linuxboost

Building Boost with crc library only

I want to build boost with crc library only, how to build in my case? I get err on Ubuntu 24

sudo apt-get install build-essential g++ autotools-dev libicu-dev libbz2-dev libboost-all-dev
git clone https://github.com/boostorg/boost.git
cd boost
git submodule update --init -- libs/crc/
./bootstrap.sh
./bootstrap.sh: 1: ./tools/build/src/engine/build.sh: not found
Building B2 engine..
./bootstrap.sh: 229: ./tools/build/src/engine/build.sh: not found

Failed to build B2 build engine

Ok, let's just check boost libraries list

./bootstrap.sh --show-libraries
./bootstrap.sh: 1: ./tools/build/src/engine/build.sh: not found
Building B2 engine..
./bootstrap.sh: 229: ./tools/build/src/engine/build.sh: not found

Failed to build B2 build engine
Solution

  • Boost.CRC is a header-only library, meaning that you do not have to build anything.

    Besides, BoostDep shows that there are few dependencies: array config integer type_traits

    This is a rare instance where BCP might actually work to extract a minimul subset. However, I'd simply use the release tarball of Boost and not build anything for simplicity.

    UPDATE: Using BCP demonstration

    So, I found the time to actually test using BCP, and it actually makes a difference:

    1. assuming an example test.cpp

      #include <boost/crc.hpp>
#include <cstring>

int main() {
  boost::crc_32_type crc;
  char const *data = "Hello, world!";
  crc.process_bytes(data, std::strlen(data));
  return crc.checksum() == 0xEBE6C6E6u ? 0 : 1;
}

    2. get the boost release tarball and extract it

      wget https://archives.boost.io/release/1.85.0/source/boost_1_85_0.tar.bz2
tar xf boost_1_85_0.tar.bz2

    3. extract all header dependencies:

      mkdir extracted
bcp --boost=./boost_1_85_0/ --scan test.cpp extracted/

    4. test compilation of the sample program

      g++ -isystem extracted/ test.cpp 
./a.out ; echo $?

    extracted.zip/tarball

    Just for reference here's the 198kB zip of extracted/ folder created above: extracted.zip.

    Alternatively as the 93kB tarball: extracted.tbz2

    Using bcp to report the 110 headers with their copyright notices: report.html