diff --git a/contact-scrape.go b/contact-scrape.go new file mode 100644 index 0000000..57eaba7 --- /dev/null +++ b/contact-scrape.go @@ -0,0 +1,422 @@ +package main + +import ( + "crypto/md5" + "encoding/json" + "fmt" + "io" + "log" + "net/http" + "os" + "regexp" + "strings" + "time" + + "github.com/fsnotify/fsnotify" + "github.com/xuri/excelize/v2" +) + +type Contact struct { + Name string `json:"name"` + Position string `json:"position"` + Phone string `json:"phone,omitempty"` + ServicePhone string `json:"service_phone,omitempty"` + Table int `json:"table"` +} + +type ContactData struct { + Contacts []Contact `json:"contacts"` + LastUpdated time.Time `json:"last_updated"` + FileHash string `json:"file_hash"` +} + +var ( + currentData *ContactData + dataFile = "data/contacts.json" + xlsxFile = "TelefonniSeznamWeb.xlsx" +) + +func startAutoReload() { + // Create new watcher + watcher, err := fsnotify.NewWatcher() + if err != nil { + log.Printf("Error creating file watcher: %v", err) + return + } + defer watcher.Close() + + // Add the xlsx file to watcher + err = watcher.Add(xlsxFile) + if err != nil { + log.Printf("Error watching file: %v", err) + return + } + + // Start watching for changes + go func() { + for { + select { + case event, ok := <-watcher.Events: + if !ok { + return + } + if event.Has(fsnotify.Write) { + log.Println("Detected file change, reloading data...") + loadData() + } + case err, ok := <-watcher.Errors: + if !ok { + return + } + log.Printf("Watcher error: %v", err) + } + } + }() +} + +func main() { + // Create data directory if it doesn't exist + if err := os.MkdirAll("data", 0755); err != nil { + log.Printf("Warning: Could not create data directory: %v", err) + } + + // Start auto-reload scheduler + startAutoReload() + + // Load existing data or parse from Excel + loadData() + + // Set up HTTP handlers + http.HandleFunc("/", serveIndex) + http.HandleFunc("/contacts", serveContacts) + http.HandleFunc("/reload", reloadData) + + // Start server + port := os.Getenv("PORT") + if port == "" { + port = "80" + } + + log.Printf("Server starting on port %s", port) + log.Printf("Access the application at: http://localhost:%s", port) + log.Fatal(http.ListenAndServe(":"+port, nil)) +} + +func loadData() { + // Check if Excel file exists + if _, err := os.Stat(xlsxFile); os.IsNotExist(err) { + log.Printf("Excel file %s not found, using empty data", xlsxFile) + currentData = &ContactData{ + Contacts: []Contact{}, + LastUpdated: time.Now(), + FileHash: "", + } + return + } + + // Calculate current file hash + currentHash, err := calculateFileHash(xlsxFile) + if err != nil { + log.Printf("Error calculating file hash: %v", err) + return + } + + // Check if cached data exists and is up to date + if cachedData, err := loadCachedData(); err == nil { + if cachedData.FileHash == currentHash { + log.Println("Using cached data (file unchanged)") + currentData = cachedData + return + } + } + + // Parse Excel file + log.Println("Parsing Excel file...") + contacts, err := parseExcelFile(xlsxFile) + if err != nil { + log.Printf("Error parsing Excel file: %v", err) + // Use empty data if parsing fails + currentData = &ContactData{ + Contacts: []Contact{}, + LastUpdated: time.Now(), + FileHash: currentHash, + } + return + } + + currentData = &ContactData{ + Contacts: contacts, + LastUpdated: time.Now(), + FileHash: currentHash, + } + + // Save to cache + if err := saveCachedData(currentData); err != nil { + log.Printf("Warning: Could not save cached data: %v", err) + } + + log.Printf("Loaded %d contacts from Excel file", len(contacts)) +} + +func calculateFileHash(filename string) (string, error) { + file, err := os.Open(filename) + if err != nil { + return "", err + } + defer file.Close() + + hash := md5.New() + if _, err := io.Copy(hash, file); err != nil { + return "", err + } + + return fmt.Sprintf("%x", hash.Sum(nil)), nil +} + +func loadCachedData() (*ContactData, error) { + file, err := os.Open(dataFile) + if err != nil { + return nil, err + } + defer file.Close() + + var data ContactData + decoder := json.NewDecoder(file) + if err := decoder.Decode(&data); err != nil { + return nil, err + } + + return &data, nil +} + +func saveCachedData(data *ContactData) error { + file, err := os.Create(dataFile) + if err != nil { + return err + } + defer