beenull/ladybird
1
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird.git synced 2024-11-22 23:50:19 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 6
ladybird/Tests/LibCrypto/TestBigInteger.cpp

1044 lines
44 KiB
C++
Raw Normal View History

LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
/*
* Copyright (c) 2021, Peter Bocan <me@pbocan.net>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include <LibCrypto/BigInt/Algorithms/UnsignedBigIntegerAlgorithms.h>
#include <LibCrypto/BigInt/SignedBigInteger.h>
#include <LibCrypto/BigInt/UnsignedBigInteger.h>
#include <LibCrypto/NumberTheory/ModularFunctions.h>
#include <LibTest/TestCase.h>
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
#include <math.h>
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
static Crypto::UnsignedBigInteger bigint_fibonacci(size_t n)
{
Crypto::UnsignedBigInteger num1(0);
Crypto::UnsignedBigInteger num2(1);
for (size_t i = 0; i < n; ++i) {
Crypto::UnsignedBigInteger t = num1.plus(num2);
num2 = num1;
num1 = t;
}
return num1;
}
static Crypto::SignedBigInteger bigint_signed_fibonacci(size_t n)
{
Crypto::SignedBigInteger num1(0);
Crypto::SignedBigInteger num2(1);
for (size_t i = 0; i < n; ++i) {
Crypto::SignedBigInteger t = num1.plus(num2);
num2 = num1;
num1 = t;
}
return num1;
}
TEST_CASE(test_bigint_fib500)
{
Vector<u32> result {
315178285, 505575602, 1883328078, 125027121, 3649625763,
347570207, 74535262, 3832543808, 2472133297, 1600064941, 65273441
};
EXPECT_EQ(bigint_fibonacci(500).words(), result);
}
TEST_CASE(test_unsigned_bigint_addition_initialization)
{
Crypto::UnsignedBigInteger num1;
Crypto::UnsignedBigInteger num2(70);
Crypto::UnsignedBigInteger num3 = num1.plus(num2);
bool pass = (num3 == num2);
pass &= (num1 == Crypto::UnsignedBigInteger(0));
EXPECT(pass);
}
TEST_CASE(test_unsigned_bigint_addition_borrow_with_zero)
{
Crypto::UnsignedBigInteger num1({ UINT32_MAX - 3, UINT32_MAX });
Crypto::UnsignedBigInteger num2({ UINT32_MAX - 2, 0 });
Vector<u32> expected_result { 4294967289, 0, 1 };
EXPECT_EQ(num1.plus(num2).words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_basic_add_to_accumulator)
{
Crypto::UnsignedBigInteger num1(10);
Crypto::UnsignedBigInteger num2(70);
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
EXPECT_EQ(num1.words(), Vector<u32> { 80 });
}
TEST_CASE(test_unsigned_bigint_basic_add_to_empty_accumulator)
{
LibCrypto: Make the constructors of (Un)SignedBigInteger templated This means it can take any (un)signed word of size at most Word. This means the constructor can be disambiguated if we were to add a double constructor :^). This requires a change in just one test.
2022-08-25 21:35:34 +00:00
Crypto::UnsignedBigInteger num1 {};
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
Crypto::UnsignedBigInteger num2(10);
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
EXPECT_EQ(num1.words(), Vector<u32> { 10 });
}
TEST_CASE(test_unsigned_bigint_basic_add_to_smaller_accumulator)
{
Crypto::UnsignedBigInteger num1(10);
Crypto::UnsignedBigInteger num2({ 10, 10 });
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
Vector<u32> expected_result { 20, 10 };
EXPECT_EQ(num1.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_add_to_accumulator_with_multiple_carry_levels)
{
Crypto::UnsignedBigInteger num1({ UINT32_MAX - 2, UINT32_MAX });
Crypto::UnsignedBigInteger num2(5);
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
Vector<u32> expected_result { 2, 0, 1 };
EXPECT_EQ(num1.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_add_to_accumulator_with_leading_zero)
{
Crypto::UnsignedBigInteger num1(1);
Crypto::UnsignedBigInteger num2({ 1, 0 });
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
EXPECT_EQ(num1.words(), Vector<u32> { 2 });
}
TEST_CASE(test_unsigned_bigint_add_to_accumulator_with_carry_and_leading_zero)
{
Crypto::UnsignedBigInteger num1({ UINT32_MAX, 0, 0, 0 });
Crypto::UnsignedBigInteger num2({ 1, 0 });
Crypto::UnsignedBigIntegerAlgorithms::add_into_accumulator_without_allocation(num1, num2);
Vector<u32> expected_result { 0, 1, 0, 0 };
EXPECT_EQ(num1.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_simple_subtraction)
{
Crypto::UnsignedBigInteger num1(80);
Crypto::UnsignedBigInteger num2(70);
EXPECT_EQ(num1.minus(num2), Crypto::UnsignedBigInteger(10));
}
TEST_CASE(test_unsigned_bigint_simple_subtraction_invalid)
{
Crypto::UnsignedBigInteger num1(50);
Crypto::UnsignedBigInteger num2(70);
EXPECT(num1.minus(num2).is_invalid());
}
TEST_CASE(test_unsigned_bigint_simple_subtraction_with_borrow)
{
Crypto::UnsignedBigInteger num1(UINT32_MAX);
Crypto::UnsignedBigInteger num2(1);
Crypto::UnsignedBigInteger num3 = num1.plus(num2);
Crypto::UnsignedBigInteger result = num3.minus(num2);
EXPECT_EQ(result, num1);
}
TEST_CASE(test_unsigned_bigint_subtraction_with_large_numbers)
{
Crypto::UnsignedBigInteger num1 = bigint_fibonacci(343);
Crypto::UnsignedBigInteger num2 = bigint_fibonacci(218);
Crypto::UnsignedBigInteger result = num1.minus(num2);
Vector<u32> expected_result {
811430588, 2958904896, 1130908877, 2830569969, 3243275482,
3047460725, 774025231, 7990
};
EXPECT_EQ(result.plus(num2), num1);
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_subtraction_with_large_numbers2)
{
Crypto::UnsignedBigInteger num1(Vector<u32> { 1483061863, 446680044, 1123294122, 191895498, 3347106536, 16, 0, 0, 0 });
Crypto::UnsignedBigInteger num2(Vector<u32> { 4196414175, 1117247942, 1123294122, 191895498, 3347106536, 16 });
Crypto::UnsignedBigInteger result = num1.minus(num2);
