It will be true if $input represents a pattern that will be a positive match against the string held in $_

The /o modifier has fallen out of favor nowadays, largely replaced by the qr// operator. But in this case what it means is that once the RE is compiled once, it won't be recompiled again, even if the value of $input changes. The choice of regexp delimiters as the '|' character seems a little ackward, as it makes alternation difficult.

perlre, perlrequick and perlretut can be helpful.

1. The m tells perl to start a match.
2. The | character is used as a delimiter. Using / is normal but any delimiter can be used. If you use / then you don't need to start with an m
3. The $input is interpolated (The content of the variable is placed in the regular expression)
   
4. The o modifier signifies that the regular expression is compiled once. Changing the value in $input will not change your regex.
5. The whole regular expression is matched against the value held in $_ by default. You would use the =~ operator to match a particular variable.

Thank you all for providing such good explanation and providing links for further study .. Thank you very much!!!

In addition to the very helpful comments you've already received, I'd like to point out that, as a rule, I'd "metaquote" the scalar variable in the regex, just in case it contains characters (e.g. +) that otherwise would have special meaning in a regular expression:

m|\Q$input|o
[download]

See the aforementioned perlre, and while you're at it also check out quotemeta.

Unless of course $input were a stored regular expression (such as qr//), where the matching behavior was expressly desired.

--
[ e d @ h a l l e y . c c ]

perldoc perlre
[download]

perlre probably contains much more than you want at this stage, but at least you would know the kind of thing that's available for when you need it in the future.