Basically, yes, though I am going to be doing it all at once (using a technique described in a later answer) - reading in the entire file, and then processing it with:

$fragment =~ s/<foo>\d+<\/foo>/'<foo>' . $i++ . '<\/foo>'/eg;
[download]

Thank you.

All that  foo foo foo stuff... That's a lot of foo-ing around for a simple substitution — and a lot of room for error to creep in. Another way (needs Perl version 5.10+ regex extensions):

c:\@Work\Perl\monks>perl -wMstrict -le
"use 5.010;
 ;;
 my $s = qq{www<foo>  3  </foo>\n<bar>999</bar>\n<foo>14  </foo>xx\n}
       . qq{<bar>998</bar>\nyyy<foo>  159</foo>z\n<bar>997</bar>\n}
       ;
 print qq{[[$s]]};
 ;;
 { my $n = 1;
   $s =~ s{ <(foo)> \s* \K \d+ (?= \s* </\g-1>) }{ $n++ }xmsge;
   }
 print qq{[[$s]]};
"
[[www<foo>  3  </foo>
<bar>999</bar>
<foo>14  </foo>xx
<bar>998</bar>
yyy<foo>  159</foo>z
<bar>997</bar>
]]
[[www<foo>  1  </foo>
<bar>999</bar>
<foo>2  </foo>xx
<bar>998</bar>
yyy<foo>  3</foo>z
<bar>997</bar>
]]
[download]

If you want to get rid of the leading/trailing whitespace, that's easily done. If you only have pre-5.10 regexes to work with, that can be handled too, although it's a bit more messy.

---

Give a man a fish:  <%-{-{-{-<

foo...