使用 Homebrew 安装 boost-bcp

项目链接

• 主页
• 仓库

命令和别名

• bcp

项目历史

The boostorg/bcp Git history contains an initial commit adding the BCP utility in June 2003, followed quickly by dependency-analysis fixes and the report mode. That places BCP in the early era when Boost was becoming a large, modular C++ library collection and users needed tooling to manage subsets.

The official BCP documentation, copyrighted 2009 by John Maddock, describes the core job that has remained stable: copy headers, library source, build files, and transitive dependencies for a named Boost component or for dependencies discovered by scanning user code. It can also generate HTML reports listing licenses, copyright holders, and dependency reasons.

BCP continued to be released with Boost version tags after Boost moved to the modular GitHub organization. The tool's repository carries Boost release tags from the 1.31 era onward, and its README points users to the canonical boost.org documentation.

采用历史

BCP is not a mainstream end-user command so much as a specialist tool for C++ packaging and vendoring. Its adoption follows Boost itself: wherever developers wanted a subset of Boost without shipping the full tree, BCP provided an official route.

Homebrew's separate boost-bcp formula reflects package-manager recognition that this utility is useful on its own, even when the complete Boost distribution is available separately. That split is exactly the sort of detail package maintainers care about: one executable from a giant upstream project, packaged for a narrow but real workflow.

使用方式

The documented examples include copying `boost/scoped_ptr.hpp` and dependencies, copying Boost.Regex with source and build files, copying whole libraries, renaming the Boost namespace for vendored builds, scanning non-Boost source files for Boost includes, and generating HTML dependency/license reports.

BCP intentionally follows broad dependency paths because a portable Boost subset must often include headers for multiple compilers, platforms, and preprocessor branches. The docs warn that dependency sets can look fat compared with what one compiler happens to use.

为什么软件包爱好者会关心

Boost BCP is deeply package-nerd significant because it is an upstream-sanctioned answer to vendoring before modern C++ dependency managers were common. It turns a huge source distribution into a smaller redistributable tree while preserving transitive headers, compiled-library source, build metadata, and license accounting.

It also exposes a tension every C++ package maintainer knows: header-only libraries are not really dependency-free once portability macros, compiler workarounds, and optional source libraries enter the graph. BCP's deliberately conservative dependency expansion is a historical artifact of Boost's promise to work across many compilers and platforms.

时间线

• 2003-06-09: Initial Git history entry adds the BCP utility.
• 2003-06-23: Dependency-analysis fixes and report mode are added.
• 2004-02-04: boost-1.31.0 tag appears in the BCP repository history.
• 2009: Official QuickBook documentation records John Maddock copyright and BCP usage model.
• 2010s: Boost's modular GitHub organization makes BCP available as boostorg/bcp while still releasing with Boost.
• 2024: Recent Boost release tags continue to include the BCP repository.

Related projects

• Boost is the parent C++ library collection and release ecosystem.
• Boost.Build is related because BCP can copy build files and the build system for selected libraries.
• Boost.Regex is a prominent documented example because BCP can copy its headers, source, build data, and dependencies.
• boostdep is related as another Boost tool for dependency work in the modular repository era.

来源

• https://github.com/boostorg/bcp#readme
• https://github.com/boostorg/boost#readme
• https://raw.githubusercontent.com/boostorg/bcp/develop/doc/bcp.qbk
• https://www.boost.org/doc/libs/release/tools/bcp/doc/html/index.html
• source_facts.package-manager