Speed up hex encode/decode with lookup tables

Data.hexEncodedString() built each byte via String(format: "%02x", _),
paying Foundation format-string parsing per byte on the hot BLE receive
path (PeerID(hexData:) is called several times per received packet).
Replace it with an ASCII lookup table writing into a preallocated
buffer; output is byte-for-byte identical (lowercase hex).

Data(hexString:) similarly allocated a two-character substring and went
through integer radix parsing per byte; replace with direct nibble
lookups over the UTF-8 bytes. Semantics are unchanged except that
sign-prefixed pairs like "+f", which the old radix parser incorrectly
accepted, are now rejected.

Add round-trip, known-vector, and invalid-input tests to
BitFoundationTests.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
jack
2026-07-01 23:26:38 +02:00
co-authored by Claude Fable 5
parent f688e529f6
commit 3f6dc7dd36
2 changed files with 124 additions and 14 deletions
@@ -8,12 +8,44 @@
import struct Foundation.Data
/// Lowercase hex digits used by `hexEncodedString()`.
private let hexDigits: [UInt8] = Array("0123456789abcdef".utf8)
/// Maps an ASCII byte to its hex nibble value, or nil for non-hex characters.
/// Accepts both lowercase and uppercase hex digits.
@inline(__always)
private func hexNibble(_ ascii: UInt8) -> UInt8? {
switch ascii {
case UInt8(ascii: "0")...UInt8(ascii: "9"):
return ascii - UInt8(ascii: "0")
case UInt8(ascii: "a")...UInt8(ascii: "f"):
return ascii - UInt8(ascii: "a") + 10
case UInt8(ascii: "A")...UInt8(ascii: "F"):
return ascii - UInt8(ascii: "A") + 10
default:
return nil
}
}
public extension Data {
/// Lowercase hex representation of the bytes.
///
/// Lookup-table based: this sits on the hot BLE receive path (it is called
/// several times per received packet via `PeerID(hexData:)`), where the
/// previous per-byte `String(format: "%02x", _)` implementation spent most
/// of its time re-parsing the format string through Foundation.
func hexEncodedString() -> String {
if self.isEmpty {
if isEmpty {
return ""
}
return self.map { String(format: "%02x", $0) }.joined()
var output = [UInt8](repeating: 0, count: count * 2)
var i = 0
for byte in self {
output[i] = hexDigits[Int(byte >> 4)]
output[i + 1] = hexDigits[Int(byte & 0x0F)]
i += 2
}
return String(decoding: output, as: UTF8.self)
}
/// Initialize Data from a hex string.
@@ -28,28 +60,28 @@ public extension Data {
hex = String(hex.dropFirst(2))
}
let ascii = Array(hex.utf8)
// Reject odd-length strings
guard hex.count % 2 == 0 else {
guard ascii.count % 2 == 0 else {
return nil
}
// Reject empty strings
guard !hex.isEmpty else {
// Accept empty strings
guard !ascii.isEmpty else {
self = Data()
return
}
let len = hex.count / 2
var data = Data(capacity: len)
var index = hex.startIndex
for _ in 0..<len {
let nextIndex = hex.index(index, offsetBy: 2)
guard let byte = UInt8(String(hex[index..<nextIndex]), radix: 16) else {
var data = Data(capacity: ascii.count / 2)
var index = 0
while index < ascii.count {
guard let high = hexNibble(ascii[index]),
let low = hexNibble(ascii[index + 1]) else {
return nil
}
data.append(byte)
index = nextIndex
data.append((high << 4) | low)
index += 2
}
self = data
@@ -0,0 +1,78 @@
//
// DataHexTests.swift
// bitchatTests
//
// This is free and unencumbered software released into the public domain.
// For more information, see <https://unlicense.org>
//
import Testing
import Foundation
@testable import BitFoundation
struct DataHexTests {
// MARK: - Encoding
@Test func encode_knownVectors() {
#expect(Data().hexEncodedString() == "")
#expect(Data([0x00]).hexEncodedString() == "00")
#expect(Data([0x0f]).hexEncodedString() == "0f")
#expect(Data([0xf0]).hexEncodedString() == "f0")
#expect(Data([0xff]).hexEncodedString() == "ff")
#expect(Data([0xde, 0xad, 0xbe, 0xef]).hexEncodedString() == "deadbeef")
#expect(Data([0x01, 0x23, 0x45, 0x67, 0x89, 0xab, 0xcd, 0xef]).hexEncodedString() == "0123456789abcdef")
}
@Test func encode_allByteValues_matchesFormatReference() {
let all = Data((0...255).map { UInt8($0) })
let reference = (0...255).map { String(format: "%02x", $0) }.joined()
#expect(all.hexEncodedString() == reference)
}
@Test func encode_worksOnDataSlices() {
let data = Data([0xaa, 0xde, 0xad, 0xbe, 0xef, 0xbb])
let slice = data.dropFirst().dropLast()
#expect(slice.hexEncodedString() == "deadbeef")
}
// MARK: - Decoding
@Test func decode_knownVectors() {
#expect(Data(hexString: "deadbeef") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: "DEADBEEF") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: "DeAdBeEf") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: "00") == Data([0x00]))
#expect(Data(hexString: "0123456789abcdef") == Data([0x01, 0x23, 0x45, 0x67, 0x89, 0xab, 0xcd, 0xef]))
}
@Test func decode_handlesPrefixAndWhitespace() {
#expect(Data(hexString: "0xdeadbeef") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: "0XDEADBEEF") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: " deadbeef\n") == Data([0xde, 0xad, 0xbe, 0xef]))
#expect(Data(hexString: "") == Data())
#expect(Data(hexString: "0x") == Data())
}
@Test func decode_rejectsInvalidInput() {
#expect(Data(hexString: "abc") == nil) // odd length
#expect(Data(hexString: "zz") == nil) // non-hex characters
#expect(Data(hexString: "0xg1") == nil) // non-hex after prefix
#expect(Data(hexString: "+f") == nil) // sign characters are not hex
#expect(Data(hexString: "-0") == nil)
#expect(Data(hexString: "a\u{00e9}") == nil) // non-ASCII
#expect(Data(hexString: "de ad") == nil) // interior whitespace
}
// MARK: - Round trip
@Test func roundTrip_randomLengths() {
for length in [0, 1, 2, 3, 8, 16, 31, 32, 33, 64, 255, 1024] {
let data = Data((0..<length).map { _ in UInt8.random(in: .min ... .max) })
let hex = data.hexEncodedString()
#expect(hex.count == length * 2)
#expect(Data(hexString: hex) == data)
#expect(Data(hexString: hex.uppercased()) == data)
}
}
}