Content Sanitization & PHP Documentation Transformer

[butschster]

This PR adds a powerful content sanitization system and a PHP-to-documentation transformer to the context generator.

1. Content Sanitization System

Introduces a comprehensive system for sanitizing and transforming content:

CommentInsertionRule: Adds structured comments to code blocks

{
  "documents": [{
    "description": "Sanitized API Code",
    "outputPath": "docs/sanitized-api.md",
    "sources": [{
      "type": "file",
      "sourcePaths": ["src/Api"],
      "modifiers": [{
        "name": "sanitizer",
        "options": {
          "rules": [{
            "type": "comment",
            "fileHeaderComment": "WARNING: This file contains sensitive operations",
            "methodComment": "Security critical method",
            "frequency": 0
          }]
        }
      }]
    }]
  }]
}

KeywordRemovalRule: Removes sensitive content based on keywords

{
  "documents": [{
    "description": "Redacted Documentation",
    "outputPath": "docs/redacted.md",
    "sources": [{
      "type": "file",
      "sourcePaths": ["config"],
      "modifiers": [{
        "name": "sanitizer",
        "options": {
          "rules": [{
            "type": "keyword",
            "keywords": ["API_KEY", "SECRET", "PASSWORD"],
            "replacement": "[REDACTED]",
            "caseSensitive": false,
            "removeLines": true
          }]
        }
      }]
    }]
  }]
}

RegexReplacementRule: Pattern-based text transformations

{
  "documents": [{
    "description": "Anonymized Content",
    "outputPath": "docs/anonymized.md",
    "sources": [{
      "type": "file",
      "sourcePaths": ["src/User"],
      "modifiers": [{
        "name": "sanitizer",
        "options": {
          "rules": [{
            "type": "regex",
            "patterns": {
              "/[\\w.+-]+@[\\w-]+\\.[\\w.-]+/": "[EMAIL]",
              "/\\d{3}-\\d{2}-\\d{4}/": "[SSN]" 
            }
          }]
        }
      }]
    }]
  }]
}

2. AST Documentation Transformer

Added AstDocTransformer that converts PHP code into structured markdown documentation:

• Parses PHP code using Abstract Syntax Tree (AST)
• Generates clean, readable markdown documentation
• Preserves method implementations in code blocks
• Extracts annotations like route information
• Supports classes, interfaces, traits, and enums

Configuration Example:

{
  "documents": [{
    "description": "API Documentation",
    "outputPath": "docs/api-docs.md",
    "sources": [{
      "type": "file",
      "description": "Controllers",
      "sourcePaths": ["src/Controller"],
      "modifiers": [{
        "name": "php-docs",
        "options": {
          "include_private_methods": false,
          "include_protected_methods": true,
          "extract_routes": true,
          "include_implementations": true,
          "class_heading_level": 1,
          "method_heading_format": "### {name}",
          "code_block_format": "php"
        }
      }]
    }]
  }]
}

Combined Usage Example:

{
  "documents": [{
    "description": "Clean API Documentation",
    "outputPath": "docs/clean-api-docs.md",
    "sources": [{
      "type": "file",
      "description": "API Controllers",
      "sourcePaths": ["src/Controller"],
      "modifiers": [
        {
          "name": "php-docs",
          "options": {
            "include_private_methods": false,
            "extract_routes": true
          }
        },
        {
          "name": "context-sanitizer",
          "options": {
            "rules": [{
              "type": "keyword",
              "keywords": ["SECRET", "INTERNAL"]
            }]
          }
        }
      ]
    }]
  }]
}

This transformer is ideal for:

• Preparing code for LLM context
• Generating API documentation
• Creating code overviews for team discussions