Sune Karlsson writes
> In which way does it fail? Is the configure script executed?
It does not look like it. Compare
oaiadm@amorp:~$ cd /tmp/ ; rm -rf ReDIF-perl-2.75/ ; tar xzf /home/adrepec/hubec.repec.org/ReDIF-perl/ReDIF-perl-2.75.tar.gz ; cd ReDIF-perl-2.75/
oaiadm@amorp:/tmp/ReDIF-perl-2.75$ perl Makefile.PL
Preparing to install ReDIF perl package.
Configure failed, so do I at Makefile.PL line 5.
to
oaiadm@amorp:~$ cd /tmp/ ; rm -rf ReDIF-perl-2.75/ ; tar xzf /home/adrepec/hubec.repec.org/ReDIF-perl/ReDIF-perl-2.75.tar.gz ; cd ReDIF-perl-2.75/
oaiadm@amorp:/tmp/ReDIF-perl-2.75$ mv Configure Configure.org
oaiadm@amorp:/tmp/ReDIF-perl-2.75$ perl Makefile.PL
Preparing to install ReDIF perl package.
Configure failed, so do I at Makefile.PL line 5.
--
Cheers,
Thomas Krichel http://openlib.org/home/krichel
skype:thomaskrichel
Perl 5.24 is walzing in.
It appears that at least in my vesion on amorp, the line
do("Configure");
fails. I had to replace it with
eval `cat Configure`;
even tough it looks that the description of do() in
man perlfunc has not changed. This may be a bug in
Perl itself.
--
Cheers,
Thomas Krichel http://openlib.org/home/krichel
skype:thomaskrichel
Kit writes
> I thought we were now also allowing .redif, but cannot find any discussion
> of that. We have a provider in BE (the stm archive) which is producing
> files that are indeed UTF-8 but with no BOM. They are being mangled (see
> 15.rdf). If I download that file, it shows up as proper UTF-8 in my text
> editor. Nevertheless, the content is mangled in both IDEAS and EconPapers.
> Whatsa hoppnin? Do we indeed still require a BOM?
It's late here.
I think the spec says or means to say that if the file is .redif, no
BOM is requride, if the file ends .rdf, a BOM is required.
I then suggested that people working in the Mircosoft wold should be
told to use .rdf, and in the *nix world should be encouraged to use
.redif.
I am not aware of a reaction to my suggestion.
'Christian Zimmermann' writes
> True. I was waiting for Thomas to write a blog post about this as a way to
> introduce the change to the world. He has not done that yet.
The change is introduced to the world with the change in the spec. But
there is still this suggestion that is not been decided on.
--
Cheers,
Thomas Krichel http://openlib.org/home/krichel
skype:thomaskrichel
Sune Karlsson, on CitEc-run, wrote
> That was not a terribly good fix. Turns out that $attribute can be undefined so
>
> $value = decode_and_strip_html( $value ) unless defined $attribute and $attribute eq 'x-bibl';
>
> is better unless you want a lot of warnings about $attribute being undefined.
What abouot disabling stripping for all x-attributes
$value = decode_and_strip_html($value)
unless defined $attribute and lc(substr($attribute,0,2)) eq 'x-');
--
Cheers,
Thomas Krichel http://openlib.org/home/krichel
skype:thomaskrichel
Jose Manuel Barrueco writes
>
> > > adnetec@shuli:~$ rere RePEc/remo/per/pers/s/psc418.rdf
> > >
> > > the output is not utf8
> >
> > Well rere is a very old executable.
>
> well, this doesn't seem to be a good answer :)
I made no claims to the quality of the answer.
> any way, I get the same result when I use the perl module.
It's an rdf file with a utf-8 BOM. It should really be a
ReDIF file.
> Any idea about how to revert to the previous status in the meanwhile...
This will not read any .redif files.
Do you have a minimal test handy? I may hit the sack though
right now.
--
Cheers,
Thomas Krichel http://openlib.org/home/krichel
skype:thomaskrichel