Convert and separate emoji from strings - php

Question

I have been searching for some time for what I am trying to accomplish, and I am not an expert in this emoji 'stuff' and I need some help
I have an application which has multiple SMS service providers attached (2 of them at the moment), which sends SMS messages to (inbound), and we send SMS (outbound) all via API.
When an SMS message is received (from either provider), our platform sees the incoming text (with emojis) as:
"hello \ud83e\udd2a"
I have already changed my database to store the emojis, and I changed my charset in the PHP application to display them correctly in HTML, including the email forward, so I am all good there
The issue I am running into is that 1 of the providers (when sending) will accept this as a vaild emoji:
"hello to you \ud83e\udd2a"
But the other will not. The second provider needs (i think) HTML Dec format, so when I sent to them, it needs to look like this:
"hello to you &#128512 ;"
I have 2 separate php functions to send to each provider, so I can do any conversion code from the front-app I need to before it sends it to the provider.
My front-end is using a jQuery emoji picker, so the PHP form post sends "hello to you \ud83e\udd2a" to the php function that calls the API.
Any insight you can give will be greatly appreciated!
thanks in advance.
$utf32 = mb_convert_encoding($message['text'], 'UTF-32', 'UTF-8' );
$hex4 = bin2hex($utf32);
$dec = hexdec($hex4);
$emoji_replaced = "&#$dec;";
echo "&#$dec;";
this was giving me a dec value of the whole string

Answer 1

As a start, when the JSON document is decoded the \ud83e\udd2a will be decoded to the standard 🤪 emoji. These escapes are part of the JSON spec.
As for the provider that "doesn't support" emojis, they likely using a text encoding other than UTF-8 that will fail to display any "special" character or non-english text, not just emojis. This is further evidenced by the fact that they can still render the emoji when you send it as an HTML entity.
While you can use the below code to convert emojis to HTML entity sequences, you should also be checking that other providers docs as to the encoding they use, and converting the messages appropriately. That said, non-UTF-8 encodings simply do not have emojis at all, so you may still want to convert to entities first.
As their name implies, HTML entities will only work when viewed as part of an HTML document. YMMV.
<?php
function emoji_to_entity($emoji) {
return sprintf(
'&#%s;',
unpack(
'Ntgt',
mb_convert_encoding($emoji, 'UTF-32', 'UTF-8')
)['tgt']
);
}
// PHP>=8.2 supports the Extended_Pictographic property
//$emoji_regex = '/\p{Extended_Pictographic}/u';
// PHP<8.2 needs to use a monstrous regex like this
// culled from: https://raw.githubusercontent.com/PCRE2Project/pcre2/master/maint/Unicode.tables/emoji-data.txt
$emoji_regex = '/[\x{23}-\x{23}\x{2a}-\x{2a}\x{30}-\x{39}\x{a9}-\x{a9}\x{ae}-\x{ae}\x{200d}-\x{200d}\x{203c}-\x{203c}\x{2049}-\x{2049}\x{20e3}-\x{20e3}\x{2122}-\x{2122}\x{2139}-\x{2139}\x{2194}-\x{2199}\x{21a9}-\x{21aa}\x{231a}-\x{231b}\x{2328}-\x{2328}\x{2388}-\x{2388}\x{23cf}-\x{23cf}\x{23e9}-\x{23f3}\x{23f8}-\x{23fa}\x{24c2}-\x{24c2}\x{25aa}-\x{25ab}\x{25b6}-\x{25b6}\x{25c0}-\x{25c0}\x{25fb}-\x{25fe}\x{2600}-\x{2605}\x{2607}-\x{2612}\x{2614}-\x{2685}\x{2690}-\x{2705}\x{2708}-\x{2712}\x{2714}-\x{2714}\x{2716}-\x{2716}\x{271d}-\x{271d}\x{2721}-\x{2721}\x{2728}-\x{2728}\x{2733}-\x{2734}\x{2744}-\x{2744}\x{2747}-\x{2747}\x{274c}-\x{274c}\x{274e}-\x{274e}\x{2753}-\x{2755}\x{2757}-\x{2757}\x{2763}-\x{2767}\x{2795}-\x{2797}\x{27a1}-\x{27a1}\x{27b0}-\x{27b0}\x{27bf}-\x{27bf}\x{2934}-\x{2935}\x{2b05}-\x{2b07}\x{2b1b}-\x{2b1c}\x{2b50}-\x{2b50}\x{2b55}-\x{2b55}\x{3030}-\x{3030}\x{303d}-\x{303d}\x{3297}-\x{3297}\x{3299}-\x{3299}\x{fe0f}-\x{fe0f}\x{1f000}-\x{1f0ff}\x{1f10d}-\x{1f10f}\x{1f12f}-\x{1f12f}\x{1f16c}-\x{1f171}\x{1f17e}-\x{1f17f}\x{1f18e}-\x{1f18e}\x{1f191}-\x{1f19a}\x{1f1ad}-\x{1f1ff}\x{1f201}-\x{1f20f}\x{1f21a}-\x{1f21a}\x{1f22f}-\x{1f22f}\x{1f232}-\x{1f23a}\x{1f23c}-\x{1f23f}\x{1f249}-\x{1f53d}\x{1f546}-\x{1f64f}\x{1f680}-\x{1f6ff}\x{1f774}-\x{1f77f}\x{1f7d5}-\x{1f7ff}\x{1f80c}-\x{1f80f}\x{1f848}-\x{1f84f}\x{1f85a}-\x{1f85f}\x{1f888}-\x{1f88f}\x{1f8ae}-\x{1f8ff}\x{1f90c}-\x{1f93a}\x{1f93c}-\x{1f945}\x{1f947}-\x{1faff}\x{1fc00}-\x{1fffd}\x{e0020}-\x{e007f}]/u';
$input = json_decode('"hello to you \ud83e\udd2a"');
var_dump(
$input,
preg_replace_callback(
$emoji_regex,
function($a) { return emoji_to_entity($a[0]); },
$input
)
);
Output:
string(17) "hello to you 🤪"
string(22) "hello to you 🤪"
And outside of a code block: string(22) "hello to you 🤪"