Search for regex terms independent of symbol-CodePudding

So I have this string right here, it's a registration number:

12.325.767/0001-90

And I want to create a regex when I type "23", return for me the "2.3" or if I type "700", return for me "7/00" or if I type "19", return "1-9".

So, I want to get the number even if there is a symbol in the middle.

I'm trying this on Javascript, here what I have:

const cnpj = "12.325.767/0001-90";
const search = "12";

const regex = new RegExp(`(${search})`, "i");

const result = cnpj.split(regex);

output:

[ '', '12', '.325.767/0001-90' ]

This output is correct, because I put a number that does not have symbols in its composition.

But when I try to search a number that contains a symbol in its composition, is not splitted.

CodePudding user response：

I used your explanation to improve my code for what I want, look:

I will try to show the cases:

12.322.362/3002-32

In this number, if I type "23", notice that I have in multiple parts of the number. So I did this:

const cnpj = "12.322.362/3002-32";
const search = "23";

const symbols = ".,/:;-";
const reInput = search.split("").join(`[${symbols}]?`);
const regex = new RegExp(`(${reInput})`, "i");

const result = cnpj.split(regex);

And the output is perfect, because he shows me every "23" of the expression, look the output:

result: ['1', '2.3', '2','2.3', '6',   '2/3', '00',  '2-3', '2']

Even if I put the complete number, he returns me the correct result, for exemple, if I type "123223", this is the output:

[ '', '12.322.3', '62/3002-32' ]

And finally, with this, I can check in the array the indexes that matches, look:

const cnpj = "12.322.362/3002-32";
const search = "23";

const symbols = ".,/:;-";
const reInput = search.split("").join(`[${symbols}]?`);
const regex = new RegExp(`(${reInput})`, "i");

const data = cnpj.split(regex);

data.map((item, index) => {
  if (item.toLowerCase().match(regex)) {
    console.log("match");
  } else {
    console.log("doesn't match");
  }
});

Output:

doesn't match
match
doesn't match
match
doesn't match
match
doesn't match
match
doesn't match

So, thank you eventHandler for your amazing explanation, the logics works fine this way.

CodePudding user response：

You could convert the user input into a RegExp that has symbols between each digit. For example, 12 becomes 1[.,/:;]?2, so you search if there is one or none of those symbols between the numbers.

If you used a function like this, you would get the result you want:

function underline(input, text) {
    const symbols = '.,/:;';
    const reInput = input.split('').join(`[${symbols}]?`);
    const re = new RegExp(reInput);

    return re.exec(text);
}

for user input 23, with your example text 12.325.767/0001-90, this function returns the following result

[ '2.3', index: 1, input: '12.325.767/0001-90', groups: undefined ]

You can substring from that, using the index and the result length, or make a more complex RegExp, like this:

function underline(input, text) {
    const symbols = '.,/:;';
    const reInput = input.split('').join(`[${symbols}]?`);
    const reText = '(.*?)('   reInput   ')(.*)';
    const re = new RegExp(reText);
    const [_, ...result] = re.exec(text);
    
    return result;
}

This returns the array you expected.

[ '1', '2.3', '25.767/0001-90' ]