Apple Tools MCP

/** * Edge case tests for message indexing * Tests NSAttributedString extraction, cloaked emails, and attachment handling */ import { describe, it, expect, beforeEach, vi } from 'vitest' import { createEmbeddingMock, MAC_ABSOLUTE_EPOCH } from '../helpers/indexing-mocks.js' describe('Message Edge Cases', () => { let mockEmbedder beforeEach(() => { vi.clearAllMocks() const mock = createEmbeddingMock() mockEmbedder = mock.mockEmbedder }) describe('NSAttributedString extraction', () => { it('should detect NSString marker in binary data', () => { // NSString marker followed by '+' byte (0x2B) const marker = Buffer.from('NSString') const plusByte = 0x2B const testData = Buffer.concat([ Buffer.from([0x00, 0x00, 0x00]), marker, Buffer.from([0x00, 0x00, plusByte]), Buffer.from([5]), // length = 5 Buffer.from('Hello') ]) const markerIndex = testData.indexOf(marker) expect(markerIndex).toBeGreaterThan(-1) // Find '+' byte after marker const plusIndex = testData.indexOf(plusByte, markerIndex + 8) expect(plusIndex).toBeGreaterThan(markerIndex) }) it('should extract text after length byte', () => { const length = 11 const text = 'Hello World' // Simulated binary structure const buffer = Buffer.concat([ Buffer.from('NSString'), Buffer.from([0x00, 0x00, 0x2B]), // '+' marker Buffer.from([length]), Buffer.from(text) ]) // Find + byte const plusIndex = buffer.indexOf(0x2B) const extractedLength = buffer[plusIndex + 1] const extractedText = buffer.slice(plusIndex + 2, plusIndex + 2 + extractedLength).toString('utf-8') expect(extractedLength).toBe(length) expect(extractedText).toBe(text) }) it('should handle multi-byte length encoding', () => { // For lengths > 127, the encoding may use multiple bytes // This tests the concept of length > 127 const shortLength = 50 const longLength = 200 expect(shortLength).toBeLessThanOrEqual(127) expect(longLength).toBeGreaterThan(127) // In real implementation, would check high bit const isMultiByte = (len) => len > 127 expect(isMultiByte(shortLength)).toBe(false) expect(isMultiByte(longLength)).toBe(true) }) it('should validate printable character ratio', () => { const validatePrintable = (text) => { const printable = text.split('').filter(c => { const code = c.charCodeAt(0) return code >= 32 && code <= 126 || code >= 128 // ASCII printable or extended }).length return (printable / text.length) > 0.8 } expect(validatePrintable('Hello World')).toBe(true) expect(validatePrintable('Normal text message')).toBe(true) // Binary garbage would fail const binaryGarbage = String.fromCharCode(0, 1, 2, 3, 4, 5) expect(validatePrintable(binaryGarbage)).toBe(false) }) it('should enforce 10000 char limit', () => { const longText = 'x'.repeat(15000) const maxLength = 10000 const truncated = longText.substring(0, maxLength) expect(truncated.length).toBe(maxLength) }) }) describe('NULL text field fallback', () => { it('should use attributedBody when text is NULL', () => { const message = { text: null, attributedBody: Buffer.from('Fallback text') } const getText = (msg) => { if (msg.text) return msg.text if (msg.attributedBody) return 'extracted from attributedBody' return '' } expect(getText(message)).toBe('extracted from attributedBody') }) it('should prefer text field when present', () => { const message = { text: 'Primary text', attributedBody: Buffer.from('Fallback text') } const text = message.text || 'extracted from attributedBody' expect(text).toBe('Primary text') }) it('should handle both NULL', () => { const message = { text: null, attributedBody: null } const text = message.text || (message.attributedBody ? 'extracted' : '') expect(text).toBe('') }) }) describe('cloaked email handling', () => { it('should extract name before "via Cloaked"', () => { const sender = 'John Doe via Cloaked' const extractName = (s) => { const match = s.match(/^(.+?)\s+via\s+Cloaked$/i) return match ? match[1] : s } expect(extractName(sender)).toBe('John Doe') }) it('should handle non-cloaked senders', () => { const sender = '+15551234567' const extractName = (s) => { const match = s.match(/^(.+?)\s+via\s+Cloaked$/i) return match ? match[1] : s } expect(extractName(sender)).toBe('+15551234567') }) it('should be case-insensitive for "via Cloaked"', () => { const variants = [ 'Name via Cloaked', 'Name via cloaked', 'Name VIA CLOAKED', 'Name Via cloaked' ] const extractName = (s) => { const match = s.match(/^(.+?)