versions are
> never found. Further more - according to Ivy documentation for version
> range
> using '[' the first version should be included !

Indeed, it should, and that's what the equality comparison at the beginning
of the comparator should make work. But if you say it doesn't, it's a bug.
Maybe the bug is in the equals implementation, I don't know. Please open a
JIRA issue, and try to give as much detail as possible on your environment.
A junit test case reproducing the issue would be great! As you can see in
our current unit test for version range matcher [1], all cases should be
tested, so I think the problem happens only in some specific case.

Xavier
[1]
https://svn.apache.org/repos/asf/incubator/ivy/core/trunk/test/java/org/apache/ivy/plugins/version/VersionRangeMatcherTest.java



>
>
> Thanks,
>       Eran B.
> -----Original Message-----
> From: Xavier Hanin [mailto:xavier.hanin@gmail.com]
> Sent: Sunday, November 11, 2007 12:50 PM
> To: ivy-user@incubator.apache.org
> Subject: Re: Ivy VersionRangeMatcher lower range bug [4.2.0.3,4.2.+]
>
> Indeed, the version range matcher has not been thought to handle another
> dynamic revision as one bound, so I'm not surprised it doesn't work. Open
> an
> issue in JIRA, but IMO this is more an undocumented  limitation than a
> bug.
>
> Xavier
>
> On Nov 11, 2007 10:50 AM, Eran Bartenstein <eranb@personeta.com> wrote:
>
> > Hi,
> >
> > In my version it seems that there is a bug in the VersionRangeMatcher.
> > In case the found version is identical to the lower version it is being
> > discarded by the matcher and the jar is not found.
> > I checked in the code and it seems that there is a bug in the following
> > code:
> >
> > In VersionRangeMatcher (L:83)
> >
> > The following code doesn't return 0 in case the modules (art1,art2) are
> > identical in content but not in instances:
> >
> >        private final Comparator COMPARATOR = new Comparator() {
> >                public int compare(Object o1, Object o2) {
> >                        if (o1.equals(o2)) {
> >                                return 0;
> >                        }
> >                        ArtifactInfo art1 = new
> > MRIDArtifactInfo((ModuleRevisionId)o1);
> >                        ArtifactInfo art2 = new
> > MRIDArtifactInfo((ModuleRevisionId)o2);
> >                        ArtifactInfo art =
> > getLatestStrategy().findLatest(new ArtifactInfo[] {art1,art2}, null);
> >                        return art == art1 ? -1 : 1;
> >                }
> >        };
> >
> > Thanks,
> >
> > Eran Bartenstein
> > SCE Team leader
> >
> >
> >        1 Hanagar Street
> >        P.O.B. 7188
> >        Hod Hasharon 45241 Israel
> > *       Phone : 972 9 7627734
> > 7       Fax     : 972 9 7412060
> > *       Eranb@personeta.com
> >      www.personeta.com <http://www.personeta.com/>
> >
> >
> > Always look on the bright side of life, for:
> > 'you come from nothing,
> > you go back to nothing,
> > what did you lose ?
> > NOTHING'
> > (Monty Python)
> >
> >
>
>
> --
> Xavier Hanin - Independent Java Consultant
> http://xhab.blogspot.com/
> http://ant.apache.org/ivy/
> http://www.xoocode.org/
>
>


-- 
Xavier Hanin - Independent Java Consultant
http://xhab.blogspot.com/
http://ant.apache.org/ivy/
http://www.xoocode.org/

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message