ExStruct Documentation¶
Welcome to the ExStruct docs. ExStruct has two primary capabilities: extracting structured data from Excel workbooks and editing workbooks through a public patch-based core.
Choose an Interface¶
| Use case | Start here | What is canonical |
|---|---|---|
| Embed extraction in Python code | API Reference | Extraction keeps the top-level Python API. ExStruct's editing API exists, but ordinary Python workbook editing is usually better served by openpyxl / xlwings. |
| Run local shell or agent workflows | CLI Guide | Editing commands such as exstruct patch and exstruct make are the canonical operational interface. |
| Run sandboxed or host-managed tools | MCP Server | MCP is the integration / compatibility layer and owns PathPolicy, transport, and artifact behavior. |
| Understand the product direction | Concept / Why ExStruct? | Extraction fidelity, layout semantics, and downstream AI use cases. |
| Check licensing details | Corporate Usage License Guide | Commercial-use guidance. |
At a Glance¶
- Extract structured JSON (default), YAML, and TOON from Excel workbooks.
- Edit workbooks primarily through the JSON-first editing CLI.
- Use
dry_run, warnings, diff output, and inverse-ops support for safer edit flows. - Keep MCP for restricted hosts that need path controls and tool transport.
Quick Links¶
- Install:
pip install exstruct - Extraction CLI:
exstruct input.xlsx --pretty --mode standard - Editing CLI:
exstruct patch --input book.xlsx --ops ops.json --dry-run