einvoice/test/suite/einvoice_performance/test.perf-10.cache-efficiency.ts

/**
 * @file test.perf-10.cache-efficiency.ts
 * @description Performance tests for cache efficiency and optimization
 */

import { tap } from '@git.zone/tstest/tapbundle';
import * as plugins from '../../plugins.js';
import { EInvoice } from '../../../ts/index.js';
import { CorpusLoader } from '../../suite/corpus.loader.js';
import { PerformanceTracker } from '../../suite/performance.tracker.js';
import { FormatDetector } from '../../../ts/formats/utils/format.detector.js';

const performanceTracker = new PerformanceTracker('PERF-10: Cache Efficiency');

tap.test('PERF-10: Cache Efficiency - should demonstrate effective caching strategies', async (t) => {
  // Test 1: Format detection cache
  const formatDetectionCache = await performanceTracker.measureAsync(
    'format-detection-cache',
    async () => {
      const results = {
        withoutCache: {
          iterations: 0,
          totalTime: 0,
          avgTime: 0
        },
        withCache: {
          iterations: 0,
          totalTime: 0,
          avgTime: 0,
          cacheHits: 0,
          cacheMisses: 0
        },
        improvement: null
      };
      
      // Test data
      const testDocuments = [
        {
          id: 'ubl-1',
          content: '<?xml version="1.0"?><Invoice xmlns="urn:oasis:names:specification:ubl:schema:xsd:Invoice-2"><ID>UBL-001</ID></Invoice>'
        },
        {
          id: 'cii-1',
          content: '<?xml version="1.0"?><rsm:CrossIndustryInvoice xmlns:rsm="urn:un:unece:uncefact:data:standard:CrossIndustryInvoice:100"><ID>CII-001</ID></rsm:CrossIndustryInvoice>'
        },
        {
          id: 'unknown-1',
          content: '<?xml version="1.0"?><UnknownFormat><ID>UNKNOWN-001</ID></UnknownFormat>'
        }
      ];
      
      // Test without cache (baseline)
      const iterations = 100;
      const startWithoutCache = Date.now();
      
      for (let i = 0; i < iterations; i++) {
        for (const doc of testDocuments) {
          FormatDetector.detectFormat(doc.content);
          results.withoutCache.iterations++;
        }
      }
      
      results.withoutCache.totalTime = Date.now() - startWithoutCache;
      results.withoutCache.avgTime = results.withoutCache.totalTime / results.withoutCache.iterations;
      
      // Implement simple cache
      const formatCache = new Map<string, { format: string; timestamp: number }>();
      const cacheMaxAge = 60000; // 1 minute
      
      const detectFormatWithCache = async (content: string) => {
        // Create cache key from content hash
        const hash = Buffer.from(content).toString('base64').slice(0, 20);
        
        // Check cache
        const cached = formatCache.get(hash);
        if (cached && Date.now() - cached.timestamp < cacheMaxAge) {
          results.withCache.cacheHits++;
          return cached.format;
        }
        
        // Cache miss
        results.withCache.cacheMisses++;
        const format = FormatDetector.detectFormat(content);
        
        // Store in cache
        formatCache.set(hash, { format: format || 'unknown', timestamp: Date.now() });
        
        return format;
      };
      
      // Test with cache
      const startWithCache = Date.now();
      
      for (let i = 0; i < iterations; i++) {
        for (const doc of testDocuments) {
          await detectFormatWithCache(doc.content);
          results.withCache.iterations++;
        }
      }
      
      results.withCache.totalTime = Date.now() - startWithCache;
      results.withCache.avgTime = results.withCache.totalTime / results.withCache.iterations;
      
      // Calculate improvement
      results.improvement = {
        speedup: (results.withoutCache.avgTime / results.withCache.avgTime).toFixed(2),
        timeReduction: ((results.withoutCache.totalTime - results.withCache.totalTime) / results.withoutCache.totalTime * 100).toFixed(2),
        hitRate: ((results.withCache.cacheHits / results.withCache.iterations) * 100).toFixed(2),
        efficiency: results.withCache.cacheHits > 0 ? 
          ((results.withCache.cacheHits / (results.withCache.cacheHits + results.withCache.cacheMisses)) * 100).toFixed(2) : '0'
      };
      
      return results;
    }
  );
  
