tm_back/cmd/scraper/bootstrap.go

package main

import (
	"context"
	"fmt"
	"os"
	"os/signal"
	"syscall"
	"time"
	"tm/internal/tender"
	"tm/pkg/config"
	"tm/pkg/logger"
	"tm/pkg/mongo"
	"tm/ted"
)

// Init Application Configuration
func initConfig() *Config {
	conf, err := config.LoadConfig(".", &Config{})
	if err != nil {
		panic(fmt.Sprintf("Failed to load config: %v", err))
	}

	return conf
}

// Init Logger Service
func initLogger(conf config.LoggingConfig) logger.Logger {
	return logger.NewLogger(&logger.Config{
		Level:  conf.Level,
		Format: conf.Format,
		Output: conf.Output,
		File: logger.FileConfig{
			Path:       conf.File.Path,
			MaxSize:    conf.File.MaxSize,
			MaxBackups: conf.File.MaxBackups,
			MaxAge:     conf.File.MaxAge,
			Compress:   conf.File.Compress,
		},
	})
}

// Init MongoDB Connection Manager
func initMongoDB(conf config.MongoConfig, log logger.Logger) *mongo.ConnectionManager {
	// Convert infra.MongoConfig to mongo.ConnectionConfig
	connectionConfig := mongo.ConnectionConfig{
		URI:                    conf.URI,
		Database:               conf.Name,
		MaxPoolSize:            uint64(conf.MaxPoolSize),
		MinPoolSize:            5,                // Default minimum pool size
		MaxConnIdleTime:        30 * time.Minute, // Default idle time
		ConnectTimeout:         conf.Timeout,
		ServerSelectionTimeout: 5 * time.Second, // Default server selection timeout
	}

	// Create MongoDB connection manager
	connectionManager, err := mongo.NewConnectionManager(connectionConfig, log)
	if err != nil {
		log.Error("Failed to initialize MongoDB connection", map[string]interface{}{
			"error": err.Error(),
			"uri":   conf.URI,
			"db":    conf.Name,
		})
		panic(fmt.Sprintf("Failed to initialize MongoDB connection: %v", err))
	}

	log.Info("MongoDB connection manager initialized successfully", map[string]interface{}{
		"database":  conf.Name,
		"uri":       conf.URI,
		"pool_size": conf.MaxPoolSize,
	})

	return connectionManager
}

// init TED scraper
func initTEDScraper(config Config, mongoManager *mongo.ConnectionManager, appLogger logger.Logger) {
	// Initialize tender repository
	tenderRepo := tender.NewTenderRepository(mongoManager, appLogger)

	// Initialize TED scraper
	tedScraper := ted.NewTEDScraper(
		&ted.Config{
			BaseURL:          config.TED.BaseURL,
			Timeout:          config.TED.Timeout,
			MaxRetries:       config.TED.MaxRetries,
			RetryDelay:       config.TED.RetryDelay,
			UserAgent:        config.TED.UserAgent,
			MaxConcurrency:   config.TED.MaxConcurrency,
			DownloadDir:      config.TED.DownloadDir,
			CleanupAfter:     config.TED.CleanupAfter,
			ScrapingInterval: config.TED.ScrapingInterval,
		},
		appLogger,
		mongoManager,
		tenderRepo,
	)

	// Create context with cancellation
	ctx, cancel := context.WithCancel(context.Background())
	defer cancel()

	// Set up signal handling for graceful shutdown
	signalChan := make(chan os.Signal, 1)
	signal.Notify(signalChan, syscall.SIGINT, syscall.SIGTERM)

	// Start the scraper in a goroutine
	scraperDone := make(chan error, 1)
	go func() {
		appLogger.Info("Starting TED scraper", map[string]interface{}{
			"interval": config.TED.ScrapingInterval.String(),
		})
		scraperDone <- tedScraper.RunPeriodicScraping(ctx)
	}()

	// Wait for shutdown signal or scraper completion
	select {
	case <-signalChan:
		appLogger.Info("Received shutdown signal, stopping scraper...", map[string]interface{}{})
		cancel()

		// Wait for scraper to finish gracefully with timeout
		shutdownTimeout := time.NewTimer(30 * time.Second)
		select {
		case <-scraperDone:
			appLogger.Info("Scraper stopped gracefully", map[string]interface{}{})
		case <-shutdownTimeout.C:
			appLogger.Warn("Scraper shutdown timeout, forcing exit", map[string]interface{}{})
		}

	case err := <-scraperDone:
		if err != nil {
			appLogger.Error("Scraper finished with error", map[string]interface{}{
				"error": err.Error(),
			})
			os.Exit(1)
		}
		appLogger.Info("Scraper finished successfully", map[string]interface{}{})
	}
}