Replace in string with regex

Question

I am struggling to modify captured value with regex.

For example, I wanna change "Hello, he is hero" to "HEllo, HE is HEro" using Regex.

I know there are ways to change this without regex, but it is just an example to show the problem. I actually use the regex instead of just he, but I cannot provide it here. That is why using regex is required.

The code below somehow does not work. Are there any ways to make it work?

"Hello, he is hero".replacingOccurrences(
                of: #"(he)"#,
                with: "$1".uppercased(), // <- uppercased is not applied
                options: .regularExpression
            )

.uppercased will make the literal "$1" uppercased, not the captured group. — Marc
– Marc, Commented Nov 1, 2021 at 7:19
If I knew I would have told you. Just wanted to let you know why your current solution doesn't work — Marc
– Marc, Commented Nov 1, 2021 at 7:23
I think you need to work with Range here so you can manually uppercase each match, see for instance this article — Joakim Danielson
– Joakim Danielson, Commented Nov 1, 2021 at 7:27
using regex is required - why? You need an output, any kind of code that accomplishes this should be ok, why do you specifically need a regex soution? — Cristik
– Cristik, Commented Nov 1, 2021 at 8:13

Joakim Danielson · Accepted Answer · 2021-11-02 10:34:54Z

2

You need to use your regex in combination with Range (range(of:)) to find matches and then replace each found range separately

Here is a function as an extension to String that does this by using range(of:) starting from the start of the string and then moving the start index to match from forward to after the last match. The actual replacement is done inside a separate function that is passed as an argument

extension String {
    func replace(regex: String, with replace: (Substring) -> String) -> String {
        var string = self
        var startIndex = self.startIndex
        let endIndex = self.endIndex

        while let range = string.range(of: regex, options: [.regularExpression] , range: startIndex..<endIndex) {

            if range.isEmpty {
                startIndex = string.index(startIndex, offsetBy: 1)
                if startIndex >= endIndex { break }
                continue
            }
            string.replaceSubrange(range, with: replace(string[range]))

            startIndex = range.upperBound
        }
        return string
    }
}

Example where we do an case insensitive search for words starting with "he" and replace each match with the uppercased version

let result = "Hello, he is hero. There he is".replace(regex: #"(?i)\bhe"#) {
    $0.uppercased() 
}

Output

HEllo, HE is HEro. There HE is

edited Nov 2, 2021 at 10:34

answered Nov 1, 2021 at 16:10

Joakim Danielson

53.1k5 gold badges36 silver badges79 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Leo Dabus Over a year ago

Your method will result in an infinite loop if the resulting range of the regex represents an empty string. Check this post for a proper implementation which supports substrings as well

Joakim Danielson Over a year ago

@LeoDabus Thanks, I didn't think about that edge case

Paul B · Accepted Answer · 2021-11-02 05:20:10Z

0

You can try NSRegularExpression. Something like:

import Foundation

var sourceStr = "Hello, he is hero"
let regex = try! NSRegularExpression(pattern: "(he)")
let matches = regex.matches(in: sourceStr, range: NSRange(sourceStr.startIndex..., in: sourceStr))
regex.enumerateMatches(in: sourceStr, range: NSRange(sourceStr.startIndex..., in: sourceStr)) { (match, _, _) in
            guard  let match = match else { return }
            guard let range = Range(match.range, in: sourceStr) else { return }
            let sub = sourceStr[range]
            sourceStr = sourceStr.replacingOccurrences(of: sub, with: sub.uppercased(), options: [], range: range)
    }

print(sourceStr)

edited Nov 2, 2021 at 5:20

answered Nov 1, 2021 at 16:28

Paul B

5,33347 silver badges57 bronze badges

Comments

kennyzi yusuf · Accepted Answer · 2021-11-01 08:34:55Z

-1

this is the solution i can provide

var string = "Hello, he is hero"
let occurrence = "he"

string = string.lowercased().replacingOccurrences(
    of: occurrence,
    with: occurrence.uppercased(),
    options: .regularExpression
)
print(string)

answered Nov 1, 2021 at 8:34

kennyzi yusuf

443 bronze badges

4 Comments

Pytan Over a year ago

thanks for answering. but if occurrence is like ^he\d+, it doesn't work.

Joakim Danielson Over a year ago

You can’t lowercase the whole string, this might change other characters that shouldn’t be replaced.

kennyzi yusuf Over a year ago

oh i seem, in that case we can change the option to .caseInsensitive and remove the .lowercased() correct?

Joakim Danielson Over a year ago

Yes but you still need to supply a proper regex, now you only have a plain string. And this is will never work using replacingOccurrenses as far as I know

Collectives™ on Stack Overflow

Replace in string with regex

3 Answers 3

2 Comments

Comments

4 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

Comments

4 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related