Help with a small regex

cspctec has asked for the wisdom of the Perl Monks concerning the following question:

I have the following array:

my @data = ('scp', 'sss40s1u01:/home/example/test.txt', '/home/example
+/testdir');
[download]

And I'm trying to create a regex to match "/home/example/testdir" and NOT any part of "sss40s1u01:/home/example/test.txt". I'm having some trouble coming up with a regex for this. So far, I have tried negative lookbehind:

foreach (@data) {
  my $match = $_ if /(?<!:)\/.+/;
}
[download]

But that matches "sss40s1u01:/home/example/test.txt" also.

I would appreciate any help.

Comment on Help with a small regex Select or Download Code

Replies are listed 'Best First'.
Re: Help with a small regex by hippo (Archbishop) on Feb 06, 2015 at 18:09 UTC
If you really only want to test for "/home/example/testdir" then this should work: `foreach (@data) { my $match = $_ if /testdir/; }` [download] OTOH, if the spec is actually to match against just a path then this should do: `foreach (@data) { my $match = $_ if /^\//; }` [download]	[reply] [d/l] [select]
Re: Help with a small regex by kennethk (Abbot) on Feb 06, 2015 at 18:14 UTC
The most obvious difference to me is the presence of a colon, which seems to be what you've used in your own logic. How about instead of returning true if it contains no colon, you try returning false if it contains one? `foreach (@data) { my $match = $_ if !/:/; }` [download] The issue you are having is that you need the entire string to contain no colons, but what you have tested for is actually that it contains non-colon characters. The closest code to your spec would be `foreach (@data) { my $match = $_ if /^[^:]$/; }` [download] which says that all characters between the start and end of the string are not colons. Advanced use case, this is exactly the sort of thing grep was made for: `my ($match) = grep !/:/, @data;` [download] Update*: Fixed typo in case two; thanks AnomalousMonk. #11929 First ask yourself `How would I do this without a computer?' Then have the computer do it the same way.	[reply] [d/l] [select]
Re: Help with a small regex by cspctec (Sexton) on Feb 06, 2015 at 21:02 UTC
Thanks for the help you two. I forgot to mention that "testdir" can and will change (the matches definitely won't contain "testdir"). Your solutions work, but they also match "scp". I just added `next if /scp/;` [download] to the loop along with if !/:/ to fix it.	[reply] [d/l]
Re^2: Help with a small regex by AnomalousMonk (Archbishop) on Feb 06, 2015 at 22:11 UTC
I don't see why hippo's second second suggestion above (the OTOH one) would not cover all your cases: `c:\@Work\Perl\monks>perl -wMstrict -le "for my $path ( 'scp', 'sss40s1u01:/home/example/test.txt', '/home/example/testdir' +, @ARGV, ) { print qq{'$path' matches} if $path =~ /^\//; } " "/any" "/other/examples" '/home/example/testdir' matches '/any' matches '/other/examples' matches` [download] (BTW: My own preference for the regex expression would be `m{ \A / }xms` rather than `/^\//`) Give a man a fish: `<%-(-(-(-<`	[reply] [d/l] [select]