Develop Reference

Replace every character but which is in my regex

I found this regex which validates Instagram usernames
/^(?!.*\.\.)(?!.*\.$)[^\W][\w.]{0,29}$/gim
What I'm trying to do is to replace all characters which not match my regex
let regex = new RegExp(/^(?!.*\.\.)(?!.*\.$)[^\W][\w.]{0,29}$/gim);
const filteredString = replace(text, regex, '');
I tried to add ?! at the start as a negative lookahead but no luck

Removing all the aprts that don't mach is the same as keeping the matches.
Instead of using replace you can use match and add all the matches to your filteredString, like shown below:
let text = `riegiejeyaranchen
riegie.jeyaranchen
_riegie.jeyaranchen
.riegie
riegie..jeyaranchen
riegie._.jeyaranchen
riegie.
riegie.__`;
let regex = new RegExp(/^(?!.*\.\.)(?!.*\.$)[^\W][\w.]{0,29}$/gim);
let filteredString = '';
text.match(regex).forEach(value =>
{
filteredString += value + '\r\n';
});
console.log(filteredString);
Of course the \r\n is optional (just places one on each line).
Now you get a string where non matches are removed.

Based on the regex /^(?!.*\.\.)(?!.*\.$)[^\W][\w.]{0,29}$/, the name must not have consecutive dots, no leading and trailing dots, and have max 30 word/dot chars. This code cleans up names accordingly:
[
'f',
'foobar',
'_foo.bar',
'_foo..bar',
'_foo...bar',
'_foo.bar.',
'.foo.bar',
'foo<$^*>bar',
'123456789012345678901234567890',
'1234567890123456789012345678901'
].forEach(name => {
let clean = name
.replace(/\.{2,}/g, '.') // reduce multiple dots to one dot
.replace(/^\.+/, '') // remove leading dots
.replace(/\.+$/, '') // remove trailing dots
.replace(/[^\w\.]/g, '') // remove non word/dot chars
.replace(/^(.{30}).+/, '$1'); // restrict to 30 chars
console.log(name + ' => ' + clean);
});
Output:
f => f
foobar => foobar
_foo.bar => _foo.bar
_foo..bar => _foo.bar
_foo...bar => _foo.bar
_foo.bar. => _foo.bar
.foo.bar => foo.bar
foo<$^*>bar => foobar
123456789012345678901234567890 => 123456789012345678901234567890
1234567890123456789012345678901 => 123456789012345678901234567890
Note that the original regex requires at least one char. You need to test for empty string after cleanup.

JavaScript regex to split camel case string

I have been trying to split a string with space when camel case happens. This is so far I can:
CamelCaseSplit = (inputString) => {
const f = str => str.match(/^[A-Z]?[^A-Z]*|[A-Z][^A-Z]*/g).join('
<CAMEL> ');
return f(inputString);
}
It can handle string likes "camelCase", "simple", "number1Case2".
But fails for "ABCWordDEF" and gives me "A B C Word D E F" instead of "ABC Word DEF".
Also the output of "abcDeF123" should be "abc De F123"
How to properly implement splitting into camel case?

You use Lodash's _.words() (https://lodash.com/docs/4.17.15#words) and say
const _ = require('lodash');
const camelCaseToWords = s => _.words(s).join(' ');

You can use
(?<=[a-z\d])(?=[A-Z])|(?<=[A-Z])(?=[A-Z][a-z])
To replace with a space. See the regex demo. Details:
(?<=[a-z\d])(?=[A-Z]) - a location between a lowercase letter or a digit, and a an uppercase letter
| - or
(?<=[A-Z])(?=[A-Z][a-z]) - a location between an uppercase letter and a sequence of an uppercase letter + a lowercase letter.
See the JavaScript demo:
CamelCaseSplit = (inputString) => {
const f = str => str.replace(/(?<=[a-z\d])(?=[A-Z])|(?<=[A-Z])(?=[A-Z][a-z])/g, ' ');
return f(inputString);
}
const texts = ['camelCase','simple','number1Case2','ABCWordDEF','abcDeF123'];
for (const s of texts) {
console.log(s, '->', CamelCaseSplit(s))
}

Convert different strings to snake_case in Javascript

I know that we have a question similar to this but not quite the same.
I'm trying to make my function work which takes in a string as an argument and converts it to snake_case . It works most of the time with all the fancy !?<>= characters but there is one case that it can't convert and its camelCase .
It fails when I'm passing strings like snakeCase. It returns snakecase instead of snake_case.
I tried to implement it but I ended up just messing it up even more..
Can I have some help please?
my code:
const snakeCase = string => {
string = string.replace(/\W+/g, " ").toLowerCase().split(' ').join('_');
if (string.charAt(string.length - 1) === '_') {
return string.substring(0, string.length - 1);
}
return string;
}

