einvoice-cli 1.0.0

package/lib/PdfFullElectronicInvoiceService.js

@@ -0,0 +1,325 @@
+const { Invoice, Detail } = require('./Invoice');
+const RegexPatterns = require('./RegexPatterns');
+const InvoiceValidator = require('./InvoiceValidator');
+const BaseInvoiceService = require('./BaseInvoiceService');
+
+/**
+ * PDF 全电发票服务 - 优化版
+ * 整合：正则表达式优化 + 字段验证
+ */
+class PdfFullElectronicInvoiceService {
+  static extract(fullText, allText, pageWidth, items) {
+    return BaseInvoiceService.safeExtract(() => {
+      const invoice = new Invoice();
+
+      this.extractBasicFields(invoice, allText);
+      this.extractAmountInfo(invoice, allText, fullText);
+      BaseInvoiceService.extractPersonInfo(invoice, allText);
+      this.extractDetails(invoice, fullText, allText, items);
+
+      BaseInvoiceService.validateInvoice(invoice);
+
+      return invoice;
+    });
+  }
+
+  static extractBasicFields(invoice, allText) {
+    // 使用基类的通用方法
+    BaseInvoiceService.extractBasicFields(invoice, allText);
+
+    // 全电发票特定的字段提取
+    const electronicsNumberMatch = allText.match(RegexPatterns.UTILITY_PATTERNS.electronicsNumber);
+    if (electronicsNumberMatch) {
+      invoice.number = electronicsNumberMatch[1];
+    }
+
+    // 提取日期（使用基类方法）
+    const date = BaseInvoiceService.extractDate(allText);
+    if (date) {
+      invoice.date = date;
+    }
+
+    // 提取购销方名称（使用基类方法）
+    const partyNames = BaseInvoiceService.extractPartyNames(allText);
+    if (partyNames.buyerName) {
+      invoice.buyerName = partyNames.buyerName;
+    }
+    if (partyNames.sellerName) {
+      invoice.sellerName = partyNames.sellerName;
+    }
+
+    // 提取纳税人识别号（使用基类方法）
+    const taxIds = BaseInvoiceService.extractTaxIds(allText);
+    if (taxIds.length > 0) {
+      invoice.buyerCode = taxIds[0];
+    }
+    if (taxIds.length > 1) {
+      invoice.sellerCode = taxIds[1];
+    }
+
+    this.detectInvoiceType(invoice, allText);
+  }
+
+  static detectInvoiceType(invoice, allText) {
+    // 优先匹配完整的发票类型（注意：allText是规范化后的文本，括号可能被替换）
+    const fullMatch = allText.match(/电子发票\s*（增值税专用发票）/) ||
+                     allText.match(/电子发票（增值税专用发票）/) ||
+                     allText.match(/电子发票\s*\(增值税专用发票\)/) ||
+                     allText.match(/电子发票\(增值税专用发票\)/) ||
+                     allText.match(/电子发票\s*（普通发票）/) ||
+                     allText.match(/电子发票（普通发票）/) ||
+                     allText.match(/电子发票\s*\(普通发票\)/) ||
+                     allText.match(/电子发票\(普通发票\)/);
+
+    if (fullMatch) {
+      // 提取完整的发票类型
+      const matchText = fullMatch[0];
+      if (matchText.includes('增值税专用发票')) {
+        invoice.title = '电子发票（增值税专用发票）';
+        invoice.type = '专用发票';
+      } else if (matchText.includes('普通发票')) {
+        invoice.title = '电子发票（普通发票）';
+        invoice.type = '普通发票';
+      }
+    } else if (allText.includes('电子发票')) {
+      invoice.title = '电子发票';
+      invoice.type = '电子发票';
+    } else if (allText.includes('普通发票')) {
+      invoice.title = '普通发票';
+      invoice.type = '普通发票';
+    } else if (allText.includes('增值税专用发票')) {
+      invoice.title = '增值税专用发票';
+      invoice.type = '专用发票';
+    } else if (allText.includes('通行费')) {
+      invoice.title = '通行费发票';
+      invoice.type = '通行费';
+    }
+  }
+
+  static extractAmountInfo(invoice, allText, fullText) {
+    // 首先尝试从"合计"行提取金额和税额
+    const amountMatch = fullText.match(/合\s*计\s*¥?\s*(\d+\.\d+)\s+¥?\s*(\d+\.\d+)/);
