einvoice/test/suite/einvoice_format-detection/test.fd-01.ubl-detection.ts

import { tap, expect } from '@git.zone/tstest/tapbundle';
import { EInvoice } from '../../../ts/index.js';
import { InvoiceFormat } from '../../../ts/interfaces/common.js';
import { FormatDetector } from '../../../ts/formats/utils/format.detector.js';
import { CorpusLoader, PerformanceTracker } from '../../helpers/test-utils.js';

/**
 * Test ID: FD-01
 * Test Description: UBL Format Detection
 * Priority: High
 * 
 * This test validates the accurate detection of UBL (Universal Business Language) format
 * from XML invoice files across different UBL versions and implementations.
 */

tap.test('FD-01: UBL Format Detection - Corpus files', async (t) => {
  // Load UBL test files from corpus
  const ublFiles = await CorpusLoader.loadCategory('UBL_XMLRECHNUNG');
  const peppolFiles = await CorpusLoader.loadCategory('PEPPOL');
  const en16931UblFiles = await CorpusLoader.loadCategory('EN16931_UBL_EXAMPLES');
  
  const allUblFiles = [...ublFiles, ...peppolFiles, ...en16931UblFiles];
  
  console.log(`Testing ${allUblFiles.length} UBL files for format detection`);
  
  let successCount = 0;
  let failureCount = 0;
  const detectionTimes: number[] = [];
  
  for (const file of allUblFiles) {
    try {
      const xmlBuffer = await CorpusLoader.loadFile(file.path);
      const xmlString = xmlBuffer.toString('utf-8');
      
      // Track performance
      const { result: detectedFormat, metric } = await PerformanceTracker.track(
        'format-detection',
        async () => FormatDetector.detectFormat(xmlString),
        { file: file.path, size: file.size }
      );
      
      detectionTimes.push(metric.duration);
      
      // UBL files can be detected as UBL or XRechnung (which is UBL-based)
      const validFormats = [InvoiceFormat.UBL, InvoiceFormat.XRECHNUNG];
      
      if (validFormats.includes(detectedFormat)) {
        successCount++;
        t.pass(`✓ ${path.basename(file.path)}: Correctly detected as ${detectedFormat}`);
      } else {
        failureCount++;
        t.fail(`✗ ${path.basename(file.path)}: Detected as ${detectedFormat}, expected UBL or XRechnung`);
      }
      
    } catch (error) {
      failureCount++;
      t.fail(`✗ ${path.basename(file.path)}: Detection failed - ${error.message}`);
    }
  }
  
  // Calculate statistics
  const avgTime = detectionTimes.length > 0 
    ? detectionTimes.reduce((a, b) => a + b, 0) / detectionTimes.length 
    : 0;
  
  console.log(`\nUBL Detection Summary:`);
  console.log(`- Files tested: ${allUblFiles.length}`);
  console.log(`- Successful detections: ${successCount} (${(successCount / allUblFiles.length * 100).toFixed(1)}%)`);
  console.log(`- Failed detections: ${failureCount}`);
  console.log(`- Average detection time: ${avgTime.toFixed(2)}ms`);
  
  // Performance assertion
  t.ok(avgTime < 10, 'Average detection time should be under 10ms');
  
  // Success rate assertion (allow some flexibility for edge cases)
  const successRate = successCount / allUblFiles.length;
  t.ok(successRate > 0.9, 'Success rate should be above 90%');
});

tap.test('FD-01: UBL Format Detection - Specific UBL elements', async (t) => {
  // Test specific UBL invoice
  const ublInvoice = `<?xml version="1.0" encoding="UTF-8"?>
<Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2"
         xmlns:cac="urn:oasis:names:specification:ubl:schema:xsd:CommonAggregateComponents-2"
         xmlns:cbc="urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2">
  <cbc:ID>INV-001</cbc:ID>
  <cbc:IssueDate>2024-01-01</cbc:IssueDate>
  <cac:AccountingSupplierParty>
    <cac:Party>
      <cac:PartyName>
        <cbc:Name>Test Supplier</cbc:Name>
      </cac:PartyName>
    </cac:Party>
  </cac:AccountingSupplierParty>
</Invoice>`;

  const format = FormatDetector.detectFormat(ublInvoice);
  t.equal(format, InvoiceFormat.UBL, 'Should detect standard UBL invoice');
  
