Removed old document scraper service

This commit is contained in:
m.nazemi
2026-05-03 13:46:38 +03:30
parent 0bd5e41b70
commit 446c11dfbf
11 changed files with 9 additions and 1371 deletions
+2 -57
View File
@@ -104,9 +104,7 @@ import (
"tm/internal/feedback"
"tm/internal/inquiry"
"tm/internal/kanban"
"tm/internal/notice"
"tm/internal/notification"
"tm/internal/scraper"
"tm/internal/tender"
"tm/internal/tender_approval"
"tm/internal/user"
@@ -118,56 +116,6 @@ import (
"tm/cmd/web/router"
)
func buildTEDConfig(c *bootstrap.TEDConfig) *scraper.TEDConfig {
if c == nil {
return &scraper.TEDConfig{
BaseURL: "https://ted.europa.eu",
Timeout: 30 * time.Second,
MaxRetries: 3,
RetryDelay: 5 * time.Second,
UserAgent: "TM-TED-Scraper/1.0",
MaxConcurrency: 5,
DownloadDir: "./downloads",
CleanupAfter: 24 * time.Hour,
ScrapingInterval: "* * 10 * * *",
AlertMail: "alerts@opplens.com",
}
}
baseURL := c.BaseURL
if baseURL == "" {
baseURL = "https://ted.europa.eu"
}
userAgent := c.UserAgent
if userAgent == "" {
userAgent = "TM-TED-Scraper/1.0"
}
downloadDir := c.DownloadDir
if downloadDir == "" {
downloadDir = "./downloads"
}
scrapingInterval := c.ScrapingInterval
if scrapingInterval == "" {
scrapingInterval = "* * 10 * * *"
}
alertMail := c.AlertMail
if alertMail == "" {
alertMail = "alerts@opplens.com"
}
return &scraper.TEDConfig{
BaseURL: baseURL,
Timeout: c.Timeout,
MaxRetries: c.MaxRetries,
RetryDelay: c.RetryDelay,
UserAgent: userAgent,
MaxConcurrency: c.MaxConcurrency,
DownloadDir: downloadDir,
CleanupAfter: c.CleanupAfter,
ScrapingInterval: scrapingInterval,
MaxNoticesPerDay: c.MaxNoticesPerDay,
AlertMail: alertMail,
}
}
func main() {
conf := bootstrap.InitConfig()
@@ -229,7 +177,7 @@ func main() {
inquiryRepository := inquiry.NewRepository(mongoManager, logger)
contactRepository := contact.NewContactRepository(mongoManager, logger)
cmsRepository := cms.NewRepository(mongoManager, logger)
noticeRepository := notice.NewRepository(mongoManager, logger)
// noticeRepository := notice.NewRepository(mongoManager, logger)
boardRepository := kanban.NewBoardRepository(mongoManager, logger)
columnRepository := kanban.NewColumnRepository(mongoManager, logger)
cardRepository := kanban.NewCardRepository(mongoManager, logger)
@@ -256,8 +204,6 @@ func main() {
notificationService := notification.NewService(notificationSDK, notificationRepository, userService, customerService, logger)
contactService := contact.NewService(contactRepository, logger, notificationSDK)
cmsService := cms.NewService(cmsRepository, logger)
tedConfig := buildTEDConfig(&conf.TED)
scraperService := scraper.NewServiceWithTED(conf.Scraper.BaseURL, conf.Scraper.Timeout, logger, mongoManager, noticeRepository, notificationSDK, tedConfig)
kanbanService := kanban.NewService(boardRepository, columnRepository, cardRepository, logger)
documentScraperService := document_scraper.NewService(tenderRepository, logger)
logger.Info("Services initialized successfully", map[string]interface{}{
@@ -277,7 +223,6 @@ func main() {
notificationHandler := notification.NewHandler(notificationService)
contactHandler := contact.NewHandler(contactService, hcaptchaVerifier)
cmsHandler := cms.NewHandler(cmsService)
scraperHandler := scraper.NewHandler(scraperService)
kanbanHandler := kanban.NewHandler(kanbanService)
fileStoreHandler := filestore.NewHandler(fileStoreService, logger)
documentScraperHandler := document_scraper.NewHandler(documentScraperService, logger)
@@ -289,7 +234,7 @@ func main() {
e := bootstrap.InitHTTPServer(conf, logger)
// Register routes
router.RegisterAdminRoutes(e, userHandler, companyHandler, customerHandler, tenderHandler, feedbackHandler, tenderApprovalHandler, inquiryHandler, categoryHandler, flagHandler, notificationHandler, contactHandler, cmsHandler, scraperHandler, kanbanHandler, fileStoreHandler, xssPolicy)
router.RegisterAdminRoutes(e, userHandler, companyHandler, customerHandler, tenderHandler, feedbackHandler, tenderApprovalHandler, inquiryHandler, categoryHandler, flagHandler, notificationHandler, contactHandler, cmsHandler, kanbanHandler, fileStoreHandler, xssPolicy)
router.RegisterPublicRoutes(e, customerHandler, tenderHandler, feedbackHandler, tenderApprovalHandler, companyHandler, inquiryHandler, categoryHandler, flagHandler, notificationHandler, contactHandler, cmsHandler, fileStoreHandler, xssPolicy)
router.RegisterDocumentScraperRoutes(e, documentScraperHandler, conf.DocumentScraper.APIKey)