Migrate ExcelConverter from Python to TypeScript

Replace pandas+openpyxl with exceljs for Excel file processing. This migration enables seamless integration with the Electron main process and maintains 1:1 functional parity with the Python implementation. Key changes: - Add exceljs dependency (v4.4.0) - Implement ExcelConverter class in TypeScript with strict types - Add comprehensive unit tests (13 test cases, all passing) - Support field name mapping, order parsing, and material extraction - Handle empty data tables and file conflicts gracefully Verification: - Successfully tested with sample Excel file (99 orders, 625 records) - All unit tests passing - Output format matches Python version Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2026-02-28 16:48:28 +08:00
parent d8709d83ae
commit 9db298c027
7 changed files with 1752 additions and 25 deletions
--- a/src/main/utils/excelConverter.ts
+++ b/src/main/utils/excelConverter.ts
@@ -0,0 +1,423 @@
+/**
+ * Excel 报表数据转换工具组件
+ * 将 Excel 报表数据转换为数据库记录形式
+ */
+
+import * as fs from 'fs';
+import * as path from 'path';
+import * as ExcelJS from 'exceljs';
+
+/**
+ * 材料数据行结构
+ */
+export interface MaterialRow {
+  '序号': string | number;
+  '材料编码': string;
+  '材料名称': string;
+  '规格': string;
+  '型号': string;
+  '图号': string;
+  '物料材质': string;
+  '计划数量': number;
+  '单位': string;
+  '需用日期': string;
+  '发料仓库': string;
+  '单位用量': number;
+  '累计出库数量': number;
+  [key: string]: any; // Allow order_info fields
+}
+
+/**
+ * 订单数据结构
+ */
+export interface OrderData {
+  order_info: Record<string, string>;
+  materials: Partial<MaterialRow>[];
+}
+
+/**
+ * 转换器配置选项
+ */
+export interface ConverterOptions {
+  output?: string;
+  verbose?: boolean;
+}
+
+/**
+ * 转换结果
+ */
+export interface ConverterResult {
+  records: MaterialRow[];
+  orderCount: number;
+  rowCount: number;
+  outputFile?: string;
+}
+
+/**
+ * Excel 报表数据转换器
+ */
+export class ExcelConverter {
+  /**
+   * 字段名称映射（解决字段名冲突）
+   */
+  private static readonly FIELD_NAME_MAPPING: Record<string, string> = {
+    计划数量: '产品计划数量',
+    单位: '产品单位',
+  };
+
+  private verbose: boolean;
+
+  constructor(verbose: boolean = true) {
+    this.verbose = verbose;
+  }
+
+  /**
+   * 打印日志（如果 verbose=true）
+   */
+  private print(...args: any[]): void {
+    if (this.verbose) {
+      console.log(...args);
+    }
+  }
+
+  /**
+   * 转换 Excel 文件
+   *
+   * @param inputPath 输入文件路径
+   * @param options 转换选项
+   * @returns 转换结果
+   */
+  public async convert(
+    inputPath: string,
+    options: ConverterOptions = {}
+  ): Promise<ConverterResult> {
+    const verbose = options.verbose ?? this.verbose;
+    const outputFile = options.output
+      ? this.handleOutputFile(options.output)
+      : undefined;
+
+    // 读取工作簿
+    const workbook = new ExcelJS.Workbook();
+    await workbook.xlsx.readFile(inputPath);
+    const worksheet = workbook.worksheets[0]; // 使用活动工作表
+
+    // 解析订单数据
+    const orders = this.parseSheet(worksheet);
+
+    // 转换为扁平化记录
+    const records = this.convertToRecords(orders);
+
+    // 保存文件
+    if (outputFile && records.length > 0) {
+      const outputWorkbook = new ExcelJS.Workbook();
+      const outputWorksheet = outputWorkbook.addWorksheet('Sheet1');
+
+      // 添加标题行
+      if (records.length > 0) {
+        const columns = Object.keys(records[0]);
+        outputWorksheet.columns = columns.map((col) => ({
+          header: col,
+          key: col,
+        }));
+
+        // 添加数据行
+        records.forEach((record) => {
+          outputWorksheet.addRow(record);
+        });
+      }
+
+      await outputWorkbook.xlsx.writeFile(outputFile);
+    }
+
+    // 打印汇总报告
+    this.print('='.repeat(60));
+    this.print('转换完成');
+    this.print('='.repeat(60));
+    this.print(`订单数: ${orders.length}`);
+    this.print(`数据行数: ${records.length}`);
+    this.print(`输出文件: ${outputFile ?? 'N/A'}`);
+    this.print('='.repeat(60));
+
+    return {
+      records,
+      orderCount: orders.length,
