util: Add consteval ""_hex[_v][_u8] literals

""_hex is a compile-time user-defined literal returning std::array<std::byte>, equivalent of ParseHex. Variants: - ""_hex_v returns std::vector<std::byte> - ""_hex_u8 returns std::array<uint8_t> - ""_hex_v_u8 returns std::vector<uint8_t> - Directly serializable as a size-prefixed OP_PUSH CScript payload using operator<<. Also extracts from_hex into shared util::ConstevalHexDigit function. Co-Authored-By: hodlinator <172445034+hodlinator@users.noreply.github.com> Co-Authored-By: MarcoFalke <*~=`'#}+{/-|&$^_@721217.xyz> Co-Authored-By: Ryan Ofsky <ryan@ofsky.org> Co-Authored-By: stickies-v <stickies-v@protonmail.com>
bitcoin · Aug 28, 2024 · 5b74a84 · 5b74a84
1 parent dc5f6f6
commit 5b74a84
Show file tree

Hide file tree

Showing 3 changed files with 108 additions and 10 deletions.
diff --git a/src/test/util_tests.cpp b/src/test/util_tests.cpp
@@ -7,6 +7,7 @@
 #include <hash.h> // For Hash()
 #include <key.h>  // For CKey
 #include <script/parsing.h>
+#include <span.h>
 #include <sync.h>
 #include <test/util/random.h>
 #include <test/util/setup_common.h>
@@ -45,6 +46,8 @@
 #include <boost/test/unit_test.hpp>
 
 using namespace std::literals;
+using namespace util::hex_literals;
+using util::ConstevalHexDigit;
 using util::Join;
 using util::RemovePrefix;
 using util::RemovePrefixView;
@@ -151,6 +154,20 @@ BOOST_AUTO_TEST_CASE(parse_hex)
 
     // Basic test vector
     std::vector<unsigned char> expected(std::begin(HEX_PARSE_OUTPUT), std::end(HEX_PARSE_OUTPUT));
+    constexpr std::array<std::byte, 65> hex_literal_array{operator""_hex<util::detail::Hex(HEX_PARSE_INPUT)>()};
+    auto hex_literal_span{MakeUCharSpan(hex_literal_array)};
+    BOOST_CHECK_EQUAL_COLLECTIONS(hex_literal_span.begin(), hex_literal_span.end(), expected.begin(), expected.end());
+
+    const std::vector<std::byte> hex_literal_vector{operator""_hex_v<util::detail::Hex(HEX_PARSE_INPUT)>()};
+    hex_literal_span = MakeUCharSpan(hex_literal_vector);
+    BOOST_CHECK_EQUAL_COLLECTIONS(hex_literal_span.begin(), hex_literal_span.end(), expected.begin(), expected.end());
+
+    constexpr std::array<uint8_t, 65> hex_literal_array_uint8{operator""_hex_u8<util::detail::Hex(HEX_PARSE_INPUT)>()};
+    BOOST_CHECK_EQUAL_COLLECTIONS(hex_literal_array_uint8.begin(), hex_literal_array_uint8.end(), expected.begin(), expected.end());
+
+    result = operator""_hex_v_u8<util::detail::Hex(HEX_PARSE_INPUT)>();
+    BOOST_CHECK_EQUAL_COLLECTIONS(result.begin(), result.end(), expected.begin(), expected.end());
+
     result = ParseHex(HEX_PARSE_INPUT);
     BOOST_CHECK_EQUAL_COLLECTIONS(result.begin(), result.end(), expected.begin(), expected.end());
 
@@ -179,6 +196,10 @@ BOOST_AUTO_TEST_CASE(parse_hex)
     BOOST_CHECK_EQUAL_COLLECTIONS(result.begin(), result.end(), expected.begin(), expected.end());
 
     // Empty string is supported
+    static_assert(""_hex.empty());
+    static_assert(""_hex_u8.empty());
+    BOOST_CHECK_EQUAL(""_hex_v.size(), 0);
+    BOOST_CHECK_EQUAL(""_hex_v_u8.size(), 0);
     BOOST_CHECK_EQUAL(ParseHex("").size(), 0);
     BOOST_CHECK_EQUAL(TryParseHex<uint8_t>("").value().size(), 0);
 
@@ -203,6 +224,14 @@ BOOST_AUTO_TEST_CASE(parse_hex)
     BOOST_CHECK(!TryParseHex("12 3").has_value());
 }
 
