Mateusz Pusz fc7eb739e7 fix: complete double_width_int operator suite so it can stand in for __int128 on MSVC (#795)
`double_width_int` was previously missing a number of operators that generic numerical code
expects from a 128-bit integer.  On platforms with native `__int128` this was harmless, but on
MSVC (where `int128_t` / `uint128_t` alias `double_width_int<(u)int64_t>`) the recent
`UsesIntegerScaling` change wired the synthetic dwint into concept checks, and `compare_quantities`
+ magnitude folding into runtime paths, exposing every gap.

This change rounds out the operator set so `double_width_int` truly behaves as an integer type:

* binary `+`, `-`, `*`, `/` between two `double_width_int`s (alongside the existing narrow-rhs
  overloads).
* unary `~` and binary `&`, `|`, `^`.
* compound assignment for arithmetic, bitwise, and shift operations.
* pre/post `++` and `--`.
* `static_cast<long double>` (and `double`/`float`) with sign-preserving conversion that avoids
  catastrophic cancellation on platforms where `long double == double`.
* `std::numeric_limits<double_width_int<T>>` specialization so generic code probing `::max()`,
  `::min()`, `::digits`, `::is_signed`, etc. gets correct answers (needed by
  `checked_int_pow`, `compute_base_power`, `safe_int::operator-`).
* fields `hi_` / `lo_` and the `(hi, lo)` ctor are public so cross-instance inline friend
  operators can access each other without further friend declarations.

While here, also:
* extract `double_width_int` (and its `std::numeric_limits` specialization) into a dedicated
  header `bits/double_width_int.h`; `fixed_point.h` now just includes it and keeps the
  `int128_t` aliases, `min_width_uint_t` / `double_width_int_for_t` / `wide_product_of`
  helpers, and the `fixed_point` class itself.
* rename a local variable `m` in `wide_product_of` to `mid` to avoid shadowing
  `si::unit_symbols::m` (MSVC C4459).
* rewrite the `lo_ > 0 ? -1 : 0` unary-minus body to use `Tl{0} - lo_` instead of `-lo_`,
  silencing MSVC C4146 about unary minus on an unsigned operand.

Tests:
* new `test/static/double_width_int_test.cpp` pins down each new operator at compile time
  (carry/borrow edges, schoolbook multiplication, narrow/wide division paths, bitwise
  identities, numeric_limits values, long-double round-trips).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 23:19:38 +02:00
2026-02-26 17:37:15 +00:00
2026-05-14 11:37:19 +02:00
2026-03-27 22:49:05 +00:00
2025-11-06 23:21:41 +01:00
2024-04-25 19:33:03 +02:00
2025-12-24 15:50:11 +01:00
2020-09-05 13:06:09 +02:00
2026-05-01 13:48:03 +02:00
2026-05-15 09:50:43 +02:00

logo

License C++ Standard

Conan CI CMake CI clang-tidy CI Freestanding CI Formatting CI

Conan Center Conan testing

mp-units The Domain-Correct Quantities and Units Library for C++

🎯 Overview

mp-units is a Modern C++ (C++20 and later) library providing the full spectrum of compiletime safety for domain-specific quantities and units — from dimensional analysis to quantity kind safety — built on the ISO 80000 International System of Quantities (ISQ). It is a candidate for C++29 standardization (P3045) — your chance to shape the future of C++.

#include <mp-units/systems/isq.h>
#include <mp-units/systems/si.h>

using namespace mp_units;
using namespace mp_units::si::unit_symbols;

// Compile-time dimensional analysis — zero runtime overhead
static_assert(1 * km / (1 * s) == 1000 * m / s);

// Function signatures encode domain/physics, not just dimensions
void calculate_trajectory(quantity<isq::kinetic_energy[J]> e);

int main()
{
  quantity<isq::potential_energy[J]> Ep = 42 * J;
  quantity<isq::kinetic_energy[J]>   Ek = 123 * J;
  calculate_trajectory(Ek);         // ✅ correct
  // calculate_trajectory(Ep);      // ❌ potential energy ≠ kinetic energy (both in J)

  // quantity<Gy> q = 42 * Sv;      // ❌ absorbed dose ≠ dose equivalent (both J/kg)
}

Try it live on Compiler Explorer

What Sets mp-units Apart?