// this test only verifies that we don't crash on an assertion
}
TEST_CASE(test_unsigned_bigint_subtraction_regression_1)
{
auto num = Crypto::UnsignedBigInteger { 1 }.shift_left(256);
Vector<u32> expected_result {
4294967295, 4294967295, 4294967295, 4294967295, 4294967295,
4294967295, 4294967295, 4294967295, 0
};
EXPECT_EQ(num.minus(1).words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_simple_multiplication)
{
Crypto::UnsignedBigInteger num1(8);
Crypto::UnsignedBigInteger num2(251);
Crypto::UnsignedBigInteger result = num1.multiplied_by(num2);
EXPECT_EQ(result.words(), Vector<u32> { 2008 });
}
TEST_CASE(test_unsigned_bigint_multiplication_with_big_numbers1)
{
Crypto::UnsignedBigInteger num1 = bigint_fibonacci(200);
Crypto::UnsignedBigInteger num2(12345678);
Crypto::UnsignedBigInteger result = num1.multiplied_by(num2);
Vector<u32> expected_result { 669961318, 143970113, 4028714974, 3164551305, 1589380278, 2 };
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_multiplication_with_big_numbers2)
{
Crypto::UnsignedBigInteger num1 = bigint_fibonacci(200);
Crypto::UnsignedBigInteger num2 = bigint_fibonacci(341);
Crypto::UnsignedBigInteger result = num1.multiplied_by(num2);
Vector<u32> expected_result {
3017415433, 2741793511, 1957755698, 3731653885, 3154681877,
785762127, 3200178098, 4260616581, 529754471, 3632684436,
1073347813, 2516430
};
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_unsigned_bigint_simple_division)
{
Crypto::UnsignedBigInteger num1(27194);
Crypto::UnsignedBigInteger num2(251);
auto result = num1.divided_by(num2);
Crypto::UnsignedDivisionResult expected = { Crypto::UnsignedBigInteger(108), Crypto::UnsignedBigInteger(86) };
EXPECT_EQ(result.quotient, expected.quotient);
EXPECT_EQ(result.remainder, expected.remainder);
}
TEST_CASE(test_unsigned_bigint_division_with_big_numbers)
{
Crypto::UnsignedBigInteger num1 = bigint_fibonacci(386);
Crypto::UnsignedBigInteger num2 = bigint_fibonacci(238);
auto result = num1.divided_by(num2);
Crypto::UnsignedDivisionResult expected = {
Crypto::UnsignedBigInteger(Vector<u32> { 2300984486, 2637503534, 2022805584, 107 }),
Crypto::UnsignedBigInteger(Vector<u32> { 1483061863, 446680044, 1123294122, 191895498, 3347106536, 16, 0, 0, 0 })
};
EXPECT_EQ(result.quotient, expected.quotient);
EXPECT_EQ(result.remainder, expected.remainder);
}
TEST_CASE(test_unsigned_bigint_division_combined_test)
{
auto num1 = bigint_fibonacci(497);
auto num2 = bigint_fibonacci(238);
auto div_result = num1.divided_by(num2);
EXPECT_EQ(div_result.quotient.multiplied_by(num2).plus(div_result.remainder), num1);
}
TEST_CASE(test_unsigned_bigint_base10_from_string)
{
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
auto result = TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(10, "57195071295721390579057195715793"sv));
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
Vector<u32> expected_result { 3806301393, 954919431, 3879607298, 721 };
EXPECT_EQ(result.words(), expected_result);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
Vector<StringView> invalid_base10_number_strings { "1A"sv, "1:"sv, "Z1"sv, "1/"sv };
for (auto invalid_base10_number_string : invalid_base10_number_strings)
EXPECT_EQ(Crypto::UnsignedBigInteger::from_base(10, invalid_base10_number_string).is_error(), true);
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
}
TEST_CASE(test_unsigned_bigint_base10_to_string)
{
LibCrypto: Define *BigInteger::to_base to convert big integers to String
2023-01-13 16:40:04 +00:00
auto bigint = Crypto::UnsignedBigInteger {
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
Vector<u32> { 3806301393, 954919431, 3879607298, 721 }
LibCrypto: Define *BigInteger::to_base to convert big integers to String
2023-01-13 16:40:04 +00:00
};
auto result = MUST(bigint.to_base(10));
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
EXPECT_EQ(result, "57195071295721390579057195715793");
}
TEST_CASE(test_bigint_modular_inverse)
{
auto result = Crypto::NumberTheory::ModularInverse(7, 87);
EXPECT_EQ(result, 25);
}
TEST_CASE(test_bigint_even_simple_modular_power)
{
Crypto::UnsignedBigInteger base { 7 };
Crypto::UnsignedBigInteger exponent { 2 };
Crypto::UnsignedBigInteger modulo { 10 };
auto result = Crypto::NumberTheory::ModularPower(base, exponent, modulo);
EXPECT_EQ(result.words(), Vector<u32> { 9 });
}
TEST_CASE(test_bigint_odd_simple_modular_power)
{
Crypto::UnsignedBigInteger base { 10 };
Crypto::UnsignedBigInteger exponent { 2 };
Crypto::UnsignedBigInteger modulo { 9 };
auto result = Crypto::NumberTheory::ModularPower(base, exponent, modulo);
EXPECT_EQ(result.words(), Vector<u32> { 1 });
}
TEST_CASE(test_bigint_large_even_fibonacci_modular_power)
{
Crypto::UnsignedBigInteger base = bigint_fibonacci(200);
Crypto::UnsignedBigInteger exponent = bigint_fibonacci(100);
Crypto::UnsignedBigInteger modulo = bigint_fibonacci(150);
// Result according to Wolfram Alpha : 7195284628716783672927396027925
auto result = Crypto::NumberTheory::ModularPower(base, exponent, modulo);
Vector<u32> expected_result { 2042093077, 1351416233, 3510104665, 90 };
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_bigint_large_odd_fibonacci_modular_power)
{
Crypto::UnsignedBigInteger base = bigint_fibonacci(200);
Crypto::UnsignedBigInteger exponent = bigint_fibonacci(100);
Crypto::UnsignedBigInteger modulo = bigint_fibonacci(149);
// Result according to Wolfram Alpha : 1136278609611966596838389694992
auto result = Crypto::NumberTheory::ModularPower(base, exponent, modulo);