You need to be able to detect the points at which an upper-case letter is in the string following another letter (that is, not following a space). You can do this with a regular expression, before you call toLowerCase on the input string:
\B(?=[A-Z])
In other words, a non-word boundary, followed by an upper case character. Split on either the above, or on a literal space, then .map the resulting array to lower case, and then you can join by underscores:
const snakeCase = string => {
return string.replace(/\W+/g, " ")
.split(/ |\B(?=[A-Z])/)
.map(word => word.toLowerCase())
.join('_');
};
console.log(snakeCase('snakeCase'));

Let's try that again Stan... this should do snake_case while realising that CamelCASECapitals = camel_case_capitals. It's basically the accepted answer with a pre-filter.
let splitCaps = string => string
.replace(/([a-z])([A-Z]+)/g, (m, s1, s2) => s1 + ' ' + s2)
.replace(/([A-Z])([A-Z]+)([^a-zA-Z0-9]*)$/, (m, s1, s2, s3) => s1 + s2.toLowerCase() + s3)
.replace(/([A-Z]+)([A-Z][a-z])/g,
(m, s1, s2) => s1.toLowerCase() + ' ' + s2);
let snakeCase = string =>
splitCaps(string)
.replace(/\W+/g, " ")
.split(/ |\B(?=[A-Z])/)
.map(word => word.toLowerCase())
.join('_');
> a = ['CamelCASERules', 'IndexID', 'CamelCASE', 'aID',
'theIDForUSGovAndDOD', 'TheID_', '_IDOne']
> _.map(a, snakeCase)
['camel_case_rules', 'index_id', 'camel_case', 'a_id', 'the_id_for_us_gov_and_dod',
'the_id_', '_id_one']
// And for the curious, here's the output from the pre-filter:
> _.map(a, splitCaps)
['Camel case Rules', 'Index Id', 'Camel Case', 'a Id', 'the id For us Gov And Dod',
'The Id_', '_id One']

Suppose the string is Hello World? and you want the returned value as hello_world? (with the character, then follow the below code)
const snakeCase = (string) => {
return string.replace(/\d+/g, ' ')
.split(/ |\B(?=[A-Z])/)
.map((word) => word.toLowerCase())
.join('_');
};
Example
snakeCase('Hello World?')
// "hello_world?"
snakeCase('Hello & World')
// "hello_&_world"

EDIT: It turns out this answer isn’t fool proof. The fool, being me ;-) Please check out a better one by Orwellophile here: https://stackoverflow.com/a/69878219/5377276
——
I think this one should cover all the bases 😄
It was inspired by #h0r53's answer to the accepted answer. However it evolved into a more complete function, as it will convert any string, camelCase, kebab-case or otherwise into snake_case the way you'd expect it, containing only a-z and 0-9 characters, usable for function and variable names:
convert_to_snake_case(string) {
return string.charAt(0).toLowerCase() + string.slice(1) // lowercase the first character
.replace(/\W+/g, " ") // Remove all excess white space and replace & , . etc.
.replace(/([a-z])([A-Z])([a-z])/g, "$1 $2$3") // Put a space at the position of a camelCase -> camel Case
.split(/\B(?=[A-Z]{2,})/) // Now split the multi-uppercases customerID -> customer,ID
.join(' ') // And join back with spaces.
.split(' ') // Split all the spaces again, this time we're fully converted
.join('_') // And finally snake_case things up
.toLowerCase() // With a nice lower case
}
Conversion examples:
'snakeCase' => 'snake_case'
'CustomerID' => 'customer_id'
'GPS' => 'gps'
'IP-address' => 'ip_address'
'Another & Another, one too' => 'another_another_one_too'
'random ----- Thing123' => 'random_thing123'
'kebab-case-example' => 'kebab_case_example'

There is method in lodash named snakeCase(). You can consider that as well.
https://lodash.com/docs/4.17.15#snakeCase

Orwellophile's answer does not work for uppercase words delimited by a space:
E.g: 'TEST CASE' => t_e_s_t_case
The following solution does not break consecutive upper case characters and is a little shorter:
const snakeCase = str =>
str &&
str
.match(/[A-Z]{2,}(?=[A-Z][a-z]+[0-9]*|\b)|[A-Z]?[a-z]+[0-9]*|[A-Z]|[0-9]+/g)
.map(x => x.toLowerCase())
.join('_');
However, trailing underscores after uppercase words (examples from Orwellophile as well), do not work properly.
E.g: 'TheID_' => the_i_d
Taken from https://www.w3resource.com/javascript-exercises/fundamental/javascript-fundamental-exercise-120.php.

