Skip to content

replit.md

Latest commit

 

History

History
107 lines (79 loc) · 4.72 KB

replit.md

File metadata and controls

107 lines (79 loc) · 4.72 KB

Menu Allergen Converter

Overview

The Menu Allergen Converter is a Streamlit-based web application designed for school meal planning. It processes school menu data and automatically generates allergen-free versions by substituting ingredients that contain specified allergens (gluten, dairy, nuts, egg products, soy, fish). The application uses AI-powered suggestions via OpenAI's API to intelligently replace problematic ingredients while maintaining nutritional value and meal structure. Users can upload menu files, select allergens to exclude, review AI-suggested substitutions, and manage custom substitution rules.

User Preferences

Preferred communication style: Simple, everyday language.

System Architecture

Frontend Architecture

Technology: Streamlit web framework

  • Rationale: Provides rapid development of data-centric applications with minimal frontend code
  • Structure: Single-page application with sidebar configuration and main content area
  • State Management: Uses Streamlit's session state to persist processed results, file hashes, and allergen selections across reruns
  • User Flow: Upload menu → Select allergens → AI processes substitutions → Review/customize → Download modified menu

Backend Architecture

Core Components:

  1. Menu Processing Pipeline (utils/menu_processor.py)

    • Parses CSV/text menu files into structured pandas DataFrames
    • Extracts breakfast, lunch, and snack meals from formatted text
    • Maintains original menu structure for output formatting
    • Design Pattern: Parser/transformer pattern for data conversion

  2. AI Substitution Engine (utils/openai_service.py)

    • Integrates with OpenAI API for intelligent ingredient substitution
    • Supports both single meal and batch processing for efficiency
    • Implements retry logic (3 attempts, 2-second delay) for API resilience
    • Trade-off: Uses external AI service (cost, latency) for better substitution quality vs rule-based approach

  3. Substitution Management (utils/substitutions.py)

    • Combines custom user-defined rules with AI-generated suggestions
    • Provides CRUD operations for substitution rules
    • Prioritizes custom rules over AI suggestions for consistency
    • Benefit: Allows organizations to enforce specific dietary guidelines

  4. Confetti Animation (utils/confetti.py)

    • Provides visual feedback using canvas-confetti JavaScript library
    • Enhances user experience upon successful processing
    • Implementation: Injects JavaScript via Streamlit's HTML component

Data Storage

Database: PostgreSQL via SQLAlchemy ORM

Schema Design:

  1. SubstitutionRule Table

    • Fields: id, allergen, original, replacement, created_at
    • Indexed on allergen for query performance
    • Stores user-defined substitution preferences

  2. Menu Table (defined but usage not shown in current code)

    • Fields: id, name, content (JSON), created_at
    • Designed for storing historical menu data
    • JSON column allows flexible menu structure storage

Database Configuration:

  • Connection pooling enabled with pre-ping health checks
  • 1-hour connection recycling to prevent stale connections
  • 10-second connection timeout for failure detection
  • Application name set for database monitoring

Alternatives Considered:

  • SQLite: Simpler but lacks concurrent access capabilities needed for multi-user scenarios
  • NoSQL: More flexible but overkill for straightforward relational data

External Dependencies

Third-Party Services:

  1. OpenAI API

    • Purpose: Generate intelligent ingredient substitutions
    • Authentication: API key via environment variable (OPENAI_API_KEY)
    • Error Handling: Raises ValueError if API key not configured
    • Rate Limiting: Implements retry mechanism for transient failures

  2. Canvas Confetti CDN

    • Purpose: Client-side animation library
    • URL: https://cdn.jsdelivr.net/npm/canvas-confetti@1.5.1/dist/confetti.browser.min.js
    • No authentication required (public CDN)

Python Dependencies:

  • streamlit: Web application framework
  • pandas: Data manipulation and CSV parsing
  • sqlalchemy: ORM for database operations
  • openai: Official OpenAI Python client

Database:

  • PostgreSQL: Primary data store
  • Connection string via DATABASE_URL environment variable
  • Automatic URL transformation for SQLAlchemy compatibility (postgres:// → postgresql://)

Environment Variables Required:

  • DATABASE_URL: PostgreSQL connection string
  • OPENAI_API_KEY: OpenAI API authentication key

Security Considerations:

  • Sensitive credentials stored in environment variables
  • No hardcoded API keys or database credentials
  • Database connection includes application identification for audit trails