Vector<u32> expected_result { 2106049040, 2169509253, 1468244710, 14 };
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_bigint_large_odd_fibonacci_with_carry_modular_power)
{
Crypto::UnsignedBigInteger base = bigint_fibonacci(200);
Crypto::UnsignedBigInteger exponent = bigint_fibonacci(100);
Crypto::UnsignedBigInteger modulo = bigint_fibonacci(185);
// Result according to Wolfram Alpha : 55094573983071006678665780782730672080
auto result = Crypto::NumberTheory::ModularPower(base, exponent, modulo);
Vector<u32> expected_result { 1988720592, 2097784252, 347129583, 695391288 };
EXPECT_EQ(result.words(), expected_result);
}
TEST_CASE(test_bigint_modular_power_extra_tests)
{
struct {
Crypto::UnsignedBigInteger base;
Crypto::UnsignedBigInteger exp;
Crypto::UnsignedBigInteger mod;
Crypto::UnsignedBigInteger expected;
} mod_pow_tests[] = {
{ "2988348162058574136915891421498819466320163312926952423791023078876139"_bigint, "2351399303373464486466122544523690094744975233415544072992656881240319"_bigint, "10000"_bigint, "3059"_bigint },
{ "24231"_bigint, "12448"_bigint, "14679"_bigint, "4428"_bigint },
{ "1005404"_bigint, "8352654"_bigint, "8161408"_bigint, "2605696"_bigint },
{ "3665005778"_bigint, "3244425589"_bigint, "565668506"_bigint, "524766494"_bigint },
{ "10662083169959689657"_bigint, "11605678468317533000"_bigint, "1896834583057209739"_bigint, "1292743154593945858"_bigint },
{ "99667739213529524852296932424683448520"_bigint, "123394910770101395416306279070921784207"_bigint, "238026722756504133786938677233768788719"_bigint, "197165477545023317459748215952393063201"_bigint },
{ "49368547511968178788919424448914214709244872098814465088945281575062739912239"_bigint, "25201856190991298572337188495596990852134236115562183449699512394891190792064"_bigint, "45950460777961491021589776911422805972195170308651734432277141467904883064645"_bigint, "39917885806532796066922509794537889114718612292469285403012781055544152450051"_bigint },
{ "48399385336454791246880286907257136254351739111892925951016159217090949616810"_bigint, "5758661760571644379364752528081901787573279669668889744323710906207949658569"_bigint, "32812120644405991429173950312949738783216437173380339653152625840449006970808"_bigint, "7948464125034399875323770213514649646309423451213282653637296324080400293584"_bigint },
};
for (auto test_case : mod_pow_tests) {
auto actual = Crypto::NumberTheory::ModularPower(
test_case.base, test_case.exp, test_case.mod);
EXPECT_EQ(actual, test_case.expected);
}
}
TEST_CASE(test_bigint_primality_test)
{
struct {
Crypto::UnsignedBigInteger candidate;
bool expected_result;
} primality_tests[] = {
{ "1180591620717411303424"_bigint, false }, // 2**70
{ "620448401733239439360000"_bigint, false }, // 25!
{ "953962166440690129601298432"_bigint, false }, // 12**25
{ "620448401733239439360000"_bigint, false }, // 25!
{ "147926426347074375"_bigint, false }, // 35! / 2**32
{ "340282366920938429742726440690708343523"_bigint, false }, // 2 factors near 2^64
{ "73"_bigint, true },
{ "6967"_bigint, true },
{ "787649"_bigint, true },
{ "73513949"_bigint, true },
{ "6691236901"_bigint, true },
{ "741387182759"_bigint, true },
{ "67466615915827"_bigint, true },
{ "9554317039214687"_bigint, true },
{ "533344522150170391"_bigint, true },
{ "18446744073709551557"_bigint, true }, // just below 2**64
};
for (auto test_case : primality_tests) {
bool actual_result = Crypto::NumberTheory::is_probably_prime(test_case.candidate);
EXPECT_EQ(test_case.expected_result, actual_result);
}
}
TEST_CASE(test_bigint_random_number_generation)
{
struct {
Crypto::UnsignedBigInteger min;
Crypto::UnsignedBigInteger max;
} random_number_tests[] = {
{ "1"_bigint, "1000000"_bigint },
{ "10000000000"_bigint, "20000000000"_bigint },
{ "1000"_bigint, "200000000000000000"_bigint },
{ "200000000000000000"_bigint, "200000000000010000"_bigint },
};
for (auto test_case : random_number_tests) {
auto actual_result = Crypto::NumberTheory::random_number(test_case.min, test_case.max);
EXPECT(!(actual_result < test_case.min));
EXPECT(actual_result < test_case.max);
}
}
TEST_CASE(test_bigint_random_distribution)
{
auto actual_result = Crypto::NumberTheory::random_number(
"1"_bigint,
"100000000000000000000000000000"_bigint); // 10**29
if (actual_result < "100000000000000000000"_bigint) { // 10**20
FAIL("Too small");
LibCrypto: Define *BigInteger::to_base to convert big integers to String
2023-01-13 16:40:04 +00:00
outln("The generated number {} is extremely small. This *can* happen by pure chance, but should happen only once in a billion times. So it's probably an error.", MUST(actual_result.to_base(10)));
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
} else if ("99999999900000000000000000000"_bigint < actual_result) { // 10**29 - 10**20
FAIL("Too large");
LibCrypto: Define *BigInteger::to_base to convert big integers to String
2023-01-13 16:40:04 +00:00
outln("The generated number {} is extremely large. This *can* happen by pure chance, but should happen only once in a billion times. So it's probably an error.", MUST(actual_result.to_base(10)));
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
}
}
TEST_CASE(test_bigint_import_big_endian_decode_encode_roundtrip)
{
u8 random_bytes[128];
u8 target_buffer[128];
AK+Everywhere: Change AK::fill_with_random to accept a Bytes object Rather than the very C-like API we currently have, accepting a void* and a length, let's take a Bytes object instead. In almost all existing cases, the compiler figures out the length.
