Skip Menu | You are currently an anonymous guest. | Login | Return to Main | About rt.cpan.org
 

Please report any issues with rt.cpan.org to rt-cpan-admin@bestpractical.com.

X Report information
Id: 29423
Status: resolved
Left: 0 min
Priority: 0/0
Queue: XML-DOM-XPath

Owner: MIROD <MIROD [...] cpan.org>
Requestors: d.paleino [...] gmail.com
Cc:
AdminCc:

Severity: Important
Broken in: 0.13
Fixed in: 0.14



X History Display mode: Brief headersFull headers
#   Mon Sep 17 09:16:29 2007 d.paleino[...]gmail.com - Ticket created  
Subject: make test - t/test_non_ascii.t fails
[text/plain 2.5k]
Hi,
I'm a member of the Debian Perl Group, and while packaging your module,
I've found that it fails the testsuite:

--8<--

t/test_non_ascii......................
# Failed test 'toString (on file)'
# in t/test_non_ascii.t at line 15.
# got: '<doc><ent>aü</ent><char>bü</char></doc>
# '
# expected: '<doc><ent>a�</ent><char>b�</char></doc>
# '
t/test_non_ascii......................ok 2/10

# Failed test 'findnodes_as_string( '//char') (on file)'
# in t/test_non_ascii.t at line 17.
# got: '<char>bü</char>'
# expected: '<char>b�</char>'
t/test_non_ascii......................NOK 3

# Failed test 'findvalue( '//ent') (on file)'
# in t/test_non_ascii.t at line 18.
# got: 'aü'
# expected: 'a

# Failed test 'findnodes_as_string( '//ent') (on file)'
# in t/test_non_ascii.t at line 19.
# got: '<ent>aü</ent>'
# expected: '<ent>a�</ent>'

# Failed test 'toString (on string)'
# in t/test_non_ascii.t at line 27.
# got: '<doc><ent>aü</ent><char>bü</char></doc>
# '
# expected: '<doc><ent>a�</ent><char>b�</char></doc>
# '
t/test_non_ascii......................NOK 6

t/test_non_ascii......................NOK 8# Failed test
'findnodes_as_string( '//char') (on string)'
# in t/test_non_ascii.t at line 29.
# got: '<char>bü</char>'
# expected: '<char>b�</char>'

# Failed test 'findvalue( '//ent') (on string)'
# in t/test_non_ascii.t at line 30.
t/test_non_ascii......................NOK 9# got: 'aü'

# expected: 'a

# Failed test 'findnodes_as_string( '//ent') (on string)'
# in t/test_non_ascii.t at line 31.
t/test_non_ascii......................NOK 10# got:
'<ent>aü</ent>'
# expected: '<ent>a�</ent>'
# Looks like you failed 8 tests of 10.
t/test_non_ascii......................dubious

Test returned status 8 (wstat 2048, 0x800)
DIED. FAILED tests 1, 3-6, 8-10
Failed 8/10 tests, 20.00% okay

--8<--

This happens with LANG=it_IT.UTF-8. We also tried other charsets, like
ISO-8859-15, ISO-8859-1 and LANG=C, but neither worked.
Is there a solution to this? This problem is stopping your module from
being included in Debian repositories. And no, skipping 'make test'
cannot be a solution, we are used to run make test when building all our
packages (this is to ensure that each package entering Debian is fully
functional).

Kindly,
David Paleino
Debian Perl Group
http://pkg-perl.alioth.debian.org/
#   Tue Sep 18 03:07:35 2007 MIROD - Status changed from 'new' to 'open'  
#   Tue Sep 18 03:07:37 2007 MIROD - Given to MIROD  
#   Tue Sep 18 03:42:41 2007 MIROD - Correspondence added  
[text/plain 1k]
On Mon Sep 17 09:16:29 2007, d.paleino[...]gmail.com wrote:

> I'm a member of the Debian Perl Group, and while packaging your module,
> I've found that it fails the testsuite:

> t/test_non_ascii......................
[...]
> DIED. FAILED tests 1, 3-6, 8-10
> Failed 8/10 tests, 20.00% okay

> This happens with LANG=it_IT.UTF-8. We also tried other charsets, like
> ISO-8859-15, ISO-8859-1 and LANG=C, but neither worked.
> Is there a solution to this? This problem is stopping your module from
> being included in Debian repositories. And no, skipping 'make test'
> cannot be a solution, we are used to run make test when building all our
> packages (this is to ensure that each package entering Debian is fully
> functional).

