I'm not sure why your code behaves this way when executed as a CGI script (I'd want to see more of the code to figure that out), but I will point out that you are not using the right content-type. You're printing a content-type of text/html, when your content is actually text/plain. | [reply] |
Here is the entire code.. The code works.. It's just that fact that it seemingly executes twice via a web browser like I mentioned before.
#!/usr/bin/perl
<BR>
use Lingua::Ispell;<BR>
print "Content-type: text/plain\n\n";<BR>
print "beginning\n\n";<BR>
$line="delte five";<BR>
spellit();<BR>
sub spellit<BR>
{<BR>
$spellline=$line;<BR>
$realine=$line;<BR>
for my $r ( Lingua::Ispell::spellcheck($spellline) ) {<BR>
print "right after <BR>$spellline\n ";<BR>
if ( $r->{'type'} eq 'miss' ) {<BR>
print "$r->{'term'}";<BR>
} <BR>
} <BR>
}<BR>
This is the output via web browser activation:
beginning Content-type: text/plain beginning right after delte five delte
Output after running via command prompt:
beginning Content-type: text/plain beginning right after delte five delte | [reply]
[d/l] |
I'm a little confused... In your original post, you said that, when run as a CGI, the program output everything twice. But in the sample output you posted in your second node, you show Content-type: text/plain beginning output twice, but the list of mispelled words output only once.
If the latter is the case, then I suspect that Lingua::Ispell is forking off a subprocess to handle the spell-checking, before the output buffers have been flushed, and so everything that was printed before the fork gets output twice.
The solution is to turn on autflushing, by setting $| = 1 at the top of your program, before you print anything.
(I believe that the latest version of Perl avoids this problem by automatically flushing output buffers when it's about to fork.)
| [reply]
[d/l] |