Decoding numeric html entities

koray

I need to bring a string back to its original format. Through a simple CMS text was entered and html markup was applied to it. I now need to use the same text to send in plain text email.

So far I have:

$text = strip_tags($text);
$text = html_entity_decode($text, ENT_QUOTES, "ISO-8859-1");

but this leaves me with numeric entities such as

&#8482;

&#8221;

I tried mb_decode_numericentity() but it does not work on the server I am using.

Any help/direction is much appreciated

bpat1434

So what does NOT decoding the html entities give you?

When you use html_entities_decode you are actually taking all of the ASCII characters used in HTML (like <>) and putting them in the &#8221 format. i don't believe that there is a way around it.

w3schools HTML Entity Character List
WebDesignGroup - HTML 4.0 Entities

~Brett

koray

here is the string

<p>Legunt saepius Claritas est etiam processus dynamicus. Sollemnes &amp;amp;
futurum. Videntur parum clari fiant <strong>sollemnes</strong> in futurum, 
<em>Typi non habent</em> claritatem insitam est&#8212;usus legentis in iis 
qui facit eorum claritatem.</p>

here it is after the strip tags()

Legunt saepius Claritas est etiam processus dynamicus. Sollemnes &amp;amp;
futurum. Videntur parum clari fiant sollemnes in futurum, Typi non habent 
claritatem insitam est&#8212;usus legentis in iis qui facit eorum claritatem.

and after html_entity_decode()

Legunt saepius Claritas est etiam processus dynamicus. Sollemnes & futurum. 
Videntur parum clari fiant sollemnes in futurum, Typi non habent claritatem 
insitam est&#8212;usus legentis in iis qui facit eorum claritatem.

This does a good job in itself with tags and char entities but there are numeric entities being left behind.

Any ideas?

bpat1434

Try using htmlentities() to encode it back to the way it was.

PHP.net HTML entities Function

$text = htmlentities($string, ENT_QUOTES, ISO-8859-1);

htmlentities is the opposite of html_entity_decode().

~Brett

koray

The original text has already been passed through htmlentities(). I'm trying to bring it all back to it's original state. Running it back through htmlentities() will only bring the character entities back that I have already decoded. I want & rather than &amp; - I want ™ rather than &#8482;

Unlike all the help I can find online - I want to decode entities.

without having to write a huge function to address each instance.

bpat1434

Here's the problem I'm seeing:

Your string already contains — and &. Perhaps the function html_entity_decode() is not catching it as an HTML entity since it was originally in the string. Have you tried your string without using the & in the string, but rather the actual '&'?

~Brett

here is the string

<p>Legunt saepius Claritas est etiam processus dynamicus. Sollemnes &amp;
    futurum. Videntur parum clari fiant <strong>sollemnes</strong> in futurum, 
    <em>Typi non habent</em> claritatem insitam est&#8212;usus legentis in iis
    qui facit eorum claritatem.</p>

koray

First off - thanks for taking the time to try and help me out.

note: using spaces in entities in parentheses so they are not recognized and converted.

Funny the quoted string in your post shows an actual ampersand. The original post that it is from shows me (& a m p 😉

Maybe that is why it seems we are not on the same page.

I want numeric entities like (& # 8 2 1 2 😉 to become an actual double quote. I want (& a m p 😉 to become an actual ampersand.

The text is already web ready. it was formatted for the web and saved. I need to now take it and send it in a plain text email. I can not have any html entities. If i can convert them to the original that would be great but if I can not convert them back (such as trademark symbol) then I would rather delete them than have them stay in the entity format.

bpat1434

Okay, are you using PHP 4.3.+? I'm guessing that you are.

Also, have you tried stripping the HTML AFTER you change the entities back to normal text?

I'm also noticing something:

with the & you have a space before & after it. Perhaps adding the proper spacing to &#8212 would allow your code to work?

~Brett

koray

yeah PHP 4.3+ - however, the module needed for the function mb_decode_numericentity() is not installed.

The string I have been using is just for demonstration. There are hundreds of articles in the database. The tags and entities were applied with PHP when they were posted through a CMS.

html_entity_decode() does not change numeric entities. It handles the character entities just fine.

Perhaps I'm not using the proper search terms in google because I have not found much of anything regarding bringing these entities back to original format. Losing hope - have already started the tedious build of the function to take care of these entities. Might actually be good - I can apply things like (c) for copyright and (tm) for trademark, etc. I will post the function here when it is complete. Might be sometime. I hope I find an easier solution though.