  // Test 2: Validation cache
  const validationCache = await performanceTracker.measureAsync(
    'validation-cache',
    async () => {
      const results = {
        cacheStrategies: [],
        optimalStrategy: null
      };
      
      // Test invoice
      const testInvoice = {
        format: 'ubl' as const,
        data: {
          documentType: 'INVOICE',
          invoiceNumber: 'CACHE-VAL-001',
          issueDate: '2024-03-05',
          seller: { name: 'Cache Test Seller', address: 'Address', country: 'US', taxId: 'US123' },
          buyer: { name: 'Cache Test Buyer', address: 'Address', country: 'US', taxId: 'US456' },
          items: Array.from({ length: 20 }, (_, i) => ({
            description: `Item ${i + 1}`,
            quantity: 1,
            unitPrice: 100,
            vatRate: 10,
            lineTotal: 100
          })),
          totals: { netAmount: 2000, vatAmount: 200, grossAmount: 2200 }
        }
      };
      
      // Cache strategies to test
      const strategies = [
        {
          name: 'No cache',
          cacheSize: 0,
          ttl: 0
        },
        {
          name: 'Small cache',
          cacheSize: 10,
          ttl: 30000
        },
        {
          name: 'Medium cache',
          cacheSize: 100,
          ttl: 60000
        },
        {
          name: 'Large cache',
          cacheSize: 1000,
          ttl: 300000
        },
        {
          name: 'LRU cache',
          cacheSize: 50,
          ttl: 120000,
          lru: true
        }
      ];
      
      for (const strategy of strategies) {
        const cache = new Map<string, { result: any; timestamp: number; accessCount: number }>();
        let cacheHits = 0;
        let cacheMisses = 0;
        
        const validateWithCache = async (invoice: any) => {
          const key = JSON.stringify(invoice).slice(0, 50); // Simple key generation
          
          // Check cache
          const cached = cache.get(key);
          if (cached && Date.now() - cached.timestamp < strategy.ttl) {
            cacheHits++;
            cached.accessCount++;
            return cached.result;
          }
          
          // Cache miss
          cacheMisses++;
          // Mock validation result for performance testing
          const result = { valid: true, errors: [] };
          
          // Cache management
          if (strategy.cacheSize > 0) {
            if (cache.size >= strategy.cacheSize) {
              if (strategy.lru) {
                // Remove least recently used
                let lruKey = '';
                let minAccess = Infinity;
                for (const [k, v] of cache.entries()) {
                  if (v.accessCount < minAccess) {
                    minAccess = v.accessCount;
                    lruKey = k;
                  }
                }
                cache.delete(lruKey);
              } else {
                // Remove oldest
                const oldestKey = cache.keys().next().value;
                cache.delete(oldestKey);
              }
            }
            
            cache.set(key, { result, timestamp: Date.now(), accessCount: 1 });
          }
          
          return result;
        };
        
        // Test with mixed workload
        const workload = [];
        
        // Repeated validations of same invoice
        for (let i = 0; i < 50; i++) {
          workload.push(testInvoice);
        }
        
        // Variations of the invoice
        for (let i = 0; i < 30; i++) {
          const variation = JSON.parse(JSON.stringify(testInvoice));
          variation.data.invoiceNumber = `CACHE-VAL-${i + 2}`;
          workload.push(variation);
        }
        
        // Repeat some variations
        for (let i = 0; i < 20; i++) {
          const variation = JSON.parse(JSON.stringify(testInvoice));
          variation.data.invoiceNumber = `CACHE-VAL-${(i % 10) + 2}`;
          workload.push(variation);
        }
        