2023-04-02 17:08:43 +00:00
fill_with_random(random_bytes);
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
auto encoded = Crypto::UnsignedBigInteger::import_data(random_bytes, 128);
encoded.export_data({ target_buffer, 128 });
EXPECT(memcmp(target_buffer, random_bytes, 128) == 0);
}
TEST_CASE(test_bigint_import_big_endian_encode_decode_roundtrip)
{
u8 target_buffer[128];
auto encoded = "12345678901234567890"_bigint;
auto size = encoded.export_data({ target_buffer, 128 });
auto decoded = Crypto::UnsignedBigInteger::import_data(target_buffer, size);
EXPECT_EQ(encoded, decoded);
}
TEST_CASE(test_bigint_big_endian_import)
{
Everywhere: Add sv suffix to strings relying on StringView(char const*) Each of these strings would previously rely on StringView's char const* constructor overload, which would call __builtin_strlen on the string. Since we now have operator ""sv, we can replace these with much simpler versions. This opens the door to being able to remove StringView(char const*). No functional changes.
2022-07-11 17:32:29 +00:00
auto number = Crypto::UnsignedBigInteger::import_data("hello"sv);
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
EXPECT_EQ(number, "448378203247"_bigint);
}
TEST_CASE(test_bigint_big_endian_export)
{
auto number = "448378203247"_bigint;
char exported[8] { 0 };
auto exported_length = number.export_data({ exported, 8 }, true);
EXPECT_EQ(exported_length, 5u);
EXPECT(memcmp(exported + 3, "hello", 5) == 0);
}
LibCrypto+LibJS: Better bigint bitwise_or binop Similar to the bitwise_and change, but we have to be careful to sign-extend two's complement numbers only up to the highest set bit in the positive number.
2022-01-18 13:46:42 +00:00
TEST_CASE(test_bigint_one_based_index_of_highest_set_bit)
{
auto num1 = "1234567"_bigint;
auto num2 = "1234567"_bigint;
EXPECT_EQ("0"_bigint.one_based_index_of_highest_set_bit(), 0u);
EXPECT_EQ("1"_bigint.one_based_index_of_highest_set_bit(), 1u);
EXPECT_EQ("7"_bigint.one_based_index_of_highest_set_bit(), 3u);
EXPECT_EQ("4294967296"_bigint.one_based_index_of_highest_set_bit(), 33u);
}
TEST_CASE(test_signed_bigint_bitwise_not_fill_to_one_based_index)
{
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(0), "0"_bigint);
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(1), "1"_bigint);
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(2), "3"_bigint);
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(4), "15"_bigint);
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(32), "4294967295"_bigint);
EXPECT_EQ("0"_bigint.bitwise_not_fill_to_one_based_index(33), "8589934591"_bigint);
}
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
TEST_CASE(test_bigint_bitwise_or)
{
auto num1 = "1234567"_bigint;
auto num2 = "1234567"_bigint;
EXPECT_EQ(num1.bitwise_or(num2), num1);
}
TEST_CASE(test_bigint_bitwise_or_different_lengths)
{
auto num1 = "1234567"_bigint;
auto num2 = "123456789012345678901234567890"_bigint;
auto expected = "123456789012345678901234622167"_bigint;
auto result = num1.bitwise_or(num2);
EXPECT_EQ(result, expected);
}
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
TEST_CASE(test_signed_bigint_bitwise_or)
{
auto num1 = "-1234567"_sbigint;
auto num2 = "1234567"_sbigint;
EXPECT_EQ(num1.bitwise_or(num1), num1);
LibCrypto+LibJS: Better bigint bitwise_or binop Similar to the bitwise_and change, but we have to be careful to sign-extend two's complement numbers only up to the highest set bit in the positive number.
2022-01-18 13:46:42 +00:00
EXPECT_EQ(num1.bitwise_or(num2), "-1"_sbigint);
EXPECT_EQ(num2.bitwise_or(num1), "-1"_sbigint);
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
EXPECT_EQ(num2.bitwise_or(num2), num2);
LibCrypto+LibJS: Better bigint bitwise_or binop Similar to the bitwise_and change, but we have to be careful to sign-extend two's complement numbers only up to the highest set bit in the positive number.
2022-01-18 13:46:42 +00:00
EXPECT_EQ("0"_sbigint.bitwise_or("-1"_sbigint), "-1"_sbigint);
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
}
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
TEST_CASE(test_bigint_bitwise_and)
{
auto num1 = "1234567"_bigint;
auto num2 = "1234561"_bigint;
EXPECT_EQ(num1.bitwise_and(num2), "1234561"_bigint);
}
TEST_CASE(test_bigint_bitwise_and_different_lengths)
{
auto num1 = "1234567"_bigint;
auto num2 = "123456789012345678901234567890"_bigint;
EXPECT_EQ(num1.bitwise_and(num2), "1180290"_bigint);
}
LibJS+LibCrypto: Fix SignedBitInteger::bitwise_not and use it in LibJS Bitwise operators are defined on two's complement, but SignedBitInteger uses sign-magnitude. Correctly convert between the two. Let LibJS delegate to SignedBitInteger for bitwise_not, like it does for all other bitwise_ operations on bigints. No behavior change (LibJS is now the only client of SignedBitInteger::bitwise_not()).
2022-01-17 17:22:51 +00:00
TEST_CASE(test_signed_bigint_bitwise_not)
{
EXPECT_EQ("3"_sbigint.bitwise_not(), "-4"_sbigint);
EXPECT_EQ("-1"_sbigint.bitwise_not(), "0"_sbigint);
}
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
TEST_CASE(test_signed_bigint_bitwise_and)
{
auto num1 = "-1234567"_sbigint;
auto num2 = "1234567"_sbigint;
EXPECT_EQ(num1.bitwise_and(num1), num1);
LibCrypto+LibJS: Better bigint bitwise_and binop Bitwise and is defined in terms of two's complement, so some converting needs to happen for SignedBigInteger's sign/magnitude representation to work out. UnsignedBigInteger::bitwise_not() is repurposed to convert all high-order zero bits to ones up to a limit, for the two's complement conversion to work. Fixes test262/test/language/expressions/bitwise-and/bigint.js.
2022-01-18 00:54:02 +00:00
EXPECT_EQ(num1.bitwise_and(num2), "1"_sbigint);
EXPECT_EQ(num2.bitwise_and(num1), "1"_sbigint);
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
EXPECT_EQ(num2.bitwise_and(num2), num2);
LibCrypto+LibJS: Better bigint bitwise_and binop Bitwise and is defined in terms of two's complement, so some converting needs to happen for SignedBigInteger's sign/magnitude representation to work out. UnsignedBigInteger::bitwise_not() is repurposed to convert all high-order zero bits to ones up to a limit, for the two's complement conversion to work. Fixes test262/test/language/expressions/bitwise-and/bigint.js.
