muyu-apiserver/pkg/excelparse/parse.go

package excelparse

import (
	"bytes"
	"context"
	"encoding/json"
	"fmt"
	"os"
	"os/exec"
	"path/filepath"
	"strings"
	"time"

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

var parsers []FormatParser

func init() {
	parsers = append(parsers, &NativeProductParser{})
	parsers = append(parsers, &MiaozhangParser{})
}

// RegisterParser adds a format parser at runtime (e.g. for plugins).
func RegisterParser(p FormatParser) {
	parsers = append(parsers, p)
}

// xlsToolPath returns the path to the xls2json.js script.
// Checks XLS_TOOL_PATH env var first, then falls back to a path
// relative to the binary's directory so it works both locally and in Docker.
func xlsToolPath() (string, error) {
	if p := os.Getenv("XLS_TOOL_PATH"); p != "" {
		return p, nil
	}
	// Docker layout: /app/tools/xls2json.js
	candidates := []string{
		"/app/tools/xls2json.js",
	}
	// Local dev: look relative to the process working directory
	if cwd, err := os.Getwd(); err == nil {
		candidates = append(candidates,
			filepath.Join(cwd, "deploy/tools/xls2json.js"),
			filepath.Join(cwd, "../deploy/tools/xls2json.js"),
		)
	}
	for _, p := range candidates {
		if _, err := os.Stat(p); err == nil {
			return p, nil
		}
	}
	return "", fmt.Errorf("xls2json.js not found; set XLS_TOOL_PATH env var or deploy to /app/tools/")
}

// ParseFile reads an Excel file, auto-detects the competitor format, and returns parsed products.
func ParseFile(filePath string) (*ParseResult, error) {
	ext := strings.ToLower(filepath.Ext(filePath))

	var headers []string
	var rows [][]string
	var err error

	switch ext {
	case ".xlsx":
		headers, rows, err = readXLSX(filePath)
	case ".xls":
		headers, rows, err = readXLS(filePath)
	default:
		return nil, ErrUnsupportedExt
	}
	if err != nil {
		return nil, fmt.Errorf("read excel: %w", err)
	}
	if len(rows) == 0 {
		return nil, ErrEmptyFile
	}

	for _, p := range parsers {
		if p.Match(headers) {
			result, err := p.ParseRows(headers, rows)
			if err != nil {
				return nil, fmt.Errorf("parse %s: %w", p.Name(), err)
			}
			result.ParserName = p.Name()
			return result, nil
		}
	}

	return nil, ErrUnknownFormat
}

func readXLSX(filePath string) ([]string, [][]string, error) {
	f, err := excelize.OpenFile(filePath)
	if err != nil {
		return nil, nil, err
	}
	defer f.Close()

	sheets := f.GetSheetList()
	if len(sheets) == 0 {
		return nil, nil, ErrEmptyFile
	}

	allRows, err := f.GetRows(sheets[0])
	if err != nil {
		return nil, nil, err
	}
	if len(allRows) < 2 {
		return nil, nil, ErrEmptyFile
	}

	return allRows[0], allRows[1:], nil
}

// readXLS delegates to the xls2json.js Node.js script via subprocess.
// This handles legacy .xls (BIFF8) files including those with embedded images,
// which pure-Go libraries cannot reliably process.
func readXLS(filePath string) ([]string, [][]string, error) {
	toolPath, err := xlsToolPath()
	if err != nil {
		return nil, nil, err
	}

	ctx, cancel := context.WithTimeout(context.Background(), 60*time.Second)
	defer cancel()

	var stdout, stderr bytes.Buffer
	cmd := exec.CommandContext(ctx, "node", toolPath, filePath)
	cmd.Stdout = &stdout
	cmd.Stderr = &stderr

	if err := cmd.Run(); err != nil {
		msg := strings.TrimSpace(stderr.String())
		if msg == "" {
			msg = err.Error()
		}
		return nil, nil, fmt.Errorf("xls2json: %s", msg)
	}

	// JSON format: [][]interface{}  (first row = headers)
	var raw [][]interface{}
	if err := json.Unmarshal(stdout.Bytes(), &raw); err != nil {
		return nil, nil, fmt.Errorf("xls2json output parse: %w", err)
	}
	if len(raw) < 2 {
		return nil, nil, ErrEmptyFile
	}

	toStrings := func(row []interface{}) []string {
		out := make([]string, len(row))
		for i, v := range row {
			if v == nil {
				out[i] = ""
			} else {
				out[i] = strings.TrimSpace(fmt.Sprintf("%v", v))
			}
		}
		return out
	}

	headers := toStrings(raw[0])
	rows := make([][]string, 0, len(raw)-1)
	for _, r := range raw[1:] {
		rows = append(rows, toStrings(r))
	}

	return headers, rows, nil
}