Develop Reference

Regex to capture all vars between delimiters

How do I capture all 1, 2 and 3 in not |1|2|3|
My regex \|(.*?)\| skips 2.
const re = /\|(.*?)\|/gi;
const text = 'not |1|2|3|'
console.log(text.match(re).map(m => m[1]));

You can use
const re = /\|([^|]*)(?=\|)/g;
const text = 'not |123|2456|37890|'
console.log(Array.from(text.matchAll(re), m => m[1]));
Details:
\| - a | char
([^|]*) - Group 1: zero or more chars other than |
(?=\|) - a positive lookahead that matches a location that is immediately followed with |.
If you do not care about matching the | on the right, you can remove the lookahead.
If you also need to match till the end of string when the trailing | is missing, you can use /\|([^|]*)(?=\||$)/g.

JS - how to replace in regex for links

I have a regex that replaces the link from the moment http to the occurrence of certain characters in [], I would like to add to these characters - that is to replace the string with the occurrence of certain characters or a hard space:
"https://test.pl'".replace(/https?:\/\/[^ "'><]+/g," ")
Works fine for the characters mentioned in [], I don't know how to add here

You can use
.replace(/https?:\/\/.*?(?: |[ '"><]|$)/g," ")
See the regex demo.
Details:
https?:\/\/ - http:// or https://
.*? - any zero or more chars other than line break chars as few as possible
(?: |[ '"><]|$) - one of the following:
- char sequence
| - or
[ "'><] - a space, ", ', > or < char
| - or
$ - end of string.
JavaScript demo:
const texts = ["https://test.pl test","https://test.pl'test"];
const re = /https?:\/\/.*?(?: |[ '"><]|$)/g;
for (const s of texts) {
console.log(s, '=>', s.replace(re, " "));
}

How can I check the last character in a string?

I have these line options:
<40m:22s - ok
<40m:22m; - not ok
<40h:22s;<40m:22m - ok
<40m:22m;<40m:22m; - not ok
I need to check for semicolons. If I have one entry, then it shouldn't be. If I have several entries in a row, then the last entry should not have a semicolon.
Now I have so far only succeeded:
([<>][1-9][0-9][hms]:[1-9][0-9][hms][;?]+)(?<!;)
I will be grateful for any help, hint

You can use
^(?:[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms](?:;(?!$)|$))+$
Or, a bit more verbose since it includes a repetition of the main pattern:
^[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms](?:;[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms])*$
See the regex #1 demo and regex #2 demo.
Details:
^ - start of string
(?:[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms](?:;(?!$)|$))+ - one or more repetitions of
[<>] - a < or > char
[1-9] - a non-zero digit
[0-9]? - an optional digit (remove ? if it must be obligatory)
[hms] -h, mors`
: - a colon
[1-9][0-9]?[hms] - a non-zero digit, an optional digit and h/m/s
(?:;(?!$)|$) - a ; not at the end of string or end of string
$ - end of string.
The ^[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms](?:;[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms])*$ pattern follows the ^<MAIN>(?:<SEP><MAIN>)*$ scheme, and this pattern can be easily built dynamically using RegExp constructor.
const texts = ['<40m:22s', '<40m:22m;', '<40h:22s;<40m:22m', '<40m:22m;<40m:22m;'];
const rx = /^(?:[<>][1-9][0-9]?[hms]:[1-9][0-9]?[hms](?:;(?!$)|$))+$/;
for (let text of texts) {
console.log(text, '=>', rx.test(text));
}

The general pattern for a delimited list is
^ item (delimiter item)* $
To avoid self-repetition and make it all more or less readable, it would make sense to use variables, template strings and whitespace. This way your regexp looks like a grammar definition (what it actually is) and not as a soup of symbols.
let term = `[1-9] [0-9] [hms]`
let item = `< ${term} : ${term}`
let list = `^ ${item} ( ; ${item} )* $`
let re = new RegExp(list.replace(/\s/g, ''))
console.log(re)
test = `
<40m:22s
<40m:22m;
<40h:22s;<40m:22m
<40m:22m;<40m:22m;
`
for (t of test.trim().split('\n'))
console.log(t, re.test(t))

Lets simplify the problem to a = [<>][1-9][0-9][hms]:[1-9][0-9][hms], so the accepted strings can be
a - ok
a;a - ok
a; - not ok
a;a; - not ok
so our regex must end with a which leads to a$
now we want to accept none or multiple a with ; between each a, the regex for that is (a;)*
combining these 2 will resut in const regex = /^(a;)*a$/;
now if we replace a with [<>][1-9][0-9][hms]:[1-9][0-9][hms] the result will be const regex = /^([<>][1-9][0-9][hms]:[1-9][0-9][hms];)*[<>][1-9][0-9][hms]:[1-9][0-9][hms]$/;
demo

JS Regex Add Spaces to string

I have a two-part string and parts always delimited by space and |. Like this:
teststring | secondstring
It's possible to add predefined count of space between parts using ONLY Javascript regex.replace()?
I tried something like this:
([^\|]+)(\s){0,17}(?(R2)\s|\s)([\|a-zA-Z0-9]+)
And Substitution:
$1$2$3
It's possible to repeat capture group in substitution e.g. $2{17} or match same space multiple times?
EDIT:
I have function
function InvokeRegexp(originalString, pattern, replaceExpr)
{
return originalString.replace(pattern, replaceExpr);
}
and i want to pass two-part text, pattern containing number of spaces or replaceExpr containin number of spaces and get result: firstpart | secondpart

