#!/usr/bin/perl -w
####################
# findlinks
#    Homework exercise to find and print links in an html file.
#    Usage:  ./findlinks file.html
#    
#    Here I have chosen to define a "link" as any string
#    of url-ish characters (as defined below) 
#    that ends in .html or .htm , regardless of whether
#    it's in an <a href=...> tag or not.  
#    
#    Jim Mahoney, Oct 2001
######################
use strict;

local $/ = undef;    # Undefining the input record seperator means we can now
my $html = <>;       # read in the entire file all at once.

# The following regular expression (stuff inside the m{} ) matches
#   \b       word boundry
#   [\w./:]   any single word character, period, slash, or colon
#   +        lots of them in a row (at least 1)
#   \.       a period
#   htm      the letters "htm"
#   l?       optionally, the letter l
# The ig at the end means "ignore case" and "global match".
# In scalar context, "global match" means that each time 
# we look, we pick up after the last successful match.
# The special variable $& is set to the string that is matched.
while ( $html =~ m{\b[\w./:]+\.html?}ig ) {
  print $&, "\n";
}