c4core is a library of low-level C utilities, written with low-latency projects in mind.

Some of the utilities provided by c4core have already equivalent functionality in the C standard, but they are provided as the existing C equivalent may be insufficient (eg, std::string_view), inefficient (eg, std::string), heavy (eg streams), or plainly unusable on some platforms/projects, (eg exceptions); some other utilities have equivalent under consideration for C standardisation; and yet some other utilities have (to my knowledge) no equivalent under consideration. Be that as it may, I've been using these utilities in this or similar forms for some years now, and I've found them incredibly useful in my projects. I'm packing these as a separate library, as all of my projects use it.

c4core is extensively unit-tested in Linux, Windows and MacOS. The tests cover x64, x86, arm, wasm (emscripten), aarch64, ppc64le and s390x architectures, and include analysing c4core with:

• valgrind
• clang-tidy
• clang sanitizers:
  • memory
  • address
  • undefined behavior
  • thread
• LGTM.com

c4core also works in bare-metal, as well as in RISC-V, and in LoongArch, but at the moment it's not easy to add automated tests to the CI, so for now these are not in the list of official architectures.

• c4core - C core utilities
  • Obtaining c4core
  • Using c4core in your project
    • CMake
    • Bazel
    • Header-only
    • Package managers
  • Quick tour
    • Writeable string views: c4::substr and c4::csubstr
    • Value <-> character interoperation
    • String formatting and parsing
    • c4::span and c4::blob
    • Enums and enum symbols
    • Bitmasks and bitmask symbols
    • Base64 encoding / decoding
    • Fuzzy float comparison
    • Multi-platform / multi-compiler utilities
    • Runtime assertions and error handling
    • Memory allocation
    • Mass initialization/construction/destruction

c4core uses git submodules. It is best to clone c4core with the --recursive option:

# using --recursive makes sure git submodules are also cloned at the same time
git clone --recursive https://github.com/biojppm/c4core

If you ommit the --recursive option, then after cloning you will have to make git checkout the current version of the submodules, using git submodule init followed by git submodule update.

c4core can be built with cmake, or can be used header only. It can also be obtained through some package managers.

The recommended way to use c4core is by making it part of your project by using add_subdirectory(${path_to_c4core_root}) in your CMakeLists.txt. Doing this is not intrusive to your cmake project because c4core is fast to build, also prefixes every cmake variable with C4CORE_. But more importantly, this will enable you to compile c4core with the exact same compile settings used by your project.

Here's a very quick complete example of setting up your project to use c4core as a cmake subproject:

project(foo)

add_subdirectory(c4core)

add_library(foo foo.cpp)
target_link_libraries(foo PUBLIC c4core) # that's it!

Note above that the call to target_link_libraries() is using PUBLIC linking. This is required to make sure the include directories from c4core are transitively used by clients of foo.

If you prefer to pick a single header to get you quickly going, there is an amalgamation tool which generates this header:

[user@host c4core]$ python tools/amalgamate.py -h
usage: amalgamate.py [-h] [--fastfloat | --no-fastfloat] [--stl | --no-stl] [output]

positional arguments:
  output          output file. defaults to stdout

options:
  -h, --help      show this help message and exit
  --fastfloat     enable fastfloat library. this is the default.
  --no-fastfloat  enable fastfloat library. the default is --fastfloat.
  --stl           enable stl interop. this is the default.
  --no-stl        enable stl interop. the default is --stl.

c4core is available through the following package managers:

• vcpkg: vcpkg install c4core
• Arch Linux/Manjaro:
  • rapidyaml

All of the utilities in this library are under the namespace c4; any exposed macros use the prefix C4_: eg C4_ASSERT().

See partial documentation in rapidyaml's doxygen docs.

Here: #include <c4/substr.hpp>.

There is a useful quickstart sample in rapidyaml's doxygen docs.

Here: #include <c4/charconv.hpp>

// TODO: elaborate on the topics:

c4::digits_dec(), c4::read_dec(), c4::write_dec()
c4::digits_hex(), c4::read_hex(), c4::write_hex()
c4::digits_oct(), c4::read_oct(), c4::write_oct()
c4::digits_bin(), c4::read_bin(), c4::write_bin()

c4::utoa(), c4::atou()
c4::itoa(), c4::atoi()
c4::ftoa(), c4::atof()
c4::dtoa(), c4::atod()
c4::xtoa(), c4::atox()

c4::to_chars(), c4::from_chars()
c4::to_chars_sub()
c4::to_chars_first()

The charconv funcions above are very fast; even faster than C 's fastest facility std::from_chars(), std::to_chars(). For continuous benchmark results, browse through c4core's github CI benchmark runs. For example, a benchmark run on Linux/g 11.2 shows that:

• c4::to_chars() can be expected to be roughly...
  • ~40% to 2x faster than std::to_chars()
  • ~10x-30x faster than sprintf()
  • ~50x-100x faster than a naive stringstream::operator<<() followed by stringstream::str()
• c4::from_chars() can be expected to be roughly...
  • ~10%-30% faster than std::from_chars()
  • ~10x faster than scanf()
  • ~30x-50x faster than a naive stringstream::str() followed by stringstream::operator>>()

Here are the results:

Or here are plots for g 12.1 and VS2019 (from the same computer):

Linux gxx12.1	Windows VS2019

• #include <c4/format.hpp>

• There is a useful quickstart sample in rapidyaml's doxygen docs.

// TODO: elaborate on the topics:

c4::cat(), c4::uncat()
c4::catsep(), c4::uncatsep()
c4::format(), c4::unformat()

c4::catrs()
c4::catseprs()
c4::formatrs()

// formatting:
c4::fmt::overflow_checked
c4::fmt::real
c4::fmt::boolalpha
c4::fmt::dec
c4::fmt::hex
c4::fmt::oct
c4::fmt::bin
c4::fmt::zpad
c4::fmt::right
c4::fmt::left
c4::fmt::raw, c4::fmt::craw
c4::fmt::base64, c4::fmt::cbase64

• #include <c4/span.hpp>
• #include <c4/blob.hpp>

#include <c4/enum.hpp>

// TODO: elaborate on the topics:

c4::e2str(), c4::str2e()

#include <c4/bitmask.hpp>

// TODO: elaborate on the topics:

c4::bm2str(), c4::str2bm()

#include <c4/base64.hpp>

// TODO: elaborate on the topics:
#include <c4/error.hpp>

C4_RESTRICT, $, c$, $$, c$$
#include <c4/restrict.hpp>
#include <c4/unrestrict.hpp>

#include <c4/windows_push.hpp>
#include <c4/windows_pop.hpp>

C4_UNREACHABLE()

c4::type_name()

// portable attributes
C4_LIKELY()/C4_UNLIKELY()
C4_ALWAYS_INLINE
C4_CONST
C4_PURE
C4_HOT
C4_COLD

// TODO: elaborate on the topics:

error callback

C4_ASSERT()
C4_XASSERT()
C4_CHECK()

C4_ERROR()
C4_NOT_IMPLEMENTED()

// TODO: elaborate on the topics:

c4::aalloc(), c4::afree() // aligned allocation

c4::MemoryResource // global and scope

c4::Allocator

// TODO: elaborate on the topics:

c4::make_room()/c4::destroy_room()
c4::construct()/c4::construct_n()
c4::destroy()/c4::destroy_n()
c4::copy_construct()/c4::copy_construct_n()
c4::copy_assign()/c4::copy_assign_n()
c4::move_construct()/c4::move_construct_n()
c4::move_assign()/c4::move_assign_n()