Blasp Laravel Package

Product Decisions This Supports

• Compliance & Safety: Enforce brand guidelines, legal requirements (e.g., COPPA, GDPR), or platform policies (e.g., "no profanity in comments").
• User Experience: Balance moderation with usability—e.g., masking mild profanity while rejecting extreme cases.
• Roadmap Priorities:
  • Build vs. Buy: Avoid reinventing a profanity filter wheel; leverage this package’s multi-language support, severity scoring, and driver architecture instead of custom regex.
  • Scalability: Use the pipeline driver to chain regex + phonetic for high accuracy without performance hits.
  • Localization: Ship with Spanish/German/French support; extend to other languages via Blasp::in().
• Use Cases:
  • Comments/Forums: Reject profanity in user-generated content (e.g., CheckProfanity middleware).
  • Messaging Apps: Mask profanity in real-time (e.g., Blasp::mask('grawlix')).
  • APIs: Validate incoming data (e.g., Profanity validation rule).
  • Eloquent Models: Auto-sanitize fields like Comment::body via Blaspable trait.
  • Analytics: Track profanity trends with ProfanityDetected events.

When to Consider This Package

• Adopt if:
  • Your app handles user-generated content (comments, reviews, chat) and needs consistent moderation.
  • You require multi-language support (e.g., global audience) or severity-based filtering (e.g., allow "damn" but block "fuck").
  • You want flexibility to customize detection (e.g., regex for obfuscation, phonetic for sound-alikes) or masking (e.g., grawlix for humor).
  • Your team lacks time/resources to maintain a custom profanity filter (this package is batteries-included).
  • You need integration with Laravel’s ecosystem (Eloquent, validation, middleware, Blade directives).
• Look elsewhere if:
  • You need real-time moderation (e.g., live chat) and require sub-10ms latency—this package’s regex driver may be too slow for high-throughput systems (consider a dedicated API like blasp.app).
  • Your use case is niche (e.g., filtering medical jargon) and the package’s word lists are insufficient.
  • You’re not using Laravel/PHP—this package is Laravel-specific (though it powers a universal API).
  • You need whitelisting for specific contexts (e.g., "allow profanity in game chat but not reviews")—this requires custom middleware logic.

How to Pitch It (Stakeholders)

For Executives: *"Blasp is a turnkey profanity filter for Laravel that saves us from building and maintaining a custom solution. It handles obfuscated slurs, multi-language content, and severity-based moderation—all while integrating seamlessly with our stack. For example:

• Comments/Forums: Auto-rejects extreme profanity (e.g., fuck) but masks mild words (e.g., damn) to keep discussions civil.
• APIs: Validates user input in real-time, reducing support costs from flagged content.
• Scalability: The pipeline driver combines regex (for evasions like f*ck) and phonetic (for sound-alikes like phuck) without sacrificing performance.
• Cost: Free to use (MIT license), with optional paid tiers for the [blasp.app API] if we need to scale beyond self-hosted limits. This aligns with our brand safety and user experience goals while reducing engineering overhead."*

For Engineering: *"Blasp is a highly extensible profanity filter with:

• Driver Architecture: Swap between regex (accurate but slower), pattern (fast exact matches), or phonetic (catches sound-alikes like sheit for shit). Chain them with the pipeline driver.
• Severity Scoring: Words are categorized as mild/moderate/high/extreme, letting us set thresholds (e.g., reject only high severity).
• Laravel Integrations:
  • Middleware: Reject or sanitize profanity in incoming requests (blasp:sanitize,mild).
  • Eloquent: Add the Blaspable trait to auto-sanitize model fields like Comment::body.
  • Validation: Use Profanity::in('spanish')->maxScore(50) in form rules.
  • Blade: @clean($comment->body) masks profanity in views.
• Performance: Caches results by default (configurable TTL) and supports batch checking for bulk operations.
• Extensibility: Add custom drivers, word lists, or masking strategies via config. Trade-offs:
• The regex driver is CPU-intensive for high-volume traffic (monitor blasp logs).
• False positives may require tuning the phonetic driver’s max_distance_ratio or adding words to false_positives. Recommendation: Start with the pipeline driver (default: regex + phonetic) and adjust severity thresholds based on A/B tests. Use the middleware for API endpoints and Blaspable for Eloquent models."*