[Apologies if this is a duplicate; my post via NNTP did not
show up here.]
I'm having the same issue with 24implicit_utf8 on HP-UX 11.11
with 32-bit Perl 5.8.8 on PA-RISC 2.0.
I tried multiple values of NLS_LANG and NLS_NCHAR, including unset.
I tried multiple versions of DBI, DBD::Oracle and the Oracle
client. I tried HP's build of 5.8.8, compiled with fairly weak
optimization. There is a post elsewhere wherein AIX exhibited this same
issue, so it doesn't seem to be confined to HP-UX.
Finally I reverted back to 5.8.7, which passed the tests.
5.6.1 also worked.
It seems that something changed in the delta between 5.8.7 and 5.8.8 which has
broken utf8 support in DBD::Oracle. Whether that triggered a bug in Perl
itself or in the DBD::Oracle module is unknown.
I would really like to get to the bottom of this so that I can use
Perl 5.8.8. Let me know if you need any further information.
| perl | DBI | DBD::Oracle | Oracle | result |
|----------+------+-------------+--------+--------|
| 5.8.8 | 1.53 | 1.19 | 9.2.0 | FAIL |
| 5.8.8 | 1.53 | 1.19 | 9.0.1 | FAIL |
| 5.8.8 | 1.53 | 1.19 | 8.1.7 | SKIP |
| 5.6.1 | 1.48 | 1.16 | 9.2.0 | OK |
| 5.8.8 | 1.53 | 1.16 | 9.0.1 | FAIL |
| 5.8.8 | 1.48 | 1.16 | 9.0.1 | FAIL |
| HP 5.8.8 | 1.50 | 1.19 | 9.2.0 | FAIL |
| HP 5.8.8 | 1.50 | 1.19 | 9.0.1 | FAIL |
| HP 5.8.7 | 1.53 | 1.19 | 9.2.0 | OK |
Jim
On Fri, 9 Mar 2007 14:01:56 -0700, "Tanner, Richard L" did write
in <[EMAIL PROTECTED]>:
> I am having a problem getting DBD-Oracle-1.19 to compile. I 'googled'
> many references to the problem identical to mine but found no solutions.
> Note We have no NLS environment variables currently set prior to running
> 'make test'.
> HP-UX 11.23 with PA-RISC2.0 Chip set
> Perl 5.8.8 configured as perl README.hpux from 5.8.8 docs and
> README.hpux.txt from Oracle DBD docs
> (non-threaded, dynamic linking, -Ae, +Z -D64 -Duselargefiles ...)
> DBI 1.54
> DBD Oracle 1.19:
> When we run 'make test' for DBD-Oracle-1.19
> t/24implicit_utf8.......
> # Failed test 'byte_string test of row 4: column: nch (uft8) smiley face'
> # in t/nchar_test_lib.pl at line 341.
> # got: '38|58'
> # expected: '226|152|186'
>
> # Failed test 'nice_string test of row 4: column: nch (uft8) smiley face'
> # in t/nchar_test_lib.pl at line 344.
> # got: '&:'
> # expected: '\x{263A}'
> # row 4: DUMP(nch) =3D Typ=3D1 Len=3D4: 0,38,0,58
>
> # Failed test 'byte_string test of row 4: column: nch (uft8) smiley face'
> # in t/nchar_test_lib.pl at line 341.
> # got: '38|58'
> # expected: '226|152|186'
>
> # Failed test 'nice_string test of row 4: column: nch (uft8) smiley face'
> # in t/nchar_test_lib.pl at line 344.
> # got: '&:'
> # expected: '\x{263A}'
> # row 4: DUMP(nch) =3D Typ=3D1 Len=3D4: 0,38,0,58
> # Looks like you failed 4 tests of 74.
--
"But when I was forced to re-examine it I saw that indeed his critique of
the argument was quite accurate. There was simply a gap in the logic of the
attempted device for assuring the avoidance of self-intersections in the
embedding of a non-compact manifold of n intrinsic dimensions in an Euclidean
space of (n+1)*(n/2)*(3*n+11) dimensions." -- Prof. John Nash
[EMAIL PROTECTED] / 0x43340710 / 517B C658 D2CB 260D 3E1F 5ED1 6DB3 FBB9 4334
0710
