使用 Homebrew, apk, apt, dnf, MacPorts, Nix, pacman, zypper, winget 安装 opencc

项目链接

• 主页
• 仓库
• 上游文档
• 软件包管理器页面

命令和别名

• opencc
• opencc_dict
• opencc_phrase_extract

项目历史

BYVoid's project page describes OpenCC as an open source Traditional/Simplified Chinese conversion project with high-quality phrase dictionaries and a C library, originally hosted on Google Code and GitHub. The source header credits Carbo Kuo and contributors from 2010 onward, giving the project a long maintenance history for what could otherwise look like a simple text filter.

The project evolved from a C/C++ conversion library and command-line tool into a multi-ecosystem package. The README lists installation paths through Debian, Ubuntu, Fedora, Arch Linux, Homebrew, WinGet, Bazel, Node.js, and Python, and includes C, C++, Node.js, and Python usage examples.

OpenCC's release history shows the project treating dictionaries and conversion behavior as first-class software. Version 1.1.0 introduced the `ocd2` dictionary format based on Marisa Trie and reduced dictionary size; version 1.3.x rewrote core conversion paths, added Jieba plugin work, and expanded diagnostics such as segmentation and inspection; version 1.4.0 raised the C++ ABI/SOVERSION to avoid silent ABI mismatches.

采用历史

OpenCC became a common dependency because script conversion is not just character substitution. The project emphasizes distinctions such as one-simplified-to-many-traditional mappings, variant characters, and regional vocabulary across Mainland China, Taiwan, Hong Kong, and Japanese Shinjitai.

The package's breadth across Python, Node.js, WinGet, Linux distributions, Homebrew, and Bazel makes it useful in desktop tools, web backends, NLP preprocessing, publishing pipelines, chat applications, and input-method ecosystems. OpenCC's own documentation states that it is not a Mandarin/Cantonese translation tool, which helps frame its adoption around orthography and vocabulary variants rather than language translation.

使用方式

Users run the `opencc` CLI with JSON configurations such as Simplified-to-Traditional, Traditional-to-Simplified, Taiwan phrase, Hong Kong phrase, or Japanese Shinjitai conversion modes. Developers embed the library through C/C++, Python, Node.js, or downstream ports when they need deterministic script conversion inside applications.

Dictionary maintainers and power users use `opencc_dict`, inline dictionaries, JSON/JSONC configuration, resource bundles, and diagnostic modes to inspect segmentation and conversion stages. Those tools matter because many conversion bugs are phrase-boundary or regional-vocabulary bugs rather than simple missing characters.

为什么软件包爱好者会关心

OpenCC is a deceptively deep package: a text converter with ABI, dictionary format, data-resource packaging, language bindings, prebuilt binaries, and region-sensitive data updates. The interesting packaging work is making the executable, shared library, dictionaries, configs, and bindings agree across platforms.

It is also a good example of data as code. Release notes are dominated by dictionary corrections, phrase additions, normalization behavior, and compatibility work, showing why downstreams need both the binary and the exact data files.

时间线

• 2010: Source headers credit the beginning of the Carbo Kuo/contributor copyright span.
• 2016: BYVoid's project page documented OpenCC as a Google Code and GitHub-hosted project with a C library and phrase dictionaries.
• 2020-05-10: OpenCC 1.1.0 introduced the `ocd2` dictionary format and smaller Marisa Trie-backed dictionaries.
• 2026-04-17: OpenCC 1.3.0 added experimental Jieba segmentation plugin work.
• 2026-07-01: OpenCC 1.4.0 raised the C++ ABI/SOVERSION to 1.4 and hardened dictionary handling.

Related projects

• Related packages include OpenCC's Python and Node.js bindings, opencc-js and other ports, Jieba segmentation integrations, Rime/librime input-method uses, and downstream operating-system packages that split OpenCC data, docs, and shared libraries.

来源

• https://byvoid.com/en/project/opencc/
• https://github.com/BYVoid/OpenCC
• https://opencc.byvoid.com/
• https://raw.githubusercontent.com/BYVoid/OpenCC/master/NEWS.md
• https://raw.githubusercontent.com/BYVoid/OpenCC/master/src/opencc.h