2022-01-18 00:54:02 +00:00
EXPECT_EQ("-3"_sbigint.bitwise_and("-2"_sbigint), "-4"_sbigint);
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
}
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
TEST_CASE(test_bigint_bitwise_xor)
{
auto num1 = "1234567"_bigint;
auto num2 = "1234561"_bigint;
EXPECT_EQ(num1.bitwise_xor(num2), 6);
}
TEST_CASE(test_bigint_bitwise_xor_different_lengths)
{
auto num1 = "1234567"_bigint;
auto num2 = "123456789012345678901234567890"_bigint;
EXPECT_EQ(num1.bitwise_xor(num2), "123456789012345678901233441877"_bigint);
}
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
TEST_CASE(test_signed_bigint_bitwise_xor)
{
auto num1 = "-3"_sbigint;
auto num2 = "1"_sbigint;
EXPECT_EQ(num1.bitwise_xor(num1), "0"_sbigint);
LibCrypto+LibJS: Better bitwise binary_xor binop We went through some trouble to make & and | work right. Reimplement ^ in terms of & and | to make ^ work right as well. This is less fast than a direct implementation, but let's get things working first.
2022-01-18 01:03:51 +00:00
EXPECT_EQ(num1.bitwise_xor(num2), "-4"_sbigint);
EXPECT_EQ(num2.bitwise_xor(num1), "-4"_sbigint);
LibCrypto: Add tests for SignedBigInteger bitwise operations
2021-06-30 22:38:42 +00:00
EXPECT_EQ(num2.bitwise_xor(num2), "0"_sbigint);
}
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
TEST_CASE(test_signed_bigint_fibo500)
{
Vector<u32> expected_result {
315178285, 505575602, 1883328078, 125027121,
3649625763, 347570207, 74535262, 3832543808,
2472133297, 1600064941, 65273441
};
auto result = bigint_signed_fibonacci(500);
EXPECT_EQ(result.unsigned_value().words(), expected_result);
}
TEST_CASE(test_signed_addition_edgecase_borrow_with_zero)
{
Crypto::SignedBigInteger num1 { Crypto::UnsignedBigInteger { { UINT32_MAX - 3, UINT32_MAX } }, false };
Crypto::SignedBigInteger num2 { Crypto::UnsignedBigInteger { UINT32_MAX - 2 }, false };
Vector<u32> expected_result { 4294967289, 0, 1 };
EXPECT_EQ(num1.plus(num2).unsigned_value().words(), expected_result);
}
TEST_CASE(test_signed_addition_edgecase_addition_to_other_sign)
{
Crypto::SignedBigInteger num1 = INT32_MAX;
Crypto::SignedBigInteger num2 = num1;
num2.negate();
EXPECT_EQ(num1.plus(num2), Crypto::SignedBigInteger { 0 });
}
TEST_CASE(test_signed_subtraction_simple_subtraction_positive_result)
{
Crypto::SignedBigInteger num1(80);
Crypto::SignedBigInteger num2(70);
EXPECT_EQ(num1.minus(num2), Crypto::SignedBigInteger(10));
}
TEST_CASE(test_signed_subtraction_simple_subtraction_negative_result)
{
Crypto::SignedBigInteger num1(50);
Crypto::SignedBigInteger num2(70);
EXPECT_EQ(num1.minus(num2), Crypto::SignedBigInteger { -20 });
}
LibCrypto: Fix subtracting two negative `SignedBigInteger`s Currently, we get the following results -1 - -2 = -1 -2 - -1 = 1 Correct would be: -1 - -2 = 1 -2 - -1 = -1 This was already attempted to be fixed in 7ed8970, but that change was incorrect. This directly translates to LibJS BigInts having the same incorrect behavior - it even was tested.
2021-11-16 09:17:55 +00:00
TEST_CASE(test_signed_subtraction_both_negative)
{
Crypto::SignedBigInteger num1(-50);
Crypto::SignedBigInteger num2(-70);
EXPECT_EQ(num1.minus(num2), Crypto::SignedBigInteger { 20 });
EXPECT_EQ(num2.minus(num1), Crypto::SignedBigInteger { -20 });
}
LibCrypto+LibTLS: Split and move test suite into Tests directory This change splits test-crypto.cpp from Userland into separate test suites located in Tests/ directory.
2021-06-18 21:28:49 +00:00
TEST_CASE(test_signed_subtraction_simple_subtraction_with_borrow)
{
Crypto::SignedBigInteger num1(Crypto::UnsignedBigInteger { UINT32_MAX });
Crypto::SignedBigInteger num2(1);
Crypto::SignedBigInteger num3 = num1.plus(num2);
Crypto::SignedBigInteger result = num2.minus(num3);
num1.negate();
EXPECT_EQ(result, num1);
}
TEST_CASE(test_signed_subtraction_with_large_numbers)
{
Crypto::SignedBigInteger num1 = bigint_signed_fibonacci(343);
Crypto::SignedBigInteger num2 = bigint_signed_fibonacci(218);
Crypto::SignedBigInteger result = num2.minus(num1);
auto expected = Crypto::UnsignedBigInteger { Vector<u32> { 811430588, 2958904896, 1130908877, 2830569969, 3243275482, 3047460725, 774025231, 7990 } };
EXPECT_EQ(result.plus(num1), num2);
EXPECT_EQ(result.unsigned_value(), expected);
}
TEST_CASE(test_signed_subtraction_with_large_numbers_check_for_assertion)
{
Crypto::SignedBigInteger num1(Crypto::UnsignedBigInteger { Vector<u32> { 1483061863, 446680044, 1123294122, 191895498, 3347106536, 16, 0, 0, 0 } });
Crypto::SignedBigInteger num2(Crypto::UnsignedBigInteger { Vector<u32> { 4196414175, 1117247942, 1123294122, 191895498, 3347106536, 16 } });
Crypto::SignedBigInteger result = num1.minus(num2);
// this test only verifies that we don't crash on an assertion
}
TEST_CASE(test_signed_multiplication_with_negative_number)
{
Crypto::SignedBigInteger num1(8);
Crypto::SignedBigInteger num2(-251);
Crypto::SignedBigInteger result = num1.multiplied_by(num2);
EXPECT_EQ(result, Crypto::SignedBigInteger { -2008 });
}
TEST_CASE(test_signed_multiplication_with_big_number)
{
Crypto::SignedBigInteger num1 = bigint_signed_fibonacci(200);
Crypto::SignedBigInteger num2(-12345678);
Crypto::SignedBigInteger result = num1.multiplied_by(num2);
Vector<u32> expected_result { 669961318, 143970113, 4028714974, 3164551305, 1589380278, 2 };
EXPECT_EQ(result.unsigned_value().words(), expected_result);
EXPECT(result.is_negative());
}
TEST_CASE(test_signed_multiplication_with_two_big_numbers)
{
Crypto::SignedBigInteger num1 = bigint_signed_fibonacci(200);
Crypto::SignedBigInteger num2 = bigint_signed_fibonacci(341);
num1.negate();
Crypto::SignedBigInteger result = num1.multiplied_by(num2);
Vector<u32> expected_results {
3017415433, 2741793511, 1957755698, 3731653885,
3154681877, 785762127, 3200178098, 4260616581,
529754471, 3632684436, 1073347813, 2516430
};
EXPECT_EQ(result.unsigned_value().words(), expected_results);
EXPECT(result.is_negative());
}
LibCrypto: Do not allow signed big integers to be negative zero If a big integer were to become negative zero, set the sign to instead be positive. This prevents odd scenarios where users of signed big ints would falsely think the result of some big int arithmetic is negative.
