Re: New test failure.

[John Darrington]

On Tue, Feb 16, 2010 at 07:31:01AM -0800, Ben Pfaff wrote:
     John Darrington <address@hidden> writes:
     
     > For some reason I'm now getting the following problem on make check:
     >
     >> error: CHISQUARE test specified 6 expected values, but 4 distinct 
values were encountered in variable y.
     > compare errors 2
     > FAILED
     > FAIL: tests/command/npar-chisquare.sh
     >
     > Does anyone else get that, or is it just me?
     
     The build last night didn't show that:
     http://benpfaff.org/~blp/pspp-master/20100216040502
     more specifically:
     http://benpfaff.org/~blp/pspp-master/20100216040502/log.html#6812

That is odd.

        activity="run program 2"
        $SUPERVISOR $PSPP -o pspp.csv $TESTFILE  > $TEMPDIR/output
        if [ $? -eq 0 ] ; then no_result ; fi

        activity="compare errors 2"
        perl -pi -e 's/^\s*$//g' $TEMPDIR/output
        diff -b  $TEMPDIR/output - << EOF
        error: CHISQUARE test specified 6 expected values, but 4 distinct 
values were encountered in variable y.
        EOF
        if [ $? -ne 0 ] ; then fail ; fi

The problem (on my machine) is that the error message (which the test expects 
to see) 
is no longer going to stdout, but to pspp.csv
Consequently, the "compare errors 2" stage fails (as it should).

I wonder why this doesn't occur on other machines?

J'

-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://pgp.mit.edu or any PGP keyserver for public key.

signature.asc
Description: Digital signature