On GitHub, idealvin/coost has picked up 4.2k stars, C++. A lightweight C++ base library with Go-style coroutines.
Snapshot summary built from the project's own GitHub metadata — there's no written TopGit review yet. The page will update automatically when a full review is published.
WHY NO REVIEW YET
TopGit writes full reviews for the most-starred, most-requested repositories. This page is a snapshot until then — see the READ ME tab for the original README in full.
coost is a lightweight, high-performance, easy-to-use C++ base library. It supports Linux, macOS, Windows, FreeBSD, and CPU architectures such as x86, x64, ARM, and ARM64.
coost, abbreviated as co, includes go-style coroutines, networking, logging, config parsing, unit testing, benchmarking, memory allocator, and other components. Its goal is to make C++ programming simple and enjoyable.
Commercial Support
coost remains open source. If you use coost in production and need in-depth support such as custom development, architecture adaptation (RISC-V / MIPS), coroutine hooks, performance optimization, or team training, the author offers paid commercial support to reduce risk and save time.
For service packages, pricing, and process, see:
👉 coost Commercial Support
Contact the author via GitHub Issues or email [email protected]. A free initial diagnosis is available to assess scope and feasibility.
Documentation
English
简体中文
Core Components
flag
flag is a command-line argument and config file parsing library. It supports flag aliases, automatic config file generation, and more.
coost provides a coroutine-based network programming framework:
Socket API. It is similar in form to the system socket API. Users familiar with socket programming can easily write high-performance network programs in a synchronous style.
TCP. Simple encapsulation of TCP server and client, compatible with IPv6.
RPC. A simple RPC framework using JSON for serialization.
Logging
log is a high-performance logging component. It supports printing stack traces when the program crashes.
#include "co/log.h"
int main(int argc, char** argv) {
flag::parse(argc, argv);
log::info("hello ", 23); // info
log::warn("hello ", 23); // warning
log::error("hello ", 23); // error
log::check(1+1==2, "xx"); // runtime assertion; prints stack trace and exits on failure
return 0;
}
log is very fast. Here are some test results:
platform
glog
co/log
speedup
win2012 HDD
1.6MB/s
180MB/s
112.5
win10 SSD
3.7MB/s
560MB/s
151.3
mac SSD
17MB/s
450MB/s
26.4
linux SSD
54MB/s
1023MB/s
18.9
The table above shows the write speed comparison between co/log and glog when continuously printing 1 million logs in a single thread. co/log is nearly two orders of magnitude faster than glog.
threads
linux co/log
linux spdlog
win co/log
win spdlog
speedup
1
0.087235
2.076172
0.117704
0.461156
23.8/3.9
2
0.183160
3.729386
0.158122
0.511769
20.3/3.2
4
0.206712
4.764238
0.316607
0.743227
23.0/2.3
8
0.302088
3.963644
0.406025
1.417387
13.1/3.5
The table above shows the time consumed to print 1 million logs with 1, 2, 4, and 8 threads respectively, in seconds. speedup is the performance improvement factor of co/log over spdlog on Linux and Windows.
Unit Testing
unitest is a simple and easy-to-use unit testing framework. Many components in coost use it to write unit test code, providing important assurance for coost's stability.
The above is a simple example. The DEF_test macro defines a test unit, which is actually a function. The DEF_case macro defines a test case, and each test case is actually a code block.
The coost/unitest directory contains the unit test code in coost. Run the following commands to build and run:
xmake b unitest
xmake r unitest # run all unit test cases
xmake r unitest -os # run only test cases in the os unit; os is the unit test name
Benchmark
benchmark is a simple and easy-to-use performance benchmark framework.
Below is the performance comparison between co/json and rapidjson:
os
co/json stringify
co/json parse
rapidjson stringify
rapidjson parse
speedup
win
569
924
2089
2495
3.6/2.7
mac
783
1097
1289
1658
1.6/1.5
linux
468
764
1359
1070
2.9/1.4
The table above shows the average time for stringify and parse measured after minifying twitter.json, in microseconds (us). speedup is the performance improvement factor of co/json over rapidjson in stringify and parse.
Code Structure
include
Header files of coost.
src
Source code of coost, which compiles into libco.
benchmark
Performance benchmark code. Each .cc file corresponds to a different test unit, and all code is compiled into a single test program.
test
Test code. Each .cc file is compiled into a separate test program.
unitest
Unit test code. Each .cc file corresponds to a different test unit, and all code is compiled into a single test program.
gen
Code generation tools.
Build
Compiler Requirements
The latest version of coost requires a compiler that supports C++17:
Linux: gcc
Mac: clang
Windows: MSVC
Build with xmake
coost recommends xmake as the build tool.
Quick Start
# All commands are executed in the coost root directory, and this will not be repeated below
xmake # build libco by default
xmake -a # build all projects (libco, benchmark, gen, test, unitest)
Enable backtrace
On Linux and macOS, printing stack traces on crashes requires libbacktrace. Newer versions of gcc on Linux already have a built-in backtrace library; on macOS it usually needs to be installed manually.
xmake f --with_backtrace=true
xmake b stack # test/stack.cc
xmake r stack # run stack test program
Install libco
xmake install -o pkg # package and install to the pkg directory
xmake i -o pkg # same as above
xmake install -o /usr/local # install to /usr/local
Build with cmake
Build libco
mkdir cmakebuild && cd cmakebuild
cmake ..
make -j8
Build all projects
mkdir cmakebuild && cd cmakebuild
cmake .. -DBUILD_ALL=ON -DCMAKE_INSTALL_PREFIX=/usr/local
make -j8
cd bin
./unitest # run unit test program
Enable backtrace
mkdir cmakebuild && cd cmakebuild
cmake .. -DWITH_BACKTRACE=ON
make -j8
License
The MIT license. coost includes code from some other projects, which may use different licenses. See LICENSE.md for details.
Special Thanks
The related code of context is taken from ruki's tbox. ruki also helped improve the xmake build script. Special thanks!
izhengfan provided the cmake build script. Special thanks!
SpaceIm improved the cmake build script and provided find_package support. Special thanks!
Leedehai and daidai21 helped translate coost's Chinese reference documentation into English in the early days. Special thanks!
No homepage URL was recorded for idealvin/coost in TopGit's last sync. The README tab above frequently contains screenshots and demo links, or check the repository description on GitHub.
How active is development on idealvin/coost?
The most recent commit recorded on idealvin/coost was 16 days ago, based on the GitHub push timestamp. The repository has 593 forks — one of the better signals of community interest.
Is idealvin/coost open source?
TopGit's metadata for idealvin/coost does not record a license. Most public repositories on GitHub ARE open source, but the exact terms vary — verify by opening the LICENSE file directly.
What topics is idealvin/coost associated with?
GitHub's repository topics for idealvin/coost: "coroutine", "flag", "goroutine", "log", "memory-allocator", "unit-test". TopGit's editorial category is open-source.
Where do I read more about idealvin/coost?
This TopGit page is a snapshot — the READ ME tab shows the project's own README content (links stripped, images preserved). The GitHub repository at github.com/idealvin/coost is the definitive source.
Read full README in the tab above.
Is coost worth your time?
ChatGPT, Claude and Perplexity can all read this page. Ask one of them what it makes of coost.