Removing three dots after or before a newline

Question

I use this function to remove three dots before or after a newline:

private function trimThreeDots(string $text): string {
    $threeDotsAtTheBeginning = '((^|[\n]{1})[\\.]{3})';
    $threeDotsAtTheEnd = '([\\.]{3}[\n]{1})';

    $pattern = '/' . $threeDotsAtTheBeginning . '|' . $threeDotsAtTheEnd . '/';

    return preg_replace_callback($pattern, function ($str) {
        return str_replace("...", "", $str[0]);
    }, $text);
}

It works, except that in case that I have a string like this one:

"Bla bla bla...
...bla bla bla"

I only get a single match. But I want to get two matches, because I need to remove both occurrences of three dots. I wonder if maybe there is a more straightforward and efficient way to write this regex and getting two matches in the above case?

You don't need regex, just put the newline character in your `str_replace()` source string. — Alex Howansky, Oct 15 '18 at 13:46
Without regex I would consider splitting into lines, `trim()` the dots, and re-join if necessary. — Andy G, Oct 15 '18 at 13:50
@AndyG `trim()` won't do the job. The second argument is a list of characters to trim. See [this](https://stackoverflow.com/questions/50174181/why-does-ltrim-remove-one-character-when-the-second-argument-contains-an-operato/50174338#50174338) — Cid, Oct 15 '18 at 14:00
This will remove any number of leading/trailing dots. Try this `echo trim("bla bla bla.......", '.');` — Cid, Oct 15 '18 at 14:16
@Cid Yes, I was assuming just a sequence of dots, not stressing the need to remove explicitly three of them. But I'm not necessarily trying to present an alternative solution, just a comment that regex might be overkill for such a task. — Andy G, Oct 15 '18 at 14:23

Wiktor Stribiżew · Accepted Answer · 2018-10-15T13:56:24.133

5

You may use

$s = "Bla bla bla...
...bla bla bla";
echo preg_replace('~(?:\.{3})?(\R)(?:\.{3})?~', '$1', $s);

See the PHP demo and a regex demo.

Details

(?:\.{3})? - an optional sequence of 3 dots
(\R) - Group 1: any line break sequence
(?:\.{3})? - an optional sequence of 3 dots

The $1 replacement replaces the match with the exact line break sequence the regex matches.

edited Oct 15 '18 at 13:56

answered Oct 15 '18 at 13:44

Wiktor Stribiżew

607,720
39
448
563

Didn't need the "\h*" because I am not removing the three dots in case that there is a space after or before a newline. But the rest is ok, thanks. – Ramy Al Zuhouri Oct 15 '18 at 13:54
@RamyAlZuhouri Then why use `\s*` in your `'([\\.]{3}[\s]*[\n]{1})'` pattern? Ok, since that was your pattern error, I removed the `\h*` part. – Wiktor Stribiżew Oct 15 '18 at 13:55

Zenoo · Answer 2 · 2018-10-15T13:53:32.973

2

You can use the RegEx \.{3}(?=\n)|(?<=\n)\.{3}

\.{3}(?=\n) matches the case where the newline is after the dots
(?<=\n)\.{3} matches the case where the newline is before the dots

Demo.

PHP:

preg_replace('~\.{3}(?=\n)|(?<=\n)\.{3}~', '', $s);

edited Oct 15 '18 at 13:53

answered Oct 15 '18 at 13:44

Zenoo

12,670
4
45
69

score 0 · Answer 3 · answered Oct 15 '18 at 13:50

0

Seems overly complex maybe? What about this?

$re = '/(^\.{3}|\.{3}$)/gm';
$str = 'Bla bla bla...
...bla bla bla';
$subst = '';

$result = preg_replace($re, $subst, $str);

echo "The result of the substitution is ".$result;

answered Oct 15 '18 at 13:50

springbo

1,096
1
9
15

Removing three dots after or before a newline

3 Answers3