Project Icon

style-dictionary-utils

Style Dictionary 的设计令牌处理工具集

style-dictionary-utils 是一个为 Style Dictionary 开发的工具集,用于处理 w3c 设计令牌。它包含解析器、过滤器、转换器和格式化器,支持颜色转换、尺寸单位转换和字体样式处理等功能。该工具兼容 ESM 和 CommonJS 模块系统,并能根据媒体查询和选择器生成高级 CSS 变量。它简化了设计系统开发流程,提高了开发效率。

Style dictionary utils

npm GitHub release (release name instead of tag name)

style-dictionary-utils is a collection of parsers, filters, transformers and formats for Style Dictionary that make working with w3c design tokens a lot easier.

Installation

Install the style-dictionary-utils as well as style-dictionary.

npm i -D style-dictionary-utils style-dictionary

If you are using .json5 files to define your design tokens install json5 as well.

npm i -D json5

How to use style dictionary version 3?

If you are not ready to upgrade to style dictinary version 3 you can continue using style-dictionary-utils by locking to v2 currently v2.4.1 version.

Getting started

The easiest way to use style-dictionary-utils is to import the prepared StyleDictionary object into your build file:

// build.ts
import StyleDictionary from 'style-dictionary-utils'

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['color/hexAlpha', 'shadow/css'],
      "files": [{
        "filter": "isSource",
        "destination": "tokens.ts",
        "format": "javascript/esm",
      }]
    }
  }
});

myStyleDictionary.buildAllPlatforms();

Now all the included utilities* are available to you via the keys mentioned in the docs below.

* You only need to register the w3cTokenJson5Parser if you want to use json5.

Extending style dictionary

You can still extend style dictionary with your own transformers and formats like before. The only difference is that you must use the StyleDictionary object that you import from style-dictionary-utils.

// build.ts
import StyleDictionary from 'style-dictionary-utils'

StyleDictionary.registerTransform({
  name: 'transform/pxToRem',
  type: `value`,
  transitive: true,
  transformer: () => // ...
})

Look at the tests to get an idea how it works.

Included utilities

📠 Parsers

w3cTokenJsonParser

This parser parses .json with w3c design tokens.

This means the following files can be used with this parser.

{
  "token": {
    "value": "#223344",
    "type": "color",
    "description": "token description"
  },
  "w3cToken": {
    "$value": "#223344",
    "$type": "color",
    "$description": "token description"
  }
}

The parser will keep most propertys as is and only change $value to value and $description or description to comment. This required for Style Dictionary.

To register the parsers add the following code to your build file.

import StyleDictionary from 'style-dictionary-utils'
import { w3cTokenJsonParser } from 'style-dictionary-utils/dist/parser/w3c-token-json-parser'

StyleDictionary.registerParser(w3cTokenJsonParser)

w3cTokenJson5Parser (not autoloaded)

If you are using .json5 or .jsonc files to define your design tokens you need to use the w3cTokenJson5Parser. This is NOT enabled by default as it requires an additonal package, json5, to work.

This parser is exactly the same as the w3cTokenJsonParser with the only difference that it can parse .json5 or .jsonc.

To register the parsers add the following code to your build file.

import StyleDictionary from 'style-dictionary-utils'
import { w3cTokenJson5Parser } from 'style-dictionary-utils/dist/parser/w3c-token-json5-parser'

StyleDictionary.registerParser(w3cTokenJson5Parser)

Make sure to install json5 by running npm i -D json5.

If you’re using Prettier, be aware that the default configuration removes quote props, which are needed in $type and $value props in order to parse the tokens.

Here’s an example of a prettier config that overrides the default:

semi: false
singleQuote: true
overrides:
  - files: '*.json[c|5]'
    options:
      quoteProps: preserve
      singleQuote: false

📑 Formats

javascript/esm

The javascript/esm format exports a token dictionary as an es6 export statement.

export default {
  colors: {
    primary: '#0D70E6'
  }
}
Usage:
const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": //...,
      "files": [{
        // ...
        "format": "javascript/esm",
      }]
    }
  }
});

javascript/commonJs

The javascript/commonJs format exports a token dictionary as an commonJs module.

exports.default = {
  colors: {
    primary: '#0D70E6'
  }
}
Usage:
const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "js": {
      "transforms": //...,
      "files": [{
        // ...
        "format": "javascript/commonJs",
      }]
    }
  }
});

css/advanced

The css/advanced format exports a token dictionary as a css file with css variables. It allows you to define media queries that can wrap specific parts of your css variables. If nothing is defined the entire file will be wrapped in a :root selector.

