breakpilot-compliance/ai-compliance-sdk/internal/roadmap/parser.go

package roadmap

import (
	"bytes"
	"encoding/csv"
	"encoding/json"
	"fmt"
	"strconv"
	"strings"
	"time"

	"github.com/xuri/excelize/v2"
)

// Parser handles file parsing for roadmap imports
type Parser struct{}

// NewParser creates a new parser
func NewParser() *Parser {
	return &Parser{}
}

// ColumnMapping defines expected column names and their variations
var ColumnMapping = map[string][]string{
	"title":          {"title", "titel", "name", "bezeichnung", "massnahme", "maßnahme", "aufgabe", "task"},
	"description":    {"description", "beschreibung", "details", "inhalt", "content"},
	"category":       {"category", "kategorie", "bereich", "type", "typ"},
	"priority":       {"priority", "priorität", "prioritaet", "prio", "dringlichkeit"},
	"status":         {"status", "stand", "zustand"},
	"control_id":     {"control_id", "control", "kontrolle", "massnahme_id", "ctrl"},
	"regulation_ref": {"regulation", "regulation_ref", "verordnung", "gesetz", "artikel", "article", "gdpr_ref"},
	"gap_id":         {"gap_id", "gap", "luecke", "lücke"},
	"effort_days":    {"effort_days", "effort", "aufwand", "tage", "days", "pt", "personentage"},
	"assignee":       {"assignee", "verantwortlich", "zustaendig", "zuständig", "owner", "responsible"},
	"department":     {"department", "abteilung", "bereich", "team"},
	"planned_start":  {"planned_start", "start", "beginn", "startdatum", "start_date"},
	"planned_end":    {"planned_end", "end", "ende", "enddatum", "end_date", "deadline", "frist"},
	"notes":          {"notes", "notizen", "bemerkungen", "kommentar", "comment", "anmerkungen"},
}

// DetectedColumn represents a detected column mapping
type DetectedColumn struct {
	Index       int    `json:"index"`
	Header      string `json:"header"`
	MappedTo    string `json:"mapped_to"`
	Confidence  float64 `json:"confidence"`
}

// ParseResult contains the result of parsing a file
type ParseResult struct {
	Format       ImportFormat        `json:"format"`
	TotalRows    int                 `json:"total_rows"`
	ValidRows    int                 `json:"valid_rows"`
	InvalidRows  int                 `json:"invalid_rows"`
	Columns      []DetectedColumn    `json:"columns"`
	Items        []ParsedItem        `json:"items"`
	Errors       []string            `json:"errors"`
}

// ParseFile detects format and parses the file
func (p *Parser) ParseFile(data []byte, filename string, contentType string) (*ParseResult, error) {
	format := p.detectFormat(filename, contentType)

	switch format {
	case ImportFormatExcel:
		return p.parseExcel(data)
	case ImportFormatCSV:
		return p.parseCSV(data)
	case ImportFormatJSON:
		return p.parseJSON(data)
	default:
		return nil, fmt.Errorf("unsupported file format: %s", filename)
	}
}

// detectFormat detects the file format
func (p *Parser) detectFormat(filename string, contentType string) ImportFormat {
	filename = strings.ToLower(filename)

	if strings.HasSuffix(filename, ".xlsx") || strings.HasSuffix(filename, ".xls") {
		return ImportFormatExcel
	}
	if strings.HasSuffix(filename, ".csv") {
		return ImportFormatCSV
	}
	if strings.HasSuffix(filename, ".json") {
		return ImportFormatJSON
	}

	// Check content type
	switch contentType {
	case "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
		"application/vnd.ms-excel":
		return ImportFormatExcel
	case "text/csv":
		return ImportFormatCSV
	case "application/json":
		return ImportFormatJSON
	}

	return ""
}

// parseExcel parses an Excel file
func (p *Parser) parseExcel(data []byte) (*ParseResult, error) {
	result := &ParseResult{
		Format: ImportFormatExcel,
	}

	f, err := excelize.OpenReader(bytes.NewReader(data))
	if err != nil {
		return nil, fmt.Errorf("failed to open Excel file: %w", err)
	}
	defer f.Close()

