docgen

Русский

Generate structured DOCX documents and PPTX presentations from validated JSON configurations.

docgen provides a small CLI and Python API for data-driven document generation without assumptions about organization, subject, or document category.

What it does

• Builds DOCX files from ordered sections, reusable content blocks and fragments.
• Builds PPTX files from typed slide configurations.
• Expands opt-in scalar variables in user-visible text fields.
• Generates charts and structure diagrams.
• Embeds images and inline or external CSV tables by stable resource IDs.
• Validates the complete project before writing output atomically.

Installation

Python 3.10 or newer is required.

python -m pip install .

For development:

python -m pip install -e '.[dev]'
python -m pytest -q

Quick start

docgen init ./my-document
docgen build ./my-document/document.json

docgen init ./my-slides --kind presentation
docgen build ./my-slides/presentation.json

Override the configured output with docgen build CONFIG -o OUTPUT. Validate without writing output with docgen validate CONFIG.

Minimal document

{
  "kind": "document",
  "output": "result.docx",
  "sections": [
    {
      "type": "content",
      "title": "Overview",
      "content": [
        {"type": "paragraph", "text": "Hello."}
      ]
    }
  ]
}

Save as document.json, then run docgen validate document.json and docgen build document.json.

Minimal presentation

{
  "kind": "presentation",
  "output": "result.pptx",
  "title": "Demo",
  "slides": [
    {"type": "title"},
    {"type": "text", "title": "One", "text": "Hello."}
  ]
}

Save as presentation.json, then run docgen validate presentation.json and docgen build presentation.json.

Documentation

• Documents: sections, blocks, rich text, citations, and formatting.
• Presentations: typed slides, rich text, image fitting, and overflow.
• Variables and fragments: authoring expansion, normalization, and limits.
• Resources: tables, figures, diagrams, images, and CSV sources.
• CLI reference: commands, validation, diagnostics, and schemas.
• Development: development commands and known limitations.