A non regex answer:
str.split("|").join(
" ".repeat(9 /*whitespaces*/) + "|"
)
Or with regex its probably:
str.replace(/\|/," ".repeat(9)+"|")

You could use padStart and padEnd. Because you said you want them to have a certain length.
const input = 'teststring | secondstring';
// Split the input variable and select spaces as well.
// 1. Select multiple spaces: \s+
// 2. Select pipe: \|
// 3. Select all following spaces: \s+
const parts = input.split( /\s+\|\s+/ );
// So every part should be at least 20 chars in this example.
const len = 20;
const output = `${ parts[ 0 ].padEnd( len ) }|${ parts[ 1 ].padStart( len )}`;
console.log( output );

Regex - collect characters between obligatory prefix and optional groups

I'm creating regex in JavaScript that find all groups occurrences, all optional.
I have collected optional groups (thanks for #wiktor-stribiżew) now. Missing thing is gathering characters between new- prefix and first occurred group.
Input:
new-rooms-3-area-50
new-poland-warsaw-rooms-3-area-50-bar
new-some-important-location-rooms-3-asdads-anything-area-50-uiop
new-another-location-area-50-else
Requested output:
["rooms-3", "area-50"]
["poland-warsaw", "rooms-3", "area-50"]
["some-important-location", "rooms-3", "area-50"]
["another-location", "area-50"]
I have now
new-(?:.*?(rooms-\d+))?.*?(area-\d+)
regex. I think that collecting .* between new- and rooms|area may be stupid solution.
Online demo: https://regex101.com/r/QvmYN0/5
Note: I created two separated questions, because it refers to 2 separately problems. I hope that somebody have similar problems in the future.

I think it is better to split by steps like this:
// Split by \n to work with each line
getArrays = input => input.split`\n`.map(x => {
// Split by your desired delimiters:
// -dashes which has "area" or "rooms" in front
return x.split(/-(?=area-|rooms-)/g).map(y => {
// remove the "new-" from start or anything in front the numbers
return y.replace(/^new-|\D+$/, '');
// make sure you don't have empty cases
}).filter(y => y);
});
var txt = `new-rooms-3-area-50
new-poland-warsaw-rooms-3-area-50-bar
new-some-important-location-rooms-3-asdads-anything-area-50-uiop
new-another-location-area-50-else`;
console.log(getArrays(txt));
EDIT:
The above code returns the requested output. However, I was thinking you should want an array of models instead:
// initial state of your model
getModel = () => ({
new: '',
area: 0,
rooms: 0,
});
// the function that will return the array of models:
getModels = input => input.split`\n`.map(line => {
var model = getModel();
// set delimiters:
var delimiters = new RegExp(
'-(?=(?:' + Object.keys(model).join`|` + ')-)', 'g');
// set the properties of your model:
line.split(delimiters).forEach(item => {
// remove non-digits after the last digit:
item.replace(/(\d)\D+$/, '$1')
// set each matched property:
.replace(/^([^-]+)-(.*)/,
(whole_match, key, val) => model[key] = val);
});
return model;
});
var txt = `new-rooms-3-area-50
new-poland-warsaw-rooms-3-area-50-bar
new-some-important-location-rooms-3-asdads-anything-area-50-uiop
new-another-location-area-50-else`;
console.log(getModels(txt));

This is the high-end solution which does it all at once.
Doesn't split or massage the data, just takes it as is (and always will be).
It may not be for beginners, but be for the more experienced.
(Note that I don't know JS, but I can tell you, this took about 20 minutes
googling about strings. This is just too easy, do people really get paid
to do this ?!)
This uses exec to push each element ( group 2 )
and create an array of records, one for each line.
( ^ new ) # (1)
|
( # (2 start)
(?: rooms | area )
- \d+
| (?:
(?:
(?!
(?: rooms | area )
- \d+
)
[a-z]
)+
(?:
-
(?:
(?!
(?: rooms | area )
- \d+
)
[a-z]
)+
)+
)
) # (2 end)
var strTarget = "\
new-rooms-3-area-50\n\
new-poland-warsaw-rooms-3-area-50-bar\n\
new-some-important-location-rooms-3-asdads-anything-area-50-uiop\n\
new-another-location-area-50-else\n\
";
var RxLine = /^new.+/mg;
var RxRecord = /(^new)|((?:rooms|area)-\d+|(?:(?:(?!(?:rooms|area)-\d+)[a-z])+(?:-(?:(?!(?:rooms|area)-\d+)[a-z])+)+))/g;
var records = [];
var matches
var match;
while( (match = RxLine.exec( strTarget )) ){
var line = match[0];
matches = [];
while( (match = RxRecord.exec( line )) ){
if ( match[2] )
matches.push( match[2] );
}
records.push( matches );
}
console.log( records );

Here you go:
new-(.*?)?-?(rooms-\d+|area-\d+).*?(area-\d+)?.*
Demo: https://regex101.com/r/Qvdkdx/1