aaronpk / emoji-detector-php Goto Github PK

This library will find all emoji in an input string and return information about each emoji character.

License: MIT License

PHP 100.00%

emoji-detector-php's Introduction

Emoji Detection

This library will find all emoji in an input string and return information about each emoji character. It supports emoji with skin tone modifiers, as well as the composite emoji that are made up of multiple people.

The current version supports Emoji version 15.1 (Sept 2023)

You can see a catalog of the emoji data here:

Installation

composer require p3k/emoji-detector

Or include src/Emoji.php in your project, and make sure the map.json and regexp.json files are available in the same folder as Emoji.php. You don't need any of the other files for use in your own projects.

Usage

Detect Emoji

$input = "Hello 👍🏼 World 👨‍👩‍👦‍👦";
$emoji = Emoji\detect_emoji($input);

print_r($emoji);

The function returns an array with details about each emoji found in the string.

Array
(
    [0] => Array
        (
            [emoji] => 👍🏼
            [short_name] => +1
            [num_points] => 2
            [points_hex] => Array
                (
                    [0] => 1F44D
                    [1] => 1F3FC
                )
            [hex_str] => 1F44D-1F3FC
            [skin_tone] => skin-tone-3
            [byte_offset] => 6
            [grapheme_offset] => 6
        )
    [1] => Array
        (
            [emoji] => 👨‍👩‍👦‍👦
            [short_name] => man-woman-boy-boy
            [num_points] => 7
            [points_hex] => Array
                (
                    [0] => 1F468
                    [1] => 200D
                    [2] => 1F469
                    [3] => 200D
                    [4] => 1F466
                    [5] => 200D
                    [6] => 1F466
                )
            [hex_str] => 1F468-200D-1F469-200D-1F466-200D-1F466
            [skin_tone] =>
            [byte_offset] => 21
            [grapheme_offset] => 14
        )
)

emoji - The emoji sequence found, as the original byte sequence. You can output this to show the original emoji.
short_name - The short name of the emoji, as defined by Slack's emoji data.
num_points - The number of unicode code points that this emoji is composed of.
points_hex - An array of each unicode code point that makes up this emoji. These are returned as hex strings. This will also include "invisible" characters such as the ZWJ character and skin tone modifiers.
hex_str - A list of all unicode code points in their hex form separated by hyphens. This string is present in the Slack emoji data array.
skin_tone - If a skin tone modifier was used in the emoji, this field indicates which skin tone, since the short_name will not include the skin tone.
byte_offset - The position of the emoji in the string in bytes, used with the plain str* functions
grapheme_offset - The position of the emoji in the string, counting each emoji as 1 char, used with the grapheme_* functions

You can use the grapheme_* functions to extract parts of the string using the grapheme_offset position returned. For example:

$string = "Trešnja 🍒";
$emoji = Emoji\detect_emoji($string);
echo '.'.grapheme_substr($string, 0, $emoji[0]['grapheme_offset']).".\n";
echo '.'.substr($string, 0, $emoji[0]['byte_offset']).".\n";
// Both output ".Trešnja ."

Replace emoji with string representations

$string = Emoji\replace_emoji('I like 🌮 and 🌯', ':', ':');
echo $string;
// I like :taco: and :burrito:

Test if a string is a single emoji

Since simply counting the number of unicode characters in a string does not tell you how many visible emoji are in the string, determining whether a single character is an emoji is more involved. This function will return the emoji data only if the string contains a single emoji character, and false otherwise.

$emoji = Emoji\is_single_emoji('👨‍👩‍👦‍👦');
print_r($emoji);

Array
(
    [emoji] => 👨‍👩‍👦‍👦
    [short_name] => man-woman-boy-boy
    [num_points] => 7
    [points_hex] => Array
        (
            [0] => 1F468
            [1] => 200D
            [2] => 1F469
            [3] => 200D
            [4] => 1F466
            [5] => 200D
            [6] => 1F466
        )

    [hex_str] => 1F468-200D-1F469-200D-1F466-200D-1F466
    [skin_tone] =>
    [byte_offset] => 0
    [grapheme_offset] => 0
)

$emoji = Emoji\is_single_emoji('😻🐈');
// false

Remove emoji from a string

You can remove all emoji from a string with this function, optionally removing trailing spaces.

$string = "I like 🌮 and 🌯";
echo Emoji\remove_emoji($string);
// "I like  and "
echo Emoji\remove_emoji($string, ['collapse' => true]);
// "I like and";

Updates

When a new emoji set is released, this library will need to be updated with the new unicode points and names. The source of the emoji data is iamcal/emoji-data, so first check there for the latest updates. You can build the new source files this library uses with the following command:

composer build

Tests

A comprehensive set of tests is available to ensure things are working as expected, including tests for the new emoji added in new emoji versions. You can run the tests with the following command:

composer test

License

Made with ❤️ by Aaron Parecki.

