hasaneyldrm · marcomarcogd · Jul 14, 2026 · Jul 14, 2026 · Jul 14, 2026
diff --git a/NOTICE.md b/NOTICE.md
@@ -1,3 +1,5 @@
+[**English**](NOTICE.md) · [简体中文](NOTICE.zh-CN.md)
+
 # Media Attribution & License
 
 The exercise **media** (thumbnail images and animation GIFs) in this repository

diff --git a/NOTICE.zh-CN.md b/NOTICE.zh-CN.md
@@ -0,0 +1,28 @@
+[English](NOTICE.md) · **简体中文**
+
+# 媒体署名与许可说明
+
+> **译文声明：** 本文件是 `NOTICE.md` 的非官方简体中文参考译文，仅为方便阅读，不构成新的授权、许可或法律意见。若本译文与英文原文存在差异，应以英文 [`NOTICE.md`](NOTICE.md)、[`LICENSE`](LICENSE) 及权利人的原始条款为准。
+
+本仓库中的动作训练媒体文件（缩略图和 GIF 动画）归 **Gym visual** 所有，并经其许可在此重新分发。
+
+> **© Gym visual — https://gymvisual.com/**
+
+## 使用条件
+
+- **分辨率：** 仅以 **180×180** 分辨率分发。
+- **署名：** 每次使用都必须保留版权标识 **© Gym visual — https://gymvisual.com/**。每条 [`data/exercises.json`](data/exercises.json) 记录也包含此 `attribution` 声明。
+
+使用这些媒体文件时，请完整保留署名并遵守 180×180 分辨率限制。
+
+## 再利用与许可
+
+这些媒体文件依据权利人给予的单独书面许可收录于本仓库。媒体文件仍归 Gym visual 所有，其使用和再次利用受 **Gym visual 使用条款与条件**约束：
+
+> **https://gymvisual.com/content/3-terms-and-conditions-of-use**
+
+若要在自己的项目中使用这些媒体文件，请先阅读上述条款，并在需要时直接向 Gym visual 获取许可。除 Gym visual 条款允许的范围外，本仓库不会授予你额外的媒体权利；克隆本仓库不等于取得媒体许可。
+
+## 数据集（不含媒体）
+
+动作训练数据（名称、分类、身体部位、器械、目标肌肉、肌群、多语言训练说明及本地化语言包）独立于媒体文件，并根据 MIT License 发布，详见 [`LICENSE`](LICENSE)。
diff --git a/README.md b/README.md
@@ -1,3 +1,5 @@
+[**English**](README.md) · [简体中文](README.zh-CN.md)
+
 <div align="center">
 
 # 💪 Exercises Dataset
@@ -33,6 +35,7 @@
 - 1,324 exercises with category, body-part, equipment, target and muscle-group data
 - an animation GIF + 180×180 thumbnail for every exercise (media © [Gym visual](https://gymvisual.com/) — see [License](#-license--use))
 - step-by-step instructions in 9 languages (🇬🇧 English, 🇪🇸 Spanish, 🇮🇹 Italian, 🇹🇷 Turkish, 🇷🇺 Russian, 🇨🇳 Chinese, 🇮🇳 Hindi, 🇵🇱 Polish, 🇰🇷 Korean)
+- an extensible browser localization system with complete Simplified Chinese exercise names, taxonomy, and reviewed instructions in a single locale pack
 - the interactive browser (`index.html`) and developer setup guide (`setup.html`)
 
 ---
@@ -42,6 +45,7 @@
 - [Data Source](#-data-source)
 - [Overview](#-overview)
 - [Interactive Browser & Developer Setup](#-interactive-browser--developer-setup)
+- [Localization Packs](#-localization-packs)
 - [File Structure](#-file-structure)
 - [Statistics](#-statistics)
 - [Data Schema](#-data-schema)
@@ -78,7 +82,7 @@ Each exercise entry contains:
 
 ## 🖥️ Interactive Browser & Developer Setup
 
-This repository includes two ready-to-use HTML tools — no server required, just open in a browser.
+This repository includes two ready-to-use HTML tools. The exercise browser can be opened directly; when deployed, the setup guide reads `data/exercises.json` from the repository when generating SQL.
 
