The mailing list for listmasters using Sympa

[Victoriano Giralt <address@concealed>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi. I am testing Sympa's SOAP and have found that the SOAP service, even
when queried with sympa_soap_client.pl, if the lists subjects or gecos
fields include UTF-8 characters. The response content-length header is
smaller than the real character count in the response document, thus,
clients create too small buffers to process the XML document and validation
of said document fails.

I have verified this capturing traffic off the wire with tcpdump, checking
that the character count of the captured document and the number in the
content-length header differed and then modifying all my lists to remove
the UTF-8 characters from config files, which eliminates the problem. As
soon a the config files are restored to their normal UTF-8 content, the
problem reappears.

I have been googling and the problem seems to be that libwww is counting
*characters* instead os *bytes*, someone proposed a patch to libwww that
some libwww guro refuses and proposes that either the data should be base64
encoded (which I think could do more harm to interoperability), or the
socket should be told to use binary in case the data is not ASCII with
binmode($sock, ":utf8").

I do not know enough neither Sympa internals nor Perl to propose a patch.
Can you please help? I'm using Sympa 6.1.4, SOAP::Lite 0.70 and the libwww
of Perl 5.805.

Thanks.
- --
Victoriano Giralt
Systems Manager
Central ICT Services
University of Malaga
SPAIN
- -
A: Yes.
| > Q: Are you sure ?
|> >> A: Because it reverses the logical flow of conversation.
|>> >>> Q: Why is top posting annoying in email ?

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with CentOS - http://enigmail.mozdev.org

iD8DBQFOYeFjV6+mDjj1PTgRAqooAKDJJQx47kciwAO/dYhJjLRK+q7ypwCeJObG
ummEyxN3UyKVSr/P9nXU0xU=
=Hugx
-----END PGP SIGNATURE-----