        // Process workload
        const startTime = Date.now();
        
        for (const invoice of workload) {
          await validateWithCache(invoice);
        }
        
        const totalTime = Date.now() - startTime;
        
        results.cacheStrategies.push({
          name: strategy.name,
          cacheSize: strategy.cacheSize,
          ttl: strategy.ttl,
          lru: strategy.lru || false,
          totalRequests: workload.length,
          cacheHits,
          cacheMisses,
          hitRate: ((cacheHits / workload.length) * 100).toFixed(2),
          totalTime,
          avgTime: (totalTime / workload.length).toFixed(2),
          finalCacheSize: cache.size,
          memoryUsage: (cache.size * 1024).toFixed(0) // Rough estimate in bytes
        });
      }
      
      // Find optimal strategy
      const validStrategies = results.cacheStrategies.filter(s => s.cacheSize > 0);
      if (validStrategies.length > 0) {
        results.optimalStrategy = validStrategies.reduce((best, current) => {
          const bestScore = parseFloat(best.hitRate) / (parseFloat(best.avgTime) + 1);
          const currentScore = parseFloat(current.hitRate) / (parseFloat(current.avgTime) + 1);
          return currentScore > bestScore ? current : best;
        });
      }
      
      return results;
    }
  );
  
  // Test 3: Schema cache efficiency
  const schemaCache = await performanceTracker.measureAsync(
    'schema-cache-efficiency',
    async () => {
      const results = {
        schemaCaching: {
          enabled: false,
          tests: []
        },
        improvement: null
      };
      
      // Simulate schema validation with and without caching
      const schemas = {
        ubl: { size: 1024 * 50, parseTime: 50 }, // 50KB, 50ms parse time
        cii: { size: 1024 * 60, parseTime: 60 }, // 60KB, 60ms parse time
        zugferd: { size: 1024 * 80, parseTime: 80 }, // 80KB, 80ms parse time
        xrechnung: { size: 1024 * 70, parseTime: 70 } // 70KB, 70ms parse time
      };
      
      const schemaCache = new Map<string, { schema: any; loadTime: number }>();
      
      const loadSchemaWithoutCache = async (format: string) => {
        const schema = schemas[format];
        if (schema) {
          await new Promise(resolve => setTimeout(resolve, schema.parseTime));
          return { format, size: schema.size };
        }
        throw new Error(`Unknown schema format: ${format}`);
      };
      
      const loadSchemaWithCache = async (format: string) => {
        const cached = schemaCache.get(format);
        if (cached) {
          results.schemaCaching.enabled = true;
          return cached.schema;
        }
        
        const schema = await loadSchemaWithoutCache(format);
        schemaCache.set(format, { schema, loadTime: Date.now() });
        return schema;
      };
      
      // Test workload
      const workload = [];
      const formats = Object.keys(schemas);
      
      // Initial load of each schema
      for (const format of formats) {
        workload.push(format);
      }
      
      // Repeated use of schemas
      for (let i = 0; i < 100; i++) {
        workload.push(formats[i % formats.length]);
      }
      
      // Test without cache
      const startWithoutCache = Date.now();
      for (const format of workload) {
        await loadSchemaWithoutCache(format);
      }
      const timeWithoutCache = Date.now() - startWithoutCache;
      
      // Test with cache
      const startWithCache = Date.now();
      for (const format of workload) {
        await loadSchemaWithCache(format);
      }
      const timeWithCache = Date.now() - startWithCache;
      
      // Calculate memory usage
      let totalCachedSize = 0;
      for (const format of schemaCache.keys()) {
        totalCachedSize += schemas[format].size;
      }
      
      results.improvement = {
        timeWithoutCache,
        timeWithCache,
        speedup: (timeWithoutCache / timeWithCache).toFixed(2),
        timeReduction: ((timeWithoutCache - timeWithCache) / timeWithoutCache * 100).toFixed(2),
        memoryCost: (totalCachedSize / 1024).toFixed(2), // KB
        schemasLoaded: workload.length,
        uniqueSchemas: schemaCache.size
      };
      
      return results;
    }
  );
  
  // Test 4: Corpus cache analysis
  const corpusCacheAnalysis = await performanceTracker.measureAsync(
    'corpus-cache-analysis',
    async () => {
      const files = await CorpusLoader.loadPattern('**/*.xml');
      const results = {
        cacheableOperations: {
          formatDetection: { count: 0, duplicates: 0 },
          parsing: { count: 0, duplicates: 0 },
          validation: { count: 0, duplicates: 0 }
        },
        potentialSavings: null
      };
      
