feat(plugin-dashboard): ObjectDataTable columns support string[] shorthand

[Copilot]

ObjectDataTable crashes when columns is a string[] (e.g. ['name', 'amount']) because derivedColumns passes strings straight through to the data-table renderer, which accesses col.accessorKey on a plain string → undefined.

Changes

• normalizeColumns helper in ObjectDataTable.tsx — maps each string entry to { header, accessorKey }, passes objects through unchanged. Handles per-element normalization so mixed arrays work too.
• Header derivation — snake_case (close_date → Close Date) and camelCase (firstName → First Name) both title-cased correctly.
• 8 unit tests for normalizeColumns, 2 integration tests for the component with string columns.

Usage

// Before: only object[] worked
columns: [
  { header: 'Name', accessorKey: 'name' },
  { header: 'Close Date', accessorKey: 'close_date' },
]

// Now: string[] shorthand also works
columns: ['name', 'close_date']

// Mixed is fine too
columns: ['name', { header: 'Amount ($)', accessorKey: 'amount' }]

Original prompt

---

This section details on the original issue you should resolve

<issue_title>ObjectDataTable: columns 应兼容 string[] 简写格式，自动 normalize 为对象数组</issue_title>
<issue_description># 渲染兼容建议：table columns 支持字符串简写格式

目前 ObjectUI 的 ObjectDataTable 在解析 columns 字段时，直接透传了 schema.columns，没有对字符串数组 such as ['name', 'amount'] 做转换，导致当业务 schema 采用字符串简写时，下游渲染器会因为访问 undefined (col.accessorKey) 而崩溃。

但根据 @objectstack/spec 协议，Widget.options 是 z.unknown()，业务层 columns 配置为 string[] 本应属于合法且更简洁、国际化友好的写法（header 可自动推导，多语言 UI 可基于 accessorKey + i18n 自动实现）。

建议修复

在 ObjectDataTable 中，对 columns 做 normalize，兼容 string[] 和对象数组：

const normalizeColumns = (columns) => {
  if (!columns || columns.length === 0) return [];
  if (typeof columns[0] === 'string') {
    return columns.map((field) => ({
      header: field.charAt(0).toUpperCase() + field.slice(1).replace(/_/g, ' '),
      accessorKey: field,
    }));
  }
  return columns;
};

这样即��无缝支持业务 schema 更简洁的 columns 字段，同时兼容国际化和完整对象写法。

参考

• Spec 协议对 columns 并无格式强约束
• ObjectDataTable 源码
• 讨论：ObjectUI Issue objectstack-ai/objectui#303

建议

• columns 支持 string[]/object[] 双语法，提升易用性和兼容性
• 推荐在文档中增加关于 columns 用法的说明示例
  </issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes ObjectDataTable: columns 应兼容 string[] 简写格式，自动 normalize 为对象数组 #1102

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
objectui-demo	Ready	Preview, Comment	Mar 21, 2026 7:59am

2 Skipped Deployments

Project	Deployment	Actions	Updated (UTC)
objectui	Ignored		Mar 21, 2026 7:59am
objectui-storybook	Ignored		Mar 21, 2026 7:59am

[Copilot]

Pull request overview

Adds string[] shorthand support for ObjectDataTable columns by normalizing entries into { header, accessorKey } objects, preventing downstream data-table crashes and improving header generation.

Changes:

• Introduced normalizeColumns() to convert string column keys into full column defs (supports mixed string | object arrays).
• Updated ObjectDataTable to pass normalized columns to the data-table renderer.
• Added unit tests for normalizeColumns plus component-level regression tests, and documented the feature in the changelog.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
packages/plugin-dashboard/src/ObjectDataTable.tsx	Normalizes schema.columns so string[] shorthand doesn’t crash downstream renderers.
packages/plugin-dashboard/src/tests/ObjectDataTable.test.tsx	Adds regression coverage for string[] columns and direct tests for normalization behavior.
CHANGELOG.md	Documents the new string[] shorthand support for ObjectDataTable columns.

Comments suppressed due to low confidence (1)

packages/plugin-dashboard/src/ObjectDataTable.tsx:143

• The auto-derived column headers (when schema.columns is empty) still don't normalize snake_case keys (e.g. close_date becomes Close_date) and use a different casing algorithm than normalizeColumns. Consider reusing normalizeColumns for the auto-derive path (e.g. by passing keys through it) to keep header derivation consistent and avoid duplicated formatting logic.

    }
    if (finalData.length === 0) return [];
    // Exclude internal/private fields (prefixed with '_') from auto-derived columns
    const keys = Object.keys(finalData[0]).filter(k => !k.startsWith('_'));
    // Convert camelCase keys to human-readable headers (e.g. firstName → First Name)