RegEx convert img tags from html to xhtml?

DssTrainer

I have a site that I'm upgrading to xhtml from html 4.1

So I got most of the changes done, with the exception of self-closing elements like img, link, etc.

But rather than change 500 spot in the code that use the non-self-closing tag... I was hoping i could do it with a regex replace in a text editor like notepad++ or pspad, ultraedit, etc.

I gather that the regex find for an img tag is
<img[^>]+>

but how can I do

FIND: <img[^>]+>
REPLACE: <img original content />

thanks.

DssTrainer

OK.. i got a bit farther:

FIND: <img([^>]+)>
REPL: <img\1 />

But i have embedded php echos so its finding the first > instead of the closing >

my string:
<img src="<?php echo $image['src'];?>" alt="<?php echo $image['alt'];?>">

Needs to be:
<img src="<?php echo $image['src'];?>" alt="<?php echo $image['alt'];?>" />

DssTrainer

Ok, I got it working with:

FIND: <img(.+)">

REPL: <img\1" />

nrg_alpha

DssTrainer;10912139 wrote:
Ok, I got it working with:

FIND: <img(.+)">

REPL: <img\1" />

That solution isn't entirely reliable.. consider the following example:

$str = '<img src="some/path/some/file"><a class="foo"></a>';
$str = preg_replace('#<img(.+)">#', '<img\1" />', $str);
echo $str;

If your image tag is the last tag prior to a newline, then it will be ok.. if not, you'll have problems..

What is important to understand is that quantifiers like + and * are greedy.. that is to say, they will keep trying to match the last instance of what it is looking for.. in this case, you are using dot match all (which be default matches anything other than a newline. So if you look at my example, (.+)"> will not stop at "> for the image tag, but rather will stop matching "> at the anchor tag.

So the general usage of .+ or .* is frowned upon for not only speed purposes (due to the engine needing to backtrack), but even worse, accuracy issues.

I gave a small example of backtracking explanations here... (but failed to include accuracy issues in that post, which is why I posted my small code snippet above.)

As for the proper solution to this issue, I'm not sure to be honest what it would be.. nested stuff gets ugly in a hurry.. I'm not sure if this can be done using DOMDocument() or XPath and / or not.

DssTrainer

luckily my code base was structured in a way for it to work, but yea I was pleasantly surprised that it didn't take more than it needed.

Weedpacket

$doc = new DOMDocument();
$doc->loadHTML($html);
$doc->getElementsByTagName('html')->item(0)
	->setAttribute('xmlns','http://www.w3.org/1999/xhtml');
$xml = $doc->saveXML();

That just leaves the doctype to be corrected (which can't be done in DOM):

$xml = preg_replace(
	'/<!DOCTYPE.*?>/',
	'<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">',
	$xml);

nrg_alpha

Hmm.. interesting.. I'll have to give that a go just for kicks. Didn't know you can convert entire files like that.

NogDog

Does that work if converting a file that contains <?php...?> code within and/or between tags? (I'm just curious and too lazy to try it myself.)

Weedpacket

Within tags I reckon it would choke ('<' is forbidden in that situation).

Between tags? Provided that the full "<?php" is used to open the PHP block, and so long as '<' or '?>' don't appear within the block .... I don't really know, to be honest: it would be legitimate in well-formed XML, but HTML doesn't have PIs.

shae_marks

Need regex: find img tags in a string
Asked by Smoerble in Regular Expressions
Tags: img, find, tags, regex, string
I have a generated HTML string where I want to strip out all HTML tags except some special ones, so I am looking for a regular expression that can help me.

So my current idea is:
1) use a string replace to make (example) <img .../> tags to [img.../]
2) use a function like stripHTML to get rid of all HTML tags
3) convert the [img.../] back into <img.../>

So I need a query that finds <img...> and <img ... /> (with and without an "/" at the end).
And the same query for [img...]. The second one needs, as far as I know, some escapes, this is why I am not sure if it can be done easily.

shae marks

nrg_alpha

You mean something like this?

Example:

$str = <<<HTML
<p>some text...</p> <img src="http://www.whatever.com/somefile.gif" /> <p>some more text</p> <a href="http://www.whoever.com">link</a> <img src="http://www.someimage.com/someimage.png">
HTML;

function imgSetup($string){
	$string = strip_tags($string, '<img>');
	$string = preg_replace('#<img[^>]+src=[\'"]([^\'"]+)[\'"][^/>]*/?>#i', '[img]$1[/img]', $string);
	return $string;
}

echo imgSetup(&$str);