lucenenet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Laimonas Simutis <lai...@gmail.com>
Subject Re: Possible Query Equals bug
Date Sun, 21 Dec 2014 22:55:18 GMT
Itamar,

I haven't run into that type of issue anywhere else, at least not yet. At
the moment looking at the other failing tests in the Facets area to try to
figure out why they are failing.


Laimonas

On Sun Dec 21 2014 at 10:21:19 AM Itamar Syn-Hershko <itamar@code972.com>
wrote:

> Yeah I can see the bug there, I was assuming GetType() will always return
> typeof(Query), which is obviously wrong.
>
> Are there other places where you see this change?
>
> --
>
> Itamar Syn-Hershko
> http://code972.com | @synhershko <https://twitter.com/synhershko>
> Freelance Developer & Consultant
> Author of RavenDB in Action <http://manning.com/synhershko/>
>
> On Sat, Dec 20, 2014 at 2:48 PM, Laimonas Simutis <laimis@gmail.com>
> wrote:
>
> > I am going through the failing tests and seeing if I can fix them. While
> > looking at the TestDrillDownQuery (
> >
> > https://github.com/apache/lucenenet/blob/master/src/
> Lucene.Net.Tests/core/Facet/TestDrillDownQuery.cs
> > )
> > I noticed that all of the failures in there are caused by a modification
> to
> > Query class implementation of Equals.
> >
> > The commit that contains possible offending code is here:
> >
> >
> > https://github.com/apache/lucenenet/commit/
> a37fd3810e630560f209175bb5897975c21c57a2#diff-
> 83ac470f4ea73edd2b7e2d6bf9190ba7
> >
> > It removes "if (this.GetType() != obj.GetType())" check that is present
> in
> > java version.
> >
> > The tests mentioned above fail because of the check not being made.
> > QueryUtils.CheckHashEquals ends up comparing queries with the same boost
> > value but of different type, expectation is for such queries to come back
> > as not equal. Without type check, they evaluate as equal.
> >
> > I added the type check back in and the tests pass.
> >
> > Should I go through branch / PR dance for such a small change? If this
> is a
> > legitimate problem and I haven't missed anything, perhaps it would be
> > easiest for one of you with commit rights to check it in.
> >
> >
> > Thank you,
> >
> > Laimonas
> >
>

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