  // Test UBL credit note
  const ublCreditNote = `<?xml version="1.0" encoding="UTF-8"?>
<CreditNote xmlns="urn:oasis:names:specification:ubl:schema:xsd:CreditNote-2"
            xmlns:cac="urn:oasis:names:specification:ubl:schema:xsd:CommonAggregateComponents-2"
            xmlns:cbc="urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2">
  <cbc:ID>CN-001</cbc:ID>
  <cbc:IssueDate>2024-01-01</cbc:IssueDate>
</CreditNote>`;

  const creditNoteFormat = FormatDetector.detectFormat(ublCreditNote);
  t.equal(creditNoteFormat, InvoiceFormat.UBL, 'Should detect UBL credit note');
});

tap.test('FD-01: UBL Format Detection - PEPPOL BIS', async (t) => {
  // Test PEPPOL BIS 3.0 (which is UBL-based)
  const peppolInvoice = `<?xml version="1.0" encoding="UTF-8"?>
<Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2"
         xmlns:cac="urn:oasis:names:specification:ubl:schema:xsd:CommonAggregateComponents-2"
         xmlns:cbc="urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2">
  <cbc:CustomizationID>urn:cen.eu:en16931:2017#compliant#urn:fdc:peppol.eu:2017:poacc:billing:3.0</cbc:CustomizationID>
  <cbc:ProfileID>urn:fdc:peppol.eu:2017:poacc:billing:01:1.0</cbc:ProfileID>
  <cbc:ID>Peppol-001</cbc:ID>
</Invoice>`;

  const format = FormatDetector.detectFormat(peppolInvoice);
  t.ok(
    [InvoiceFormat.UBL, InvoiceFormat.XRECHNUNG].includes(format),
    'Should detect PEPPOL BIS as UBL or specialized format'
  );
});

tap.test('FD-01: UBL Format Detection - Edge cases', async (t) => {
  // Test with minimal UBL
  const minimalUBL = '<Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2"></Invoice>';
  const minimalFormat = FormatDetector.detectFormat(minimalUBL);
  t.equal(minimalFormat, InvoiceFormat.UBL, 'Should detect minimal UBL invoice');
  
  // Test with different namespace prefix
  const differentPrefix = `<?xml version="1.0"?>
<ubl:Invoice xmlns:ubl="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2">
  <ubl:ID>123</ubl:ID>
</ubl:Invoice>`;
  
  const prefixFormat = FormatDetector.detectFormat(differentPrefix);
  t.equal(prefixFormat, InvoiceFormat.UBL, 'Should detect UBL with different namespace prefix');
  
  // Test without XML declaration
  const noDeclaration = `<Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2">
  <cbc:ID xmlns:cbc="urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2">456</cbc:ID>
</Invoice>`;
  
  const noDecFormat = FormatDetector.detectFormat(noDeclaration);
  t.equal(noDecFormat, InvoiceFormat.UBL, 'Should detect UBL without XML declaration');
});

tap.test('FD-01: UBL Format Detection - Performance benchmarks', async (t) => {
  // Test detection speed with various file sizes
  const testCases = [
    { name: 'Small UBL', size: 1000, content: generateUBLInvoice(5) },
    { name: 'Medium UBL', size: 10000, content: generateUBLInvoice(50) },
    { name: 'Large UBL', size: 100000, content: generateUBLInvoice(500) }
  ];
  
  for (const testCase of testCases) {
    const times: number[] = [];
    
    // Run multiple iterations for accuracy
    for (let i = 0; i < 100; i++) {
      const start = performance.now();
      FormatDetector.detectFormat(testCase.content);
      times.push(performance.now() - start);
    }
    
    const avgTime = times.reduce((a, b) => a + b, 0) / times.length;
    
    console.log(`${testCase.name} (${testCase.content.length} bytes): avg ${avgTime.toFixed(3)}ms`);
    t.ok(avgTime < 5, `${testCase.name} detection should be under 5ms`);
  }
});

// Helper function to generate UBL invoice with specified number of line items
function generateUBLInvoice(lineItems: number): string {
  let invoice = `<?xml version="1.0" encoding="UTF-8"?>
<Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2"
         xmlns:cac="urn:oasis:names:specification:ubl:schema:xsd:CommonAggregateComponents-2"
         xmlns:cbc="urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2">
  <cbc:ID>TEST-${Date.now()}</cbc:ID>
  <cbc:IssueDate>2024-01-01</cbc:IssueDate>`;
  
  for (let i = 1; i <= lineItems; i++) {
    invoice += `
  <cac:InvoiceLine>
    <cbc:ID>${i}</cbc:ID>
    <cbc:InvoicedQuantity unitCode="EA">${i}</cbc:InvoicedQuantity>
    <cbc:LineExtensionAmount currencyID="EUR">${i * 100}</cbc:LineExtensionAmount>
  </cac:InvoiceLine>`;
  }
  
  invoice += '\n</Invoice>';
  return invoice;
}

// Generate performance report at the end
tap.teardown(async () => {
  const stats = PerformanceTracker.getStats('format-detection');
  if (stats) {
    console.log('\nPerformance Summary:');
    console.log(`- Total detections: ${stats.count}`);
    console.log(`- Average time: ${stats.avg.toFixed(2)}ms`);
    console.log(`- Min/Max: ${stats.min.toFixed(2)}ms / ${stats.max.toFixed(2)}ms`);
    console.log(`- P95: ${stats.p95.toFixed(2)}ms`);
  }
});

// Import path for basename
import * as path from 'path';

tap.start();