2022-02-06 13:17:34 +00:00
TEST_CASE(test_negative_zero_is_not_allowed)
{
Crypto::SignedBigInteger zero(Crypto::UnsignedBigInteger(0), true);
EXPECT(!zero.is_negative());
zero.negate();
EXPECT(!zero.is_negative());
Crypto::SignedBigInteger positive_five(Crypto::UnsignedBigInteger(5), false);
Crypto::SignedBigInteger negative_five(Crypto::UnsignedBigInteger(5), true);
zero = positive_five.plus(negative_five);
EXPECT(zero.unsigned_value().is_zero());
EXPECT(!zero.is_negative());
}
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
TEST_CASE(double_comparisons)
{
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
#define EXPECT_LESS_THAN(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleGreaterThanBigInt)
#define EXPECT_GREATER_THAN(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleLessThanBigInt)
#define EXPECT_EQUAL_TO(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleEqualsBigInt)
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
Crypto::SignedBigInteger zero { 0 };
EXPECT_EQUAL_TO(zero, 0.0);
EXPECT_EQUAL_TO(zero, -0.0);
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
Crypto::SignedBigInteger one { 1 };
EXPECT_EQUAL_TO(one, 1.0);
EXPECT_GREATER_THAN(one, -1.0);
EXPECT_GREATER_THAN(one, 0.5);
EXPECT_GREATER_THAN(one, -0.5);
EXPECT_LESS_THAN(one, 1.000001);
one.negate();
auto const& negative_one = one;
EXPECT_EQUAL_TO(negative_one, -1.0);
EXPECT_LESS_THAN(negative_one, 1.0);
EXPECT_LESS_THAN(one, 0.5);
EXPECT_LESS_THAN(one, -0.5);
EXPECT_GREATER_THAN(one, -1.5);
EXPECT_LESS_THAN(one, 1.000001);
EXPECT_GREATER_THAN(one, -1.000001);
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
double double_infinity = HUGE_VAL;
VERIFY(isinf(double_infinity));
Crypto::SignedBigInteger one { 1 };
EXPECT_LESS_THAN(one, double_infinity);
EXPECT_GREATER_THAN(one, -double_infinity);
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
double double_max_value = NumericLimits<double>::max();
double double_below_max_value = nextafter(double_max_value, 0.0);
VERIFY(double_below_max_value < double_max_value);
VERIFY(double_below_max_value < (double_max_value - 1.0));
auto max_value_in_bigint = TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "fffffffffffff800000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
auto max_value_plus_one = max_value_in_bigint.plus(Crypto::SignedBigInteger { 1 });
auto max_value_minus_one = max_value_in_bigint.minus(Crypto::SignedBigInteger { 1 });
auto below_max_value_in_bigint = TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "fffffffffffff000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
EXPECT_EQUAL_TO(max_value_in_bigint, double_max_value);
EXPECT_LESS_THAN(max_value_minus_one, double_max_value);
EXPECT_GREATER_THAN(max_value_plus_one, double_max_value);
EXPECT_LESS_THAN(below_max_value_in_bigint, double_max_value);
EXPECT_GREATER_THAN(max_value_in_bigint, double_below_max_value);
EXPECT_GREATER_THAN(max_value_minus_one, double_below_max_value);
EXPECT_GREATER_THAN(max_value_plus_one, double_below_max_value);
EXPECT_EQUAL_TO(below_max_value_in_bigint, double_below_max_value);
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
double double_min_value = NumericLimits<double>::lowest();
double double_above_min_value = nextafter(double_min_value, 0.0);
VERIFY(double_above_min_value > double_min_value);
VERIFY(double_above_min_value > (double_min_value + 1.0));
auto min_value_in_bigint = TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-fffffffffffff800000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
auto min_value_plus_one = min_value_in_bigint.plus(Crypto::SignedBigInteger { 1 });
auto min_value_minus_one = min_value_in_bigint.minus(Crypto::SignedBigInteger { 1 });
auto above_min_value_in_bigint = TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-fffffffffffff000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
EXPECT_EQUAL_TO(min_value_in_bigint, double_min_value);
EXPECT_LESS_THAN(min_value_minus_one, double_min_value);
EXPECT_GREATER_THAN(min_value_plus_one, double_min_value);
EXPECT_GREATER_THAN(above_min_value_in_bigint, double_min_value);
EXPECT_LESS_THAN(min_value_in_bigint, double_above_min_value);
EXPECT_LESS_THAN(min_value_minus_one, double_above_min_value);
EXPECT_LESS_THAN(min_value_plus_one, double_above_min_value);
EXPECT_EQUAL_TO(above_min_value_in_bigint, double_above_min_value);
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
{
double just_above_255 = bit_cast<double>(0x406fe00000000001ULL);
double just_below_255 = bit_cast<double>(0x406fdfffffffffffULL);
double double_255 = 255.0;
Crypto::SignedBigInteger bigint_255 { 255 };
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
Everywhere: Run clang-format The following command was used to clang-format these files: clang-format-18 -i $(find . \ -not \( -path "./\.*" -prune \) \ -not \( -path "./Base/*" -prune \) \ -not \( -path "./Build/*" -prune \) \ -not \( -path "./Toolchain/*" -prune \) \ -not \( -path "./Ports/*" -prune \) \ -type f -name "*.cpp" -o -name "*.mm" -o -name "*.h") There are a couple of weird cases where clang-format now thinks that a pointer access in an initializer list, e.g. `m_member(ptr->foo)`, is a lambda return statement, and it puts spaces around the `->`.
