mkdocs-material/docs/blog/posts/chinese-search-support.md
2024-04-25 10:51:05 +07:00

84 lines
2.9 KiB
Markdown
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

---
date: 2022-05-05
authors: [squidfunk]
title: Chinese search support
description: >
Insiders adds Chinese language support for the built-in search plugin a
feature that has been requested many times
categories:
- Search
links:
- blog/posts/search-better-faster-smaller.md
- plugins/search.md#segmentation
- insiders/index.md#how-to-become-a-sponsor
---
# Chinese search support 中文搜索​支持
__Insiders adds experimental Chinese language support for the [built-in search
plugin] a feature that has been requested for a long time given the large
number of Chinese users.__
After the United States and Germany, the third-largest country of origin of
Material for MkDocs users is China. For a long time, the [built-in search plugin]
didn't allow for proper segmentation of Chinese characters, mainly due to
missing support in [lunr-languages] which is used for search tokenization and
stemming. The latest Insiders release adds long-awaited Chinese language support
for the built-in search plugin, something that has been requested by many users.
<!-- more -->
_Material for MkDocs終於支持中文文本正確分割並且容易找到。_
{ style="display: inline" }
_This article explains how to set up Chinese language support for the built-in
search plugin in a few minutes._
{ style="display: inline" }
[built-in search plugin]: ../../plugins/search.md
[lunr-languages]: https://github.com/MihaiValentin/lunr-languages
## Configuration
Chinese language support for Material for MkDocs is provided by [jieba], an
excellent Chinese text segmentation library. If [jieba] is installed, the
built-in search plugin automatically detects Chinese characters and runs them
through the segmenter. You can install [jieba] with:
```
pip install jieba
```
The next step is only required if you specified the [`separator`][separator]
configuration in `mkdocs.yml`. Text is segmented with [zero-width whitespace]
characters, so it renders exactly the same in the search modal. Adjust
`mkdocs.yml` so that the [`separator`][separator] includes the `\u200b`
character:
``` yaml
plugins:
- search:
separator: '[\s\u200b\-]'
```
That's all that is necessary.
## Usage
If you followed the instructions in the configuration guide, Chinese words will
now be tokenized using [jieba]. Try searching for
[:octicons-search-24: 支持][q=支持] to see how it integrates with the
built-in search plugin.
---
Note that this is an experimental feature, and I, @squidfunk, am not
proficient in Chinese (yet?). If you find a bug or think something can be
improved, please [open an issue].
[jieba]: https://pypi.org/project/jieba/
[zero-width whitespace]: https://en.wikipedia.org/wiki/Zero-width_space
[separator]: ../../plugins/search.md#config.separator
[q=支持]: ?q=支持
[open an issue]: https://github.com/squidfunk/mkdocs-material/issues/new/choose