+    if (amountMatch) {
+      invoice.amount = amountMatch[1];
+      invoice.taxAmount = amountMatch[2];
+    }
+
+    // 如果没找到，尝试其他模式
+    if (!invoice.amount) {
+      const amountMatch2 = RegexPatterns.tryPatterns(allText, [
+        RegexPatterns.AMOUNT_FIELDS.amount,
+        RegexPatterns.AMOUNT_FIELDS.amountWithSpace,
+      ]);
+      if (amountMatch2) {
+        const parsed = this.parseAmountString(amountMatch2.match[0]);
+        if (parsed.amount) {
+          invoice.amount = parsed.amount;
+        }
+        if (parsed.taxAmount) {
+          invoice.taxAmount = parsed.taxAmount;
+        }
+      }
+    }
+
+    // 提取价税合计
+    const totalMatch = allText.match(
+      /价税合计\s*\(?大写\)?\s*([壹贰叁肆伍陆柒捌玖拾佰仟万亿圆整]+)\s*\(?小写\)?\s*¥?\s*(\d+\.\d+)\s*(?=[合计]|$)/
+    );
+    if (totalMatch) {
+      invoice.totalAmountString = totalMatch[1];
+      invoice.totalAmount = totalMatch[2];
+    }
+
+    // 如果还没找到价税合计，尝试更宽松的模式
+    if (!invoice.totalAmount) {
+      const totalMatch2 = fullText.match(/价税合计.*?¥?\s*(\d+\.\d+)/);
+      if (totalMatch2) {
+        invoice.totalAmount = totalMatch2[1];
+      }
+    }
+
+    // 如果还没找到大写金额，尝试提取
+    if (!invoice.totalAmountString) {
+      const chineseMatch = allText.match(/价税合计.*?([壹贰叁肆伍陆柒捌玖拾佰仟万亿圆整]+)/);
+      if (chineseMatch) {
+        invoice.totalAmountString = chineseMatch[1];
+      }
+    }
+
+    // 验证金额一致性
+    if (invoice.amount && invoice.taxAmount && invoice.totalAmount) {
+      const amount = parseFloat(invoice.amount);
+      const taxAmount = parseFloat(invoice.taxAmount);
+      const totalAmount = parseFloat(invoice.totalAmount);
+
+      if (!isNaN(amount) && !isNaN(taxAmount) && !isNaN(totalAmount)) {
+        const diff = Math.abs(totalAmount - (amount + taxAmount));
+        if (diff > 0.01) {
+          invoice.amountWarning = `金额不匹配: ${amount} + ${taxAmount} ≠ ${totalAmount}`;
+        }
+      }
+    }
+  }
+
+  static parseAmountString(amountStr) {
+    const parts = amountStr.match(/(\d+\.?\d*)/g) || [];
+    return {
+      amount: parts[0] || '',
+      taxAmount: parts[1] || '',
+    };
+  }
+
+
+  static extractDetails(invoice, fullText, allText, items) {
+    const details = [];
+    const lines = fullText.split('\n');
+
+    let taxRateLineIdx = -1;
+    let mergeLineIdx = -1;
+
+    for (let i = 0; i < lines.length; i++) {
+      const line = lines[i];
+      if (line.includes('税率') || line.includes('合计') || line.includes('销售额')) {
+        if (line.includes('税率')) {
+          taxRateLineIdx = i;
+        }
+        if (line.includes('合计') && i > taxRateLineIdx) {
+          mergeLineIdx = i;
+          break;
+        }
+      }
+    }
+
+    if (taxRateLineIdx > 0 && mergeLineIdx > taxRateLineIdx) {
+      for (let i = taxRateLineIdx + 1; i < mergeLineIdx; i++) {
+        const line = lines[i].trim();
+        if (!line || this.isLineIgnorable(line)) continue;
+
+        if (this.isDetailLine(line)) {
+          const detail = this.parseDetailLine(line);
+          if (detail && detail.amount) {
+            details.push(detail);
+          }
+        }
+      }
+    }
+
+    invoice.details = details;
+  }
+
+  static isLineIgnorable(line) {
+    return (
+      line.length < 2 ||
+      /^[-\s=*]*$/.test(line) ||
+      /^\s*$/.test(line) ||
+      line.startsWith('商品')
+    );
+  }
+
+  static isDetailLine(line) {
+    // 排除包含"合计"的行