+      rowCount: records.length,
+      outputFile,
+    };
+  }
+
+  /**
+   * 处理输出文件，如果文件存在则尝试删除
+   *
+   * @param outputPath 输出文件路径
+   * @returns 实际使用的输出文件路径
+   */
+  private handleOutputFile(outputPath: string): string {
+    if (fs.existsSync(outputPath)) {
+      try {
+        fs.unlinkSync(outputPath);
+      } catch (error: any) {
+        if (error.code === 'EPERM' || error.code === 'EACCES') {
+          this.print(
+            `警告: 无法删除 ${outputPath}，可能文件被其他程序打开`
+          );
+          // 修改文件名
+          const parsedPath = path.parse(outputPath);
+          outputPath = path.join(
+            parsedPath.dir,
+            `${parsedPath.name}_new${parsedPath.ext}`
+          );
+        }
+      }
+    }
+    return outputPath;
+  }
+
+  /**
+   * 解析一个工作表，返回所有订单的数据
+   *
+   * 每个订单包含:
+   * - order_info: 订单头信息（包括页脚）
+   * - materials: 物料数据列表
+   *
+   * @param worksheet ExcelJS 工作表对象
+   * @returns 订单列表
+   */
+  private parseSheet(worksheet: ExcelJS.Worksheet): OrderData[] {
+    const orders: OrderData[] = [];
+
+    // 获取所有行数据
+    const allRows: any[][] = [];
+    worksheet.eachRow((row, rowNumber) => {
+      const values = row.values as any[];
+      allRows.push(values);
+    });
+
+    // 逐行扫描，按订单结构解析
+    let i = 0;
+    while (i < allRows.length) {
+      const row = allRows[i];
+
+      // 检查是否是订单标题行
+      // ExcelJS的row.values数组第一个元素（索引0）通常是undefined，实际数据从索引1开始
+      const firstCell = row[1];
+      if (firstCell && String(firstCell).includes('离散备料计划')) {
+        this.print(`找到订单标题行，行号: ${i + 1}`);
+
+        // 解析订单头信息（接下来的4行）
+        const orderInfo: Record<string, string> = {};
+        for (let j = 1; j <= 4; j++) {
+          if (i + j < allRows.length && allRows[i + j]) {
+            this.parseHeaderRow(allRows[i + j], orderInfo);
+          }
+        }
+
+        this.print(`订单信息: ${JSON.stringify(orderInfo)}`);
+
+        // 查找表格标题行（搜索"序号"）
+        let tableRow = i + 5;
+        while (
+          tableRow < allRows.length &&
+          allRows[tableRow] &&
+          allRows[tableRow][1] !== '序号'
+        ) {
+          tableRow += 1;
+        }
+
+        this.print(`表格标题行: ${tableRow + 1}, 第2列值: "${allRows[tableRow]?.[1]}"`);
+
+        // 检查是否是表格标题行
+        if (
+          tableRow < allRows.length &&
+          allRows[tableRow] &&
+          allRows[tableRow][1] === '序号'
+        ) {
+          // 检查表头下一行是否为空，判断是否存在数据
+          const nextRow = tableRow + 1;
+          const nextRowData = nextRow < allRows.length ? allRows[nextRow] : null;
+          const isEmptyRow =
+            nextRowData &&
+            !this.isRowWithData(nextRowData);
+
+          this.print(`下一行: ${nextRow + 1}, 是否为空: ${!!isEmptyRow}`);
+
+          if (isEmptyRow) {
+            // 没有数据，查找页脚信息
+            this.print('订单无物料数据（空表格）');
+            const materials: Partial<MaterialRow>[] = [];
+            const footerInfo: Record<string, string> = {};
+            let dataRow = nextRow + 1;
+            while (dataRow < allRows.length && allRows[dataRow]) {
+              const firstCell = allRows[dataRow][1];
+              if (
+                firstCell &&
+                (String(firstCell).includes('制单人') ||
+                  String(firstCell).includes('打印人'))
+              ) {
+                this.parseHeaderRow(allRows[dataRow], footerInfo);
+                if (dataRow + 1 < allRows.length && allRows[dataRow + 1]) {
+                  this.parseHeaderRow(allRows[dataRow + 1], footerInfo);
+                }
+                break;
+              }
+              dataRow += 1;
+            }
+
+            orders.push({
+              order_info: { ...orderInfo, ...footerInfo },
+              materials,
+            });
+          } else {
+            // 有数据，开始提取物料
+            this.print('开始提取物料数据...');
+            const materials: Partial<MaterialRow>[] = [];
+            const footerInfo: Record<string, string> = {};
+            let dataRow = tableRow + 1;
