So I have a RegExp regex = /asd/
I am storing it as a as a key in my key-val store system.
So I say str = String(regex) which returns "/asd/".
Now I need to convert that string back to a RegExp.
So I try: RegExp(str) and I see /\/asd\//
this is not what I want. It is not the same as /asd/
Should I just remove the first and last characters from the string before converting it to regex? That would get me the desired result in this situation, but wouldn't necessarily work if the RegExp had modifiers like /i or /g
Is there a better way to do this?
If you don't need to store the modifiers, you can use Regexp#source to get the string value, and then convert back using the RegExp constructor.
var regex = /abc/g;
var str = regex.source; // "abc"
var restoreRegex = new RegExp(str, "g");
If you do need to store the modifiers, use a regex to parse the regex:
var regex = /abc/g;
var str = regex.toString(); // "/abc/g"
var parts = /\/(.*)\/(.*)/.exec(str);
var restoredRegex = new RegExp(parts[1], parts[2]);
This will work even if the pattern has a / in it, because .* is greedy, and will advance to the last / in the string.
If performance is a concern, use normal string manipulation using String#lastIndexOf:
var regex = /abc/g;
var str = regex.toString(); // "/abc/g"
var lastSlash = str.lastIndexOf("/");
var restoredRegex = new RegExp(str.slice(1, lastSlash), str.slice(lastSlash + 1));
const regex = /asd/gi;
converting RegExp to String
const obj = {flags: regex.flags, source: regex.source};
const string = JSON.stringify(obj);
then back to RegExp
const obj2 = JSON.parse(string);
const regex2 = new RegExp(obj2.source, obj2.flags);
Requires ES6+.
You can use the following before storage of your regex literal:
(new RegExp(regex)).source
See https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/RegExp/source
Example:
regex = /asd/
string = (new RegExp(regex)).source
// string is now "asd"
regex = RegExp(string)
// regex has the original value /asd/
let rx = RegExp.apply(RegExp, str.match(/\/(.*)\/(.*)/).slice(1));
A modified version of #PegasusEpsilon answer
StackOverflow saves the day again, thanks #4castle! I wanted to store some regex rules in a JS file, and some in a DB, combine them into an array of objects like so:
module.exports = {
[SETTINGS.PRODUCTION_ENV]: [
{
"key": /<meta name="generator"[\s\S]*?>/gmi,
"value": "",
"regex": true
},
...
]
}
Then, loop through each environment's objects and apply it to a string of text. This is for a node/lambda project, so I wanted to use ES6. I used #4castle's code, with some destructuring, and I ended up with this:
let content = body;
const regexString = replacement.key.toString();
const regexParts = /\/(.*)\/(.*)/.exec(regexString);
const {1: source, 2: flags} = regexParts;
const regex = new RegExp(source, flags);
content = content.replace(regex, replacement.value);
return content;
Works a treat!
Related
I want to add a (variable) tag to values with regex, the pattern works fine with PHP but I have troubles implementing it into JavaScript.
The pattern is (value is the variable):
/(?!(?:[^<]+>|[^>]+<\/a>))\b(value)\b/is
I escaped the backslashes:
var str = $("#div").html();
var regex = "/(?!(?:[^<]+>|[^>]+<\\/a>))\\b(" + value + ")\\b/is";
$("#div").html(str.replace(regex, "" + value + ""));
But this seem not to be right, I logged the pattern and its exactly what it should be.
Any ideas?
To create the regex from a string, you have to use JavaScript's RegExp object.
If you also want to match/replace more than one time, then you must add the g (global match) flag. Here's an example:
var stringToGoIntoTheRegex = "abc";
var regex = new RegExp("#" + stringToGoIntoTheRegex + "#", "g");
// at this point, the line above is the same as: var regex = /#abc#/g;
var input = "Hello this is #abc# some #abc# stuff.";
var output = input.replace(regex, "!!");
alert(output); // Hello this is !! some !! stuff.
JSFiddle demo here.
In the general case, escape the string before using as regex:
Not every string is a valid regex, though: there are some speciall characters, like ( or [. To work around this issue, simply escape the string before turning it into a regex. A utility function for that goes in the sample below:
function escapeRegExp(stringToGoIntoTheRegex) {
return stringToGoIntoTheRegex.replace(/[-\/\\^$*+?.()|[\]{}]/g, '\\$&');
}
var stringToGoIntoTheRegex = escapeRegExp("abc"); // this is the only change from above
var regex = new RegExp("#" + stringToGoIntoTheRegex + "#", "g");
// at this point, the line above is the same as: var regex = /#abc#/g;
var input = "Hello this is #abc# some #abc# stuff.";
var output = input.replace(regex, "!!");
alert(output); // Hello this is !! some !! stuff.
