How to use metacharacters in combination with cell arrays to build up a pattern for regexp?

I just started coding, and I can't manage regexp properly and I guess my codes are not that efficient, but here is what I would need:

I wrote a code like this, where X and Y are both cell arrays (where every cell consists of string arrays).

Y{1} is defined.

for i=1:length(X)
    for k=1:length(Y{i})
    Z{i}(k+1)=max(find(~cellfun(@isempty,regexp(X{i}, Y{i}{k}))));
    ...

    end
end

This works fine cause I expect Z to be a cell array of indices, indices of max position of occurence of the string contained in Y{i}{k}.

The problem is that this string (let's say for example "22") is contained also in other strings within Y{i}{k} (let's say "3422"), but my goal is to get the max index of occurence of the string "22" with no other digits or characters before or after it.

What I ask now is a method, possibly using metacharacters, to get the max index of occurence of just the string Y{i}{k} with no digits or characters behind or after.

I thought of using something like --> '(?<!\d)(\d)(?!\d)' matches single-digit numbers (digits that do not precede or follow other digits),

but it seems hard to match metacharacters and Y{i}{k} in the same pattern. I got a snapshot below of what I attempted with no success, it might help.

Any idea of how to solve this? Thanks in advance!!

Answer 1

[Stephen23]

"my goal is to get the max index of occurence of the string "22" with no other digits or characters before or after it."

I assume for now that by "no characters" you actually mean no non-whitespace characters, or perhaps no letter characters. Clarification would be helpful on this..

Possibly a good appraoch would be to use anchors:

https://www.mathworks.com/help/matlab/matlab_prog/regular-expressions.html#f0-42983

and you can trivially join these with the Y string either concatenation or sprintf, e.g.:

rgx = sprintf('\\<%s\\>',Y{i}{k});
regex(X{i},rgx)

Most likely you can remove on of the loops by constructing one regular expression for all of the k strings, e.g.:

rgx = '\<(22|23|99|123)\>';

You can gemerate that quite easily using sprintf or strjoin or similar.

"No Characters": your description that you want "no other digits or characters before or after it" literally means that you would only match this string "22", because whitespace , ounctuation, and non-printing characters are also characters, so by your description must be excluded... leaving literally no characters.

I assumed above that you actually mean to match the substring

"... 22 ..."
%   ^  ^ whitespace, not letter, not digit

Are punctuation characters allowed or not? If you really do mean "no characters" then using regular expressions is an overly complex way to match literal strings.

[Davide Festa]

Sorry for not being clear and thanks for the answer.

I show you more in detail how the data looks like.

For example Y{1}{1} is "22".

Instead, X{1} is a 55×1 string array where the elements looks like

    "22"
    "asa"
    "22"
    "33"
    "2245"
    "stan"

I'm interested in having the max index position of "22" in that string array. As you can see the string "22" appears two times (which is good), the trouble is caused by the string "2245", which contains "22" within itself.

Because of this, when I run this code

    Z{i}(k+1)=max(find(~cellfun(@isempty,regexp(X{i}, Y{i}{k}))));

the max index that I get is referred to string "2245" (because "22" is contained in it). But my goal is to get the index of the max occurrence of the string "22" (which is the third position in the above example).

That's why I'm asking for metacharacters that helps me to pick up only the max occurence of string "22" with no other digits or characters attached.

I hope it is clear, otherwise you can ask me again :)

[Stephen23]

A simple string comparison will do that task simply and efficiently:

>> C = {'22';'asa';'22';'33';'2245';'stan'};
>> X = strcmp(C,'22')
X =
  1
  0
  1
  0
  0
  0
>> max(find(X))
ans =  3

Regular expressions are just a red-herring: they will be much less efficient than a simple string comaprison. But if you really want to use a regular expression:

X = ~cellfun(@isempty,regexp(C,'^22$'))

[Davide Festa]

Thanks a lot, it works perfectly by using strcmp.

Now I see that by using regexp I made the problem more complicated than it was!