At 07:53 AM 1/2/2001 -0500, lane @ DUPHY4.Physics.Drexel.Edu wrote:
>Here's a different patch for the same problem...it looked to me that
>the line-length problem was not occuring for the dependancy line
>(after all, that's parsed by MMS/K and not by DCL), but rather for the
>Perl command line. BUT, Perl has no problem doing wildcard expansion,
Is that true on all platforms?
>so I give it a wildcard and let it do the work. The result is that
>all of the ISO8859*.enc files are compiled into ISO8859.c...
>
>...and just now trying the Encode test shows no "loading" messages...
>
>(this is a patch to Perl@8269, a bit simpler than Craig's but to
>accomplish the same overall task)
Thanks, I had seen your earlier patch but the ground had shifted under us in the
meantime. This one here has the advantage of simplicity. I've sent Nick a revision
of mine that dumps the list of encoding files to a temp file and then has "compile"
read them in. It's more complex than yours but has no $^O eq 'VMS' blocks and should
allow the list of encodings to grow arbitrarily long. In any case we've got more than
one way to get the thing built and Nick has options on how to make it work on VMS.
>diff -uBb ext/Encode/Makefile.PL-orig ext/Encode/Makefile.PL
>--- ext/Encode/Makefile.PL-orig Sun Dec 31 00:35:25 2000
>+++ ext/Encode/Makefile.PL Sun Dec 31 00:55:27 2000
>@@ -82,7 +82,12 @@
> $str .= "\n\t\$(PERL) compile \$\@";
> foreach my $file (@{$tables{$table}})
> {
>+ next if $^O eq 'VMS' && $file =~ /iso8859.*\.enc/i;
> $str .= ' '.$self->catfile($dir,$file);
>+ }
>+ if ($^O eq 'VMS' && $table eq 'iso8859')
>+ {
>+ $str .= ' [.Encode]iso8859*.enc';
> }
> $str .= "\n\n";
> }