Emoji data sourced from iamcal/emoji-data under the MIT license.

Emoji parsing regex sourced from EmojiOne under the MIT license.

emoji-detector-php's People

Contributors

Stargazers

Watchers

emoji-detector-php's Issues

add `strlen()`/`mb_strlen()` to `is_single_emoji()`

Before throwing in my whole blogpost to see if it's a single emoji, I first checked if strlen($txt) < 10. It felt like a good thing to do with performance in mind. It turned out that the strlen of 🏳️‍🌈 is actually 14, so I changed it into mb_strlen(), which has 4 for rainbow_flag.

It might be nice to add that check to is_single_emoji() itself.

It now starts with a detect_emoji(), which uses the new monster regex on the full string. If a blogpost is 5000+ chars, it's surely not a single emoji.

I don't know if adding if(mb_strlen($string) > 10) return false; is the right way to go here. What is the longest emoji there is now? I found the following very long ones in the map.json:

"1F469-200D-2764-FE0F-200D-1F48B-200D-1F469":"woman-kiss-woman"
"1F468-200D-2764-FE0F-200D-1F48B-200D-1F468":"man-kiss-man"

Both are 8 in mb_strlen. So >= 8 should be fine already.

Not all emojis.

Great library!

I have only one issue: some "older" emojis such as ❤ are not detected. I know this is a general problem of the definition "emoji", but what is the difference between ❤ and 🏰💍💑? I think none, they're all emojis to me. However:

🏰💍💑 are detected
❤ is not detected

Non-qualified skin variations are not recognized

I saw that support for non-qualified emojis was added recently, but it appears that non-qualified skin variations are still unrecognized.

For example, 1F937 1F3FD 200D 2642 is defined in https://unicode.org/Public/emoji/14.0/emoji-test.txt, but 1F937-1F3FD-200D-2642 is not in map.json.

I tried replacing the following code in the generator, which did regenerate the map properly.

  if(isset($emoji['skin_variations'])) {
    foreach($emoji['skin_variations'] as $key=>$var) {
      $map[$var['unified']] = $short_name;
    }
  }

  if(isset($emoji['skin_variations'])) {
    foreach($emoji['skin_variations'] as $key=>$var) {
      $map[$var['unified']] = $short_name;

      if(isset($var['non_qualified'])) {
        $map[$var['non_qualified']] = $short_name;
      }
    }
  }

However, I started getting errors about the regex being too large.

Fails to factor in gender

💂🏿‍♀️ should return:
1f482-1f3ff-200d-2640-fe0f (woman guard: dark skin tone)
https://abs.twimg.com/emoji/v2/72x72/1f482-1f3ff-200d-2640-fe0f.png

instead it returns 2 separate emojis:

guardsman: dark skin tone
female_sign

Emoji 14.0 Update please?

Just found this good script. It works pretty well, but, sadly, doesn't support Emoji 14.0. Can you update it please? https://www.unicode.org/emoji/charts/full-emoji-list.html

Refactor `detect_emoji` function to not use a regex

As of Unicode 15.1, there are too many code points for the (ugly hack) regex now. We need to switch to a different method of finding the emoji.

New Version with current Emojis

@aaronpk the last version from this lib is quite old and there are some new commits with new emojis already merged.

A new version of this library would be much appreciated.

Error when passing in unrecognizable emojis

\Emoji\detect_emoji($string) seems to return an error if there are some number of unrecognizable emojis (or so thats my assumption).

Unfortunately, i do not know what emojis are causing this, but from our logs we can see here which strings are failing if that gives you any clues.

Add a function to strip all emojis

I'm using this library to strip all emojis from a string.

Here's the code, if you want to create a function for it in this package.

$emojis = \Emoji\detect_emoji($string);

foreach (array_reverse($emojis) as $emoji) {
  $length = strlen($emoji['emoji']);
  $start = substr($string, 0, $emoji['byte_offset']);
  $end = substr($string, $emoji['byte_offset'] + $length, strlen($string) - ($emoji['byte_offset'] + $length));
  $string = $start . $end;
}

Some emojis are not detected properly

☝ - this is not recognised as emoji

© - but this on is recognised as emoji, is it mistaken with ©️
® - but this on is recognised as emoji, is it mistaken with ®️
™ - but this on is recognised as emoji, is it mistaken with ™️

Update to Emoji 15.1

PHP 5.3 Compatibility

We're using this now in Semantic Linkbacks. This library uses short array notation. Being as this is the only issue for PHP 5.3 compatibility, would you consent to not using short array notation in order to encourage compatibility as it is not a breaking change?

Keep up the good work