Hi Scott
Scott E Robinson wrote:
> Dear Perl experts,
>
> I'm trying to find the right regular expressions to do some simple (?)
> string processing. Can anyone tell me how to do these?
>
> 1. Given a string consisting of substrings delimited by colons, such
> as :B520:L201:M:M260:8:G607:,
my $string = ':B520:L201:M:M260:8:G607:';
> how can I
> a. remove the single-character and all-numeric substrings (M
> and 8 in this example), leaving the rest alone?
This is an array of all the wanted substrings:
my @wanted = $string =~ /[a-z]\w+/ig;
print "@wanted\n";
output
B520 L201 M260 G607
> b. remove all but the single-character and all-numeric
> substrings and leave the rest alone?
This is an array of all the wanted substrings:
my @wanted = $string =~ /\b\w\b/ig;
print "@wanted\n";
output
M 8
> c. remove all but single- or double-character substrings and
> all-numeric substrings and leave the rest alone?
This is an array of all the wanted substrings:
my @wanted = $string =~ /\b[a-z]\w?\b/ig;
print "@wanted\n";
output
M
(Note that this won't find '8M', but your data doesn't look like
that will come up.)
> The string will never have regex metacharacters in it, just mixed
> alpha and numeric, all-alpha, or all-numeric.
That makes things a lot easier. I'm also assuming they won't contain
underscores.
> The colons can stay.
I've removed them. If you want the array turned back into a
string like you started with, use
my @wanted = $string =~ /[a-z]\w+/ig;
print join ':', '', @wanted, '';
output
:B520:L201:M260:G607:
> 2. Is there an easy way to count the number of substrings (the
> "chunks" between colons)?
The length of one of the arrays is
my $length = @wanted;
or, from one of the original strings
my $string = ':B520:L201:M:M260:8:G607:';
my $length = () = $string =~ /\w+/g;
print $length;
output
6
> 3. This one is probably a bit difficult. I don't need to have it,
> but it would save me lots of effort if I had it. Given two strings
> of the same form as in no. 1, is there a regular expression which can
> compare the two and return the number of substring positions which
> have exact matches? I.e., given string 1 = :L000:W000:M:M260:G607:
> and string 2 = :L001:W000:M:M261:M260: can match the substrings and
> their positions and return the result "2" in this case? The M260
> substring is present in both but in different positions and shouldn't
> be counted as a match.
my $string1 = ':L000:W000:M:M260:G607:';
my $string2 = ':L001:W000:M:M261:M260:';
my $matches = do {
my %pair;
@pair{$string1 =~ /\w+/g} = $string2 =~ /\w+/g;
grep $pair{$_} eq $_, keys %pair;
};
print $matches;
output
2
HTH,
Rob
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