 > **Note:** the browser displays each exercise's 180×180 thumbnail and animation GIF alongside its metadata and instructions.
 
@@ -88,41 +92,69 @@ A fully client-side exercise explorer with:
 - Live search across all 1,324 exercises
 - Filter by category, equipment, and target muscle
 - Infinite scroll grid
+- Extensible interface localization that follows the browser language on first visit and remembers later choices
+- Complete Simplified Chinese names and fitness terminology, with bilingual search and English fallback
 - Click any card to see full details and instructions in English, Spanish, Italian, Turkish, Russian, Chinese, Hindi, Polish, or Korean
 
 ### `setup.html` — Developer Setup Guide
 
 A step-by-step guide for integrating the dataset into your own application:
 
-1. **Database Setup** — `CREATE TABLE` SQL for SQL Server, PostgreSQL, MySQL, and SQLite. Generate a ready-to-run `.sql` file with all 1,324 INSERT statements, built entirely in your browser.
+1. **Database Setup** — `CREATE TABLE` SQL for SQL Server, PostgreSQL, MySQL, and SQLite. Generate a ready-to-run `.sql` file with all 1,324 exercises and every registered non-base locale. Localized records use the generic `exercise_translations` table.
 2. **API Integration** — Copy-paste client code in **JavaScript, Python, C#, Java, PHP, Go, and cURL** showing how to call your backend API. Enter your base URL and all examples update live.
 3. **Ask Your LLM** — A structured prompt (choose your framework + database) that you can paste into ChatGPT, Claude, or Gemini to generate a complete, production-ready REST API in one shot. Supports Express.js, FastAPI, ASP.NET Core, Spring Boot, Laravel, and Gin.
 
 ---
 
+## 🌐 Localization Packs
+
+Runtime translations are separate from the canonical dataset. `data/exercises.json` remains unchanged, while each UI locale is a classic UTF-8 JavaScript file that works when `index.html` is opened through `file://`.
+
+Each locale pack contains both pages' UI messages, instruction-language labels, taxonomy mappings, and optional per-exercise `{ name, steps }` translations. English uses the canonical dataset directly; `locales/zh-CN.js` supplies complete Simplified Chinese translations for all 1,324 records.
+
+The Simplified Chinese pack has been reviewed record by record in two independent passes: all 1,324 exercise names, 7,662 instruction steps, 107 taxonomy labels, and 90 interface messages were checked against the canonical English data and, where the source text was ambiguous, the corresponding image or GIF. Only clear or reviewer-agreed corrections were applied; unresolved source-data ambiguities were left unchanged rather than guessed.
+
+To add a language:
+
+1. Copy `locales/en.js` to a BCP 47 filename such as `locales/fr-FR.js`.
+2. Translate that one file. Partial exercise translations are allowed and fall back to English.
+3. Add one metadata entry to `locales/manifest.js`. The HTML files do not need to change.
+
+The setup guide exports explicit non-base translations into a generic table keyed by `(exercise_id, locale)`, so adding a locale never requires a new database column.
+
+---
+
 ## 📂 File Structure
 
 ```
 exercises-dataset/
 ├── data/
 │   ├── exercises.json        # Full dataset — 1,324 exercise records (JSON array)
 │   └── exercises.schema.json # JSON Schema (2020-12) describing every record
+├── locales/
+│   ├── runtime.js            # Shared browser i18n runtime
+│   ├── manifest.js           # Registered locale metadata
+│   ├── en.js                 # English interface locale
+│   └── zh-CN.js              # Complete Simplified Chinese locale pack
 ├── images/                  # 1,324 × 180×180 thumbnails  (© Gym visual)
 ├── videos/                  # 1,324 × 180×180 animation GIFs  (© Gym visual)
 ├── index.html               # Interactive exercise browser (client-side, no server needed)
 ├── setup.html               # Developer setup guide (DB import + API integration)
 ├── NOTICE.md                # Media attribution & license terms
+├── NOTICE.zh-CN.md          # Informational Simplified Chinese translation
+├── README.zh-CN.md          # Simplified Chinese documentation
 └── README.md
 ```
 