You can change the selector by defining it in file.options.selector.

You can define rules on a file level using file.options.rules. If one or more rules are defined, only tokens within any of the rules will be output. You can define as many rule objects within file.options.rules as you want. Tokens can be part of one or multiple rules.

A rule object may have any or all of the three properties atRule, selector and matcher.

  • selector is a string that is wrapped around your css. If the selector is undefined, the default selector or one define at file.options.selector will be used. If you don't want a selector, set it to false.
  • atRule can be a string or array of strings, that are wrapped around the css and selector with the first being the outer layer.
  • matcher is a filter function that returns true for tokens that should be included in the query. If you want to match all tokens, just return true from the matcher.
body[theme="dark"] {
  --color-background-primary: #ff0000;
  --color-background-secondary: #0000ff;
}
@media (min-width: 768px) {
  body[theme="dark"] {
    --color-button-primary: #c1c1c1;
    --color-button-secondary: #007D79;
  }
}
Usage:
const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "css": {
      "transforms": //...,
      "files": [{
        // ...
        "format": "css/advanced",
        "options": {
          selector: `body[theme="dark"]`, // defaults to :root; set to false to disable
          rules: [
          {
            atRule: '@media (min-width: 768px)',
            selector: `body[size="medium"]` // this will be used instead of body[theme="dark"]`
            matcher: (token: StyleDictionary.TransformedToken) => token.filePath.includes('tablet'), // tokens that match this filter will be added inside the media query
          }]
        }
      }]
    }
  }
});

🤖 Transformers

Transforms change the value or name of a token. You can use transforms by refering the name in the array value of the transforms property of a platform.

Transform group

If you want to use the same transformers multiple times you can create a transform group for easy reference.

StyleDictionary.registerTransformGroup({
  name: 'webHex',
  transforms: [
    'color/hexAlpha',
    'dimension/pixelToRem',
    'font/css'
  ]
});

css/extended transform group

This packages ships a predefined transform group, called css/extended. It includes all transforms from the original css transform group as well as the following transforms: color/rgbAlpha, shadow/css, font/css, fontFamily/css, fontWeight/number, name/pathToDotNotation, cubicBezier/css, border/css.

You can use it like any other transform Group:

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "css": {
      "transformGroup": 'css/extended',
      "files": [{
        // ...
      }]
    }
  }
});

name/pathToDotNotation

This name transformer replaces the token name with the entire path of the token in dot.notation. This is especially useful for flat .js or .json files.

To use it simply add name/pathToDotNotation to the transforms array.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['name/pathToDotNotation'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    red: {
      100: {
        // ...
      }
    }
  }
}
After transformation
{
  "colors.red.100": {
    // ...
  }
}

name/pathToCamelCase

This name transformer replaces the token name with the entire path of the token in camelCase notation.

To use it simply add name/pathToCamelCase to the transforms array.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['name/pathToCamelCase'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    bg: {
      default: {
        // ...
      }
    }
  }
}
After transformation
{
  "colorsBgDefault": {
    // ...
  }
}

color/rgbAlpha

This value transformer replaces the value of a token with a $type or type of color with an rgba string. If the token has an alpha value, it will be used as the alpha of the rgba string.

Note: If your initial color value has an alpha value (e.g. hex8) AND you add an alpha property, the alpha property will simply replace the previous alpha value.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['color/rgbAlpha'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    blue: {
      500: {
        value: "#0D70E6",
        $type: "color",
        alpha: 0.4
      }
    }
  }
}
After transformation
{
  colors: {
    blue: {
      500: {
        value: "rgba(13, 112, 230, 0.4)",
        $type: "color",
        alpha: 0.4
      }
    }
  }
}

color/hexAlpha

This value transformer replaces the value of a token with a $type or type of color with a hex string. If the token has an alpha value, it will be used as the alpha of the hex8 string.