	// Get the first sheet
	sheets := f.GetSheetList()
	if len(sheets) == 0 {
		return nil, fmt.Errorf("no sheets found in Excel file")
	}

	rows, err := f.GetRows(sheets[0])
	if err != nil {
		return nil, fmt.Errorf("failed to read rows: %w", err)
	}

	if len(rows) < 2 {
		return nil, fmt.Errorf("file must have at least a header row and one data row")
	}

	// Detect column mappings from header
	headers := rows[0]
	result.Columns = p.detectColumns(headers)

	// Parse data rows
	for i, row := range rows[1:] {
		rowNum := i + 2 // 1-based, skip header
		item := p.parseRow(row, result.Columns, rowNum)
		result.Items = append(result.Items, item)
		result.TotalRows++
		if item.IsValid {
			result.ValidRows++
		} else {
			result.InvalidRows++
		}
	}

	return result, nil
}

// parseCSV parses a CSV file
func (p *Parser) parseCSV(data []byte) (*ParseResult, error) {
	result := &ParseResult{
		Format: ImportFormatCSV,
	}

	reader := csv.NewReader(bytes.NewReader(data))
	reader.LazyQuotes = true
	reader.TrimLeadingSpace = true

	// Try different delimiters
	delimiters := []rune{',', ';', '\t'}
	var records [][]string
	var err error

	for _, delim := range delimiters {
		reader = csv.NewReader(bytes.NewReader(data))
		reader.Comma = delim
		reader.LazyQuotes = true

		records, err = reader.ReadAll()
		if err == nil && len(records) > 0 && len(records[0]) > 1 {
			break
		}
	}

	if err != nil {
		return nil, fmt.Errorf("failed to parse CSV: %w", err)
	}

	if len(records) < 2 {
		return nil, fmt.Errorf("file must have at least a header row and one data row")
	}

	// Detect column mappings from header
	headers := records[0]
	result.Columns = p.detectColumns(headers)

	// Parse data rows
	for i, row := range records[1:] {
		rowNum := i + 2
		item := p.parseRow(row, result.Columns, rowNum)
		result.Items = append(result.Items, item)
		result.TotalRows++
		if item.IsValid {
			result.ValidRows++
		} else {
			result.InvalidRows++
		}
	}

	return result, nil
}

// parseJSON parses a JSON file
func (p *Parser) parseJSON(data []byte) (*ParseResult, error) {
	result := &ParseResult{
		Format: ImportFormatJSON,
	}

	// Try parsing as array of items
	var items []map[string]interface{}
	if err := json.Unmarshal(data, &items); err != nil {
		// Try parsing as object with items array
		var wrapper struct {
			Items []map[string]interface{} `json:"items"`
		}
		if err := json.Unmarshal(data, &wrapper); err != nil {
			return nil, fmt.Errorf("failed to parse JSON: %w", err)
		}
		items = wrapper.Items
	}

	if len(items) == 0 {
		return nil, fmt.Errorf("no items found in JSON file")
	}

	// Detect columns from first item
	headers := make([]string, 0)
	for key := range items[0] {
		headers = append(headers, key)
	}
	result.Columns = p.detectColumns(headers)

	// Parse items
	for i, itemMap := range items {
		rowNum := i + 1

		// Convert map to row slice
		row := make([]string, len(result.Columns))
		for j, col := range result.Columns {
			if val, ok := itemMap[col.Header]; ok {
				row[j] = fmt.Sprintf("%v", val)
			}
		}

		item := p.parseRow(row, result.Columns, rowNum)
		result.Items = append(result.Items, item)
		result.TotalRows++
		if item.IsValid {
			result.ValidRows++
		} else {
			result.InvalidRows++
		}
	}

	return result, nil
}

// detectColumns detects column mappings from headers
func (p *Parser) detectColumns(headers []string) []DetectedColumn {
	columns := make([]DetectedColumn, len(headers))

	for i, header := range headers {
		columns[i] = DetectedColumn{
			Index:      i,
			Header:     header,
			Confidence: 0,
		}

		headerLower := strings.ToLower(strings.TrimSpace(header))

