text-to-excel

Text-to-Excel

将结构化文本转换为格式规整的Excel文件。

Workflow

Step 1: Parse User Input

Analyze user-provided structured text. Supported input formats:

• Markdown tables (| col1 | col2 |)
• Plaintext tables (tab/space-separated)
• CSV/TSV data
• Lists (bulleted, numbered)
• Key-value pairs
• JSON/dict-like structures
• Free-form text with identifiable tabular structure
• Descriptions of desired table content

Identify: headers, data rows, data types per column, and any special formatting requests.

Step 2: Ask for Output Path

Ask the user for the save path. Suggest a reasonable default filename based on the content (e.g. ~/Desktop/sales_data.xlsx).

Step 3: Build JSON Config

Construct a JSON config object following the schema in references/config_schema.md.

Minimal example:

{
  "sheets": [{
    "name": "Sheet1",
    "headers": ["Name", "Age", "City"],
    "data": [
      ["Alice", 30, "Beijing"],
      ["Bob", 25, "Shanghai"]
    ]
  }]
}

Full-featured example with styles, charts, and validation:

{
  "sheets": [{
    "name": "Sales",
    "title_row": {"text": "Q1 Sales Report", "style": {"font": {"bold": true, "size": 16}}},
    "headers": ["Month", "Revenue", "Growth"],
    "data": [
      ["January", 50000, 0.12],
      ["February", 62000, 0.24],
      ["March", 58000, 0.16]
    ],
    "column_types": ["text", "currency_cny", "percent"],
    "freeze_pane": "A2",
    "auto_filter": true,
    "charts": [{
      "type": "bar",
      "title": "Monthly Revenue",
      "x_column": 1,
      "y_columns": [2],
      "position": "E2"
    }],
    "conditional_formats": [{
      "type": "color_scale",
      "range": "B2:B4"
    }]
  }]
}

Step 4: Generate Excel

1. Write the JSON config to a temp file
2. Run: python3 scripts/generate_excel.py <config.json> <output_path>
3. Verify the file was created
4. Report success with the file path

Style Defaults

The script applies professional defaults automatically:

• Headers: Blue background (#4472C4), white bold text, centered
• Data rows: Alternating gray/white bands, thin borders
• Column widths: Auto-calculated based on content length
• Title row (optional): Large bold text, merged across all columns