regex to skip a tag if it is inside of another tag

Question

I would like to remove anchor tag using php regex from the given string if it's not inside of another tag.

Input:

Hi Hello <a href="#">World</a>. This is <div class="some">testing <a href="#">content</a>. some more content</div>

Output:

Hi Hello. This is <div class="some">testing <a href="#">content</a>. some more content</div>

Thanks in advance.

Where is the anchor tag? – Crunch Much Jun 09 '16 at 07:28 — Crunch Much, Jun 09 '16 at 07:28

LukStorms · Accepted Answer · 2016-06-10T20:29:09.590

4

Something like this :

$string = 'replace <a href="x">A</a> but not <div> <a>B</a> in tag </div> but also <a>C</a><div></div>';

echo preg_replace('/<a[^>]*?>([^<]*)<\/a>(?![^<]*<\/)/gi', '', $string);

A negative lookahead ensures that the anchor tag isn't followed by </.
Thus it's not enclosed by another tag.

The content of the tag is in capture group 1, incase you want to replace by '\1' instead of by ''.

If it's about the div tags specifically then this one will ignore the div's :

echo preg_replace('/<div.*?>.*?<\/div>\K|<a[^>]*?>([^<]*)<\/a>/gi', '\1', $string);

edited Jun 10 '16 at 20:29

answered Jun 09 '16 at 08:04

LukStorms

28,916
5
31
45

1

Just a warning, This regex might help with the example you gave. But for more complex html it would also replace some anchors in a div with multiple anchors.. – LukStorms Jun 09 '16 at 10:08
Imho recursion is needed [(example)](https://regex101.com/r/dR6yY5/1) but as you say for some input it might be convenient. – bobble bubble Jun 09 '16 at 10:42
hi @LukStorms, I have similar question. How can remove the 'a' tag except inside of one class? eg.. 'a' tag should not removed if it is inside class 'tweet'(
) else 'a' tag will be removed.
– Ramesh Jun 10 '16 at 05:58
Perhaps with a negative lookbehind. Which isn't supported in javascript, but should be in the php regex flavor. But it's a bit inflexible to use. For example `(?<!
)(]*?>([^<]*)<\/a>)` would remove all anchors, enclosed or not. Except when it's proceeded by literally `
`. (not that the space matters). [regex101 test](https://regex101.com/r/pE2zH2/2).
– LukStorms Jun 10 '16 at 07:38
Could also try this `(
.*?<\/div>)|]*?>([^<]*)<\/a>` and replace it by `\1\2`. It's with an OR. If it's a tweet div it will get in capture group one, else the content of the tag gets in capture group one. [regex101 test](https://regex101.com/r/sI8yT0/1)
– LukStorms Jun 10 '16 at 07:56
hi @LukStorms, thank you very much for the valuable time. Using (?<!
)(]*?>([^<]*)<\/a>) is working fine but it fails whend when 'a' is not proceeded by
... eg..
testing Tweet
– Ramesh Jun 10 '16 at 09:19
Yeah, cos that div got more then just the class. So the 2nd regex would need to be slightly modified : `(
.*?<\/div>)|]*?>([^<]*)<\/a>`. With the look behind that's impossible because look behinds have issues with stuff like `.*`
– LukStorms Jun 10 '16 at 09:50
Thanks. Last one. Apparently there's a \K switch that resets a match. Which seems useful here. `preg_replace('/
.*?<\/div>\K|]*?>([^<]*)<\/a>/gsi', '\1', $string);`
– LukStorms Jun 10 '16 at 15:23

score 1 · Answer 2 · edited May 23 '17 at 12:08

1

I think this is not a job for regex but also tried by use of a common trick and (*SKIP)(*FAIL)

'~(<(?!a\b)(\w+)\b(?>(?:(?!</?\2\b).)+(?1)?)*</\2>)(*SKIP)(*F)|<a\b.*?</a>\s*~si'

The first part before (*SKIP)(*F) matches and skips any tags that are not <a recursively.
The second part after pipe | is the part to be matched with optional whitespace at the end.
Flags used: s (PCRE_DOTALL), i (PCRE_CASELESS)

Try pattern at regex101 or see eval.in for PHP Demo

There are probably better solutions by use of DOMDocument or other parser.

edited May 23 '17 at 12:08

Community

1
1

answered Jun 09 '16 at 10:33

bobble bubble

16,888
3
27
46

1

First time I see those *SKIP and *FAIL verbs used. Nice. I'll have to experiment with those a bit. – LukStorms Jun 09 '16 at 11:05

regex to skip a tag if it is inside of another tag

2 Answers2