 ### Key Files
 
 - **`data/exercises.json`** — The primary data file. A JSON array of 1,324 exercise objects with all metadata. `image` / `gif_url` point to the local 180×180 assets, and each record carries an `attribution` field; `media_id` holds the original media reference id.
 - **`data/exercises.schema.json`** — A [JSON Schema](https://json-schema.org/) (Draft 2020-12) that formally describes every field, its type and constraints. Use it to validate the dataset or your own additions with any standard JSON Schema validator.
+- **`locales/`** — Shared i18n runtime and one self-contained translation file per locale. Browser localization is an overlay and does not modify the canonical JSON schema.
 - **`images/`, `videos/`** — 180×180 thumbnails and animation GIFs (© [Gym visual](https://gymvisual.com/), used with permission).
 - **`index.html`** — Standalone exercise browser. Open directly in any modern browser.
-- **`setup.html`** — Developer guide for DB setup, API integration, and LLM-assisted backend generation.
-- **`LICENSE`, `NOTICE.md`** — MIT (code/data) + the Gym visual media terms.
+- **`setup.html`** — Developer guide for DB setup, API integration, and LLM-assisted backend generation. SQL generation loads the external dataset at runtime instead of embedding a duplicate copy in the page.
+- **`LICENSE`, `NOTICE.md`** — MIT (code/data) + the Gym visual media terms. [`NOTICE.zh-CN.md`](NOTICE.zh-CN.md) is an informational translation; the English notice remains authoritative.
 
