/** * Written in the D programming language. * This module provides functions to uniform calculating hash values for different types * * Copyright: Copyright Igor Stepanov 2013-2013. * License: $(HTTP www.boost.org/LICENSE_1_0.txt, Boost License 1.0). * Authors: Igor Stepanov * Source: $(DRUNTIMESRC core/internal/_hash.d) */ module core.internal.hash; import core.internal.traits : Unconst; // If true ensure that positive zero and negative zero have the same hash. // Historically typeid(float).getHash did this but hashOf(float) did not. private enum floatCoalesceZeroes = true; // If true ensure that all NaNs of the same floating point type have the same hash. // Historically typeid(float).getHash didn't do this but hashOf(float) did. private enum floatCoalesceNaNs = true; // If either of the above are true then no struct or array that contains the // representation of a floating point number may be hashed with `bytesHash`. @nogc nothrow pure @safe unittest { static if (floatCoalesceZeroes) assert(hashOf(+0.0) == hashOf(-0.0)); // Same hash for +0.0 and -0.0. static if (floatCoalesceNaNs) assert(hashOf(double.nan) == hashOf(-double.nan)); // Same hash for different NaN. } private enum hasCallableToHash(T) = __traits(compiles, { size_t hash = ((T* x) => (*x).toHash())(null); }); @nogc nothrow pure @safe unittest { static struct S { size_t toHash() { return 4; } } assert(hasCallableToHash!S); assert(!hasCallableToHash!(shared const S)); } private enum isFinalClassWithAddressBasedHash(T) = __traits(isFinalClass, T) // Use __traits(compiles, ...) in case there are multiple overloads of `toHash`. && __traits(compiles, {static assert(&Object.toHash is &T.toHash);}); @nogc nothrow pure @safe unittest { static class C1 {} final static class C2 : C1 {} final static class C3 : C1 { override size_t toHash() const nothrow { return 1; }} static assert(!isFinalClassWithAddressBasedHash!Object); static assert(!isFinalClassWithAddressBasedHash!C1); static assert(isFinalClassWithAddressBasedHash!C2); static assert(!isFinalClassWithAddressBasedHash!C3); } private template isCppClassWithoutHash(T) { static if (!is(T == class) && !is(T == interface)) enum isCppClassWithoutHash = false; else enum bool isCppClassWithoutHash = __traits(getLinkage, T) == "C++" && !is(immutable T* : immutable Object*) && !hasCallableToHash!T; } /+ Is it valid to calculate a hash code for T based on the bits of its representation? Always false for interfaces, dynamic arrays, and associative arrays. False for all classes except final classes that do not override `toHash`. Note: according to the spec as of https://github.com/dlang/dlang.org/commit/d66eff16491b0664c0fc00ba80a7aa291703f1f2 the contents of unnamed paddings between fields is undefined. Currently this hashing implementation assumes that the padding contents (if any) for all instances of `T` are the same. The correctness of this assumption is yet to be verified. +/ private template canBitwiseHash(T) { static if (is(T EType == enum)) enum canBitwiseHash = .canBitwiseHash!EType; else static if (__traits(isFloating, T)) enum canBitwiseHash = !(floatCoalesceZeroes || floatCoalesceNaNs); else static if (__traits(isScalar, T)) enum canBitwiseHash = true; else static if (is(T == class)) { enum canBitwiseHash = isFinalClassWithAddressBasedHash!T || isCppClassWithoutHash!T; } else static if (is(T == interface)) { enum canBitwiseHash = isCppClassWithoutHash!T; } else static if (is(T == struct)) { static if (hasCallableToHash!T || __traits(isNested, T)) enum canBitwiseHash = false; else { import core.internal.traits : allSatisfy; enum canBitwiseHash = allSatisfy!