2024-04-24 10:53:44 +00:00
EXPECT_EQUAL_TO(bigint_255, double_255);
EXPECT_GREATER_THAN(bigint_255, just_below_255);
EXPECT_LESS_THAN(bigint_255, just_above_255);
}
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
#undef EXPECT_LESS_THAN
#undef EXPECT_GREATER_THAN
#undef EXPECT_EQUAL_TO
}
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
TEST_CASE(to_double)
{
#define EXPECT_TO_EQUAL_DOUBLE(bigint, double_value) \
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
EXPECT_EQ((bigint).to_double(Crypto::UnsignedBigInteger::RoundingMode::RoundTowardZero), double_value)
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
EXPECT_TO_EQUAL_DOUBLE(Crypto::UnsignedBigInteger(0), 0.0);
// Make sure we don't get negative zero!
EXPECT_EQ(signbit(Crypto::UnsignedBigInteger(0).to_double()), 0);
{
Crypto::SignedBigInteger zero { 0 };
EXPECT(!zero.is_negative());
EXPECT_TO_EQUAL_DOUBLE(zero, 0.0);
EXPECT_EQ(signbit(zero.to_double()), 0);
zero.negate();
EXPECT(!zero.is_negative());
EXPECT_TO_EQUAL_DOUBLE(zero, 0.0);
EXPECT_EQ(signbit(zero.to_double()), 0);
}
EXPECT_TO_EQUAL_DOUBLE(Crypto::UnsignedBigInteger(9682), 9682.0);
EXPECT_TO_EQUAL_DOUBLE(Crypto::SignedBigInteger(-9660), -9660.0);
double double_max_value = NumericLimits<double>::max();
double infinity = INFINITY;
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "fffffffffffff800000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "ffffffffffffff00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "10000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
infinity);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-fffffffffffff800000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
-double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-ffffffffffffff00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
-double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
-double_max_value);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(16, "-10000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
-infinity);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "ffffffffffffffff"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
18446744073709549568.0);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "fffffffffffff800"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
18446744073709549568.0);
EXPECT_TO_EQUAL_DOUBLE(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "fffffffffffff8ff"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
18446744073709549568.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_TO_EQUAL_DOUBLE(TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(10, "1234567890123456789"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
1234567890123456800.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_TO_EQUAL_DOUBLE(TRY_OR_FAIL(Crypto::SignedBigInteger::from_base(10, "2345678901234567890"sv)),
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
2345678901234567680.0);
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
EXPECT_EQ(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "1fffffffffffff00"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
2305843009213693696.0);
EXPECT_EQ(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "1fffffffffffff00"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::RoundTowardZero),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
2305843009213693696.0);
EXPECT_EQ(
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "1fffffffffffff80"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
2305843009213693952.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000001"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740992.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000002"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740994.0);
// 2^53 = 20000000000000, +3 Rounds up because of tiesRoundToEven
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000003"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740996.0);
// +4 is exactly 9007199254740996
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000004"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740996.0);
// +5 rounds down because of tiesRoundToEven
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000005"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740996.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "20000000000006"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
9007199254740998.0);
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
EXPECT_EQ(TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(10, "98382635059784269824"sv)).to_double(Crypto::UnsignedBigInteger::RoundingMode::IEEERoundAndTiesToEvenMantissa),
LibCrypto: Add a rounding mode to UnsignedBigInteger::to_double This allows using different options for rounding, like IEEE roundTiesToEven, which is the mode that JS requires. Also fix that the last word read from the bigint for the mantissa could be shifted incorrectly leading to incorrect results.
2022-08-25 15:56:34 +00:00
bit_cast<double>(0x4415555555555555ULL));
LibCrypto: Implement a (mostly) proper to_double for UnsignedBigInteger SignedBigInteger can immediately use this by just negating the double if the sign bit is set. For simple cases (below 2^53) we can just convert via an u64, however above that we need to extract the top 53 bits and use those as the mantissa. This function currently does not behave exactly as the JS spec specifies however it is much less naive than the previous implementation.
2022-08-24 08:13:16 +00:00
#undef EXPECT_TO_EQUAL_DOUBLE
}
LibCrypto: Add a constructor to (Un)SignedBigInteger taking a double For now this will assume that the double given is exactly representable as an integer, so no NaN, infinity or rounding.
