How would I replace this function with a regex replace

Question

I have a file name with this format yy_MM_someRandomString_originalFileName.

example:

02_01_fEa3129E_my Pic.png

I want replace the first 2 underscores with / so that the example becomes:

02/01/fEa3129E_my Pic.png

That can be done with replaceAll, but the problem is that files may contain underscores as well.

@Test
void test() {

    final var input = "02_01_fEa3129E_my Pic.png";

    final var formatted = replaceNMatches(input, "_", "/", 2);

    assertEquals("02/01/fEa3129E_my Pic.png", formatted);
}

private String replaceNMatches(String input, String regex,
                               String replacement, int numberOfTimes) {
    for (int i = 0; i < numberOfTimes; i++) {
        input = input.replaceFirst(regex, replacement);
    }
    return input;
}

I solved this using a loop, but is there a pure regex way to do this?

EDIT: this way should be able to let me change a parameter and increase the amount of underscores from 2 to n.

input.replaceFirst("^([^_]*)_([^_]*)_", "$1/$2/")

Wiktor Stribiżew
– Wiktor Stribiżew

2019-07-27 15:24:23 +00:00
Commented Jul 27, 2019 at 15:24 — Wiktor Stribiżew
– Wiktor Stribiżew, Commented Jul 27, 2019 at 15:24

The fourth bird · Accepted Answer · 2019-07-27 15:22:59Z

3

You could use 2 capturing groups and use those in the replacement where the match of the _ will be replaced by /

^([^_]+)_([^_]+)_

Replace with:

$1/$2/

Regex demo | Java demo

For example:

String regex = "^([^_]+)_([^_]+)_";
String string = "02_01_fEa3129E_my Pic.png";
String subst = "$1/$2/";

Pattern pattern = Pattern.compile(regex);
Matcher matcher = pattern.matcher(string);
String result = matcher.replaceFirst(subst);

System.out.println(result);

Result

02/01/fEa3129E_my Pic.png

answered Jul 27, 2019 at 15:22

The fourth bird

165k16 gold badges61 silver badges75 bronze badges

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

2 Comments

Ryan S. Over a year ago

Wouldn't I have to completely rewrite the regex if I wanted the first 3 underscores . instead?

The fourth bird Over a year ago

@RyanS. You could add a third group and also refer to that in the replacement regex101.com/r/3l0fMR/2

Pshemo · Accepted Answer · 2019-07-28 15:15:56Z

2

Your current solution has few problems:

It is inefficient - because each replaceFirst need to start from beginning of string so it needs to iterate over same starting characters many times.
It has a bug - because of point 1. while iterating from beginning instead of last modified place, we can replace value which was inserted previously.

For instance if we want to replace single character two times, each with X like abc -> XXc after code like
```
String input = "abc";
input = input.replaceFirst(".", "X"); // replaces a with X -> Xbc
input = input.replaceFirst(".", "X"); // replaces X with X -> Xbc
```
we will end up with Xbc instead of XXc because second replaceFirst will replace X with X instead of b with X.

To avoid that kind of problems you can rewrite your code to use Matcher#appendReplacement and Matcher#appendTail methods which ensures that we will iterate over input once and can replace each matched part with value we want

private static String replaceNMatches(String input, String regex,
                               String replacement, int numberOfTimes) {

    Matcher m = Pattern.compile(regex).matcher(input);
    StringBuilder sb = new StringBuilder();
    int i = 0;
    while(i++ < numberOfTimes && m.find() ){
        m.appendReplacement(sb, replacement); // replaces currently matched part with replacement, 
                                              // and writes replaced version to StringBuilder 
                                              // along with text before the match
    }
    m.appendTail(sb); //lets add to builder text after last match
    return sb.toString();
}

Usage example:

System.out.println(replaceNMatches("abcdefgh", "[efgh]", "X", 2)); //abcdXXgh

edited Jul 28, 2019 at 15:15

answered Jul 27, 2019 at 15:39

Pshemo

125k25 gold badges194 silver badges280 bronze badges

2 Comments

Ryan S. Over a year ago

Brilliant! I was looking at this exact problem as well(looking for a generic solution). I should probably do research into the Matcher class

Pshemo Over a year ago

BTW appendReplacement also supports replacement notation supported by String#replaceAll and String#replaceFirst like $x to reuse match from group x which makes $ special character there. To escape it you would need to use \$ (written in string literal as "\\$"), or let Matcher generate escaped string with Matcher.quoteReplacement.

Collectives™ on Stack Overflow

How would I replace this function with a regex replace

2 Answers 2

2 Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related