You need to read the regex documentation. perlretut is a good place to start.
Your immediate problem is using [...] instead of (...) for capturing. The square bracket version matches any one of the characters in the set of characters given.
Premature optimization is the root of all job security
| [reply]
[d/l]
[select] |
You haven't asked a question, nor explained what isn't working, so this is a guess.
my $row = 'We_need_feed';
#my ($last) = $row =~ /[^_]$/g; #match we
# [^_] matches everything except underscore, but only one character;
# Plus you've anchored it to the last character before the end of the
+string,
# so it will always and only match the last character of the string;
# unless it is an underscore, in which case it fails to match at all.
#my ($last) = $row =~ /^\w?[^_]/g;#match feed
# ^\w? matches a single non-whitespace character at the start of the s
+tring,
# if the first character isn't whitespace, otherwise it matches nothin
+g.
# [^_] matches a single non-underscore as above.
# Ie. You matched two characters "We"
my ($last) = $row =~ /[need]+$/g ;
# [need]+ matches a one or more characters, so long as they are either
+ 'd' or 'e' or 'n'; other wise nothing
# but the $ means only at the of the line.
# so this matched the 3-char string 'eed'.
- To match the first word, delimited by underscore (_), you could use /^[^_]+/.
- To match the last word delimited by underscore, you could use /[^_]+$/.
- To match the word 'need', somewhere in the line you could use /need/.
You'll get better answers if you tell us what you need to know in words as well as code. Relying on us to read between the lines of your non-functional code and comments to extract the meaning and question, is unlikely to get the best answers. Indeed, If I wasn't bored out of my mind waiting for a process that has been running for 50+ hours to finish, I wouldn't have bothered answering your lazy post at all.
With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
In the absence of evidence, opinion is indistinguishable from prejudice.
| [reply]
[d/l]
[select] |
thank you very much but:
To match the word 'need', somewhere in the line you could use /need/. would just return 1 or 0, so /(need)/ would match the word need.Thank you for your explanation :)
| [reply] |
so /(need)/ would match the word need.
... and return the matched word 'need'; but yes.
However, what is the purpose of capturing the matched word when you already know what it is?
Ie. If your intent is to do:
my( $foundWord ) = $string =~ /(need)/;
That is equivalent to doing: my $foundWord = $string =~ /need/ ? 'need ' : undef;
The point being that capturing is expensive and the need to capture a known constant string is strongly indicative of flawed logic.
With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
In the absence of evidence, opinion is indistinguishable from prejudice.
| [reply]
[d/l]
[select] |
I can offer one additional comment. The /g is unnecessary for your example. (The anchors ^ and $ each refer to a single place in the string.) It may be needed if your real data is a multi-line string, but then you would have other issues to consider.
For your example, you can use:
use strict;
use warnings;
my $row = 'We_need_feed';
my($first_word, $last_word) = $row =~ /^(\w+)_.*_(\w+)$/;
print $first_word, ' ', $last_word;
| [reply]
[d/l] |
All the regex operators you're using in the OPed examples are supported by Perl version 5.6 and before, so you could also have used YAPE::Regex::Explain (which only supports regexes through Perl version 5.6) to give you some insight:
c:\@Work\Perl>perl -wMstrict -le
"use YAPE::Regex::Explain;
;;
print YAPE::Regex::Explain->new(qr/[need]+$/)->explain;
"
The regular expression:
(?-imsx:[need]+$)
matches as follows:
NODE EXPLANATION
----------------------------------------------------------------------
(?-imsx: group, but do not capture (case-sensitive)
(with ^ and $ matching normally) (with . not
matching \n) (matching whitespace and #
normally):
----------------------------------------------------------------------
[need]+ any character of: 'n', 'e', 'e', 'd' (1 or
more times (matching the most amount
possible))
----------------------------------------------------------------------
$ before an optional \n, and the end of the
string
----------------------------------------------------------------------
) end of grouping
----------------------------------------------------------------------
(Note, however, that this explanation does not make clear the point already made by BrowserUk ++here that the [need] character set only encompasses three characters n e d and has nothing to do with a sequence 'need'.)
Give a man a fish: <%-{-{-{-<
| [reply]
[d/l]
[select] |
Sometimes using a split can work out better than a full blown regex. Consider this:
#!/usr/bin/perl
use strict;
use warnings;
my $row = 'We_need_feed';
my ($first, $middle, $last) = split /_/, $row;
print "first=$first, middle=$middle, last=$last\n";
#Update: just demo of indices:
($first, $middle, $last) = (split (/_/,$row))[0,1,-1];
print "middle=$middle\n";
__END__
Prints:
first=We, middle=need, last=feed
middle=need
| [reply]
[d/l] |