Note: If your initial color value has an alpha value (e.g. rgba) AND you add an alpha property, the alpha property will simply replace the previous alpha value.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['color/hexAlpha'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    blue: {
      500: {
        value: "rgba(13, 112, 230, 0.4)",
        $type: "color",
        alpha: 0.2
      }
    }
  }
}
After transformation
{
  colors: {
    blue: {
      500: {
        value: "#0D70E633", // prev alpha value is replaced with 0.2 from alpha property
        $type: "color",
        alpha: 0.2
      }
    }
  }
}

color/hex

This value transformer replaces the value of a token with a $type or type of color with a hex string.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['color/hex'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    blue: {
      500: {
        value: "rgba(13, 112, 230, 0.4)",
        $type: "color"
      }
    }
  }
}
After transformation
{
  colors: {
    blue: {
      500: {
        value: "#0D70E666",
        $type: "color"
      }
    }
  }
}

color/rgba

This value transformer replaces the value of a token with a $type or type of color with an rgba string.

const myStyleDictionary = StyleDictionary.extend({
  "platforms": {
    "ts": {
      "transforms": ['color/rgba'],
      "files": [{
        // ...
      }]
    }
  }
});
Before transformation
{
  colors: {
    blue: {
      500: {
        value: "#0D70E666",
        $type: "color"
      }
    }
  }
}
After
项目侧边栏1项目侧边栏2
推荐项目
Project Cover

豆包MarsCode

豆包 MarsCode 是一款革命性的编程助手,通过AI技术提供代码补全、单测生成、代码解释和智能问答等功能,支持100+编程语言,与主流编辑器无缝集成,显著提升开发效率和代码质量。

Project Cover

问小白

问小白是一个基于 DeepSeek R1 模型的智能对话平台,专为用户提供高效、贴心的对话体验。实时在线,支持深度思考和联网搜索。免费不限次数,帮用户写作、创作、分析和规划,各种任务随时完成!

Project Cover

白日梦AI

白日梦AI提供专注于AI视频生成的多样化功能,包括文生视频、动态画面和形象生成等,帮助用户快速上手,创造专业级内容。

Project Cover

有言AI

有言平台提供一站式AIGC视频创作解决方案,通过智能技术简化视频制作流程。无论是企业宣传还是个人分享,有言都能帮助用户快速、轻松地制作出专业级别的视频内容。

Project Cover

讯飞绘镜

讯飞绘镜是一个支持从创意到完整视频创作的智能平台,用户可以快速生成视频素材并创作独特的音乐视频和故事。平台提供多样化的主题和精选作品,帮助用户探索创意灵感。

Project Cover

讯飞文书

讯飞文书依托讯飞星火大模型,为文书写作者提供从素材筹备到稿件撰写及审稿的全程支持。通过录音智记和以稿写稿等功能,满足事务性工作的高频需求,帮助撰稿人节省精力,提高效率,优化工作与生活。

Project Cover

阿里绘蛙

绘蛙是阿里巴巴集团推出的革命性AI电商营销平台。利用尖端人工智能技术,为商家提供一键生成商品图和营销文案的服务,显著提升内容创作效率和营销效果。适用于淘宝、天猫等电商平台,让商品第一时间被种草。

Project Cover

Trae

Trae是一种自适应的集成开发环境(IDE),通过自动化和多元协作改变开发流程。利用Trae,团队能够更快速、精确地编写和部署代码,从而提高编程效率和项目交付速度。Trae具备上下文感知和代码自动完成功能,是提升开发效率的理想工具。

Project Cover

AIWritePaper论文写作

AIWritePaper论文写作是一站式AI论文写作辅助工具,简化了选题、文献检索至论文撰写的整个过程。通过简单设定,平台可快速生成高质量论文大纲和全文,配合图表、参考文献等一应俱全,同时提供开题报告和答辩PPT等增值服务,保障数据安全,有效提升写作效率和论文质量。

投诉举报邮箱: service@vectorlightyear.com
@2024 懂AI·鲁ICP备2024100362号-6·鲁公网安备37021002001498号