+    if (line.includes('合计') || line.includes('合 计')) {
+      return false;
+    }
+    return /\d+%/.test(line) || /免税|不征税|出口零税/.test(line) || /\d+\.\d+/.test(line);
+  }
+
+  static parseDetailLine(line) {
+    const detail = new Detail();
+
+    line = line.replace(/　/g, ' ').replace(/\s+/g, ' ').trim();
+    const tokens = line.split(/\s+/);
+
+    if (tokens.length < 2) {
+      return null;
+    }
+
+    // 特殊处理：如果行以*开头，可能是项目名称包含*
+    if (tokens[0] === '*' && tokens.length >= 4) {
+      // 尝试重建项目名称
+      let nameParts = [];
+      let i = 0;
+
+      // 收集直到遇到数字的部分作为项目名称
+      while (i < tokens.length && !/^\d+(\.\d+)?$/.test(tokens[i])) {
+        nameParts.push(tokens[i]);
+        i++;
+      }
+
+      if (nameParts.length > 0) {
+        detail.name = nameParts.join('');
+      }
+
+      // 剩下的tokens从i开始是数字部分
+      const numberTokens = tokens.slice(i);
+
+      // 查找税率位置
+      const taxRateIdx = numberTokens.findIndex((t) => /\d+%|免税|不征税/.test(t));
+
+      // 金额应该是税率前面的一个数字
+      if (taxRateIdx > 0) {
+        detail.amount = numberTokens[taxRateIdx - 1];
+
+        // 税率
+        if (taxRateIdx >= 0) {
+          const taxRateStr = numberTokens[taxRateIdx];
+          detail.taxRate = RegexPatterns.extractTaxRate(taxRateStr);
+        }
+
+        // 税额是税率后面的数字
+        if (taxRateIdx + 1 < numberTokens.length) {
+          const nextToken = numberTokens[taxRateIdx + 1];
+          if (/^\d+(\.\d+)?$/.test(nextToken)) {
+            detail.taxAmount = nextToken;
+          }
+        }
+      }
+
+      return Object.keys(detail).length > 1 ? detail : null;
+    }
+
+    // 原来的逻辑（用于非*开头的行）
+    const amountIdx = tokens.findIndex((t) => /^\d+(\.\d+)?$/.test(t));
+    const taxRateIdx = tokens.findIndex((t) => /\d+%|免税|不征税/.test(t));
+
+    if (amountIdx >= 0) {
+      detail.amount = tokens[amountIdx];
+
+      let numCount = 0;
+      for (let i = 0; i < amountIdx; i++) {
+        if (/^-?\d+(\.\d+)?$/.test(tokens[i])) {
+          if (numCount === 0) {
+            detail.count = tokens[i];
+          } else if (numCount === 1) {
+            detail.price = tokens[i];
+          }
+          numCount++;
+        } else {
+          if (!detail.name) {
+            detail.name = tokens[i];
+          } else if (!detail.model) {
+            detail.model = tokens[i];
+          } else if (!detail.unit) {
+            detail.unit = tokens[i];
+          }
+        }
+      }
+
+      if (taxRateIdx >= 0) {
+        const taxRateStr = tokens[taxRateIdx];
+        detail.taxRate = RegexPatterns.extractTaxRate(taxRateStr);
+      }
+
+      if (taxRateIdx >= 0 && taxRateIdx + 1 < tokens.length) {
+        const nextToken = tokens[taxRateIdx + 1];
+        if (/^\d+(\.\d+)?$/.test(nextToken)) {
+          detail.taxAmount = nextToken;
+        }
+      }
+    }
+
+    return Object.keys(detail).length > 1 ? detail : null;
+  }
+}
+
+module.exports = PdfFullElectronicInvoiceService;

package/lib/PdfInvoiceExtractor.js

@@ -0,0 +1,124 @@
+const fs = require('fs');
+const pdfjs = require('pdfjs-dist/build/pdf');
+const { Invoice } = require('./Invoice');
+const PdfRegularInvoiceService = require('./PdfRegularInvoiceService');
+const PdfFullElectronicInvoiceService = require('./PdfFullElectronicInvoiceService');
+const PdfFinancialInvoiceService = require('./PdfFinancialInvoiceService');
+const StringUtils = require('./StringUtils');
