breakpilot-pwa/edu-search-service/internal/config/config.go

package config

import (
	"os"
	"strconv"
)

type Config struct {
	// Server
	Port string

	// OpenSearch
	OpenSearchURL      string
	OpenSearchUsername string
	OpenSearchPassword string
	IndexName          string

	// Crawler
	UserAgent       string
	RateLimitPerSec float64
	MaxDepth        int
	MaxPagesPerRun  int

	// Paths
	SeedsDir string
	RulesDir string

	// API
	APIKey string

	// Backend Integration
	BackendURL   string // URL to Python Backend for Seeds API
	SeedsFromAPI bool   // If true, fetch seeds from API instead of files

	// Embedding/Semantic Search
	EmbeddingProvider   string // "openai", "ollama", or "none"
	OpenAIAPIKey        string // API Key for OpenAI embeddings
	EmbeddingModel      string // Model name (e.g., "text-embedding-3-small")
	EmbeddingDimension  int    // Vector dimension (1536 for OpenAI small)
	OllamaURL           string // Ollama base URL for local embeddings
	SemanticSearchEnabled bool  // Enable semantic search features

	// Scheduler
	SchedulerEnabled  bool   // Enable automatic crawl scheduling
	SchedulerInterval string // Crawl interval (e.g., "24h", "168h" for weekly)

	// PostgreSQL (for Staff/Publications database)
	DBHost     string
	DBPort     string
	DBUser     string
	DBPassword string
	DBName     string
	DBSSLMode  string

	// Staff Crawler
	StaffCrawlerEmail string // Contact email for CrossRef polite pool
}

func Load() *Config {
	return &Config{
		Port:               getEnv("PORT", "8084"),
		OpenSearchURL:      getEnv("OPENSEARCH_URL", "http://opensearch:9200"),
		OpenSearchUsername: getEnv("OPENSEARCH_USERNAME", "admin"),
		OpenSearchPassword: getEnv("OPENSEARCH_PASSWORD", "admin"),
		IndexName:          getEnv("INDEX_NAME", "bp_documents_v1"),
		UserAgent:          getEnv("USER_AGENT", "BreakpilotEduCrawler/1.0 (+contact: security@breakpilot.com)"),
		RateLimitPerSec:    getEnvFloat("RATE_LIMIT_PER_SEC", 0.2),
		MaxDepth:           getEnvInt("MAX_DEPTH", 4),
		MaxPagesPerRun:     getEnvInt("MAX_PAGES_PER_RUN", 500),
		SeedsDir:           getEnv("SEEDS_DIR", "./seeds"),
		RulesDir:           getEnv("RULES_DIR", "./rules"),
		APIKey:             getEnv("EDU_SEARCH_API_KEY", ""),
		BackendURL:         getEnv("BACKEND_URL", "http://backend:8000"),
		SeedsFromAPI:       getEnvBool("SEEDS_FROM_API", true),
		// Embedding/Semantic Search
		EmbeddingProvider:     getEnv("EMBEDDING_PROVIDER", "none"), // "openai", "ollama", or "none"
		OpenAIAPIKey:          getEnv("OPENAI_API_KEY", ""),
		EmbeddingModel:        getEnv("EMBEDDING_MODEL", "text-embedding-3-small"),
		EmbeddingDimension:    getEnvInt("EMBEDDING_DIMENSION", 1536),
		OllamaURL:             getEnv("OLLAMA_URL", "http://ollama:11434"),
		SemanticSearchEnabled: getEnvBool("SEMANTIC_SEARCH_ENABLED", false),
		// Scheduler
		SchedulerEnabled:  getEnvBool("SCHEDULER_ENABLED", false),
		SchedulerInterval: getEnv("SCHEDULER_INTERVAL", "24h"),
		// PostgreSQL
		DBHost:     getEnv("DB_HOST", "postgres"),
		DBPort:     getEnv("DB_PORT", "5432"),
		DBUser:     getEnv("DB_USER", "postgres"),
		DBPassword: getEnv("DB_PASSWORD", "postgres"),
		DBName:     getEnv("DB_NAME", "breakpilot"),
		DBSSLMode:  getEnv("DB_SSLMODE", "disable"),
		// Staff Crawler
		StaffCrawlerEmail: getEnv("STAFF_CRAWLER_EMAIL", "crawler@breakpilot.de"),
	}
}

func getEnvBool(key string, fallback bool) bool {
	if value := os.Getenv(key); value != "" {
		return value == "true" || value == "1" || value == "yes"
	}
	return fallback
}

func getEnv(key, fallback string) string {
	if value := os.Getenv(key); value != "" {
		return value
	}
	return fallback
}

func getEnvInt(key string, fallback int) int {
	if value := os.Getenv(key); value != "" {
		if i, err := strconv.Atoi(value); err == nil {
			return i
		}
	}
	return fallback
}

func getEnvFloat(key string, fallback float64) float64 {
	if value := os.Getenv(key); value != "" {
		if f, err := strconv.ParseFloat(value, 64); err == nil {
			return f
		}
	}
	return fallback
}