lucene-pylucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "herb@orcatec.com" <h...@orcatec.com>
Subject Re: Test failures
Date Fri, 21 May 2010 18:05:23 GMT
Thanks, Andi.
Here is a piece of the test output.  Maybe I'm misreading the separations
between tests.  I re-ran the tests.

Ran 1 test in 0.035s

OK
/usr/bin/python samples/LuceneInAction/CompoundVersusMultiFileIndexTest.py
Compound Time : 0:00:00.686435
Multi-file Time: 0:00:00.826449
F
======================================================================
FAIL: testTiming
(lia.indexing.CompoundVersusMultiFileIndexTest.CompoundVersusMultiFileIndexTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/home/herb/pylucene-2.9.2-1/samples/LuceneInAction/lia/indexing/CompoundVersusM
ultiFileIndexTest.py", line 61, in testTiming
    self.assert_(cTiming > mTiming)
AssertionError

----------------------------------------------------------------------
Ran 1 test in 1.516s

FAILED (failures=1)
/usr/bin/python samples/LuceneInAction/DistanceSortingTest.py
Sorted by: <custom:"location":
org.apache.pylucene.search.PythonSortComparatorSource@4a53fb57>
  Los Betos @ (9,6) -> 4.12310562562
  Cafe Poca Cosa @ (5,9) -> 5.09901951359
  Nico's Taco Shop @ (3,8) -> 7.28010988928
..
----------------------------------------------------------------------
Ran 2 tests in 0.086s

OK
/usr/bin/python samples/LuceneInAction/DocumentDeleteTest.py
......
----------------------------------------------------------------------
Ran 6 tests in 1.068s

OK

 

 

On May 21, 2010 at 5:37 PM Andi Vajda <vajda@apache.org> wrote:

>
> On Fri, 21 May 2010, Herbert Roitblat wrote:
>
> > Thanks, Andi.
> > The distance test did say that it failed.
>
> The test output you sent for the distance test doesn't seem to imply that it
> failed:
>    >> DistanceSorting Test
>    >> Sorted by: <custom: "Location":
>    >> org.apache.pylucene.search.PythonSortComparatorSource@4f9a32e0>
>    >> Los Betos @ (9,6) -> 4.12310562562
>    >> Cafe Poca @ (5,9) -> 5.09901951359
>    >> Nico's Taco Shop @ (3,8) -> 7.28010988928
>
> I could be missing something but the output looks correct, the restaurants
> are sorted in increasing distance.
>
> The final "make: *** [test] Error 123" you got just indicates that not all
> tests passed. Each individual test reports its own failure details at the
> time of running unless something else is seriously wrong just as core
> dumping.
>
> Andi..
>
> >
> > ----- Original Message ----- From: "Andi Vajda" <vajda@apache.org>
> > To: <pylucene-dev@lucene.apache.org>
> > Sent: Thursday, May 20, 2010 11:12 PM
> > Subject: Re: Test failures
> >
> >
> >>
> >> On May 20, 2010, at 21:21, Herbert Roitblat <herb@orcatec.com> wrote:
> >>
> >>> I'm trying to set up PyLucene on Ubuntu 10.04 with Python 2.6.2.  I
> >>> managed to get it to build, but the test revealed some errors.  Are  they
> >>> significant?  Do they need some adjustment to run under Python  2.6 or
> >>> Ubuntu 10.04?
> >>>
> >>> Thanks,
> >>> Herb
> >>>
> >>> CompoundTime : 0:00:00.763478
> >>> Multi-file Time : 0:00:00.818167
> >>> FAIL testTiming
> >>> (lia.indexing.CompoundVersusMultiFileIndexTest.CompoundVersusMultiFileInde
> >>> xTest
> >>
> >>
> >>> self.assert_(cTiming > mTiming)
> >>> Assertion Error
> >>
> >> That test is known to fail at times since it is expected for the multi file
> >> test to win, yet the timings are so close that any disturbance on the
> >> machine can slow the test down and skew the timing. The failure is
> >> ignorable. Running the test again usually passes it too.
> >>
> >>> DistanceSorting Test
> >>> Sorted by: <custom: "Location":
> >>> org.apache.pylucene.search.PythonSortComparatorSource@4f9a32e0>
> >>> Los Betos @ (9,6) -> 4.12310562562
> >>> Cafe Poca @ (5,9) -> 5.09901951359
> >>> Nico's Taco Shop @ (3,8) -> 7.28010988928
> >>
> >> This test didn't fail, did it ?
> >>
> >> Andi..
> >>
> >>> make: *** [test] Error 123
> >>>
> >>
> >
Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message