Javascript Regular expression with replace

I'm working on a project on search about emoji and replace it with icon
but I have some problem on regular expression, Below mentioned is my code for reference:
var f = ["( :3 )" , "( :P )","\(:star:\)"];
var re = function(s){return new RegExp(s, 'g');};
now when I'm going to search about emoji and replace it as shown below:
s = "hello :D how are you :P dwdwd";
for(var n in f){
var m;
if ((m = re(f[n]).exec(s)) !== null) {
m.forEach((match, groupIndex) => {
s = s.replace(match,"<img src='http://abs.twimg.com/emoji/v1/72x72/"+ r[n] +".png'>");
});
}}
In this case, it works well and replace the emoji. But it only replace when there are space before and after emoji what should i do to replace the emoji in the begin of string or end !
s = ":D hello how are you :)";
This case is not working. How can i edit my regular expression for being able to replace emoji at begin and end of string and at the same time if its found in middle of string & have space between word and emoji
My 2nd problem with regular expression is "\(:star:\)" it never replaces. While it must replace word :star: with an emoji but i think i miss some thing on regular expression for it.

You can use beginning & ending anchors along with pipe to achieve this. For example:
/(^:3\s)|(\s:3\s)|(\s:3$)/g
^ is an anchor which matches :3\s to the beginning of the string.
$ is an anchor which matches \s:3 to the end of the string.
\s matches whitespace.
| is the pipe operator which acts as a logical OR operator between the different capture groups.

Just remove the spaces from your emoji regex.
var f = ["(:3)", "(:P)", "\(:star:\)"];
var r = ["[sadface]", "[toungeface]", "[staremoji]"];
var re = function(s) {
return new RegExp(s, 'g');
};
s = "hello :3 how are you :P dwdwd :star: :3";
console.log(s);
for (var n in f) {
var m;
if ((m = re(f[n]).exec(s)) !== null) {
m.forEach((match, groupIndex) => {
s = s.replace(match, r[n]);
});
}
}
console.log(s);

var content = "hello :D how are you :P dwdwd";
content = content.replace(/((:D|:P))/g,function(match){
var result = "";
var index = -1;
switch(match)
{
case ":D":
result = "happy";
index = 0
break;
case ":P":
result = "smilie";
index = 1
break;
}
if(index != -1)
{
return "<img src='http://abs.twimg.com/emoji/v1/72x72/"+index+".png'>";
}
return result;
});
console.log(content);
Please try this.

I created a more generic solution, starting with the mapping of emojis to the relevant names. Rather than two lists that need to be kept in synch, I used a single object:
const emojis = {
'(c)': 'a9',
'(r)': 'ae',
'(tm)': '2122'
//, ...
};
This strikes me as a much more useful structure to work with, but the code below could easily be altered to deal with the two-lists version.
Then I use a helper function to escape characters which are not allowed as plain text in Regular Expression by prepending them with \:
const escapeSpecials = (() => {
const specials = ['/', '.', '*', '+', '?', '|', '(', ')', '[', ']', '{', '}', '\\'];
const reg = new RegExp('(\\' + specials.join('|\\') + ')', 'g');
return str => str.replace(reg, '\\$1');
})();
Then I have the key function:
const replaceStringsWith = (emojis, convert) => str => Object.keys(emojis).reduce(
(str, em) => str.replace(new RegExp(`(^|\\s+)(${escapeSpecials(em)})($|\\s+)`, 'g'),
(m, a, b, c) => `${a}${convert(emojis[b], b)}${c}`),
str
);
This takes an object containing string/replacement pairs and a converter function which accepts the replacement and gives you back the final form. It returns a function which takes a string, and then searches for any matches on the keys of the object (properly checked for strings or string ends), replacing them with the result of calling the converter on the object's value for the particular key.
Thus we can do:
const toUrl = (name) => `<img src='http://abs.twimg.com/emoji/v1/72x72/${name}.png'>`;
const replaceEmojis = replaceStringsWith(emojis, toUrl)
and call it as
const s = "This is Copyright (c) 2017, FooBar is (tm) BazCo (r)";
replaceEmojis(s); //=>
// `This is Copyright <img src='http://abs.twimg.com/emoji/v1/72x72/a9.png'>
// 2017, FooBar is <img src='http://abs.twimg.com/emoji/v1/72x72/2122.png'>
// BazCo <img src='http://abs.twimg.com/emoji/v1/72x72/ae.png'>`
Note that the converter also takes a second parameter. So you could instead use
const toUrl = (name, emoji) =>
`<img src='http://abs.twimg.com/emoji/v1/72x72/${name}.png' title='${emoji}'>`;
to get
//=> `This is Copyright <img
// src='http://abs.twimg.com/emoji/v1/72x72/a9.png' title='(c)'>
// 2017, FooBar is <img src='http://abs.twimg.com/emoji/v1/72x72/2122.png'
// title='(tm)'> BazCo <img src='http://abs.twimg.com/emoji/v1/72x72/ae.png' title='(r)'>"