 ---
 

diff --git a/README.zh-CN.md b/README.zh-CN.md
@@ -0,0 +1,208 @@
+[English](README.md) · **简体中文**
+
+<div align="center">
+
+# 💪 Exercises Dataset 动作训练数据集
+
+<p>
+  <img src="videos/0025-EIeI8Vf.gif" width="120" alt="杠铃卧推" />
+  <img src="videos/0043-qXTaZnJ.gif" width="120" alt="杠铃全蹲" />
+  <img src="videos/0032-ila4NZS.gif" width="120" alt="杠铃硬拉" />
+  <img src="videos/0652-lBDjFxJ.gif" width="120" alt="引体向上" />
+  <img src="videos/0294-NbVPDMW.gif" width="120" alt="哑铃肱二头肌弯举" />
+  <img src="videos/0334-DsgkuIt.gif" width="120" alt="哑铃侧平举" />
+</p>
+
+**一套包含 1,324 个健身动作的开箱即用数据集。每个动作均提供 GIF 动画、180×180 缩略图、分类、身体部位、器械、目标肌肉、肌群信息，以及英语、西班牙语、意大利语、土耳其语、俄语、中文、印地语、波兰语和韩语共 9 种语言的训练说明。**
+
+[![动作数量](https://img.shields.io/badge/动作数量-1324-blue?style=flat-square)](data/exercises.json)
+[![GIF 动画](https://img.shields.io/badge/GIF%20动画-1324-brightgreen?style=flat-square)](videos/)
+[![缩略图](https://img.shields.io/badge/缩略图-1324-orange?style=flat-square)](images/)
+[![说明语言](https://img.shields.io/badge/说明语言-9-green?style=flat-square)](#数据内容)
+[![许可](https://img.shields.io/badge/License-MIT%20%2B%20媒体条款-blue?style=flat-square)](LICENSE)
+
+</div>
+
+本仓库同时提供可扩展的浏览器多语言系统。简体中文动作名、健身术语、页面文案和精修训练步骤全部集中在 [`locales/zh-CN.js`](locales/zh-CN.js)，不会向基础 JSON 增加 `_zh` 字段，也不会改变原有 Schema。
+
+简体中文语言包已完成两轮独立逐条复核：对照规范英文数据检查全部 1,324 个动作名、7,662 条训练步骤、107 个术语标签和 90 条页面文案；英文来源存在歧义时再核对对应图片或 GIF。只有语义明确或两轮结论一致的问题才会修正，尚不能确定的源数据歧义不作猜测性改写。
+
+## 数据内容
+
+- 1,324 个动作及其分类、身体部位、器械、目标肌肉和协同肌群
+- 每个动作一张 180×180 缩略图和一个 GIF 动画
+- 9 种语言的完整说明与分步说明
+- 可直接打开的动作浏览器 `index.html`
+- 数据库导入、API 示例和大模型提示词页面 `setup.html`
+- 覆盖全部 1,324 个动作的简体中文语言包
+
+媒体文件 © [Gym visual](https://gymvisual.com/)，许可和署名要求请阅读 [`NOTICE.md`](NOTICE.md)；[`NOTICE.zh-CN.md`](NOTICE.zh-CN.md) 仅为辅助译文。
+
+## 动作浏览器与开发接入
+
+### `index.html`
+
+直接双击打开即可使用，支持：
+
+- 英文和简体中文界面，首次访问跟随浏览器语言，之后记住用户选择
+- 使用中英文动作名、分类、器械和肌肉术语搜索
+- 按分类、器械和目标肌肉筛选，切换语言不会丢失筛选状态
+- 无限滚动、响应式卡片和动作详情弹窗
+- 查看数据集中已有的 9 种语言训练步骤
+- 中文模式优先显示语言包中精修后的中文动作名和步骤
+
+### `setup.html`
+
+`setup.html` 按上游方式在生成 SQL 时读取 `data/exercises.json`，因此应与仓库中的数据、语言包和媒体文件一起部署到 Web 服务器。页面不再重复内嵌整份动作数据。
+
+setup 支持 SQL Server、PostgreSQL、MySQL 和 SQLite，可导出：
+
+- 1,324 条基础动作记录
+- manifest 中所有已注册非基础语言包的显式翻译
+- 使用 `(exercise_id, locale)` 作为主键的通用 `exercise_translations` 表
+
+SQL Server 使用 `NVARCHAR` 和 `N'…'` Unicode 字面量，MySQL 使用 `utf8mb4`，PostgreSQL 使用 `JSONB`，SQLite 使用 UTF-8 文本和 JSON 字符串。
+
+## 可扩展语言包
+
+```text
+locales/
+├── runtime.js    # 共享 i18n 运行时、回退和语言记忆
+├── manifest.js   # 已注册语言及文件路径
+├── en.js         # 英文页面文案，动作内容使用基础数据
+└── zh-CN.js      # 完整简体中文语言包
+```
+
+每种语言只维护一个经典 UTF-8 JavaScript 文件，结构如下：
+
+```js
+window.ExercisesI18n.registerLocale({
+  code: 'zh-CN',
+  label: '简体中文',
+  instructionCode: 'zh',
+  dir: 'ltr',
+  complete: true,
+  ui: {
+    index: {},
+    setup: {}
+  },
+  instructionLabels: {},
+  taxonomy: {
