feat(detection): add centralized protocol detection module

- Created ts/detection module for unified protocol detection
- Implemented TLS and HTTP detectors with fragmentation support
- Moved TLS detection logic from existing code to centralized module
- Updated RouteConnectionHandler to use ProtocolDetector for both TLS and HTTP
- Refactored ACME HTTP parsing to use detection module
- Added comprehensive tests for detection functionality
- Eliminated duplicate protocol detection code across codebase

This centralizes all non-destructive protocol detection into a single module,
improving code organization and reducing duplication between ACME and routing.

ts/detection/utils/buffer-utils.ts

@@ -0,0 +1,174 @@
+/**
+ * Buffer manipulation utilities for protocol detection
+ */
+
+/**
+ * BufferAccumulator class for handling fragmented data
+ */
+export class BufferAccumulator {
+  private chunks: Buffer[] = [];
+  private totalLength = 0;
+  
+  /**
+   * Append data to the accumulator
+   */
+  append(data: Buffer): void {
+    this.chunks.push(data);
+    this.totalLength += data.length;
+  }
+  
+  /**
+   * Get the accumulated buffer
+   */
+  getBuffer(): Buffer {
+    if (this.chunks.length === 0) {
+      return Buffer.alloc(0);
+    }
+    if (this.chunks.length === 1) {
+      return this.chunks[0];
+    }
+    return Buffer.concat(this.chunks, this.totalLength);
+  }
+  
+  /**
+   * Get current buffer length
+   */
+  length(): number {
+    return this.totalLength;
+  }
+  
+  /**
+   * Clear all accumulated data
+   */
+  clear(): void {
+    this.chunks = [];
+    this.totalLength = 0;
+  }
+  
+  /**
+   * Check if accumulator has minimum bytes
+   */
+  hasMinimumBytes(minBytes: number): boolean {
+    return this.totalLength >= minBytes;
+  }
+}
+
+/**
+ * Read a big-endian 16-bit integer from buffer
+ */
+export function readUInt16BE(buffer: Buffer, offset: number): number {
+  if (offset + 2 > buffer.length) {
+    throw new Error('Buffer too short for UInt16BE read');
+  }
+  return (buffer[offset] << 8) | buffer[offset + 1];
+}
+
+/**
+ * Read a big-endian 24-bit integer from buffer
+ */
+export function readUInt24BE(buffer: Buffer, offset: number): number {
+  if (offset + 3 > buffer.length) {
+    throw new Error('Buffer too short for UInt24BE read');
+  }
+  return (buffer[offset] << 16) | (buffer[offset + 1] << 8) | buffer[offset + 2];
+}
+
+/**
+ * Find a byte sequence in a buffer
+ */
+export function findSequence(buffer: Buffer, sequence: Buffer, startOffset = 0): number {
+  if (sequence.length === 0) {
+    return startOffset;
+  }
+  
+  const searchLength = buffer.length - sequence.length + 1;
+  for (let i = startOffset; i < searchLength; i++) {
+    let found = true;
+    for (let j = 0; j < sequence.length; j++) {
+      if (buffer[i + j] !== sequence[j]) {
+        found = false;
+        break;
+      }
+    }
+    if (found) {
+      return i;
+    }
+  }
+  return -1;
+}
+
+/**
+ * Extract a line from buffer (up to CRLF or LF)
+ */
+export function extractLine(buffer: Buffer, startOffset = 0): { line: string; nextOffset: number } | null {
+  let lineEnd = -1;
+  let skipBytes = 1;
+  
+  // Look for CRLF first
+  const crlfPos = findSequence(buffer, Buffer.from('\r\n'), startOffset);
+  if (crlfPos !== -1) {
+    lineEnd = crlfPos;
+    skipBytes = 2;
+  } else {
+    // Look for LF only
+    for (let i = startOffset; i < buffer.length; i++) {
+      if (buffer[i] === 0x0A) { // LF
+        lineEnd = i;
+        break;
+      }
+    }
+  }
+  
+  if (lineEnd === -1) {
+    return null;
+  }
+  
+  const line = buffer.slice(startOffset, lineEnd).toString('utf8');
+  return {
+    line,
+    nextOffset: lineEnd + skipBytes
+  };
+}
+
+/**
+ * Check if buffer starts with a string (case-insensitive)
+ */
+export function startsWithString(buffer: Buffer, str: string, offset = 0): boolean {
+  if (offset + str.length > buffer.length) {
+    return false;
+  }
+  
+  const bufferStr = buffer.slice(offset, offset + str.length).toString('utf8');
+  return bufferStr.toLowerCase() === str.toLowerCase();
+}
+
+/**
+ * Safe buffer slice that doesn't throw on out-of-bounds
+ */
+export function safeSlice(buffer: Buffer, start: number, end?: number): Buffer {
+  const safeStart = Math.max(0, Math.min(start, buffer.length));
+  const safeEnd = end === undefined 
+    ? buffer.length 
+    : Math.max(safeStart, Math.min(end, buffer.length));
+    
+  return buffer.slice(safeStart, safeEnd);
+}
+
+/**
+ * Check if buffer contains printable ASCII
+ */
+export function isPrintableAscii(buffer: Buffer, length?: number): boolean {
+  const checkLength = length || buffer.length;
+  
+  for (let i = 0; i < checkLength && i < buffer.length; i++) {
+    const byte = buffer[i];
+    // Check if byte is printable ASCII (0x20-0x7E) or tab/newline/carriage return
+    if (byte < 0x20 || byte > 0x7E) {
+      if (byte !== 0x09 && byte !== 0x0A && byte !== 0x0D) {
+        return false;
+      }
+    }
+  }
+  
+  return true;
+}