Get first letter of each word in a string, in JavaScript

How would you go around to collect the first letter of each word in a string, as in to receive an abbreviation?
Input: "Java Script Object Notation"
Output: "JSON"

I think what you're looking for is the acronym of a supplied string.
var str = "Java Script Object Notation";
var matches = str.match(/\b(\w)/g); // ['J','S','O','N']
var acronym = matches.join(''); // JSON
console.log(acronym)
Note: this will fail for hyphenated/apostrophe'd words Help-me I'm Dieing will be HmImD. If that's not what you want, the split on space, grab first letter approach might be what you want.
Here's a quick example of that:
let str = "Java Script Object Notation";
let acronym = str.split(/\s/).reduce((response,word)=> response+=word.slice(0,1),'')
console.log(acronym);

I think you can do this with
'Aa Bb'.match(/\b\w/g).join('')
Explanation: Obtain all /g the alphanumeric characters \w that occur after a non-alphanumeric character (i.e: after a word boundary \b), put them on an array with .match() and join everything in a single string .join('')
Depending on what you want to do you can also consider simply selecting all the uppercase characters:
'JavaScript Object Notation'.match(/[A-Z]/g).join('')

Easiest way without regex
var abbr = "Java Script Object Notation".split(' ').map(function(item){return item[0]}).join('');

This is made very simple with ES6
string.split(' ').map(i => i.charAt(0)) //Inherit case of each letter
string.split(' ').map(i => i.charAt(0)).toUpperCase() //Uppercase each letter
string.split(' ').map(i => i.charAt(0)).toLowerCase() //lowercase each letter
This ONLY works with spaces or whatever is defined in the .split(' ') method
ie, .split(', ') .split('; '), etc.
string.split(' ') .map(i => i.charAt(0)) .toString() .toUpperCase().split(',')

To add to the great examples, you could do it like this in ES6
const x = "Java Script Object Notation".split(' ').map(x => x[0]).join('');
console.log(x); // JSON
and this works too but please ignore it, I went a bit nuts here :-)
const [j,s,o,n] = "Java Script Object Notation".split(' ').map(x => x[0]);
console.log(`${j}${s}${o}${n}`);

#BotNet flaw:
i think i solved it after excruciating 3 days of regular expressions tutorials:
==> I'm a an animal
(used to catch m of I'm) because of the word boundary, it seems to work for me that way.
/(\s|^)([a-z])/gi

Try -
var text = '';
var arr = "Java Script Object Notation".split(' ');
for(i=0;i<arr.length;i++) {
text += arr[i].substr(0,1)
}
alert(text);
Demo - http://jsfiddle.net/r2maQ/

Using map (from functional programming)
'use strict';
function acronym(words)
{
if (!words) { return ''; }
var first_letter = function(x){ if (x) { return x[0]; } else { return ''; }};
return words.split(' ').map(first_letter).join('');
}

Alternative 1:
you can also use this regex to return an array of the first letter of every word
/(?<=(\s|^))[a-z]/gi
(?<=(\s|^)) is called positive lookbehind which make sure the element in our search pattern is preceded by (\s|^).
so, for your case:
// in case the input is lowercase & there's a word with apostrophe
const toAbbr = (str) => {
return str.match(/(?<=(\s|^))[a-z]/gi)
.join('')
.toUpperCase();
};
toAbbr("java script object notation"); //result JSON
(by the way, there are also negative lookbehind, positive lookahead, negative lookahead, if you want to learn more)
Alternative 2:
match all the words and use replace() method to replace them with the first letter of each word and ignore the space (the method will not mutate your original string)
// in case the input is lowercase & there's a word with apostrophe
const toAbbr = (str) => {
return str.replace(/(\S+)(\s*)/gi, (match, p1, p2) => p1[0].toUpperCase());
};
toAbbr("java script object notation"); //result JSON
// word = not space = \S+ = p1 (p1 is the first pattern)
// space = \s* = p2 (p2 is the second pattern)