(.canBitwiseHash, typeof(T.tupleof)); } } else static if (is(T == union)) { // Right now we always bytewise hash unions that lack callable `toHash`. enum canBitwiseHash = !hasCallableToHash!T; } else static if (is(T E : E[])) { static if (__traits(isStaticArray, T)) enum canBitwiseHash = (T.length == 0) || .canBitwiseHash!E; else enum canBitwiseHash = false; } else static if (__traits(isAssociativeArray, T)) { enum canBitwiseHash = false; } else { static assert(is(T == delegate) || is(T : void) || is(T : typeof(null)), "Internal error: unanticipated type "~T.stringof); enum canBitwiseHash = true; } } //enum hash. CTFE depends on base type size_t hashOf(T)(auto ref T val, size_t seed = 0) if (is(T == enum) && !__traits(isScalar, T)) { static if (is(T EType == enum)) {} //for EType return hashOf(cast(EType) val, seed); } //CTFE ready (depends on base type). size_t hashOf(T)(scope const auto ref T val, size_t seed = 0) if (!is(T == enum) && __traits(isStaticArray, T) && canBitwiseHash!T) { import core.internal.convert : toUbyte; // FIXME: // We would like to to do this: // //static if (T.length == 0) // return seed; //else static if (T.length == 1) // return hashOf(val[0], seed); //else // return bytesHashWithExactSizeAndAlignment!T(toUbyte(val), seed); // // ... but that's inefficient when using a runtime TypeInfo (introduces a branch) // and PR #2243 wants typeid(T).getHash(&val) to produce the same result as // hashOf(val). static if (T.length == 0) { return bytesHashAlignedBy!size_t((ubyte[]).init, seed); } static if (is(typeof(toUbyte(val)) == const(ubyte)[])) { return bytesHashAlignedBy!T(toUbyte(val), seed); } else //Other types. CTFE unsupported { assert(!__ctfe, "unable to compute hash of "~T.stringof~" at compile time"); return bytesHashAlignedBy!T((cast(const(ubyte)*) &val)[0 .. T.sizeof], seed); } } //CTFE ready (depends on base type). size_t hashOf(T)(auto ref T val, size_t seed = 0) if (!is(T == enum) && __traits(isStaticArray, T) && !canBitwiseHash!T) { // FIXME: // We would like to to do this: // //static if (T.length == 0) // return seed; //else static if (T.length == 1) // return hashOf(val[0], seed); //else // /+ hash like a dynamic array +/ // // ... but that's inefficient when using a runtime TypeInfo (introduces a branch) // and PR #2243 wants typeid(T).getHash(&val) to produce the same result as // hashOf(val). return hashOf(val[], seed); } //dynamic array hash size_t hashOf(T)(scope const T val, size_t seed = 0) if (is(T == S[], S) && (__traits(isScalar, S) || canBitwiseHash!S)) // excludes enum types { import core.internal.convert : toUbyte; alias ElementType = typeof(val[0]); static if (!canBitwiseHash!ElementType) { size_t hash = seed; foreach (ref o; val) { hash = hashOf(hashOf(o), hash); // double hashing to match TypeInfo.getHash } return hash; } else static if (is(typeof(toUbyte(val)) == const(ubyte)[])) //ubyteble array (arithmetic types and structs without toHash) CTFE ready for arithmetic types and structs without reference fields { return bytesHashAlignedBy!ElementType(toUbyte(val), seed); } else //Other types. CTFE unsupported { assert(!__ctfe, "unable to compute hash of "~T.stringof~" at compile time"); return bytesHashAlignedBy!ElementType((cast(const(ubyte)*) val.ptr)[0 .. ElementType.sizeof*val.length], seed); } } //dynamic array hash size_t hashOf(T)(T val, size_t seed = 0) if (is(T == S[], S) && !(__traits(isScalar, S) || canBitwiseHash!S)) // excludes enum types { size_t hash = seed; foreach (ref o; val) { hash = hashOf(hashOf(o), hash); // double hashing because TypeInfo.getHash doesn't allow to pass seed value } return hash; } // Indicates if F is a built-in complex number type. private F coalesceFloat(F)(const F val) if (__traits(isFloating, val) && !is(F == __vector) && !is(F : creal)) { static if (floatCoalesceZeroes) if (val == cast(F) 0) return cast(F) 0; static if (floatCoalesceNaNs) if (val != val) return F.nan; return val; } //scalar type hash @trusted @nogc nothrow pure size_t hashOf(T)(scope const T val) if (__traits(isScalar, T) && !is(T == __vector)) { static if (is(T V : V*)) { if (__ctfe) { if (val is null) return 0; assert(0, "Unable to calculate hash of non-null pointer at compile time"); } size_t result = cast(size_t) val; return result ^ (result >> 4); } else static if (is(T EType == enum) && is(typeof(val[0]))) { // Enum type whose base type is vector. return hashOf(cast(EType) val); } else static if (__traits(isIntegral, T)) { static if (T.sizeof <= size_t.sizeof) return val; else return cast(size_t) (val ^ (val >>> (size_t.sizeof * 8))); } else static if (is(T : creal)) { return hashOf(coalesceFloat(val.re), hashOf(coalesceFloat(val.im))); } else { static assert(__traits(isFloating, T)); auto data = coalesceFloat(val); static if (T.sizeof == float.sizeof && T.mant_dig == float.mant_dig) return *cast(const uint*) &data; else static if (T.sizeof == double.sizeof && T.mant_dig == double.mant_dig) return hashOf(*cast(const ulong*) &data); else { import core.internal.convert : floatSize, toUbyte; return bytesHashWithExactSizeAndAlignment!T(toUbyte(data)[0 .. floatSize!T], 0); } } } //scalar type hash @trusted @nogc nothrow pure size_t hashOf(T)(scope const T val, size_t seed) if (__traits(isScalar, T) && !is(T == __vector)) { static if (is(T V : V*)) { if (__ctfe) { if (val is null) return hashOf(size_t(0), seed); assert(0, "Unable to calculate hash of non-null pointer at compile time"); } return hashOf(cast(size_t) val, seed); } else static if (is(T EType == enum) && is(typeof(val[0]))) { // Enum type whose base type is vector. return hashOf(cast(EType) val, seed); } else static if (__traits(isIntegral, val) && T.sizeof <= size_t.sizeof) { static if (size_t.sizeof < ulong.sizeof) { //MurmurHash3 32-bit single round enum uint c1 = 0xcc9e2d51; enum uint c2 = 0x1b873593; enum uint c3 = 0xe6546b64; enum uint r1 = 15; enum uint r2 = 13; } else { //Half of MurmurHash3 64-bit single round //(omits second interleaved update) enum ulong c1 = 0x87c37b91114253d5; enum ulong c2 = 0x4cf5ad432745937f; enum ulong c3 = 0x52dce729; enum uint r1 = 31; enum uint r2 = 27; } size_t h = c1 * val; h = (h << r1) | (h >>> (size_t.sizeof * 8 - r1)); h = (h * c2) ^ seed; h = (h << r2) | (h >>> (size_t.sizeof * 8 - r2)); return h * 5 + c3; } else static if (__traits(isIntegral, val) && T.sizeof > size_t.sizeof) { static foreach (i; 0 .. T.sizeof / size_t.sizeof) seed = hashOf(cast(size_t) (val >>> (size_t.sizeof * 8 * i)), seed); return seed; } else static if (is(T : creal)) { return hashOf(val.re, hashOf(val.im, seed)); } else static if (__traits(isFloating, T)) { auto data = coalesceFloat(val); static if (T.sizeof == float.sizeof && T.mant_dig == float.mant_dig) return hashOf(*cast(const uint*) &data, seed); else static if (T.sizeof == double.sizeof && T.mant_dig == double.mant_dig) return hashOf(*cast(const ulong*) &data, seed); else { import core.internal.convert : floatSize, toUbyte; return bytesHashWithExactSizeAndAlignment!T(toUbyte(data)[0 .. floatSize!T], seed); } } else { static assert(0); } } size_t hashOf(T)(scope const T val, size_t seed = 0) @safe @nogc nothrow pure if (is(T == __vector)) // excludes enum types { static if (__traits(isFloating, T) && (floatCoalesceZeroes || floatCoalesceNaNs)) { if (__ctfe) { // Workaround for CTFE bug. static if (is(immutable typeof(val[0]) == immutable E, E)) {} // Get E. E[T.sizeof / E.sizeof] array; foreach (i; 0 .. T.sizeof / E.sizeof) array[i] = val[i]; return hashOf(array, seed); } return hashOf(val.array, seed); } else { import core.internal.convert : toUbyte; return bytesHashAlignedBy!T(toUbyte(val), seed); } } //typeof(null) hash. CTFE supported @trusted @nogc nothrow pure size_t hashOf(T)(scope const T val) if (!is(T == enum) && is(T : typeof(null))) { return 0; } //typeof(null) hash. CTFE supported @trusted @nogc nothrow pure size_t hashOf(T)(scope const T val, size_t seed) if (!is(T == enum) && is(T : typeof(null))) { return hashOf(size_t(0), seed); } private enum _hashOfStruct = q{ enum bool isChained = is(typeof(seed) : size_t); static if (!isChained) enum size_t seed = 0; static if (hasCallableToHash!