		// Try to match against known column names
		for fieldName, variations := range ColumnMapping {
			for _, variation := range variations {
				if headerLower == variation || strings.Contains(headerLower, variation) {
					if headerLower == variation {
						columns[i].MappedTo = fieldName
						columns[i].Confidence = 1.0
					} else if columns[i].Confidence < 0.8 {
						columns[i].MappedTo = fieldName
						columns[i].Confidence = 0.8
					}
					break
				}
			}
			if columns[i].Confidence >= 1.0 {
				break
			}
		}
	}

	return columns
}

// parseRow parses a single row into a ParsedItem
func (p *Parser) parseRow(row []string, columns []DetectedColumn, rowNum int) ParsedItem {
	item := ParsedItem{
		RowNumber: rowNum,
		IsValid:   true,
		Data:      RoadmapItemInput{},
	}

	// Build a map for easy access
	values := make(map[string]string)
	for i, col := range columns {
		if i < len(row) && col.MappedTo != "" {
			values[col.MappedTo] = strings.TrimSpace(row[i])
		}
	}

	// Extract title (required)
	if title, ok := values["title"]; ok && title != "" {
		item.Data.Title = title
	} else {
		item.IsValid = false
		item.Errors = append(item.Errors, "Titel/Title ist erforderlich")
	}

	// Extract optional fields
	if desc, ok := values["description"]; ok {
		item.Data.Description = desc
	}

	// Category
	if cat, ok := values["category"]; ok && cat != "" {
		item.Data.Category = p.parseCategory(cat)
		if item.Data.Category == "" {
			item.Warnings = append(item.Warnings, fmt.Sprintf("Unbekannte Kategorie: %s", cat))
			item.Data.Category = ItemCategoryTechnical
		}
	}

	// Priority
	if prio, ok := values["priority"]; ok && prio != "" {
		item.Data.Priority = p.parsePriority(prio)
		if item.Data.Priority == "" {
			item.Warnings = append(item.Warnings, fmt.Sprintf("Unbekannte Priorität: %s", prio))
			item.Data.Priority = ItemPriorityMedium
		}
	}

	// Status
	if status, ok := values["status"]; ok && status != "" {
		item.Data.Status = p.parseStatus(status)
		if item.Data.Status == "" {
			item.Warnings = append(item.Warnings, fmt.Sprintf("Unbekannter Status: %s", status))
			item.Data.Status = ItemStatusPlanned
		}
	}

	// Control ID
	if ctrl, ok := values["control_id"]; ok {
		item.Data.ControlID = ctrl
	}

	// Regulation reference
	if reg, ok := values["regulation_ref"]; ok {
		item.Data.RegulationRef = reg
	}

	// Gap ID
	if gap, ok := values["gap_id"]; ok {
		item.Data.GapID = gap
	}

	// Effort
	if effort, ok := values["effort_days"]; ok && effort != "" {
		if days, err := strconv.Atoi(effort); err == nil {
			item.Data.EffortDays = &days
		}
	}

	// Assignee
	if assignee, ok := values["assignee"]; ok {
		item.Data.AssigneeName = assignee
	}

	// Department
	if dept, ok := values["department"]; ok {
		item.Data.Department = dept
	}

	// Dates
	if startStr, ok := values["planned_start"]; ok && startStr != "" {
		if start := p.parseDate(startStr); start != nil {
			item.Data.PlannedStart = start
		}
	}

	if endStr, ok := values["planned_end"]; ok && endStr != "" {
		if end := p.parseDate(endStr); end != nil {
			item.Data.PlannedEnd = end
		}
	}

	// Notes
	if notes, ok := values["notes"]; ok {
		item.Data.Notes = notes
	}

	return item
}

// parseCategory converts a string to ItemCategory
func (p *Parser) parseCategory(s string) ItemCategory {
	s = strings.ToLower(strings.TrimSpace(s))

	switch {
	case strings.Contains(s, "tech"):
		return ItemCategoryTechnical
	case strings.Contains(s, "org"):
		return ItemCategoryOrganizational
	case strings.Contains(s, "proz") || strings.Contains(s, "process"):
		return ItemCategoryProcessual
	case strings.Contains(s, "dok") || strings.Contains(s, "doc"):
		return ItemCategoryDocumentation
	case strings.Contains(s, "train") || strings.Contains(s, "schul"):
		return ItemCategoryTraining
	default:
		return ""
	}
}

