a300614308
test: a few more quantity construction unit tests added
2026-05-21 18:45:22 +02:00
e2e490f0cd
refactor: scaling algorithm and concepts simplified
2026-04-11 19:00:33 +02:00
d8b72fc431
refactor: scaling the value by magnitude heavily refactored
2026-03-19 00:28:15 +01:00
cdb228d5f5
fix: cartesian_vector fixed to work properly as a representation type and enable multiply syntax
2026-03-15 07:04:02 +01:00
6cad11a3b9
test: remove redundant floating-point cross-unit comparison assertion
...
Floating-point comparisons always compile for any unit ratio; the
purpose of this block is to verify integer comparisons are SFINAE'd
out, which is already covered by no_cross_unit_int_comparison.
2026-03-08 22:19:12 +01:00
0eea272ece
fix: overflow-safe cross-unit integer comparisons using double-width arithmetic
...
- Add `comparable_in_wide_type` / `CommonlyComparableQuantities` concept and
`compare_quantities` helper in `quantity.h` that widens int reps to
`double_width_int_for_t<Rep>` (__int128 for int64) before scaling, preventing
silent overflow in cross-unit `operator==` / `operator<=>`
- Add `try_get_value<T>` to `unit_magnitude` returning `std::optional<T>`
(nullopt on overflow) and refactor `get_value<T>` unsigned path to delegate
to it; extract logic into free `try_accumulate_element<T,M>` helper taking M
as NTTP so GCC 12 treats it as a constant expression
- Wire `scaling_overflows_non_zero_values` in `value_cast.h` through
`try_get_value` so extreme ratios (e.g. km³ vs pm³) no longer abort at
compile time when evaluated inside `requires`
- Add positive tests (lb vs kg, km vs mi) exercising the int64→int128 widening
path in `yard_pound_test.cpp`
- Add negative tests using `no_cross_unit_int_comparison` template wrapper
(necessary for soft SFINAE failure on GCC) for cubic(km)/cubic(pm),
Zm/am (linear, factor=10^39>2^127) and cubic(mi)/cubic(pm)
2026-03-08 20:15:35 +01:00
6f3ef66d4b
feat: comparisons against literal 0
...
Resolves #487
2026-02-23 11:07:22 +01:00
04e3d08cb9
fix: dimensionless quantitites truncation prevention fixed
2026-02-12 14:07:56 +01:00
d194af7b97
fix: some stupid MSVC errors workarounds
2026-02-04 23:06:03 +01:00
d74f4acc57
feat: unit_for, reference_for, and rep_for added
2026-01-28 19:28:31 +01:00
916a6c1a16
feat: dimensionless quantities with unit one can be created with quantity_spec::op(Val) explicit conversions
2026-01-27 20:05:58 +01:00
63f55959a2
feat: explicit construction from number enabled for dimensionless subkinds
2026-01-21 18:38:24 +01:00
9a374a5646
refactor: isq::time swapped with isq::duration
2026-01-09 09:22:37 +01:00
85e9a722a9
test: multiply syntax uint tests added
2026-01-07 13:29:50 +01:00
86e3bf5a06
fix: recursive concepts evaluation fixed
...
Resolves #740
2026-01-07 13:26:30 +01:00
cc25814e82
formatting
2025-10-09 16:35:52 -07:00
ef0e653dfe
fix: seconds literal has unspecified type and depends on implementation
2025-10-09 16:35:06 -07:00
9c140f4b3a
feat: explicit constructor now can perform an explicit quantity_spec conversion
2025-06-20 17:10:34 +02:00
f18aa18869
refactor: compund assignment and pre-/post-increment operators are now lvalue-qualified member functions
2025-06-19 18:49:21 +02:00
e9c5f7236e
feat: ConvertibleWithNumber introduced to improve convertibility of unit one with raw numbers
...
Resolves #675
2025-02-11 21:28:15 +01:00
47a82f466c
refactor: scalar and complex renamed to real_scalar and complex_scalar respectively + concepts refactoring + electromagnetism fixes
2025-02-11 17:26:19 +01:00
5d4e1aa740
feat: might_store_converted_common_value added to detect conversions overflowing rep
...