+const ErrorHandler = require('./ErrorHandler');
+
+// 设置 worker - Node.js 环境
+const path = require('path');
+pdfjs.GlobalWorkerOptions.workerSrc = path.join(
+  path.dirname(require.resolve('pdfjs-dist/package.json')),
+  'build/pdf.worker.js'
+);
+
+/**
+ * PDF 发票提取器
+ */
+class PdfInvoiceExtractor {
+  static async extract(filePath) {
+    try {
+      const fileBuffer = fs.readFileSync(filePath);
+      // 转换 Buffer 为 Uint8Array
+      const pdfData = new Uint8Array(fileBuffer);
+      
+      // 配置 PDF.js 参数
+      const pdfjsOptions = {
+        data: pdfData,
+        cMapUrl: path.join(
+          path.dirname(require.resolve('pdfjs-dist/package.json')),
+          'cmaps/'
+        ),
+        cMapPacked: true,
+        standardFontDataUrl: path.join(
+          path.dirname(require.resolve('pdfjs-dist/package.json')),
+          'standard_fonts/'
+        ),
+      };
+      
+      const pdf = await pdfjs.getDocument(pdfjsOptions).promise;
+
+      // 只处理第一页
+      const page = await pdf.getPage(1);
+      const textContent = await page.getTextContent();
+      const pageWidth = Math.round(page.view[2]);
+
+      // 获取所有文本 - 按行分组重建文本结构
+      let fullText = '';
+      const items = [];
+
+      // 按Y坐标排序文本项（从上到下）
+      const sortedItems = textContent.items
+        .filter(item => item.str && item.str.trim())
+        .sort((a, b) => b.transform[5] - a.transform[5]); // Y坐标从大到小
+
+      // 按行分组（Y坐标相近的视为同一行）
+      const lines = [];
+      let currentLine = [];
+      let lastY = null;
+
+      for (const item of sortedItems) {
+        const y = Math.round(item.transform[5]);
+        if (lastY === null || Math.abs(y - lastY) > 5) { // 5像素容差
+          if (currentLine.length > 0) {
+            lines.push(currentLine);
+          }
+          currentLine = [item];
+          lastY = y;
+        } else {
+          currentLine.push(item);
+        }
+      }
+      if (currentLine.length > 0) {
+        lines.push(currentLine);
+      }
+
+      // 在每行内按X坐标排序并拼接
+      for (const line of lines) {
+        line.sort((a, b) => a.transform[4] - b.transform[4]); // X坐标从小到大
+
+        let lineText = '';
+        for (const item of line) {
+          if (lineText && !lineText.endsWith(' ') && !item.str.startsWith(' ')) {
+            lineText += ' ';
+          }
+          lineText += item.str;
+
+          items.push({
+            text: item.str,
+            x: item.transform[4],
+            y: item.transform[5],
+            width: item.width,
+            height: item.height,
+          });
+        }
+
+        if (fullText && !fullText.endsWith('\n')) {
+          fullText += '\n';
+        }
+        fullText += lineText;
+      }
+
+      // 规范化文本
+      let allText = StringUtils.normalize(fullText)
+        .replace(/（/g, '(')
+        .replace(/）/g, ')')
+        .replace(/￥/g, '¥');
+
+      // 判断发票类型
+      if (allText.includes('福建省社会团体会员费统一收据') || allText.includes('财政票据')) {
+        return PdfFinancialInvoiceService.extract(fullText, allText, pageWidth, items);
+      } else if (allText.includes('电子发票')) {
+        return PdfFullElectronicInvoiceService.extract(fullText, allText, pageWidth, items);
+      } else {
+        return PdfRegularInvoiceService.extract(fullText, allText, pageWidth, items);
+      }
+    } catch (error) {
+      return ErrorHandler.createErrorInvoice(error, 'pdf');
+    }
+  }
+}
+
+module.exports = PdfInvoiceExtractor;