(typeof(val))) //CTFE depends on toHash() { static if (!__traits(isSame, typeof(val), __traits(parent, val.toHash)) && is(typeof(val is null))) { static if (isChained) return hashOf(__traits(getMember, val, __traits(getAliasThis, typeof(val))), seed); else return hashOf(__traits(getMember, val, __traits(getAliasThis, typeof(val)))); } else { static if (isChained) return hashOf(cast(size_t) val.toHash(), seed); else return val.toHash(); } } else { import core.internal.convert : toUbyte; static if (__traits(hasMember, T, "toHash") && is(typeof(T.toHash) == function)) { // TODO: in the future maybe this should be changed to a static // assert(0), because if there's a `toHash` the programmer probably // expected it to be called and a compilation failure here will // expose a bug in his code. // In the future we also might want to disallow non-const toHash // altogether. pragma(msg, "Warning: struct "~__traits(identifier, T) ~" has method toHash, however it cannot be called with " ~typeof(val).stringof~" this."); static if (__traits(compiles, __traits(getLocation, T.toHash))) { enum file = __traits(getLocation, T.toHash)[0]; enum line = __traits(getLocation, T.toHash)[1].stringof; pragma(msg, " ",__traits(identifier, T),".toHash defined here: ",file,"(",line,")"); } } static if (T.tupleof.length == 0) { return seed; } else static if ((is(T == struct) && !canBitwiseHash!T) || T.tupleof.length == 1) { static if (isChained) size_t h = seed; static foreach (i, F; typeof(val.tupleof)) { static if (__traits(isStaticArray, F)) { static if (i == 0 && !isChained) size_t h = 0; static if (F.sizeof > 0 && canBitwiseHash!F) // May use smallBytesHash instead of bytesHash. h = bytesHashWithExactSizeAndAlignment!F(toUbyte(val.tupleof[i]), h); else // We can avoid the "double hashing" the top-level version uses // for consistency with TypeInfo.getHash. foreach (ref e; val.tupleof[i]) h = hashOf(e, h); } else static if (is(F == struct) || is(F == union)) { static if (hasCallableToHash!F) { static if (!__traits(isSame, F, __traits(parent, val.tupleof[i].toHash)) && is(typeof(val.tupleof[i] is null))) { static if (i == 0 && !isChained) size_t h = hashOf(__traits(getMember, val.tupleof[i], __traits(getAliasThis, F))); else h = hashOf(__traits(getMember, val.tupleof[i], __traits(getAliasThis, F)), h); } else { static if (i == 0 && !isChained) size_t h = val.tupleof[i].toHash(); else h = hashOf(cast(size_t) val.tupleof[i].toHash(), h); } } else static if (F.tupleof.length == 1) { // Handle the single member case separately to avoid unnecessarily using bytesHash. static if (i == 0 && !isChained) size_t h = hashOf(val.tupleof[i].tupleof[0]); else h = hashOf(val.tupleof[i].tupleof[0], h); } else static if (canBitwiseHash!F) { // May use smallBytesHash instead of bytesHash. static if (i == 0 && !isChained) size_t h = 0; h = bytesHashWithExactSizeAndAlignment!F(toUbyte(val.tupleof[i]), h); } else { // Nothing special happening. static if (i == 0 && !isChained) size_t h = hashOf(val.tupleof[i]); else h = hashOf(val.tupleof[i], h); } } else { // Nothing special happening. static if (i == 0 && !isChained) size_t h = hashOf(val.tupleof[i]); else h = hashOf(val.tupleof[i], h); } } return h; } else static if (is(typeof(toUbyte(val)) == const(ubyte)[]))//CTFE ready for structs without reference fields { // Not using bytesHashWithExactSizeAndAlignment here because // the result may differ from typeid(T).hashOf(&val). return bytesHashAlignedBy!T(toUbyte(val), seed); } else // CTFE unsupported { assert(!