Beyond standard dimensional analysis and automatic unit conversions, mp-units provides additional safety levels:

  • Quantity Kind Safetymp-units pioneered this safety level: distinguishes quantities that share the same dimension but represent fundamentally different physical concepts: frequency (Hz) ≠ radioactive activity (Bq), absorbed dose (Gy) ≠ dose equivalent (Sv), plane angle (rad) ≠ solid angle (sr). Dimensional analysis alone cannot catch these errors — mp-units prevents them at compile time.

  • ISO 80000 (ISQ) Support — Built on the International System of Quantities, functions can require specific quantities: isq::height (not just any isq::length), isq::kinetic_energy (not just any isq::energy). The physics of your domain becomes part of the type system.

  • Strongly-Typed Numerics for Any Domain — The quantity framework extends beyond physics: define semantically distinct types for item counts, financial values, identifiers, or any numeric abstraction that should never be silently mixed at compile time.

Key Features

Safety

  • Quantity kind safety: same dimension, different meaning → compile-time error
  • Affine space strong types (quantity and quantity_point)
  • Value-preserving conversions

Performance

  • All dimensional analysis at compile time — zero runtime overhead
  • Performance on par with (sometimes even better than) fundamental types

User Experience

  • Optimized for readable, actionable compilation errors
  • Expressive, composable unit expressions

Feature Rich

  • Systems of Quantities and Units; scalar, vector, and tensor quantities
  • Affine space, natural units, strong angular system
  • Highly adjustable text output formatting

Easy to Extend

  • Custom dimensions, quantities, and units in a single line of code

Low Adoption Cost

  • No external dependencies · macro-free API · C++20 modules-ready · freestanding-capable

Full feature overview

📚 Documentation

Extensive project documentation covers everything from getting started to advanced usage:

Explore the full documentation

🔍 Try It Out

For advanced development or contributions, we provide a fully configured cloud development environment with GitHub Codespaces:

Open in GitHub Codespaces

Alternatives:

  1. Navigate to the repository → "Code""Codespaces""Create codespace on master"
  2. Use the preconfigured devcontainer and Docker image manually in your IDE

For detailed environment documentation, see .devcontainer/README.md.

🚀 Help Shape the Future of C++

mp-units is a candidate for ISO standardization for C++29 — the future of dimensional analysis in C++! The technical case is documented in:

🤝 We are actively seeking organizations and individuals interested in fieldtrialing the library!

Your experience matters. Real-world testimonials demonstrate value to the ISO C++ Committee and help potential adopters decide. Whether you're using mp-units in production, research, or education:

  • Organizations: Share your production deployments and success stories
  • Academics: Report research applications and teaching experiences
  • Developers: Tell us about your innovative use cases and benefits

Share Experience

🤝 Contributors

mp-units is made possible by our community of contributors! 💪

Contributors Commits Stars

🏆 Core Team

🙏 All Contributors

We appreciate every contribution, from code to documentation to community support!

🌟 See our Contributors Page for the complete list and recognition details.

Ready to contribute? Check out our Contributing Guide to get started! 🚀

💝 Support the Project

mp-units is developed as open source with the ambitious goal of C++29 standardization. Your support helps maintain development momentum and accelerate standardization efforts!

Ways to support:

  • Star the repository Show your appreciation and help others discover mp-units

  • 💰 Become a sponsor Financial support enables continued development

    Sponsor

  • 📢 Share your success story Help demonstrate real-world value for standardization and other potential users

  • 🤝 Contribute Code, documentation, feedback, and community support

Languages
C++ 71.4%
CMake 19%
Python 8.3%
C 0.8%
Shell 0.5%