+    categories: {},
+    bodyParts: {},
+    equipment: {},
+    targets: {},
+    muscles: {}
+  },
+  records: {
+    '0001': {
+      name: '3/4 仰卧起坐',
+      steps: ['平躺，膝盖弯曲，双脚平放在地上。']
+    }
+  }
+});
+```
+
+新增语言时：
+
+1. 复制 `locales/en.js`，文件名使用 BCP 47 代码，例如 `fr-FR.js`。
+2. 只在该文件中翻译 index、setup、术语和动作内容。
+3. 在 `locales/manifest.js` 增加一行语言代码、显示名和文件路径。
+
+不需要修改 `index.html` 或 `setup.html`。语言包可以逐步完成；缺失项目会逐项回退英文，SQL 只导出明确提供的翻译，不会把英文回退内容标记成目标语言。
+
+## 文件结构
+
+```text
+exercises-dataset/
+├── data/
+│   ├── exercises.json        # 1,324 条规范基础记录
+│   └── exercises.schema.json # JSON Schema 2020-12
+├── locales/                  # 浏览器 i18n 运行时和语言包
+├── images/                   # 1,324 张 180×180 缩略图
+├── videos/                   # 1,324 个 GIF 动画
+├── index.html                # 动作浏览器
+├── setup.html                # 数据库和 API 接入指南
+├── README.md                 # 英文说明
+├── README.zh-CN.md           # 本文件
+├── NOTICE.md                 # 权威英文媒体声明
+├── NOTICE.zh-CN.md           # 中文辅助译文
+└── LICENSE
+```
+
+## 基础数据结构
+
+[`data/exercises.json`](data/exercises.json) 是规范数据源，保留上游原始字段：
+
+| 字段 | 说明 |
+|---|---|
+| `id` | 四位动作 ID，例如 `"0001"` |
+| `name` | 规范英文动作名 |
+| `category`、`body_part` | 英文身体部位分类 |
+| `equipment` | 英文器械名称 |
+| `instructions.<lang>` | 各语言完整训练说明 |
+| `instruction_steps.<lang>` | 各语言分步说明数组 |
+| `muscle_group` | 主要协同肌群 |
+| `secondary_muscles` | 辅助肌群数组 |
+| `target` | 主要目标肌肉 |
+| `media_id` | 原始媒体引用 ID |
+| `image`、`gif_url` | 本地缩略图和 GIF 路径 |
+| `attribution` | 媒体版权署名 |
+| `created_at` | ISO 8601 创建时间 |
+
+中文动作名和术语不写入基础 JSON，而是在浏览器运行时按动作 ID 和英文 canonical value 叠加。因此现有 JSON 使用方、Schema 和英文筛选值都保持兼容。
+
+## 通用翻译表
+
+setup 生成的数据库结构在上游 `exercises` 表之外增加：
+
+```text
+exercise_translations
+- exercise_id
+- locale
+- name
+- category
+- body_part
+- equipment
+- target
+- muscle_group
+- secondary_muscles
+- instructions
+- instruction_steps
+- PRIMARY KEY (exercise_id, locale)
+- FOREIGN KEY (exercise_id) REFERENCES exercises(id)
+```
+
+英文仍由 `exercises` 表承载。当前完整中文包会写入 1,324 条 `locale = 'zh-CN'` 的翻译记录；以后新增语言不会要求增加新的数据库列。
+
+## 使用基础 JSON
+
+Python：
+
+```python
+import json
+
+with open("data/exercises.json", "r", encoding="utf-8") as file:
+    exercises = json.load(file)
+
+print(len(exercises))
+print(exercises[0]["name"])
+print(exercises[0]["instructions"]["zh"])
+```
+
+Node.js：
+
+```js
+const exercises = require('./data/exercises.json');
+
+console.log(exercises.length);
+console.log(exercises[0].name);
+console.log(exercises[0].instruction_steps.zh);
+```
+
+## 许可与使用
+
+- 代码、数据结构和训练说明采用 [`LICENSE`](LICENSE) 中的 MIT License。
+- 图片和 GIF 动画归 [Gym visual](https://gymvisual.com/) 所有，经许可仅以 180×180 分辨率重新分发。
+- 使用媒体时必须保留 `© Gym visual — https://gymvisual.com/` 署名，并遵守 Gym visual 的条款。
+- 中文 [`NOTICE.zh-CN.md`](NOTICE.zh-CN.md) 仅供参考；涉及许可时以英文 [`NOTICE.md`](NOTICE.md)、[`LICENSE`](LICENSE) 和权利人的原始条款为准。