__ctfe, "unable to compute hash of "~T.stringof~" at compile time"); const(ubyte)[] bytes = (() @trusted => (cast(const(ubyte)*)&val)[0 .. T.sizeof])(); // Not using bytesHashWithExactSizeAndAlignment here because // the result may differ from typeid(T).hashOf(&val). return bytesHashAlignedBy!T(bytes, seed); } } }; //struct or union hash size_t hashOf(T)(scope const auto ref T val, size_t seed = 0) if (!is(T == enum) && (is(T == struct) || is(T == union)) && !is(T == const) && !is(T == immutable) && canBitwiseHash!T) { mixin(_hashOfStruct); } //struct or union hash size_t hashOf(T)(auto ref T val) if (!is(T == enum) && (is(T == struct) || is(T == union)) && !canBitwiseHash!T) { mixin(_hashOfStruct); } //struct or union hash size_t hashOf(T)(auto ref T val, size_t seed) if (!is(T == enum) && (is(T == struct) || is(T == union)) && !canBitwiseHash!T) { mixin(_hashOfStruct); } //struct or union hash - https://issues.dlang.org/show_bug.cgi?id=19332 (support might be removed in future) size_t hashOf(T)(scope auto ref T val, size_t seed = 0) if (!is(T == enum) && (is(T == struct) || is(T == union)) && (is(T == const) || is(T == immutable)) && canBitwiseHash!T && !canBitwiseHash!(Unconst!T)) { mixin(_hashOfStruct); } //delegate hash. CTFE only if null. @trusted @nogc nothrow pure size_t hashOf(T)(scope const T val, size_t seed = 0) if (!is(T == enum) && is(T == delegate)) { if (__ctfe) { if (val is null) return hashOf(size_t(0), hashOf(size_t(0), seed)); assert(0, "unable to compute hash of "~T.stringof~" at compile time"); } return hashOf(val.ptr, hashOf(cast(void*) val.funcptr, seed)); } //address-based class hash. CTFE only if null. @nogc nothrow pure @trusted size_t hashOf(T)(scope const T val) if (!is(T == enum) && (is(T == interface) || is(T == class)) && canBitwiseHash!T) { if (__ctfe) if (val is null) return 0; return hashOf(cast(const void*) val); } //address-based class hash. CTFE only if null. @nogc nothrow pure @trusted size_t hashOf(T)(scope const T val, size_t seed) if (!is(T == enum) && (is(T == interface) || is(T == class)) && canBitwiseHash!T) { if (__ctfe) if (val is null) return hashOf(size_t(0), seed); return hashOf(cast(const void*) val, seed); } //class or interface hash. CTFE depends on toHash size_t hashOf(T)(T val) if (!is(T == enum) && (is(T == interface) || is(T == class)) && !canBitwiseHash!T) { static if (__traits(compiles, {size_t h = val.toHash();})) { static if (is(__traits(parent, val.toHash) P) && !is(immutable T* : immutable P*) && is(typeof((ref P p) => p is null))) return val ? hashOf(__traits(getMember, val, __traits(getAliasThis, T))) : 0; else return val ? val.toHash() : 0; } else return val ? (cast(Object)val).toHash() : 0; } //class or interface hash. CTFE depends on toHash size_t hashOf(T)(T val, size_t seed) if (!is(T == enum) && (is(T == interface) || is(T == class)) && !canBitwiseHash!T) { static if (__traits(compiles, {size_t h = val.toHash();})) { static if (is(__traits(parent, val.toHash) P) && !is(immutable T* : immutable P*) && is(typeof((ref P p) => p is null))) return hashOf(val ? hashOf(__traits(getMember, val, __traits(getAliasThis, T))) : size_t(0), seed); else return hashOf(val ? cast(size_t) val.toHash() : size_t(0), seed); } else return hashOf(val ? (cast(Object)val).toHash() : 0, seed); } //associative array hash. CTFE depends on base types size_t hashOf(T)(T aa) if (!is(T == enum) && __traits(isAssociativeArray, T)) { static if (is(typeof(aa) : V[K], K, V)) {} // Put K & V in scope. static if (__traits(compiles, (ref K k, ref V v) nothrow => .hashOf(k) + .hashOf(v))) scope (failure) assert(0); // Allow compiler to infer nothrow. if (!aa.length) return 0; size_t h = 0; // The computed hash is independent of the foreach traversal order. foreach (ref key, ref val; aa) h += hashOf(hashOf(val), hashOf(key)); return h; } //associative array hash. CTFE depends on base types size_t hashOf(T)(T aa, size_t seed) if (!is(T == enum) && __traits(isAssociativeArray, T)) { return hashOf(hashOf(aa), seed); } // MurmurHash3 was written by Austin Appleby, and is placed in the public // domain. The author hereby disclaims copyright to this source code. // This overload is for backwards compatibility. @system pure nothrow @nogc size_t bytesHash()(scope const(void)* buf, size_t len, size_t seed) { return bytesHashAlignedBy!ubyte((cast(const(ubyte)*) buf)[0 .. len], seed); } private template bytesHashAlignedBy(AlignType) { alias bytesHashAlignedBy = bytesHash!