      // Track unique content hashes
      const contentHashes = new Map<string, number>();
      const formatResults = new Map<string, string>();
      
      // Sample corpus files
      const sampleFiles = files.slice(0, 100);
      
      for (const file of sampleFiles) {
        try {
          const content = await plugins.fs.readFile(file.path, 'utf-8');
          const hash = Buffer.from(content).toString('base64').slice(0, 32);
          
          // Track content duplicates
          const count = contentHashes.get(hash) || 0;
          contentHashes.set(hash, count + 1);
          
          if (count > 0) {
            results.cacheableOperations.formatDetection.duplicates++;
            results.cacheableOperations.parsing.duplicates++;
            results.cacheableOperations.validation.duplicates++;
          }
          
          // Perform operations
          const format = FormatDetector.detectFormat(content);
          results.cacheableOperations.formatDetection.count++;
          
          if (format && format !== 'unknown') {
            formatResults.set(hash, format);
            
            const invoice = await EInvoice.fromXml(content);
            results.cacheableOperations.parsing.count++;
            
            await invoice.validate();
            results.cacheableOperations.validation.count++;
          }
          
        } catch (error) {
          // Skip failed files
        }
      }
      
      // Calculate potential savings
      const avgFormatDetectionTime = 5; // ms
      const avgParsingTime = 20; // ms
      const avgValidationTime = 50; // ms
      
      results.potentialSavings = {
        formatDetection: {
          duplicateRatio: (results.cacheableOperations.formatDetection.duplicates / 
                          results.cacheableOperations.formatDetection.count * 100).toFixed(2),
          timeSavings: results.cacheableOperations.formatDetection.duplicates * avgFormatDetectionTime
        },
        parsing: {
          duplicateRatio: (results.cacheableOperations.parsing.duplicates / 
                          results.cacheableOperations.parsing.count * 100).toFixed(2),
          timeSavings: results.cacheableOperations.parsing.duplicates * avgParsingTime
        },
        validation: {
          duplicateRatio: (results.cacheableOperations.validation.duplicates / 
                          results.cacheableOperations.validation.count * 100).toFixed(2),
          timeSavings: results.cacheableOperations.validation.duplicates * avgValidationTime
        },
        totalTimeSavings: results.cacheableOperations.formatDetection.duplicates * avgFormatDetectionTime +
                         results.cacheableOperations.parsing.duplicates * avgParsingTime +
                         results.cacheableOperations.validation.duplicates * avgValidationTime,
        memoryCost: contentHashes.size * 100 // Rough estimate: 100 bytes per cached item
      };
      
      return results;
    }
  );
  
  // Test 5: Cache invalidation strategies
  const cacheInvalidation = await performanceTracker.measureAsync(
    'cache-invalidation-strategies',
    async () => {
      const results = {
        strategies: [],
        bestStrategy: null
      };
      
      // Test different invalidation strategies
      const strategies = [
        {
          name: 'TTL only',
          ttl: 60000,
          maxSize: Infinity,
          policy: 'ttl'
        },
        {
          name: 'Size limited',
          ttl: Infinity,
          maxSize: 50,
          policy: 'fifo'
        },
        {
          name: 'LRU with TTL',
          ttl: 120000,
          maxSize: 100,
          policy: 'lru'
        },
        {
          name: 'Adaptive',
          ttl: 60000,
          maxSize: 100,
          policy: 'adaptive'
        }
      ];
      
      for (const strategy of strategies) {
        const cache = new Map<string, { 
          data: any; 
          timestamp: number; 
          accessCount: number;
          lastAccess: number;
          size: number;
        }>();
        
        let hits = 0;
        let misses = 0;
        let evictions = 0;
        
        const cacheGet = (key: string) => {
          const entry = cache.get(key);
          
          if (!entry) {
            misses++;
            return null;
          }
          
          // Check TTL
          if (strategy.ttl !== Infinity && Date.now() - entry.timestamp > strategy.ttl) {
            cache.delete(key);
            evictions++;
            misses++;
            return null;
          }
          