+
+            while (dataRow < allRows.length && allRows[dataRow]) {
+              const currentRow = allRows[dataRow];
+              const firstCell = currentRow[1];
+
+              // 检查是否是页脚信息（制单人、打印人）
+              if (firstCell && String(firstCell).includes('制单人')) {
+                // 解析页脚信息
+                this.parseHeaderRow(currentRow, footerInfo);
+                if (dataRow + 1 < allRows.length && allRows[dataRow + 1]) {
+                  this.parseHeaderRow(allRows[dataRow + 1], footerInfo);
+                }
+                this.print('找到页脚信息，停止提取物料');
+                break;
+              }
+
+              // 提取物料数据
+              if (this.isRowWithData(currentRow)) {
+                const material: Partial<MaterialRow> = {
+                  序号: currentRow[1],
+                  材料编码: currentRow[2],
+                  材料名称: currentRow[3],
+                  规格: currentRow[4],
+                  型号: currentRow[5],
+                  图号: currentRow[6],
+                  物料材质: currentRow[7],
+                  计划数量: currentRow[8],
+                  单位: currentRow[9],
+                  需用日期: currentRow[10],
+                  发料仓库: currentRow[11],
+                  单位用量: currentRow[12],
+                  累计出库数量: currentRow[13],
+                };
+
+                this.print(`  提取物料: ${material.序号} - ${material.材料名称}`);
+                materials.push(material);
+              }
+
+              dataRow += 1;
+            }
+
+            this.print(`共提取 ${materials.length} 条物料数据`);
+            orders.push({
+              order_info: { ...orderInfo, ...footerInfo },
+              materials,
+            });
+          }
+        }
+      }
+
+      i += 1;
+    }
+
+    return orders;
+  }
+
+  /**
+   * 检查行是否有数据（非空）
+   */
+  private isRowWithData(row: any[]): boolean {
+    if (!row || row.length === 0) return false;
+    // 检查是否有任何非undefined/null/空字符串的值
+    return row.some(
+      (cell) => cell !== undefined && cell !== null && String(cell).trim() !== ''
+    );
+  }
+
+  /**
+   * 解析订单头信息的一行（字段名和值交错排列）
+   *
+   * @param row 行数据
+   * @param info 存储解析结果的字典
+   */
+  private parseHeaderRow(row: any[], info: Record<string, string>): void {
+    let i = 0;
+    while (i < row.length) {
+      const cell = row[i];
+      if (cell && String(cell).trim() && String(cell).includes('：')) {
+        // 找到字段名
+        let fieldName = String(cell).replace(/：/g, '').trim();
+
+        // 应用字段名映射
+        if (fieldName in ExcelConverter.FIELD_NAME_MAPPING) {
+          fieldName = ExcelConverter.FIELD_NAME_MAPPING[fieldName];
+        }
+
+        // 跳过空单元格，找到第一个非字段名的值
+        let j = i + 1;
+        while (
+          j < row.length &&
+          (!row[j] ||
+            !String(row[j]).trim() ||
+            String(row[j]).includes('：'))
+        ) {
+          j += 1;
+        }
+        if (
+          j < row.length &&
+          row[j] &&
+          !String(row[j]).includes('：')
+        ) {
+          info[fieldName] = String(row[j]).trim();
+        }
+        // 跳过已处理的值，继续找下一个字段名
+        i = j + 1;
+      } else {
+        i += 1;
+      }
+    }
+  }
+
+  /**
+   * 将订单数据转换为扁平化的记录数组
+   *
+   * @param orders 订单列表
+   * @returns 扁平化的记录数组
+   */
+  private convertToRecords(orders: OrderData[]): MaterialRow[] {
+    const allRecords: MaterialRow[] = [];
+
+    for (const order of orders) {
+      const { order_info: orderInfo, materials } = order;
+
+      for (const material of materials) {
+        const record: MaterialRow = {
+          序号: '',
+          材料编码: '',
+          材料名称: '',
+          规格: '',
+          型号: '',
+          图号: '',
+          物料材质: '',
+          计划数量: 0,
+          单位: '',
+          需用日期: '',
+          发料仓库: '',
+          单位用量: 0,
+          累计出库数量: 0,
+          ...orderInfo,
+          ...material,
+        } as MaterialRow;
+        allRecords.push(record);
+      }
+    }
+
+    return allRecords;
+  }
+}