split query on tags

[sicario]

Hi,

givens lines of text such as:

the cat sat on the mat
the <var> sat on the mat
the <var> sat on the <my object>
the <var> sat on the <my object> and <action taken by cat>

I want to be able to return all instances of my variables (anything within <>). Sometimes the variable will be one word only (i.e <var>, but it may be several (i.e <action taken by cat>)


I can successfuly split the text for one word variables, and return it as follows:(where $array[$j] is a line of text)

@words= split /(<\w+>)/,$array[$j];


To make multiple variable matches I have tried

@words= split /(<(\w+\s*)+>)/,$array[$j];

so I'm saying word ,0 or more spaces, repeat multiple times. I've tried multiple variants but can't nail this one down.


any suggestions please?

[Rincevent]

Not sure I follow the example, do you mean all the information you'll want to retrieve is between < and >?

because the split you're making should retrieve anything but the < and >
That's how i understand @words= split /(<(\w+\s*)+>)/,$array[$j] anyway?

From my attempts at split (not used to it, usually I use classical match),
seems anything you put in a group ( ) is considered as "not matching", or something.

so something very simple like /^the (.+) sat on the (.+) and (.+)$/ could achieve what you want? (when I tried it, I did get an empty first entry in the array, and I don't know what caused it).

Otherwise, going through a classical match:
@words = ($var =~ /^the (.+) sat on the (.+) and (.+)$/)
would get you the parameters in one version, or for the other version:
@words = ($var =~ /<([^>]+)>/g);

But maybe I'm not getting what you're trying to do altogether?

[KevinADC]

Using capturing parenthesis in the split functions regexp will capture the pattern along wth the rest of the data, but it's not much value unless you can differentiate the captured groups from the captured data, for example:

$_ = 'this <is> a test <to see> what happens';
@d = split(/<([\w ]+)>/);
print "$_\n" for @d; 

returns:

this 
is
 a test 
to see
 what happens

So an expression like Rincevent posted should be more like what you want:

@words = ($var =~ /<([^>]+)>/g);

or without the extra parenthesis:

@words = $var =~ /<([^>]+)>/g;