markdown-kroki 0.3.0

Diagram extension for Python-Markdown using Kroki server.

markdown-kroki

Diagram extension for Python-Markdown using Kroki server.

This extension converts various diagram code blocks into Base64 encoded data: URI or direct image link. This enables PDF generation with tools like MkDocs to PDF/WeasyPrint without requiring JavaScript(e.g. Mermaid)

Install

pip install markdown-kroki

Requirements

Internet access to the public Kroki server

Default setting with no options.

Self-Managed Kroki server (recommended)

Here is a sample Docker Compose file.

ref. Kroki.io > Install > Using Docker or Podman

docker compose up -d

The default port used by MkDocs (mkdocs serve) may conflict with the default port of a Dockerized Kroki instance. Consequently, you will need to change the port configuration for one of them.

Usage

```{diagram language} formant=[svg|png] {img tag attribute}="value" {diagram option}="value"
```

• format (optional): Output image format default to svg
• img tag attribute (optional): alt, width, height, class, id, style, title
• diagram option (optional): refer to Diagram options

MkDocs Integration

# mkdocs.yml
markdown_extensions:
  - markdown_kroki:
      kroki_url: http://localhost:18000  # default: https://kroki.io
      img_src: link                      # default: data/data URI, link/direct link

Pelican Integration

# pelicanconf.py
MARKDOWN = {
    'extension_configs': {
        'markdown.extensions.codehilite': {'css_class': 'highlight'},
        'markdown.extensions.extra': {},
        'markdown_kroki': {                         # Add these
            'kroki_url': 'http://localhost:18000',
            'img_src': 'link'
        },
    },
    'output_format': 'html5',
}

Python code

import markdown
from markdown_kroki import KrokiDiagramExtension

markdown_text = """```plantuml format="svg" theme="sketchy-outline" width="300"
@startuml
Alice -> Bob: Authentication Request
Bob --> Alice: Authentication Response
@enduml
```"""

html_output = markdown.markdown(markdown_text, extensions=[
                                KrokiDiagramExtension(kroki_url='https://kroki.io')])

print(html_output)

<p><img src="data:image/svg+xml;base64,PHN2ZyBhcmlhLXJvbGVkZXNjcmlwdGlvbj0ic2VxdWVuY2UiIHJvbGU
9ImdyYXBoaWNzLWRvY3VtZW50IGRvY3VtZW50IiB2aWV3Qm94PSItNTAgLTEwIDc1MCA1NzQiIHN0eWxlPSJtYXgtd2lkd
Gg6IDc1MHB4OyBiYWNrZ3JvdW5kLWNvbG9yOiB3aGl0ZTsiIHhtbG5zOnhsaW5rPSJodHRwOi8vd3d3LnczLm9yZy8xOTk
...
...
...
IHgxPSIyNzYiLz48L3N2Zz4=" width="300" ></p>

Process flow

data URI

Direct link (GET API)