2022-08-25 21:37:09 +00:00
TEST_CASE(bigint_from_double)
{
{
Crypto::UnsignedBigInteger from_zero { 0.0 };
EXPECT(from_zero.is_zero());
EXPECT(!from_zero.is_invalid());
}
#define SURVIVES_ROUND_TRIP_UNSIGNED(double_value) \
{ \
Crypto::UnsignedBigInteger bigint { (double_value) }; \
EXPECT_EQ(bigint.to_double(), (double_value)); \
}
SURVIVES_ROUND_TRIP_UNSIGNED(0.0);
SURVIVES_ROUND_TRIP_UNSIGNED(1.0);
SURVIVES_ROUND_TRIP_UNSIGNED(100000.0);
SURVIVES_ROUND_TRIP_UNSIGNED(1000000000000.0);
SURVIVES_ROUND_TRIP_UNSIGNED(10000000000000000000.0);
SURVIVES_ROUND_TRIP_UNSIGNED(NumericLimits<double>::max());
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x4340000000000002ULL));
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x4340000000000001ULL));
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x4340000000000000ULL));
// Failed on last bits of mantissa
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x7EDFFFFFFFFFFFFFULL));
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x7ed5555555555555ULL));
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x7EDCBA9876543210ULL));
// Has exactly exponent of 32
SURVIVES_ROUND_TRIP_UNSIGNED(bit_cast<double>(0x41f22f74e0000000ULL));
#define SURVIVES_ROUND_TRIP_SIGNED(double_value) \
{ \
Crypto::SignedBigInteger bigint_positive { (double_value) }; \
EXPECT_EQ(bigint_positive.to_double(), (double_value)); \
Crypto::SignedBigInteger bigint_negative { -(double_value) }; \
EXPECT_EQ(bigint_negative.to_double(), -(double_value)); \
EXPECT(bigint_positive != bigint_negative); \
bigint_positive.negate(); \
EXPECT(bigint_positive == bigint_negative); \
}
{
// Negative zero should be converted to positive zero
double const negative_zero = bit_cast<double>(0x8000000000000000);
// However it should give a bit exact +0.0
Crypto::SignedBigInteger from_negative_zero { negative_zero };
EXPECT(from_negative_zero.is_zero());
EXPECT(!from_negative_zero.is_negative());
double result = from_negative_zero.to_double();
EXPECT_EQ(result, 0.0);
EXPECT_EQ(bit_cast<u64>(result), 0ULL);
}
SURVIVES_ROUND_TRIP_SIGNED(1.0);
SURVIVES_ROUND_TRIP_SIGNED(100000.0);
SURVIVES_ROUND_TRIP_SIGNED(-1000000000000.0);
SURVIVES_ROUND_TRIP_SIGNED(10000000000000000000.0);
SURVIVES_ROUND_TRIP_SIGNED(NumericLimits<double>::max());
SURVIVES_ROUND_TRIP_SIGNED(NumericLimits<double>::lowest());
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x4340000000000002ULL));
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x4340000000000001ULL));
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x4340000000000000ULL));
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x7EDFFFFFFFFFFFFFULL));
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x7ed5555555555555ULL));
SURVIVES_ROUND_TRIP_SIGNED(bit_cast<double>(0x7EDCBA9876543210ULL));
#undef SURVIVES_ROUND_TRIP_SIGNED
#undef SURVIVES_ROUND_TRIP_UNSIGNED
}
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
TEST_CASE(unsigned_bigint_double_comparisons)
{
#define EXPECT_LESS_THAN(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleGreaterThanBigInt)
#define EXPECT_GREATER_THAN(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleLessThanBigInt)
#define EXPECT_EQUAL_TO(bigint, double_value) EXPECT_EQ(bigint.compare_to_double(double_value), Crypto::UnsignedBigInteger::CompareResult::DoubleEqualsBigInt)
{
Crypto::UnsignedBigInteger zero { 0 };
EXPECT_EQUAL_TO(zero, 0.0);
EXPECT_EQUAL_TO(zero, -0.0);
}
{
Crypto::UnsignedBigInteger one { 1 };
EXPECT_EQUAL_TO(one, 1.0);
EXPECT_GREATER_THAN(one, -1.0);
EXPECT_GREATER_THAN(one, 0.5);
EXPECT_GREATER_THAN(one, -0.5);
EXPECT_LESS_THAN(one, 1.000001);
}
{
double double_infinity = HUGE_VAL;
VERIFY(isinf(double_infinity));
Crypto::UnsignedBigInteger one { 1 };
EXPECT_LESS_THAN(one, double_infinity);
EXPECT_GREATER_THAN(one, -double_infinity);
}
{
double double_max_value = NumericLimits<double>::max();
double double_below_max_value = nextafter(double_max_value, 0.0);
VERIFY(double_below_max_value < double_max_value);
VERIFY(double_below_max_value < (double_max_value - 1.0));
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
auto max_value_in_bigint = TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "fffffffffffff800000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
auto max_value_plus_one = max_value_in_bigint.plus(Crypto::UnsignedBigInteger { 1 });
auto max_value_minus_one = max_value_in_bigint.minus(Crypto::UnsignedBigInteger { 1 });
LibCrypto: Make constructing a BigInteger from string fallible Previously, constructing a `UnsignedBigInteger::from_base()` could produce an incorrect result if the input string contained a valid Base36 digit that was out of range of the given base. The same method would also crash if the input string contained an invalid Base36 digit. An error is now returned in both these cases. Constructing a BigFraction from string is now also fallible, so that we can handle the case where we are given an input string with invalid digits.
2024-01-12 21:34:23 +00:00
auto below_max_value_in_bigint = TRY_OR_FAIL(Crypto::UnsignedBigInteger::from_base(16, "fffffffffffff000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"sv));
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
EXPECT_EQUAL_TO(max_value_in_bigint, double_max_value);
EXPECT_LESS_THAN(max_value_minus_one, double_max_value);
EXPECT_GREATER_THAN(max_value_plus_one, double_max_value);
EXPECT_LESS_THAN(below_max_value_in_bigint, double_max_value);
EXPECT_GREATER_THAN(max_value_in_bigint, double_below_max_value);
EXPECT_GREATER_THAN(max_value_minus_one, double_below_max_value);
EXPECT_GREATER_THAN(max_value_plus_one, double_below_max_value);
EXPECT_EQUAL_TO(below_max_value_in_bigint, double_below_max_value);
}
{
double just_above_255 = bit_cast<double>(0x406fe00000000001ULL);
double just_below_255 = bit_cast<double>(0x406fdfffffffffffULL);
double double_255 = 255.0;
Crypto::UnsignedBigInteger bigint_255 { 255 };
EXPECT_EQUAL_TO(bigint_255, double_255);
EXPECT_GREATER_THAN(bigint_255, just_below_255);
EXPECT_LESS_THAN(bigint_255, just_above_255);
}
#undef EXPECT_LESS_THAN
#undef EXPECT_GREATER_THAN
#undef EXPECT_EQUAL_TO
}
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
namespace AK {
template<>
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
struct Formatter<Crypto::UnsignedBigInteger::CompareResult> : Formatter<StringView> {
ErrorOr<void> format(FormatBuilder& builder, Crypto::UnsignedBigInteger::CompareResult const& compare_result)
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
{
switch (compare_result) {
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
case Crypto::UnsignedBigInteger::CompareResult::DoubleEqualsBigInt:
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
return builder.put_string("Equals"sv);
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
case Crypto::UnsignedBigInteger::CompareResult::DoubleLessThanBigInt:
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
return builder.put_string("LessThan"sv);
LibCrypto: Add a way to compare UnsignedBigInteger with double This patch also make SignedBigInteger::compare_to_double make use of the new function.
2022-10-23 15:46:35 +00:00
case Crypto::UnsignedBigInteger::CompareResult::DoubleGreaterThanBigInt:
LibCrypto: Add a way to compare a SignedBigInteger with a double This supports any double value (except for NaNs) instead of having to cast the double to some smaller type which doesn't work for very large values.
2022-08-19 21:45:36 +00:00
return builder.put_string("GreaterThan"sv);
default:
return builder.put_string("???"sv);
}
}
};
}
Reference in a new issue Copy permalink