in reply to feedback on successful module loading

Well, this is what I've settled on. Works great. Now I can upload variations of this to the several ISPs I'm involved with and be able to tell which modules they have available.
Many thanks to you all! 
#!/opt/bin/perl -w
print "running $0 in Perl $]\n";
 
my $result = eval "use CGI; 1";
if ($result) {
  print "CGI loaded\n";
} else {
  print "CGI failed\n";
}
print "$@\n";
 
my $result = eval "use HTML::Template; 1";
if ($result) {
  print "Tmplt loaded\n";
} else {
  print "Tmplt failed\n";
}
print "$@\n";

[download]
ggg

Replies are listed 'Best First'.
•Re^2: feedback on successful module loading
by merlyn (Sage) on Oct 07, 2004 at 21:22 UTC
    You should wrap all that in a BEGIN block if you want the imports from those "use" to act in a traditional fashion. Otherwise, you've essentially turned a use into a require, sorta.

    -- Randal L. Schwartz, Perl hacker
    Be sure to read my standard disclaimer if this is a reply.

[reply]
Re^2: feedback on successful module loading
by diotalevi (Canon) on Oct 07, 2004 at 22:34 UTC

    You can make your code shorter by writing your code only once and putting your modules to test into a list.

    #!/opt/bin/perl -w
print "running $0 in Perl $]\n";
for ( qw( CGI
         HTML::Template ) )
{
    print eval "require $_"
             ? "$_ ok\n"
             : "$_ failed: $@\n"
}

    [download]
[reply]
[d/l]
      I was going to work on putting the module names into a DATA file and building the working script with another script. The result would have been just as long, but I wouldn't have written it by hand. I like your idea better. It has class!
      ggg
[reply]

In Section Seekers of Perl Wisdom