Laravel Grammar Laravel Package

Getting Started

Minimal Setup

1. Installation:

   composer require djunehor/laravel-grammar
   

   Publish the config (if needed):

   php artisan vendor:publish --provider="Djunehor\Grammar\GrammarServiceProvider"
   

2. First Use Case: Detect the part of speech for a single word:

   use Djunehor\Grammar\Facades\Grammar;
   
   $pos = Grammar::detect('running');
   // Returns: ['VERB', 'NOUN', 'ADJECTIVE', 'PRESENT_PARTICIPLE']
   

3. Where to Look First:

   • Facade API in the README.
   • config/grammar.php for customization (e.g., language support).
   • app/Providers/GrammarServiceProvider.php for service binding overrides.

Implementation Patterns

Core Workflows

1. Basic Detection:

   $word = 'quickly';
   $partsOfSpeech = Grammar::detect($word);
   // Use in validation, content analysis, or NLP pipelines.
   

2. Batch Processing:

   $words = ['run', 'jumps', 'happily'];
   $results = Grammar::detectBatch($words);
   // Returns associative array: ['run' => ['VERB', 'NOUN'], ...]
   

3. Integration with Laravel Features:

   • Form Request Validation:

     public function rules()
     {
         return [
             'sentence' => [
                 'required',
                 function ($attribute, $value, $fail) {
                     $words = explode(' ', $value);
                     $adjectives = Grammar::detectBatch($words);
                     if (!array_filter($adjectives, fn($pos) => in_array('ADJECTIVE', $pos))) {
                         $fail('Sentence must contain at least one adjective.');
                     }
                 },
             ],
         ];
     }
     

   • Eloquent Model Observers:

     public function saving(Model $model)
     {
         if ($model->isTaggable()) {
             $tags = explode(',', $model->tags);
             $tagPos = Grammar::detectBatch($tags);
             $model->grammar_tags = json_encode($tagPos);
         }
     }
     

4. Custom Logic with Results:

   $pos = Grammar::detect('darkness');
   if (in_array('NOUN', $pos)) {
       // Trigger noun-specific logic (e.g., semantic analysis).
   }
   

5. Language Switching:

   Grammar::setLanguage('es'); // Spanish
   $pos = Grammar::detect('correr'); // Returns Spanish POS tags.
   

Performance Tips

• Cache Results:

  $cacheKey = 'grammar:'.md5($word);
  $pos = Cache::remember($cacheKey, now()->addHours(1), function() use ($word) {
      return Grammar::detect($word);
  });
  

• Batch Processing: Use detectBatch() for multiple words to reduce overhead.

Gotchas and Tips

Pitfalls

1. Outdated Data:

   • The package was last updated in 2019. Test thoroughly for edge cases (e.g., slang, proper nouns, or domain-specific terms).
   • Example: Grammar::detect('AI') may return unexpected results (e.g., no POS match).

2. Language Limitations:

   • Defaults to English. Other languages (e.g., es, fr) may have incomplete support.
   • Fix: Extend the Djunehor\Grammar\Contracts\GrammarDetector interface to add custom detectors.

3. False Positives:

   • Words like "run" may return multiple POS tags (VERB, NOUN). Validate logic to handle ambiguity:

     $pos = Grammar::detect('run');
     if (count($pos) > 1) {
         // Implement context-aware resolution (e.g., check surrounding words).
     }
     

4. Case Sensitivity:

   • Input is case-insensitive, but ensure consistency in your pipeline (e.g., trim() and strtolower() inputs).

Debugging

• Log Results:

  \Log::debug('POS for "'.$word.'":', Grammar::detect($word));
  

• Inspect Underlying Data: The package uses the MIT License-compatible djunehor/grammar PHP library. Review its source for edge cases.

Extension Points

1. Custom Detectors: Override the default detector in GrammarServiceProvider:

   $this->app->bind(
       \Djunehor\Grammar\Contracts\GrammarDetector::class,
       \App\Services\CustomGrammarDetector::class
   );
   

2. Add New Languages: Extend the GrammarDetector to support unsupported languages by integrating with APIs like WordNet or spaCy.

3. Mocking for Tests:

   $this->app->instance(
       \Djunehor\Grammar\Contracts\GrammarDetector::class,
       Mockery::mock(\Djunehor\Grammar\Contracts\GrammarDetector::class)
   );
   

Config Quirks

• Default Language: Set in config/grammar.php (default_language). Override via Facade:

  Grammar::setLanguage('de'); // German
  

• Performance Tuning: Disable caching in config/grammar.php if using external detectors:

  'cache_enabled' => false,
  

Pro Tips

• Combine with NLP: Use results to enhance search (e.g., Elasticsearch filters) or sentiment analysis.

  if (in_array('ADJECTIVE', Grammar::detect($word))) {
      $sentimentScore = analyzeSentiment($word); // Custom logic.
  }
  

• Leverage for SEO: Analyze content for keyword POS distribution (e.g., prioritize nouns/verbs for meta tags).