Does the XML document define what code page it's in?
John Goodyear
ATS Benchmarking Team
Gaithersburg, MD.
| [EMAIL PROTECTED]
12/12/2003 09:39 AM
|
To: ANDREW BOYNE <[EMAIL PROTECTED]> cc: [EMAIL PROTECTED] Subject: Re: XML::SAX, XML::SIMPLE on os390(Z/OS) |
ANDREW BOYNE <[EMAIL PROTECTED]> wrote on 12/10/2003 06:51:25 AM:
> Hi,
>
> We are having difficulty configuring XML-SAX-0.12 on Zos 1.3. I
> believe it may be down to lack of support for the EBCDIC codepage.
> As this perl module is required for XML-Simple it is essential we
> get it working to be able to continue our work. Could you please
> give us some help?
>
>
> Below are the results of the 'make test' :
>
> make[1]: Entering directory `/u/boyne/perl/perl-5.8.1/mods/XML-
> SAX-0.12/XML-SAX-Base'
> make[1]: Leaving directory `/u/boyne/perl/perl-5.8.1/mods/XML-SAX-0.
> 12/XML-SAX-Base'
> PERL_DL_NONLAZY=1 ../../perl "-MExtUtils::Command::MM" "-e"
> "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
> t/00basic........ok
> t/01known........ok
> t/10xmldecl1.....ok 3/5Cannot find encoding "EBCDIC" at
> /u/boyne/perl/perl-5.8.1/mods/XML-SAX-0.12
/blib/lib/XML/SAX/PurePerl/Reader/
> UnicodeExt.pm line 16.
> Document requires an element [Ln: 1, Col: 5]
This test failure message certainly looks like XML::SAX is having
trouble with EBCDIC. It is interesting to note that it is smart
enough to recognize that it is running on an EBCDIC coded
character set, but not smart enough to do anything clever about it.
My suspicion is that someone will have to make an important
decision regarding coded character sets acceptable to be
called "XML" and where the translation boundaries (if any)
are to be for processing the XML on z/OS.
With that decision in hand it would then be possible to
treat XML::SAX. For example if it were decided that
XML on z/OS is ok to be in code page 1047 and treated like
ISO-Latin on an ASCII platform arrangements could be made to
translate where necessary between those code pages.
If it needed to be UTF-EBCDIC and treated analogous to
UTF-8 on ASCII platforms then a different translation
would be needed. If it were decided that all XML on z/OS
was to literally be in UTF-8 or UCS-2 then a different
translation table would be needed.
Which character set encoding(s) is(are) the prevailing
preference(s) for XML on z/OS?
Peter Prymmer