// parsePriority converts a string to ItemPriority
func (p *Parser) parsePriority(s string) ItemPriority {
	s = strings.ToLower(strings.TrimSpace(s))

	switch {
	case strings.Contains(s, "crit") || strings.Contains(s, "krit") || s == "1":
		return ItemPriorityCritical
	case strings.Contains(s, "high") || strings.Contains(s, "hoch") || s == "2":
		return ItemPriorityHigh
	case strings.Contains(s, "med") || strings.Contains(s, "mitt") || s == "3":
		return ItemPriorityMedium
	case strings.Contains(s, "low") || strings.Contains(s, "nied") || s == "4":
		return ItemPriorityLow
	default:
		return ""
	}
}

// parseStatus converts a string to ItemStatus
func (p *Parser) parseStatus(s string) ItemStatus {
	s = strings.ToLower(strings.TrimSpace(s))

	switch {
	case strings.Contains(s, "plan") || strings.Contains(s, "offen") || strings.Contains(s, "open"):
		return ItemStatusPlanned
	case strings.Contains(s, "progress") || strings.Contains(s, "lauf") || strings.Contains(s, "arbeit"):
		return ItemStatusInProgress
	case strings.Contains(s, "block") || strings.Contains(s, "wart"):
		return ItemStatusBlocked
	case strings.Contains(s, "complet") || strings.Contains(s, "done") || strings.Contains(s, "fertig") || strings.Contains(s, "erledigt"):
		return ItemStatusCompleted
	case strings.Contains(s, "defer") || strings.Contains(s, "zurück") || strings.Contains(s, "verschob"):
		return ItemStatusDeferred
	default:
		return ""
	}
}

// parseDate attempts to parse various date formats
func (p *Parser) parseDate(s string) *time.Time {
	s = strings.TrimSpace(s)
	if s == "" {
		return nil
	}

	formats := []string{
		"2006-01-02",
		"02.01.2006",
		"2.1.2006",
		"02/01/2006",
		"2/1/2006",
		"01/02/2006",
		"1/2/2006",
		"2006/01/02",
		time.RFC3339,
	}

	for _, format := range formats {
		if t, err := time.Parse(format, s); err == nil {
			return &t
		}
	}

	return nil
}

// ValidateAndEnrich validates parsed items and enriches them with mappings
func (p *Parser) ValidateAndEnrich(items []ParsedItem, controls []string, regulations []string, gaps []string) []ParsedItem {
	// Build lookup maps
	controlSet := make(map[string]bool)
	for _, c := range controls {
		controlSet[strings.ToLower(c)] = true
	}

	regSet := make(map[string]bool)
	for _, r := range regulations {
		regSet[strings.ToLower(r)] = true
	}

	gapSet := make(map[string]bool)
	for _, g := range gaps {
		gapSet[strings.ToLower(g)] = true
	}

	for i := range items {
		item := &items[i]

		// Validate control ID
		if item.Data.ControlID != "" {
			if controlSet[strings.ToLower(item.Data.ControlID)] {
				item.MatchedControl = item.Data.ControlID
				item.MatchConfidence = 1.0
			} else {
				item.Warnings = append(item.Warnings, fmt.Sprintf("Control '%s' nicht im Katalog gefunden", item.Data.ControlID))
			}
		}

		// Validate regulation reference
		if item.Data.RegulationRef != "" {
			if regSet[strings.ToLower(item.Data.RegulationRef)] {
				item.MatchedRegulation = item.Data.RegulationRef
			}
		}

		// Validate gap ID
		if item.Data.GapID != "" {
			if gapSet[strings.ToLower(item.Data.GapID)] {
				item.MatchedGap = item.Data.GapID
			} else {
				item.Warnings = append(item.Warnings, fmt.Sprintf("Gap '%s' nicht im Mapping gefunden", item.Data.GapID))
			}
		}
	}

	return items
}