Remove CSS classes without properties via regular expression

Question

I have these CSS codes.

.class-1,
.class-2,
class-3 {
    }
.class-4,
class-5 {
}
.class-6,
class-7 {}
.not-empty-1 {font-size: 12;}
.not-empty-2 {
    font-size: 12;
}
.not-empty-3 {
font-size: 12;
}

I need to remove the CSS classes that don't have any properties.

I have tried this

.+\s{(\n.*|)}

https://regex101.com/r/8xFUDl/2

but it does not remove the previous classes like .class-1, .class-2, as you see in live example.

Any solution is appreciated.

I think that can be tricky, but for the current example perhaps like [^{}\r\n]+(?:\r?\n[^{}\r\n]+)*{\s*} regex101.com/r/QTzEu7/1 — The fourth bird
– The fourth bird, Commented Jan 27, 2020 at 7:55
Are you set on using regex for this? It might be easier using a tokenizer. Also, likely easier to debug and extend/change later. — Cully
– Cully, Commented Jan 27, 2020 at 7:55
@Cully Would you explain more. I am up for any solution which is possible in PHP. — user3631047
– user3631047, Commented Jan 27, 2020 at 7:56
@user3631047 I could post it, but looking at the pattern of anubhava I think that would be the better answer if that also works for you. — The fourth bird
– The fourth bird, Commented Jan 27, 2020 at 8:06
@Thefourthbird: ^(?:[^{}\r\n]*\R)*+[^{]+{\s*}\s* will be more efficient than what I posted earlier. Please post an answer as you were first to solve this. Making a pattern more efficient is a different issue :) — anubhava
– anubhava, Commented Jan 27, 2020 at 8:07

The fourth bird · Accepted Answer · 2020-01-27 11:42:24Z

1

From the comments and credits for the final pattern to @anubhava, for the example data you could use:

^(?:[^{}\r\n]*\R)*+[^{}]+{\s*}\s*

Explanation

^ Start of string
(?: Non capture group
- [^{}\r\n]*\R Match 0+ times any char except { } or a newline and Unicode newline sequence
)*+ Close group and repeat 0+ times using a possessive quantifier
[^{}]+ Match 1+ times any char other than { or }
{\s*} Match { and } with whitespaces in between (Note that \s also matches the newlines)
\s* Match possible trailing whitespace chars

Regex demo

edited Jan 27, 2020 at 11:42

answered Jan 27, 2020 at 8:17

The fourth bird

165k16 gold badges61 silver badges75 bronze badges

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

1 Comment

The fourth bird Over a year ago

@vivek_23 That is why I added a comment in the answer to use [^{}]+ regex101.com/r/DcDFE2/1 I will update the whole pattern, thank you for noticing.

mighty_dev · Accepted Answer · 2020-01-27 08:25:54Z

Since you tagged PHP, we can do a bit of string processing to make regex a bit simpler where we could just match a single .some_class{some_props} style and ignore all others who don't follow this property.

We can ignore all whitespaces, new lines and collect just text in a string with a space as a delimiter between each style. Now, we can explode on each one and see if it matches the style we want which has properties and collect only those and implode them later.

Snippet:

<?php

$str = '.class-1,
.class-2,
class-3 {
    }
.class-4,
class-5 {
}
.class-6,
class-7 {}
.not-empty-1 {font-size: 12;}
.not-empty-2 {
    font-size: 12;
}
.not-empty-3 {
font-size: 12;margin-top:25px;
}';

$new_str = '';

for($i=0;$i<strlen($str);++$i){
    if(strlen(trim($str[$i])) === 0) continue; // skip new lines,spaces, line feeds etc
    $new_str .= $str[$i];
    if($str[$i] == '}') $new_str .= ' ';
}

$results = [];

foreach(explode(" ",$new_str) as $css_style){
    if(preg_match('/^(.+?)\{(.+?)\}$/',$css_style,$matches)){
        $results[] = implode("\n",[$matches[1],"{\n ".$matches[2]."\n}"]);
    }
}


echo implode("\n\n",$results);

Demo: https://3v4l.org/67mXE

Collectives™ on Stack Overflow

Remove CSS classes without properties via regular expression

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related