vpremiss/arabicable

Advanced techniques for handling Arabic text in Laravel apps.

v3.2.0 2024-05-09 15:02 UTC

README

بسم الله الرحمن الرحيم

Arabicable

Several effective strategies for managing Arabic text.

Latest Version on Packagist GitHub Tests Action Status Codecov Total Downloads

Description

The primary approach here is to rely on the database to store variances of each Arabic or Indian numeral column. Having their dedicated columns makes indexing and searching efficient; combined with the appropriate choice among all available migration types based on character capactiy. And while utilizing Laravel's Eloquent observers, we can process only what's necessary again during updates.

So based on the model's property length requirement, we've added migration blueprint macros for Arabic string, tinyText, text, mediumText, and longText; plus a date one for Indian numerals. Using those string or text-type ones would generate 2 extra columns for each arabic column (a configurable affix, that is). And the indian date one would generate a column_indian to hold that in; where column is an example property name.

Finally, take a look at the list of offered methods below (the API section) to understand what kind of processing we're doing on the text in order to essentially preserve the column (without harakat), the column_with_harakat, and the column_searchable where each is prepared exactly for that...

Installation

  1. Install the package via composer:

    composer require vpremiss/arabicable
  2. Add these package service providers manually to your [bootstrap/providers.php] file:

    return [
        App\Providers\AppServiceProvider::class,
        VPremiss\Crafty\CraftyServiceProvider::class, // this as the first
        VPremiss\Arabicable\ArabicableServiceProvider::class, // and this as the second
    ];
  3. Run the package Artisan installer using this command:

    php artisan arabicable:install

[!NOTE]The config file as well as the migration table will be published automatically.

Upgrading

  1. Backup your current config, as well as the common-Arabic-text migration and seeder.

  2. Republish the package stuff using these Artisan commands:

    php artisan vendor:publish --tag="arabicable-config" --force
    php artisan vendor:publish --tag="arabicable-migrations" --force
    php artisan vendor:publish --tag="arabicable-seeders" --force

Usage

Arabicable

Alright, so let's imagine we have Note(s) and we want to have their content arabicable!

  • First, create the migration and add an arabicable column to it:

    use Illuminate\Database\Schema\Blueprint;
    use Illuminate\Support\Facades\Schema; 
    // ...
    Schema::create('notes', function (Blueprint $table) {
        $table->id();
        $table->arabicText('content'); // this also creates `content_searchable` and `content_with_harakat` of the same type
        $table->timestamps();
    });
  • Then let's make the model "arabicable", which activates the observer:

    use Illuminate\Database\Eloquent\Model;
    use VPremiss\Arabicable\Traits\Arabicable;
    
    class Note extends Model
    {
        use Arabicable;
    
        protected $fillable = ['content']; // or you'd guard the property differently
    }
  • Finally, the moment we create a new note and passing it some Arabic content (presumably with harakat), it will process its other columns automatically:

    // When spacing_after_punctuation_only is set to `false` in configuration (default)
    
    $note = Note::create([
        'content' => '"الجَمَاعَةُ مَا وَافَقَ الحَقّ، أَنْتَ الجَمَاعَةُ وَلَو كُنْتَ وَحْدَكْ."',
    ]);
    
    echo $note->content; // "الجماعة ما وافق الحق ، أنت الجماعة ولو كنت وحدك ."
    echo $note->{ar_with_harakat('content')}; // "الجَمَاعَةُ مَا وَافَقَ الحَقّ ، أَنْتَ الجَمَاعَةُ وَلَو كُنْتَ وَحْدَكْ ."
    echo $note->{ar_searchable('content')}; // "الجماعة ما وافق الحق انت الجماعة ولو كنت وحدك"
    
    // When spacing_after_punctuation_only is set to `true` in configuration
    
    $seriousContentPoliticiansDoNotLike = <<<Arabic
    - قال المُزني: سألتُ الشافعي عن مسألة في "الكلام"، فقال: سَلني عن شيء إذا أخطأتَ فيه قُلتُ "أخطأتَ!"، ولا تسألني عن شيء إذا أخطأتَ فيه قُلتُ "كفرتَ".
    Arabic;
    
    $note->update(['content' => $seriousContentPoliticiansDoNotLike]);
    
    echo $note->content;
    // - قال المزني: سألت الشافعي عن مسألة في "الكلام"، فقال: سلني عن شيء إذا أخطأت فيه قلت "أخطأت!"، ولا تسألني عن شيء إذا أخطأت فيه قلت "كفرت".
    echo $note->{ar_with_harakat('content')};
    // - قال المُزني: سألتُ الشافعي عن مسألة في "الكلام"، فقال: سَلني عن شيء إذا أخطأتَ فيه قُلتُ "أخطأتَ!"، ولا تسألني عن شيء إذا أخطأتَ فيه قُلتُ "كفرتَ".
    echo $note->{ar_searchable('content')};
    // قال المزني سالت الشافعي عن مسالة في الكلام فقال سلني عن شيء اذا اخطات فيه قلت اخطات ولا تسالني عن شيء اذا اخطات فيه قلت كفرت

Note

Notice how we can use the global helper functions (ar_with_harakat, ar_searchable, and ar_indian) to get the corresponding property name quickly.