JSFiddle demo here.
Note: the regex in the question uses the s modifier, which didn't exist at the time of the question, but does exist -- a s (dotall) flag/modifier in JavaScript -- today.
If you are trying to use a variable value in the expression, you must use the RegExp "constructor".
var regex = "(?!(?:[^<]+>|[^>]+<\/a>))\b(" + value + ")\b";
new RegExp(regex, "is")
I found I had to double slash the \b to get it working. For example to remove "1x" words from a string using a variable, I needed to use:
str = "1x";
var regex = new RegExp("\\b"+str+"\\b","g"); // same as inv.replace(/\b1x\b/g, "")
inv=inv.replace(regex, "");
You don't need the " to define a regular expression so just:
var regex = /(?!(?:[^<]+>|[^>]+<\/a>))\b(value)\b/is; // this is valid syntax
If value is a variable and you want a dynamic regular expression then you can't use this notation; use the alternative notation.
String.replace also accepts strings as input, so you can do "fox".replace("fox", "bear");
Alternative:
var regex = new RegExp("/(?!(?:[^<]+>|[^>]+<\/a>))\b(value)\b/", "is");
var regex = new RegExp("/(?!(?:[^<]+>|[^>]+<\/a>))\b(" + value + ")\b/", "is");
var regex = new RegExp("/(?!(?:[^<]+>|[^>]+<\/a>))\b(.*?)\b/", "is");
Keep in mind that if value contains regular expressions characters like (, [ and ? you will need to escape them.
I found this thread useful - so I thought I would add the answer to my own problem.
I wanted to edit a database configuration file (datastax cassandra) from a node application in javascript and for one of the settings in the file I needed to match on a string and then replace the line following it.
This was my solution.
dse_cassandra_yaml='/etc/dse/cassandra/cassandra.yaml'
// a) find the searchString and grab all text on the following line to it
// b) replace all next line text with a newString supplied to function
// note - leaves searchString text untouched
function replaceStringNextLine(file, searchString, newString) {
fs.readFile(file, 'utf-8', function(err, data){
if (err) throw err;
// need to use double escape '\\' when putting regex in strings !
var re = "\\s+(\\-\\s(.*)?)(?:\\s|$)";
var myRegExp = new RegExp(searchString + re, "g");
var match = myRegExp.exec(data);
var replaceThis = match[1];
var writeString = data.replace(replaceThis, newString);
fs.writeFile(file, writeString, 'utf-8', function (err) {
if (err) throw err;
console.log(file + ' updated');
});
});
}
searchString = "data_file_directories:"
newString = "- /mnt/cassandra/data"
replaceStringNextLine(dse_cassandra_yaml, searchString, newString );
After running, it will change the existing data directory setting to the new one:
config file before:
data_file_directories:
- /var/lib/cassandra/data
config file after:
data_file_directories:
- /mnt/cassandra/data
Much easier way: use template literals.
var variable = 'foo'
var expression = `.*${variable}.*`
var re = new RegExp(expression, 'g')
re.test('fdjklsffoodjkslfd') // true
re.test('fdjklsfdjkslfd') // false
Using string variable(s) content as part of a more complex composed regex expression (es6|ts)
This example will replace all urls using my-domain.com to my-other-domain (both are variables).
You can do dynamic regexs by combining string values and other regex expressions within a raw string template. Using String.raw will prevent javascript from escaping any character within your string values.
// Strings with some data
const domainStr = 'my-domain.com'
const newDomain = 'my-other-domain.com'
// Make sure your string is regex friendly
// This will replace dots for '\'.
const regexUrl = /\./gm;
const substr = `\\\.`;
const domain = domainStr.replace(regexUrl, substr);
// domain is a regex friendly string: 'my-domain\.com'
console.log('Regex expresion for domain', domain)
// HERE!!! You can 'assemble a complex regex using string pieces.
const re = new RegExp( String.raw `([\'|\"]https:\/\/)(${domain})(\S+[\'|\"])`, 'gm');
// now I'll use the regex expression groups to replace the domain
const domainSubst = `$1${newDomain}$3`;
// const page contains all the html text
const result = page.replace(re, domainSubst);
note: Don't forget to use regex101.com to create, test and export REGEX code.
var string = "Hi welcome to stack overflow"
var toSearch = "stack"
//case insensitive search
var result = string.search(new RegExp(toSearch, "i")) > 0 ? 'Matched' : 'notMatched'
https://jsfiddle.net/9f0mb6Lz/
Hope this helps
I am writing js code to get array of elements after splitting using regular expression.
var data = "ABCXYZ88";
var regexp = "([A-Z]{3})([A-Z]{3}d{2})";
console.log(data.split(regexp));
It returns
[ 'ABCXYZ88' ]
But I am expecting something like
['ABC','XYZ','88']
Any thoughts?