I HATE unicode related bugs! I can't seem to replicate that bug, as I
only have en locales installed (ironic considering I am in Italy).

Could you try adding the -Cio flag at the top of t/test_non_ascii.t and
let me know if it improves things by any chance? Thanks.

BTW, I would never have suggested that you skipped the tests! ;--)

__
mirod
#   Tue Sep 18 04:25:55 2007 d.paleino[...]gmail.com - Correspondence added  
Subject: Re: [rt.cpan.org #29423] make test - t/test_non_ascii.t fails
Date: Tue, 18 Sep 2007 10:20:03 +0200
To: bug-XML-DOM-XPath[...]rt.cpan.org
From: David Paleino <d.paleino[...]gmail.com>
[text/plain 1.2k]
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi :),

Il giorno Tue, 18 Sep 2007 03:42:45 -0400
" via RT" <bug-XML-DOM-XPath[...]rt.cpan.org> ha scritto:

> I HATE unicode related bugs! I can't seem to replicate that bug, as I
> only have en locales installed (ironic considering I am in Italy).

Can't you try installing them just to try? :)

> Could you try adding the -Cio flag at the top of t/test_non_ascii.t and
> let me know if it improves things by any chance? Thanks.

Did you mean changing:

#!/usr/bin/perl -w

to

#!/usr/bin/perl -w -Cio

? No, it didn't help at all. :(

But... I have a question: since you have the en locales, shouldn't our LANG=C
avoid the error?

> BTW, I would never have suggested that you skipped the tests! ;--)

Well, most users would have done that ;)

> __
> mirod

Have a nice day,
David

- --
. ''`. Debian maintainer | http://snipurl.com/gofoxygo/
: :' : Linuxer #334216 | http://www.hanskalabs.net/
`. `'` GPG: 1392B174 | http://www.debianizzati.org/
`- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFG74pC5qqQFxOSsXQRAoRoAJ4xduzULkClDnWVtoodVpgO+4W35gCgkjKh
3yHAIkfvSXwCYzeUUUDAiX8=
=R/fF
-----END PGP SIGNATURE-----

#   Tue Sep 18 05:25:44 2007 d.paleino[...]gmail.com - Correspondence added  
Subject: Re: [rt.cpan.org #29423] make test - t/test_non_ascii.t fails
Date: Tue, 18 Sep 2007 11:23:53 +0200
To: bug-XML-DOM-XPath[...]rt.cpan.org
From: David Paleino <d.paleino[...]gmail.com>
[text/plain 1k]
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Il giorno Tue, 18 Sep 2007 03:42:45 -0400
" via RT" <bug-XML-DOM-XPath[...]rt.cpan.org> ha scritto:

>
> ...
>
> I HATE unicode related bugs! I can't seem to replicate that bug, as I
> only have en locales installed (ironic considering I am in Italy).

I've just tried:

LANG=en_US.UTF-8 make test
LANG=en_US.ISO-8859-1 make test
LANG=en_US.ISO-8859-15 make test

(and also with LANG_ALL=)

but neither worked. Which locale are you actually using (I'm supposing that
the test is not failing for you)?


Kindly,
David

- --
. ''`. Debian maintainer | http://snipurl.com/gofoxygo/
: :' : Linuxer #334216 | http://www.hanskalabs.net/
`. `'` GPG: 1392B174 | http://www.debianizzati.org/
`- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFG75ks5qqQFxOSsXQRAsBHAJ4/fBbWs6Z52QtApOQwCMIHHuB1nQCguUds
Sw9H3rpnJHM9u1weV2Vo6KU=
=p4fF
-----END PGP SIGNATURE-----

#   Wed Sep 19 17:11:55 2007 d.paleino[...]gmail.com - Correspondence added  
From: d.paleino[...]gmail.com
[text/plain 253b]
A member of the group (Gregor Herrmann <gregor+debian[...]comodo.priv.at>)
solved this bug, just use:

use encoding 'utf8';

instead of:

use utf8;

in t/test_non_ascii.t to make the test run correctly.

Please, fix this in the next release.

Kindly,
David

#   Thu May 07 03:14:51 2009 MIROD - Status changed from 'open' to 'resolved'  
#   Thu May 07 03:14:51 2009 MIROD - Fixed in 0.14 added