Read data from string

I have string line:

 x='abc123(xyz456)'

How to read information only in brackets, to have result:

 y='xyz456'.

Answer 1

[Evan]

>> x='abc123(xyz456)';
>> regexp(x,'(?<=\().+(?=\))','match')
ans = 
    'xyz456'

This command uses regexp and, specifically, lookaround assertions. It's basically saying, if you find a group of characters, look behind to see if there is an "open parenthesis" character and look ahead to see if there is a "close parenthesis" character. If so, return all the characters between them.

[Azzi Abdelmalek]

This will not work if

x='abc123(xyz456)ab(cde)';

[Evan]

True. It assumes only one set of parenthesis. I used the character identifier in case of spaces, but that might not be a needed flexibility. Regardless, I suppose a better version to account for that would be similar to yours:

regexp(x,'(?<=\()[\w\s]+(?=\))','match')

[per isakson]

or

    x='abc123(xyz456)aaa(123)mmmm';
    regexp(x,'(?<=\().+?(?=\))','match')

where the added ? stands for

    Lazy expression: match as few characters as necessary.

[Cedric]

My preference goes to Per lazy expr., but I wanted to mention the following, counter-intuitive behavior:

 >> tok = regexp('_A_A-', '_(.*?)-', 'tokens')
 tok = 
    {1x1 cell}
 >> tok{1}
 ans = 
    'A_A'

and not just A, as many people would expect.

[per isakson]

Surprise!

    regexp('_A_A-', '(?<=_)[^_]+?(?=-)', 'match' )
    ans = 
        'A'

Thus, doc should say something like

    Lazy expression: match as few characters as necessary **downstream**.

[Cedric]

Yep, in other words, it stops when it matches the last part of the pattern for the first time (lazy), but it doesn't pull back the starting point (the tail? ;-)) to minimize the match (not that lazy finally, or really really lazy in fact). Thankfully, you see/understand this once and you never forget it!

Answer 2

[Azzi Abdelmalek]

 y=regexp(x,'(?<=\()[\w]+(?=\))','match')

[Azzi Abdelmalek]

%or

x=x='abc123 (xyz 45_6) ddd (rtr)ccc'
y=regexp(x,'\(([\w\s]+)\)','tokens');
celldisp(y)

Answer 3

[Jan]

x = 'abc123(xyz456)';
ini = strfind(x, '(');
fin = strfind(x, ')');
key = x(ini(1) + 1:fin(1) - 1);