elvanto / litemoji Goto Github PK

Hi! It is possible to convert cp1251 string "🔥 - это огонь" into "🔥 - это огонь"?
Now LitEmoji::encodeHtml("🔥 - это огонь") returns " - это огонь" :(

// text before:
// <p>А ведь тут совсем другой смысл заложен. Mdash — это черточка шириной с букву М. В русской типографике ее называют длинным тире. Ndash — соответственно более короткая черточка, часто даже уже, чем буква N.</p>\n

$content = self::unicodeToShortcode($content);

// text after
// "<p>А ведь тут совсем другой смысл заложен. Mdash  это черточка шириной с букву М. В русской типографике ее называют длинным тире. Ndash  соответственно более короткая черточка, часто даже уже, чем буква N.</p>\n"

Mdash symbol copypasted from this article: https://medium.com/@sergeisoloviev/mdash-31c331397e46 (2nd paragraph)

It would be nice if it would be possible to exclude a list of Emojis from being removed via method removeEmoji.

is there somewhere a list of full shortcodes with a sample of the output?

Emoji «🦄» do not convert by \LitEmoji\LitEmoji::encodeShortcode('🦄')
https://emojipedia.org/unicorn-face/

Hello.
Code: LitEmoji::encodeUnicode(':kz:');
Expected: 🇰🇿
Received: :kz:

Version litemoji: 4.3.0
PHP 8.1

Hi,
there is missing emoji Shopping trolley (https://unicode-table.com/en/1F6D2/) in regex:
https://github.com/elvanto/litemoji/blob/master/src/LitEmoji.php
const MB_REGEX = '/( \x23\xE2\x83\xA3 # Digits [\x30-\x39]\xE2\x83\xA3 | \xE2[\x9C-\x9E][\x80-\xBF] # Dingbats | \xF0\x9F[\x85-\x88][\xA6-\xBF] # Enclosed characters | \xF0\x9F[\x8C-\x97][\x80-\xBF] # Misc | \xF0\x9F\x98[\x80-\xBF] # Smilies | \xF0\x9F\x99[\x80-\x8F] | \xF0\x9F\x9A[\x80-\xBF] # Transport and map symbols | \xF0\x9F[\xA4-\xA7][\x80-\xBF] # Supplementary symbols and pictographs )/x';

UTF-8 (hex) for shopping trolley is 0xF0 0x9F 0x93 0xA6

The following 2 emojis are not matched by MB_REGEX:
https://emojipedia.org/balance-scale/
https://emojipedia.org/hammer-and-pick/

Unfortunately, this lib has nearly the same problem as every alternative: It doesn't handle skin tones correctly:

This: Hello 👼🏿
Results in Hello :angel::tone5:
But is expected to be: Hello :angel_tone5:

which is obviously not correct.

But it's even worse with this:

This: Hello 👷🏿‍♀️
Results in: Hello :construction_worker::tone5:‍♀️
But is expected to be: Hello :woman_construction_worker_tone5:

The shortcodes-array has these items:

'angel_tone5' => '1F47C-1F3FF',
// ...
'woman_construction_worker_dark_skin_tone' => '1F477-1F3FF-200D-2640-FE0F',
// ...
'woman_construction_worker_tone5' => '1F477-1F3FF-200D-2640-FE0F',

LitEmoji's composer.json says it requires PHP 5.4+:

However v1.4 is using the ARRAY_FILTER_USE_KEY constant, which is only available in PHP 5.6+ (7c0276d).

So either that code should be updated to not use a PHP 5.6+ constant, or composer.json should be updated to require PHP 5.6+.

How to remove found emoji ?

Looks like there's lack of support for a few geometric emojis in the MB_REGEX constant. For reference - https://unicode.org/emoji/charts/full-emoji-list.html#geometric
🟠 🟡 🟢 🟣 🟤 ⚫ ⚪ 🟥 🟧 🟨 🟩 🟦 🟪 🟫 ⬛ ⬜ ◼ ◻

The following are included already:
🔴 🔵

This is despite all these already appearing in the shortcodes-array file, it just doesn't match against the regex pattern.

I'd love to put a PR together, but I'm a novice with Regex at best - sorry!.

Additionally, it'd be neat if this constant could be changed to a static property to allow other to modify the regex to save you having to update it all the time. So we could do:

LitEmoji::$MB_REGEX = ...

I'm using this library to encode emojis in HTML. We experience the problem, that emojis in links are encoded as well, which leads to faulty links.

This is a sample input:

<h1>This is a heading :party:</h1>
<a href="https://dbsw.sharepoint.com/:x:/r/teams/Data/123">This is a link!</a>

This is the outcome:

<h1>This is a heading 🎉</h1>
<a href="https://dbsw.sharepoint.com/❌/r/teams/Data/123">This is a link!</a>

This is the expected outcome:

<h1>This is a heading 🎉</h1>
<a href="https://dbsw.sharepoint.com/:x:/r/teams/Data/123">This is a link!</a>

Is there any way to limit the conversion to shortcodes, which have a space before or after the shortcode? E.g. :x:, :x: . This would prevent the conversion of "shortcodes" in links, which should not include unescaped spaces!

Of course, I could parse the html and iterate through the filtered texts, which seems quite complex. Maybe there is a better / simpler solution for this problem?

At the moment I help myself by parsing the html and iterate through all links after the first conversion:

$html = LitEmoji::encodeUnicode($html);
$dom = new Dom();
$dom->loadStr($html);
$dom->find('a')->each(function ($a) {
    $a->setAttribute('href', LitEmoji::unicodeToShortcode($a->getAttribute('href')));
});
$html = $dom;

Thanks for any help in advance! :)

This PR replaces a call to array_filter with a standard foreach loop in order to maintain compatibility with PHP 5.4, per #7.