          // Update access info
          entry.accessCount++;
          entry.lastAccess = Date.now();
          hits++;
          
          return entry.data;
        };
        
        const cacheSet = (key: string, data: any, size: number = 1) => {
          // Check size limit
          if (cache.size >= strategy.maxSize) {
            let keyToEvict = '';
            
            switch (strategy.policy) {
              case 'fifo':
                keyToEvict = cache.keys().next().value;
                break;
                
              case 'lru':
                let oldestAccess = Infinity;
                for (const [k, v] of cache.entries()) {
                  if (v.lastAccess < oldestAccess) {
                    oldestAccess = v.lastAccess;
                    keyToEvict = k;
                  }
                }
                break;
                
              case 'adaptive':
                // Evict based on access frequency and age
                let lowestScore = Infinity;
                for (const [k, v] of cache.entries()) {
                  const age = Date.now() - v.timestamp;
                  const score = v.accessCount / (age / 1000);
                  if (score < lowestScore) {
                    lowestScore = score;
                    keyToEvict = k;
                  }
                }
                break;
            }
            
            if (keyToEvict) {
              cache.delete(keyToEvict);
              evictions++;
            }
          }
          
          cache.set(key, {
            data,
            timestamp: Date.now(),
            accessCount: 0,
            lastAccess: Date.now(),
            size
          });
        };
        
        // Simulate workload with temporal locality
        const workloadSize = 500;
        const uniqueItems = 200;
        const workload = [];
        
        // Generate workload with patterns
        for (let i = 0; i < workloadSize; i++) {
          if (i < 100) {
            // Initial unique accesses
            workload.push(`item-${i % uniqueItems}`);
          } else if (i < 300) {
            // Repeated access to popular items
            workload.push(`item-${Math.floor(Math.random() * 20)}`);
          } else {
            // Mixed access pattern
            if (Math.random() < 0.3) {
              // Access recent item
              workload.push(`item-${Math.floor(Math.random() * 50)}`);
            } else {
              // Access any item
              workload.push(`item-${Math.floor(Math.random() * uniqueItems)}`);
            }
          }
        }
        
        // Process workload
        const startTime = Date.now();
        
        for (const key of workload) {
          const cached = cacheGet(key);
          if (!cached) {
            // Simulate data generation
            const data = { key, value: Math.random() };
            cacheSet(key, data);
          }
        }
        
        const totalTime = Date.now() - startTime;
        
        results.strategies.push({
          name: strategy.name,
          policy: strategy.policy,
          ttl: strategy.ttl,
          maxSize: strategy.maxSize,
          hits,
          misses,
          hitRate: ((hits / (hits + misses)) * 100).toFixed(2),
          evictions,
          evictionRate: ((evictions / workloadSize) * 100).toFixed(2),
          finalCacheSize: cache.size,
          totalTime,
          avgAccessTime: (totalTime / workloadSize).toFixed(2)
        });
      }
      
      // Find best strategy
      results.bestStrategy = results.strategies.reduce((best, current) => {
        const bestScore = parseFloat(best.hitRate) - parseFloat(best.evictionRate);
        const currentScore = parseFloat(current.hitRate) - parseFloat(current.evictionRate);
        return currentScore > bestScore ? current : best;
      });
      
      return results;
    }
  );

  // Summary
  console.log('\n=== PERF-10: Cache Efficiency Test Summary ===');
  
  console.log('\nFormat Detection Cache:');
  console.log(`  Without cache: ${formatDetectionCache.withoutCache.totalTime}ms for ${formatDetectionCache.withoutCache.iterations} ops`);
  console.log(`  With cache: ${formatDetectionCache.withCache.totalTime}ms for ${formatDetectionCache.withCache.iterations} ops`);
  console.log(`  Cache hits: ${formatDetectionCache.withCache.cacheHits}, misses: ${formatDetectionCache.withCache.cacheMisses}`);
  console.log(`  Speedup: ${formatDetectionCache.improvement.speedup}x`);
  console.log(`  Hit rate: ${formatDetectionCache.improvement.hitRate}%`);
  console.log(`  Time reduction: ${formatDetectionCache.improvement.timeReduction}%`);
  