Resolves #670 , resolves #671 , resolves #658 , resolves #303
2025-02-06 15:28:55 +01:00
0b14d69539
feat: quantity::one() removed
...
Related to #408
2025-02-05 12:08:32 +01:00
06cbfae3b3
feat: quantity_spec conversions improved
2024-12-28 14:02:03 +01:00
465f88d500
refactor: 💥 is_XXX customization points for representation types removed
2024-11-26 14:48:08 +01:00
9a0f7a25dd
test: tests for dimensionless quantities static data members
2024-11-23 19:35:59 +01:00
4f8e959b5c
fix(test): shadowing warnings on gcc fixed
2024-11-16 08:28:39 +01:00
08e94171cb
test: vector quantities unit tests addded
2024-11-15 12:59:18 +01:00
5097096915
test: lots of tests for complex quantities
2024-11-07 18:42:20 +01:00
da50348115
feat: 💥 RepresentationOf concept now also accepts a QuantitySpec and accepts any representation character for quantity kinds
2024-11-07 18:39:32 +01:00
75b50b8d2c
feat: complex.h added
2024-11-07 10:58:51 +01:00
6896d8e086
fix: velocity is now defined in terms of displacement instead of position_vector
2024-11-06 18:46:01 +01:00
c7303cc5fb
refactor: representation concepts refactored + some quantities switched to complex
2024-11-05 19:09:16 +01:00
f461cf479b
test: test for lack of consitency between op+ and op+= for quantity_point
2024-11-04 09:41:11 +01:00
8ca8270d2f
feat: from now on truncation is not allowed in compound assignment
2024-10-30 08:40:49 +01:00
18525beb31
test: small refactoring of compund assignment tests
2024-10-29 10:20:11 +01:00
7445585db8
fix: compound assignment operations on quantities now behave the same as on the underying representation types
...
Resolves #137
2024-10-29 09:18:32 +01:00
00847ab558
fix(tests): freestanding build fixed
2024-10-16 17:22:59 +02:00
21784b125e
refactor: derived_quantity refactored to child_quantity
2024-10-15 21:06:52 +02:00
3eecdf3946
test: commutativity tests added to ISQ
2024-10-15 21:06:18 +02:00
569f27af8b
fix: common_unit handling fixed for some corner cases
2024-10-05 17:52:52 +02:00
7ce8d695b1
test: std::complex-based quantities tests added
2024-09-17 12:48:13 -06:00
0670fbdd9f
feat: value_cast<Representation, Unit>() complementary conversion function added
2024-09-13 21:38:59 +02:00
3d39472f46
test: unit tests for creating a quantity from a volatile variable
2024-09-06 18:07:33 +02:00
1570bda905
fix: quantity scaling between different prefixes improved
...
Resolves #608
2024-09-06 12:28:11 +02:00
2e840cfdb4
refactor: Repetitive inline constexpr removed as no longer needed
...
Not needed anymore as stated in cplusplus/draft#4601
2024-09-05 08:43:36 +02:00
3a6fac9858
Move tests to runtime
2024-07-24 20:04:15 -04:00
f6696e8f68
Fix formatting
2024-07-24 19:48:05 -04:00
7e894788d7
Compute values for rational magnitude powers
...
Since this will only ever be done at compile time (as guaranteed by
using `consteval`), we can afford to prioritize precision over speed.
To compute an Nth root, we simply do a binary search over representable
floating point numbers, looking for the number whose Nth power most
closely matches the original number.
Fixes #494 . We have included a test case reproducing the original
problem exactly. All tests use "within 4 ULPs" as the criterion, which
is (I believe) equivalent to the googletest `EXPECT_DOUBLE_EQ`
criterion.
2024-07-24 10:31:44 -04:00
b870b85c25
feat: import std; support added
...
Resolves #595
2024-07-16 17:36:00 +02:00
Mateusz Pusz
Radnyx
Chip Hogg