Re: MARC::File::XML 0.85

On 4/13/07, Joshua M. Ferraro <[email protected]> wrote:
> Hi folks,
>
> A new version of MARC::File::XML has been uploaded to CPAN. It's
> a one character fix that solves a bug for UNIMARC users who don't
> have MARC-8 encoded records. Thanks to Paul Poulain for pointing
> it out.

Unfortunately, this

    $parser->{ Handler }{ toMARC8 } = (lc($format) eq 'unimarc' &&
$enc && lc($enc) =~ /^utf-?8$/o) ? 0 : 1;

would require that BOTH the format of UNIMARC and the encoding of
UTF-8 be be used in order to write UTF-8 encoded binary MARC records.
That is, of course, unacceptable -- we need to be able to write USMARC
records in UTF-8.  Before the code gets used too widely, I suggest we
change the above line to

  $parser->{ Handler }{ toMARC8 } = (lc($format) =~ /^unimarc/o || (
$enc && lc($enc) =~ /^utf-?8$/o )) ? 0 : 1;

which translates to "use MARC-8 encoding if the format is NOT UNIMARC
or UNIMARCAUTH, or if the (USMARC) encoding is not UTF-8".  This
change is in CVS.  I'll wait for testing on this change as I'd rather
not rebreak things as the 0.85 change happened to do, though obviously
in an unintended manner.

On a related note, there are several other changes intended improve
support for UNIMARC authority records.  IIRC, I asked Paul to test
this, but it's been very hectic recently and I don't recall if he
responded to me or not.

One last thing: it would be nice for the current primary developer of
the module to get a heads up when changes a are made and a release is
planned.  I know we're all busy, but because it can take quite a while
for a release to actually go out after we (or, in this case, one of
us) decides that it probably should it would be good to get other eyes
on the changes one last time (or one first time, in this case) to make
sure nothing is being overlooked.  Thanks.

-- 
Mike Rylander
[email protected]
GPLS -- PINES Development
Database Developer
http://open-ils.org