file.Close() + + encoder := json.NewEncoder(file) + encoder.SetIndent("", " ") + return encoder.Encode(data) +} + +func parseExcelFile(filename string) ([]Contact, error) { + f, err := excelize.OpenFile(filename) + if err != nil { + return nil, fmt.Errorf("failed to open Excel file: %v", err) + } + defer f.Close() + + // Get the first sheet name + sheets := f.GetSheetList() + if len(sheets) == 0 { + return nil, fmt.Errorf("no sheets found in Excel file") + } + + sheetName := sheets[0] + + // Parse single table (A-E columns) + contacts := parseTable(f, sheetName, "A", "E", 1) + + return contacts, nil +} + +func parseTable(f *excelize.File, sheetName, startCol, endCol string, tableNum int) []Contact { + var contacts []Contact + + // Get all rows in the sheet + rows, err := f.GetRows(sheetName) + if err != nil { + log.Printf("Error getting rows: %v", err) + return contacts + } + + // Skip header rows (first row is header) + startRow := 1 + if len(rows) <= startRow { + return contacts + } + + // Column indices (A-E: name, position, phone, service_phone, extension) + nameCol, positionCol, phoneCol, servicePhoneCol, extensionCol := 0, 1, 2, 3, 4 + + for i := startRow; i < len(rows); i++ { + row := rows[i] + + // Skip if row is too short + if len(row) <= nameCol { + continue + } + + name := strings.TrimSpace(row[nameCol]) + position := "" + phone := "" + servicePhone := "" + extension := "" + + if len(row) > positionCol { + position = strings.TrimSpace(row[positionCol]) + } + if len(row) > phoneCol { + phone = strings.TrimSpace(row[phoneCol]) + } + if len(row) > servicePhoneCol { + servicePhone = strings.TrimSpace(row[servicePhoneCol]) + } + if len(row) > extensionCol { + extension = strings.TrimSpace(row[extensionCol]) + } + + // Clean phone numbers + phone = cleanPhoneNumber(phone) + servicePhone = cleanPhoneNumber(servicePhone) + + // Combine extension with service phone if both exist + if servicePhone != "" && extension != "" { + servicePhone = fmt.Sprintf("%s (%s)", servicePhone, extension) + } else if extension != "" { + servicePhone = extension + } + + // If we have a name, create new contact + if name != "" { + contacts = append(contacts, Contact{ + Name: name, + Position: position, + Phone: phone, + ServicePhone: servicePhone, + Table: tableNum, + }) + } + } + + return contacts +} + +func cleanPhoneNumber(phone string) string { + if phone == "" { + return "" + } + + // Remove extra whitespace + phone = strings.TrimSpace(phone) + + // Remove common formatting characters + re := regexp.MustCompile(`[^\d+\-\s()]`) + phone = re.ReplaceAllString(phone, "") + + // If it's just a short number (internal extension), keep as is + if len(phone) <= 3 { + return phone + } + + // If it looks like a Czech number without country code, add it + if regexp.MustCompile(`^[67]\d{8}$`).MatchString(strings.ReplaceAll(phone, " ", "")) { + return "+420 " + phone + } + + return phone +} + +func serveIndex(w http.ResponseWriter, r *http.Request) { + if r.URL.Path != "/" { + http.NotFound(w, r) + return + } + + // Check if index.html exists + if _, err := os.Stat("index.html"); os.IsNotExist(err) { + // Serve embedded HTML if file doesn't exist + w.Header().Set("Content-Type", "text/html; charset=utf-8") + fmt.Fprint(w, getEmbeddedHTML()) + return + } + + http.ServeFile(w, r, "index.html") +} + +func serveContacts(w http.ResponseWriter, r *http.Request) { + w.Header().Set("Content-Type", "application/json") + w.Header().Set("Access-Control-Allow-Origin", "*") + + if currentData == nil { + http.Error(w, `{"error": "No data available"}`, http.StatusInternalServerError) + return + } + + encoder := json.NewEncoder(w) + encoder.SetIndent("", " ") + encoder.Encode(currentData) +} + +func reloadData(w http.ResponseWriter, r *http.Request) { + if r.Method != "POST" { + http.Error(w, "Method not allowed", http.StatusMethodNotAllowed) + return + } + + log.Println("Manual reload requested") + loadData() + + w.Header().Set("Content-Type", "application/json") + fmt.Fprintf(w, `{"status": "reloaded", "contacts_count": %d}`, len(currentData.Contacts)) +} + +func getEmbeddedHTML() string { + return ` + + + + + Kontakty + + + + +
+
+

📞 Firemní telefonní seznam

+

Poppe + Potthoff kontakty

+
+
+ +
+
+
+ +
+ +
+
+ + +
+
+ +
+
+

Načítání kontaktů...

+
+ + +
+
+ + + +` +}