(AlignType.alignof >= uint.alignof); } private template bytesHashWithExactSizeAndAlignment(SizeAndAlignType) { static if (SizeAndAlignType.alignof < uint.alignof ? SizeAndAlignType.sizeof <= 12 : SizeAndAlignType.sizeof <= 10) alias bytesHashWithExactSizeAndAlignment = smallBytesHash; else alias bytesHashWithExactSizeAndAlignment = bytesHashAlignedBy!SizeAndAlignType; } // Fowler/Noll/Vo hash. http://www.isthe.com/chongo/tech/comp/fnv/ private size_t fnv()(scope const(ubyte)[] bytes, size_t seed) @nogc nothrow pure @safe { static if (size_t.max <= uint.max) enum prime = (1U << 24) + (1U << 8) + 0x93U; else static if (size_t.max <= ulong.max) enum prime = (1UL << 40) + (1UL << 8) + 0xb3UL; else enum prime = (size_t(1) << 88) + (size_t(1) << 8) + size_t(0x3b); foreach (b; bytes) seed = (seed ^ b) * prime; return seed; } private alias smallBytesHash = fnv; //----------------------------------------------------------------------------- // Block read - if your platform needs to do endian-swapping or can only // handle aligned reads, do the conversion here private uint get32bits()(scope const(ubyte)* x) @nogc nothrow pure @system { version (BigEndian) { return ((cast(uint) x[0]) << 24) | ((cast(uint) x[1]) << 16) | ((cast(uint) x[2]) << 8) | (cast(uint) x[3]); } else { return ((cast(uint) x[3]) << 24) | ((cast(uint) x[2]) << 16) | ((cast(uint) x[1]) << 8) | (cast(uint) x[0]); } } /+ Params: dataKnownToBeAligned = whether the data is known at compile time to be uint-aligned. +/ @nogc nothrow pure @trusted private size_t bytesHash(bool dataKnownToBeAligned)(scope const(ubyte)[] bytes, size_t seed) { auto len = bytes.length; auto data = bytes.ptr; auto nblocks = len / 4; uint h1 = cast(uint)seed; enum uint c1 = 0xcc9e2d51; enum uint c2 = 0x1b873593; enum uint c3 = 0xe6546b64; //---------- // body auto end_data = data+nblocks*uint.sizeof; for (; data!=end_data; data += uint.sizeof) { static if (dataKnownToBeAligned) uint k1 = __ctfe ? get32bits(data) : *(cast(const uint*) data); else uint k1 = get32bits(data); k1 *= c1; k1 = (k1 << 15) | (k1 >> (32 - 15)); k1 *= c2; h1 ^= k1; h1 = (h1 << 13) | (h1 >> (32 - 13)); h1 = h1*5+c3; } //---------- // tail uint k1 = 0; switch (len & 3) { case 3: k1 ^= data[2] << 16; goto case; case 2: k1 ^= data[1] << 8; goto case; case 1: k1 ^= data[0]; k1 *= c1; k1 = (k1 << 15) | (k1 >> (32 - 15)); k1 *= c2; h1 ^= k1; goto default; default: } //---------- // finalization h1 ^= len; // Force all bits of the hash block to avalanche. h1 = (h1 ^ (h1 >> 16)) * 0x85ebca6b; h1 = (h1 ^ (h1 >> 13)) * 0xc2b2ae35; h1 ^= h1 >> 16; return h1; } // Check that bytesHash works with CTFE pure nothrow @system @nogc unittest { size_t ctfeHash(string x) { return bytesHash(x.ptr, x.length, 0); } enum test_str = "Sample string"; enum size_t hashVal = ctfeHash(test_str); assert(hashVal == bytesHash(&test_str[0], test_str.length, 0)); // Detect unintended changes to bytesHash on unaligned and aligned inputs. version (BigEndian) { const ubyte[7] a = [99, 4, 3, 2, 1, 5, 88]; const uint[2] b = [0x04_03_02_01, 0x05_ff_ff_ff]; } else { const ubyte[7] a = [99, 1, 2, 3, 4, 5, 88]; const uint[2] b = [0x04_03_02_01, 0xff_ff_ff_05]; } // It is okay to change the below values if you make a change // that you expect to change the result of bytesHash. assert(bytesHash(&a[1], a.length - 2, 0) == 2727459272); assert(bytesHash(&b, 5, 0) == 2727459272); assert(bytesHashAlignedBy!uint((cast(const ubyte*) &b)[0 .. 5], 0) == 2727459272); }