\s+via\s+Cloaked$/i) return match ? match[1] : s } for (const v of variants) { expect(extractName(v)).toBe('Name') } }) }) describe('chatId and chatIdentifier extraction', () => { it('should extract chatId from message', () => { const message = { ROWID: 12345, chat_id: 67 } expect(message.chat_id).toBe(67) }) it('should extract chatIdentifier', () => { const chat = { chat_identifier: '+15551234567', display_name: 'John Doe' } expect(chat.chat_identifier).toBe('+15551234567') }) it('should handle group chat identifiers', () => { const groupChat = { chat_identifier: 'chat123456789', display_name: 'Family Group', is_group: 1 } expect(groupChat.chat_identifier).toMatch(/^chat/) }) it('should handle email-based chat identifiers', () => { const emailChat = { chat_identifier: 'john@example.com', display_name: '' } expect(emailChat.chat_identifier).toContain('@') }) }) describe('hasAttachment via join table', () => { it('should detect attachment count > 0', () => { const message = { ROWID: 123, attachment_count: 2 } const hasAttachment = (message.attachment_count || 0) > 0 expect(hasAttachment).toBe(true) }) it('should handle zero attachments', () => { const message = { ROWID: 123, attachment_count: 0 } const hasAttachment = (message.attachment_count || 0) > 0 expect(hasAttachment).toBe(false) }) it('should handle NULL attachment count', () => { const message = { ROWID: 123, attachment_count: null } const hasAttachment = (message.attachment_count || 0) > 0 expect(hasAttachment).toBe(false) }) it('should simulate JOIN with message_attachment_join', () => { // Simulated SQL result const messageWithJoin = { ROWID: 123, text: 'Check this out', // From LEFT JOIN message_attachment_join attachment_id: 456 // Non-null means has attachment } const hasAttachment = messageWithJoin.attachment_id !== null expect(hasAttachment).toBe(true) }) }) describe('group chat detection edge cases', () => { it('should detect by participant count alone', () => { const message = { participantCount: 5, chatName: null } const isGroupChat = (message.participantCount || 0) > 2 || Boolean(message.chatName) expect(isGroupChat).toBe(true) }) it('should detect by chat name alone', () => { const message = { participantCount: 2, // Only 2 people chatName: 'Project Team' } const isGroupChat = (message.participantCount || 0) > 2 || Boolean(message.chatName) expect(isGroupChat).toBe(true) }) it('should handle 1-on-1 correctly', () => { const message = { participantCount: 2, chatName: '' } const isGroupChat = (message.participantCount || 0) > 2 || Boolean(message.chatName && message.chatName.length > 0) expect(isGroupChat).toBe(false) }) it('should handle self-messages', () => { const message = { participantCount: 1, // Just self chatName: '' } const isGroupChat = (message.participantCount || 0) > 2 || Boolean(message.chatName) expect(isGroupChat).toBe(false) }) }) describe('Mac Absolute Time edge cases', () => { it('should convert epoch correctly', () => { // Mac epoch: Jan 1, 2001 00:00:00 UTC const macTime = 0 const unixTime = macTime + MAC_ABSOLUTE_EPOCH expect(unixTime).toBe(978307200) // Unix timestamp for 2001-01-01 }) it('should handle recent timestamps', () => { // Example: Jan 1, 2024 00:00:00 UTC // Unix: 1704067200 // Mac: 1704067200 - 978307200 = 725760000 const macTime = 725760000 const unixTime = macTime + MAC_ABSOLUTE_EPOCH expect(unixTime).toBe(1704067200) }) it('should handle nanoseconds in macOS Sonoma+', () => { // Newer macOS uses nanoseconds in date field const macTimeNanos = 725760000000000000 // Nanoseconds // Detect if nanoseconds (very large number) const isNanos = macTimeNanos > 1e15 expect(isNanos).toBe(true) // Convert to seconds const macTimeSeconds = macTimeNanos / 1e9 expect(macTimeSeconds).toBeCloseTo(725760000, 0) }) it('should handle negative values (pre-2001)', () => { // Year 2000: negative Mac time const macTime = -31536000 // -1 year const unixTime = macTime + MAC_ABSOLUTE_EPOCH // Should be year 2000 const date = new Date(unixTime * 1000) expect(date.getFullYear()).toBe(2000) }) }) describe('message text truncation', () => { it('should truncate to 500 chars for search', () => { const longMessage = 'x'.repeat(1000) const truncated = longMessage.substring(0, 500) expect(truncated.length).toBe(500) }) it('should preserve short messages', () => { const shortMessage = 'Hi there!' const maxLength = 500 const result = shortMessage.length > maxLength ? shortMessage.substring(0, maxLength) : shortMessage expect(result).toBe(shortMessage) }) }) describe('sender normalization', () => { it('should use "Me" for is_from_me=1', () => { const message = { is_from_me: 1, handle_id: '+15551234567' } const sender = message.is_from_me === 1 ? 'Me' : message.handle_id expect(sender).toBe('Me') }) it('should use handle_id for received', () => { const message = { is_from_me: 0, handle_id: '+15551234567' } const sender = message.is_from_me === 1 ? 'Me' : message.handle_id expect(sender).toBe('+15551234567') }) it('should handle NULL handle_id', () => { const message = { is_from_me: 0, handle_id: null } const sender = message.is_from_me === 1 ? 'Me' : (message.handle_id || 'Unknown') expect(sender).toBe('Unknown') }) }) })