Hai Jeff,
The solution you have provided is working fine. Thanks a lot.
Regards,
Ganesh
Jeff 'japhy' Pinyan wrote:
On Jun 1, N. Ganesh Babu said:
<o><r>C.A. Eschenbach, F. Hall, C.R. Johnson, Z. Li, The graphs of
unambiguous entries in the product of two <m>MATH0810075.dat:1</m>
sign patterns, Linear Algebra Appl. 260 (1997) 95--112.</r></o>
while($line=~m!MATH(\d{7}\.dat):(\d{1,2})!g)
{
$math=&math($1, $2);
$line=~s!MATH$1:$2!$math!;
print $line;
}
Two things to mention...
First, the '^' character in a regex does not match a '^', but instead
means "match the beginning of the string". So in your regex, you want
to backslash it.
m!\^AMATH\^D(\d{7}\.dat):(\d{1,2})\^B!
Second, there's a way to do:
while (m/XXX/g) {
YYY = some_function();
s/XXX/YYY/;
}
all at once. It's like this:
s/XXX/some_function()/eg;
The /e modifier means "execute the right-hand side of the s/// and use
its return value as the replacement text". So your regex would be
s!\^AMATH\^D(\d{7}\.dat):(\d{1,2})\^B!math($1,$2)!eg;
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
<http://learn.perl.org/> <http://learn.perl.org/first-response>