  console.log('\nValidation Cache Strategies:');
  console.log('  Strategy     | Size | TTL    | Requests | Hits | Hit Rate | Avg Time | Memory');
  console.log('  -------------|------|--------|----------|------|----------|----------|--------');
  validationCache.cacheStrategies.forEach((strategy: any) => {
    console.log(`  ${strategy.name.padEnd(12)} | ${String(strategy.cacheSize).padEnd(4)} | ${String(strategy.ttl).padEnd(6)} | ${String(strategy.totalRequests).padEnd(8)} | ${String(strategy.cacheHits).padEnd(4)} | ${strategy.hitRate.padEnd(8)}% | ${strategy.avgTime.padEnd(8)}ms | ${strategy.memoryUsage}B`);
  });
  if (validationCache.optimalStrategy) {
    console.log(`  Optimal strategy: ${validationCache.optimalStrategy.name}`);
  }
  
  console.log('\nSchema Cache Efficiency:');
  console.log(`  Without cache: ${schemaCache.improvement.timeWithoutCache}ms`);
  console.log(`  With cache: ${schemaCache.improvement.timeWithCache}ms`);
  console.log(`  Speedup: ${schemaCache.improvement.speedup}x`);
  console.log(`  Time reduction: ${schemaCache.improvement.timeReduction}%`);
  console.log(`  Memory cost: ${schemaCache.improvement.memoryCost}KB`);
  console.log(`  Schemas loaded: ${schemaCache.improvement.schemasLoaded}, unique: ${schemaCache.improvement.uniqueSchemas}`);
  
  console.log('\nCorpus Cache Analysis:');
  console.log('  Operation        | Count | Duplicates | Ratio  | Time Savings');
  console.log('  -----------------|-------|------------|--------|-------------');
  ['formatDetection', 'parsing', 'validation'].forEach(op => {
    const stats = corpusCacheAnalysis.cacheableOperations[op];
    const savings = corpusCacheAnalysis.potentialSavings[op];
    console.log(`  ${op.padEnd(16)} | ${String(stats.count).padEnd(5)} | ${String(stats.duplicates).padEnd(10)} | ${savings.duplicateRatio.padEnd(6)}% | ${savings.timeSavings}ms`);
  });
  console.log(`  Total potential time savings: ${corpusCacheAnalysis.potentialSavings.totalTimeSavings}ms`);
  console.log(`  Estimated memory cost: ${(corpusCacheAnalysis.potentialSavings.memoryCost / 1024).toFixed(2)}KB`);
  
  console.log('\nCache Invalidation Strategies:');
  console.log('  Strategy      | Policy   | Hits | Hit Rate | Evictions | Final Size');
  console.log('  --------------|----------|------|----------|-----------|------------');
  cacheInvalidation.strategies.forEach((strategy: any) => {
    console.log(`  ${strategy.name.padEnd(13)} | ${strategy.policy.padEnd(8)} | ${String(strategy.hits).padEnd(4)} | ${strategy.hitRate.padEnd(8)}% | ${String(strategy.evictions).padEnd(9)} | ${strategy.finalCacheSize}`);
  });
  if (cacheInvalidation.bestStrategy) {
    console.log(`  Best strategy: ${cacheInvalidation.bestStrategy.name} (${cacheInvalidation.bestStrategy.hitRate}% hit rate)`);
  }
  
  // Performance targets check
  console.log('\n=== Performance Targets Check ===');
  const cacheSpeedup = parseFloat(formatDetectionCache.improvement.speedup);
  const targetSpeedup = 2; // Target: >2x speedup with caching
  
  console.log(`Cache speedup: ${cacheSpeedup}x ${cacheSpeedup > targetSpeedup ? '✅' : '⚠️'} (target: >${targetSpeedup}x)`);
  
  // Overall performance summary
  console.log('\n=== Overall Performance Summary ===');
  console.log(performanceTracker.getSummary());
});

tap.start();