Important

A validation method is employed during text processing to ensure that the text is free of punctuation anomalies that could impact spacing adjustments.

Common Arabic Text

You must ensure that the migration, model, factory, and seeder are all set in place in order for this feature to be utilized.

Among many other filtering methods that Arabic facade provides, there is a removeCommons one. Use it to filter those out to help you search for more focused context.

You can combine that with whole filtered ArabicFilter::forSearch searches ahead to ensure that you didn't miss the quote itself first, and so on...

API

  • Here is a table of all the available custom migration blueprint macro columns:

    Macro Name MySQL Type Maximum Characters or Size
    indianDate date, varchar Varchar: 10
    arabicString varchar 255 - 65,535 characters
    arabicTinyText tinytext ~255 characters (equivalent to VARCHAR(255))
    arabicText text ~65,535 characters
    arabicMediumText mediumtext ~16,777,215 characters
    arabicLongText longtext ~4,294,967,295 characters
    • And keep in mind the following:

      • Each can be passed an $isNullable boolean argument, which affects all columns.
      • Each can be passed an $isUnique boolean argument, which affects the original column.
      • arabicString can be passed a $length integer argument.
      • Both arabicString and arabicTinyText can be passed a $supportsFullSearch argument, affecting their 'searchable' column.
      • Finally arabicText, arabicMediumText, and arabicLongText all do have full-text search index set on their 'searchable' column.
  • Below are the tables of all the Arabicable package helpers:

    ArabicFilter Facade Methods Description
    withHarakat(string $text): string Enhances Arabic text by converting numerals to Indian, normalizing spaces, converting punctuation to Arabic, and refining spaces around punctuation marks. Configurable to add spaces before marks based on application settings.
    withoutHarakat(string $text): string Applies the withHarakat enhancements and then removes diacritic marks from the text.
    forSearch(string $text): string Prepares text for search by removing diacritics, all punctuation, converting numerals to Arabic and Indian sequences, deduplicating these sequences, normalizing letters, and spaces.

    Arabic Facade Methods Description
    removeHarakat(string $text): string Removes diacritic marks from Arabic text.
    normalizeHuroof(string $text): string Normalizes Arabic letters to a consistent form by standardizing various forms of similar letters.
    removeCommons(string&#124;array $words): string&#124;array Removes common Arabic phrases and unnecessary single characters. It works with both strings and arrays, returning the cleaned text in the same format as the input.
    convertNumeralsToIndian(string $text): string Converts Arabic numerals to their Indian numeral equivalents.
    convertNumeralsToArabicAndIndianSequences(string $text): string Converts sequences of numerals in text to both Arabic and Indian numerals, presenting both versions side by side.
    deduplicateArabicAndIndianNumeralSequences(string $text): string Removes duplicate numeral sequences, keeping unique ones at the end of the text.
    convertPunctuationMarksToArabic(string $text): string Converts common foreign punctuation marks to their Arabic equivalents.
    removeAllPunctuationMarks(string $text): string Removes all and every punctuation mark there is; including enclosings and everything.
    validateForTextSpacing(string $text): void Prepares the text for proper spacing by ensuring there is no inconsistency when it comes to the couples of enclosing marks and so on...
    normalizeSpaces(string $text): string Gets read of all the extra spacing (consecutives) in between or around the text.
    addSpacesBeforePunctuationMarks(string $text, array $inclusions = [], array $exclusions = []): string Adds spaces before punctuation marks unless the mark is preceded by another mark or whitespace.
    addSpacesAfterPunctuationMarks(string $text, array $inclusions = [], array $exclusions = []): string Adds spaces after punctuation marks unless the mark is followed by another mark.
    removeSpacesAroundPunctuationMarks(string $text, array $inclusions = [], array $exclusions = []): string Removes spaces around punctuation marks.
    removeSpacesWithinEnclosingMarks(string $text, array $exclusions = []): string Removes spaces immediately inside enclosing marks.
    refineSpacesBetweenPunctuationMarks(string $text): string Refines spacing around punctuation marks based on configurations and special rules.

Changelogs

You can check out the package's changelogs online via WTD.

Progress

You can also checkout the project's roadmap among others in the organization's dedicated section for projects.

Support

Support ongoing package maintenance as well as the development of other projects through sponsorship or one-time donations if you prefer.

And may Allah accept your strive; aameen.

License

This package is open-sourced software licensed under the MIT license.

Credits

Inspiration


والحمد لله رب العالمين