+BOOST_AUTO_TEST_CASE(consteval_hex_digit)
+{
+    BOOST_CHECK_EQUAL(ConstevalHexDigit('0'), 0);
+    BOOST_CHECK_EQUAL(ConstevalHexDigit('9'), 9);
+    BOOST_CHECK_EQUAL(ConstevalHexDigit('a'), 0xa);
+    BOOST_CHECK_EQUAL(ConstevalHexDigit('f'), 0xf);
+}
+
 BOOST_AUTO_TEST_CASE(util_HexStr)
 {
     BOOST_CHECK_EQUAL(HexStr(HEX_PARSE_OUTPUT), HEX_PARSE_INPUT);

diff --git a/src/uint256.h b/src/uint256.h
@@ -127,19 +127,11 @@ class base_blob
 template <unsigned int BITS>
 consteval base_blob<BITS>::base_blob(std::string_view hex_str)
 {
-    // Non-lookup table version of HexDigit().
-    auto from_hex = [](const char c) -> int8_t {
-        if (c >= '0' && c <= '9') return c - '0';
-        if (c >= 'a' && c <= 'f') return c - 'a' + 0xa;
-
-        throw "Only lowercase hex digits are allowed, for consistency";
-    };
-
     if (hex_str.length() != m_data.size() * 2) throw "Hex string must fit exactly";
     auto str_it = hex_str.rbegin();
     for (auto& elem : m_data) {
-        auto lo = from_hex(*(str_it++));
-        elem = (from_hex(*(str_it++)) << 4) | lo;
+        auto lo = util::ConstevalHexDigit(*(str_it++));
+        elem = (util::ConstevalHexDigit(*(str_it++)) << 4) | lo;
     }
 }
 

diff --git a/src/util/strencodings.h b/src/util/strencodings.h
@@ -13,6 +13,8 @@
 #include <span.h>
 #include <util/string.h>
 
+#include <array>
+#include <bit>
 #include <charconv>
 #include <cstddef>
 #include <cstdint>
@@ -365,4 +367,79 @@ std::string Capitalize(std::string str);
  */
 std::optional<uint64_t> ParseByteUnits(std::string_view str, ByteUnit default_multiplier);
 
+namespace util {
+/** consteval version of HexDigit() without the lookup table. */
+consteval uint8_t ConstevalHexDigit(const char c)
+{
+    if (c >= '0' && c <= '9') return c - '0';
+    if (c >= 'a' && c <= 'f') return c - 'a' + 0xa;
+
+    throw "Only lowercase hex digits are allowed, for consistency";
+}
+
+/**
+ * ""_hex is a compile-time user-defined literal returning a
+ * `std::array<std::byte>`, equivalent to ParseHex(). Variants provided:
+ *
+ * - ""_hex_v: Returns `std::vector<std::byte>`, useful for heap allocation or
+ *   variable-length serialization.
+ *
+ * - ""_hex_u8: Returns `std::array<uint8_t>`, for cases where `std::byte` is
+ *   incompatible.
+ *
+ * - ""_hex_v_u8: Returns `std::vector<uint8_t>`, combining heap allocation with
+ *   `uint8_t`.
+ *
+ * @warning It could be necessary to use vector instead of array variants when
+ *   serializing, or vice versa, because vectors are assumed to be variable-
+ *   length and serialized with a size prefix, while arrays are considered fixed
+ *   length and serialized with no prefix.
+ *
+ * @warning It may be preferable to use vector variants to save stack space when
+ *   declaring local variables if hex strings are large. Alternatively variables
+ *   could be declared constexpr to avoid using stack space.
+ *
+ * @warning Avoid `uint8_t` variants when not necessary, as the codebase
+ *   migrates to use `std::byte` instead of `unsigned char` and `uint8_t`.
+ *
+ * @note One reason ""_hex uses `std::array` instead of `std::vector` like
+ *   ParseHex() does is because heap-based containers cannot cross the compile-
+ *   time/runtime barrier.
+ */
+inline namespace hex_literals {
+namespace detail {
+
+template <size_t N>
+struct Hex {
+    std::array<std::byte, N / 2> bytes{};
+    consteval Hex(const char (&hex_str)[N])
+        // 2 hex digits required per byte + implicit null terminator
+        requires(N % 2 == 1)
+    {
+        if (hex_str[N - 1]) throw "null terminator required";
+        for (std::size_t i = 0; i < bytes.size(); ++i) {
+            bytes[i] = static_cast<std::byte>(
+                (ConstevalHexDigit(hex_str[2 * i]) << 4) |
+                 ConstevalHexDigit(hex_str[2 * i + 1]));
+        }
+    }
+};
+
+} // namespace detail
+
+template <util::detail::Hex str>
+constexpr auto operator""_hex() { return str.bytes; }
+
+template <util::detail::Hex str>
+constexpr auto operator""_hex_u8() { return std::bit_cast<std::array<uint8_t, str.bytes.size()>>(str.bytes); }
+
+template <util::detail::Hex str>
+constexpr auto operator""_hex_v() { return std::vector<std::byte>{str.bytes.begin(), str.bytes.end()}; }
+
+template <util::detail::Hex str>
+inline auto operator""_hex_v_u8() { return std::vector<uint8_t>{UCharCast(str.bytes.data()), UCharCast(str.bytes.data() + str.bytes.size())}; }
+
+} // inline namespace hex_literals
+} // namespace util
+
 #endif // BITCOIN_UTIL_STRENCODINGS_H