It's important to trim the word before splitting it, otherwise, we'd lose some letters.
const getWordInitials = (word: string): string => {
const bits = word.trim().split(' ');
return bits
.map((bit) => bit.charAt(0))
.join('')
.toUpperCase();
};
$ getWordInitials("Java Script Object Notation")
$ "JSON"

How about this:
var str = "", abbr = "";
str = "Java Script Object Notation";
str = str.split(' ');
for (i = 0; i < str.length; i++) {
abbr += str[i].substr(0,1);
}
alert(abbr);
Working Example.

If you came here looking for how to do this that supports non-BMP characters that use surrogate pairs:
initials = str.split(' ')
.map(s => String.fromCodePoint(s.codePointAt(0) || '').toUpperCase())
.join('');
Works in all modern browsers with no polyfills (not IE though)

Getting first letter of any Unicode word in JavaScript is now easy with the ECMAScript 2018 standard:
/(?<!\p{L}\p{M}*)\p{L}/gu
This regex finds any Unicode letter (see the last \p{L}) that is not preceded with any other letter that can optionally have diacritic symbols (see the (?<!\p{L}\p{M}*) negative lookbehind where \p{M}* matches 0 or more diacritic chars). Note that u flag is compulsory here for the Unicode property classes (like \p{L}) to work correctly.
To emulate a fully Unicode-aware \b, you'd need to add a digit matching pattern and connector punctuation:
/(?<!\p{L}\p{M}*|[\p{N}\p{Pc}])\p{L}/gu
It works in Chrome, Firefox (since June 30, 2020), Node.js, and the majority of other environments (see the compatibility matrix here), for any natural language including Arabic.
Quick test:
const regex = /(?<!\p{L}\p{M}*)\p{L}/gu;
const string = "Żerard Łyżwiński";
// Extracting
console.log(string.match(regex)); // => [ "Ż", "Ł" ]
// Extracting and concatenating into string
console.log(string.match(regex).join("")) // => ŻŁ
// Removing
console.log(string.replace(regex, "")) // => erard yżwiński
// Enclosing (wrapping) with a tag
console.log(string.replace(regex, "<span>$&</span>")) // => <span>Ż</span>erard <span>Ł</span>yżwiński
console.log("_Łukasz 1Żukowski".match(/(?<!\p{L}\p{M}*|[\p{N}\p{Pc}])\p{L}/gu)); // => null

In ES6:
function getFirstCharacters(str) {
let result = [];
str.split(' ').map(word => word.charAt(0) != '' ? result.push(word.charAt(0)) : '');
return result;
}
const str1 = "Hello4 World65 123 !!";
const str2 = "123and 456 and 78-1";
const str3 = " Hello World !!";
console.log(getFirstCharacters(str1));
console.log(getFirstCharacters(str2));
console.log(getFirstCharacters(str3));
Output:
[ 'H', 'W', '1', '!' ]
[ '1', '4', 'a', '7' ]
[ 'H', 'W', '!' ]

This should do it.
var s = "Java Script Object Notation",
a = s.split(' '),
l = a.length,
i = 0,
n = "";
for (; i < l; ++i)
{
n += a[i].charAt(0);
}
console.log(n);

The regular expression versions for JavaScript is not compatible with Unicode on older than ECMAScript 6, so for those who want to support characters such as "å" will need to rely on non-regex versions of scripts.
Event when on version 6, you need to indicate Unicode with \u.
More details: https://mathiasbynens.be/notes/es6-unicode-regex

Yet another option using reduce function:
var value = "Java Script Object Notation";
var result = value.split(' ').reduce(function(previous, current){
return {v : previous.v + current[0]};
},{v:""});
$("#output").text(result.v);
<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.1/jquery.min.js"></script>
<pre id="output"/>

This is similar to others, but (IMHO) a tad easier to read:
const getAcronym = title =>
title.split(' ')
.map(word => word[0])
.join('');

ES6 reduce way:
const initials = inputStr.split(' ').reduce((result, currentWord) =>
result + currentWord.charAt(0).toUpperCase(), '');
alert(initials);

Try This Function
const createUserName = function (name) {
const username = name
.toLowerCase()
.split(' ')
.map((elem) => elem[0])
.join('');
return username;
};
console.log(createUserName('Anisul Haque Bhuiyan'));