I fixed your regex, then matched it against your string and extracted the relevant capturing groups:
var regex = /([A-Z]{3})([A-Z]{3})(\d{2})/g;
var str = 'ABCXYZ88';
let m = regex.exec(str);
if (m !== null) {
console.log(m.slice(1)); // prints ["ABC", "XYZ", "88"]
}
In your case, I don't think you can split using a regex as you were trying, as there don't seem to be any delimiting characters to match against. For this to work, you'd have to have a string like 'ABC|XYZ|88'; then you could do 'ABC|XYZ|88'.split(/\|/g). (Of course, you wouldn't use a regex for such a simple case.)
Your regexp is not a RegExp object but a string.
Your capturing groups are not correct.
String.prototype.split() is not the function you need. What split() does:
var myString = 'Hello World. How are you doing?';
var splits = myString.split(' ', 3);
console.log(splits); // ["Hello", "World.", "How"]
What you need:
var data = 'ABCXYZ88';
var regexp = /^([A-Z]{3})([A-Z]{3})(\d{2})$/;
var match = data.match(regexp);
console.log(match.slice(1)); // ["ABC", "XYZ", "88"]
Try this. I hope this is what you are looking for.
var reg = data.match(/^([A-Z]{3})([A-Z]{3})(\d{2})$/).slice(1);
https://jsfiddle.net/m5pgpkje/1/
I am using the function match for a search engine, so whenever a user types a search-string I take that string and use the match function on an array containing country names, but it doesn't seem to work.
For example if I do :
var string = "algeria";
var res = string.match(/alge/g); //alge is what the user would have typed in the search bar
alert(res);
I get a string res = "alge": //thus verifying that alge exists in algeria
But if I do this, it returns null, why? and how can I make it work?
var regex = "/alge/g";
var string = "algeria";
var res = string.match(regex);
alert(res);
To make a regex from a string, you need to create a RegExp object:
var regex = new RegExp("alge", "g");
(Beware that unless your users will be typing actual regular expressions, you'll need to escape any characters that have special meaning within regular expressions - see Is there a RegExp.escape function in Javascript? for ways to do this.)
You don't need quotes around the regex:
var regex = /alge/g;
Remove the quotes around the regex.
var regex = /alge/g;
var string = "algeria";
var res = string.match(regex);
alert(res);
found the answer, the match function takes a regex object so have to do
var regex = new RegExp(string, "g");
var res = text.match(regex);
This works fine
How to do this in a more "elegant way"?
var str = obj.property.toString(); //str value = (55.930385, -3.118425)
var remesp = str.replace(" ","");
var rempar1 = remesp.replace("(","");
var rempar2 = rempar1.replace(")",""); //rempar2 value = "55.930385,-3.118425"
Try using a regular expression character class (aka "character set") [...]:
var str = "(55.930385, -3.118425)";
str.replace(/[ ()]/g, ''); // => "55.930385,-3.118425"
In the example above, the regex /[ ()]/ matches any space or open/close parenthesis character.
The more elegant way would be using proper regex syntax in replace().
Switch ( and ) with [ and ] and use JSON.parse() to get the numbers into your new array.
var str = "(55.930385, -3.118425)".replace("(","[").replace(")","]");
var coords = JSON.parse(str);
If you need to support older browsers, you will need to use a JSON.parse shim or another solution.
How to create regex pattern which is concatenate with variable, something like this:
var test ="52";
var re = new RegExp("/\b"+test+"\b/");
alert('51,52,53'.match(re));
Thanks
var re = new RegExp("/\b"+test+"\b/");
\b in a string literal is a backspace character. When putting a regex in a string literal you need one more round of escaping:
var re = new RegExp("\\b"+test+"\\b");
(You also don't need the // in this context.)
With ES2015 (aka ES6) you can use template literals when constructing RegExp:
let test = '53'
const regexp = new RegExp(`\\b${test}\\b`, 'gi') // showing how to pass optional flags
console.log('51, 52, 53, 54'.match(regexp))
you can use
/(^|,)52(,|$)/.test('51,52,53')
but i suggest to use
var list = '51,52,53';
function test2(list, test){
return !((","+list+",").indexOf(","+test+",") === -1)
}
alert( test2(list,52) )