mddoco 2.0.1

Markdown to HTML/PDF document converter

mddoco

A CLI tool that converts markdown files to a single HTML (or future PDF) document.

Installation

pip install mddoco
playwright install chromium

Or for development:

pip install -e .
playwright install chromium

Playwright (Chromium) is required for PDF output only. HTML output works without it.

Usage

mddoco [OPTIONS] INPUT_PATH

INPUT_PATH can be a directory (all *.md and *.md.j2 files are found recursively and sorted together) or a single .md or .md.j2 file.

Options

Option	Short	Default	Description
--output PATH	-o	.	Directory to write the output file
--format [html|pdf]	-f	html	Output format
--title TEXT	-t	(none)	Document title shown at the top of the page
--theme NAME		default	Theme to use for rendering
--toc / --no-toc		off	Generate a table of contents
--toc-depth N		3	Maximum heading depth included in the TOC (1–6)

Examples

Convert all markdown files in a directory:

mddoco ./docs

Convert a single file:

mddoco README.md

With a title, TOC, and custom output directory:

mddoco ./docs --title "My Project" --toc --output ./out

Output

All matched markdown files are combined into a single HTML document in the output directory. The filename is derived from the input folder or file name.

Jinja2 templates

Files ending in .md.j2 are Jinja2 templates that produce Markdown. They are sorted alongside regular .md files (numeric prefixes such as 01_, 02_ determine order) and converted through the same pipeline.

Context variables (JSON and CSV)

Place *.json or *.csv files in the same input directory. Each file's stem becomes a top-level variable in the Jinja2 context — report.json is available as report, people.csv as people, and so on. If both data.json and data.csv exist, the tool will exit with an error.

docs/
  01_intro.md
  02_summary.md.j2    ← Jinja2 template
  report.json         ← available as {{ report }}
  people.csv          ← available as {{ people }}

JSON files are loaded as-is; the variable holds whatever structure the JSON contains.

CSV files are loaded as a list of row dicts, one dict per row:

{% for person in people %}
- {{ person.name }} ({{ person.role }})
{% endfor %}

Cell values containing ; are automatically split into a list:

name,skills
Alice,python;flask;sql
Bob,java

{{ people[0].skills }}   {# → ["python", "flask", "sql"] #}
{{ people[1].skills }}   {# → "java" (plain string — no ;) #}

Quoting a field in the CSV prevents ; splitting — "python;flask" remains a single string.

Excluding files

Files whose name starts with _ are excluded from scanning. Use this for Jinja2 macro files that should be imported but not rendered as documents:

docs/
  _macros.j2          ← excluded; safe to use with {% import %}
  01_intro.md
  02_report.md.j2

Using macros

Use {% import %} or {% from ... import %} to make macros from another file callable — not {% include %}. {% include %} injects rendered output only; macros defined in an included file are not visible to the calling template and will raise an undefined error.

{# correct — macro is callable after this #}
{% import "_macros.j2" as macros %}
{{ macros.val(item) }}

{# also correct #}
{% from "_macros.j2" import val %}
{{ val(item) }}

{# wrong — val will be undefined #}
{% include "_macros.j2" %}
{{ val(item) }}

Themes

Themes are self-contained Jinja2 HTML files with embedded CSS. Pass a theme name with --theme NAME.

Theme	Description
default	Clean GitHub-style layout, 860 px centred column
default-wide	Same as default but full viewport width
professional	Corporate blue palette, dark title banner, card layout, 900 px centred
professional-wide	Same as professional but full viewport width
dark	Dark background, blue accent, light text — good for technical docs, 860 px centred
dark-wide	Same as dark but full viewport width
academic	Serif (Georgia) typography, justified text, print-optimised, 720 px centred
academic-wide	Same as academic but full viewport width
vanilla	Minimal HTML with no styling — useful for testing or custom CSS
paged-professional	Corporate style with Paged.js pagination for print-ready PDFs

All themes support:

• Optional document title
• Optional table of contents
• Mermaid.js diagram support (loaded only when diagrams are present)
• Graph diagram blocks

Graph diagrams

Use ```graph fenced blocks with a JSON payload. The only required field is data.

Minimal example — all defaults applied:

```graph
{
  "data": {
    "x": ["Jan", "Feb", "Mar"],
    "Sales": [100, 150, 120]
  }
}
```

Series are inferred from the data keys (everything except x). Each series defaults to a line chart in blue.

Full schema:

Field	Required	Default	Description
data.x	yes	—	Category labels
data.<name>	yes	—	Values for each series (one key per series)
title	no	(none)	Chart title
orientation	no	vertical	vertical or horizontal
show_legend	no	true	Show the legend
width_px	no	640	Width in pixels
height_px	no	480	Height in pixels
min / max	no	—	Axis bounds
series	no	(auto)	Override series definitions (see below)

Series fields (all optional when auto-generated):

Field	Default	Description
label	(data key)	Must match a key in data
type	line	bar, line (solid), or line2 (dotted)
colour	(palette)	Colour string, or list of colours per bar
marker	false	Show point markers (line types only)

Full example:

```graph
{
  "title": "Sales vs Target",
  "orientation": "vertical",
  "show_legend": true,
  "data": {
    "x": ["Jan", "Feb", "Mar", "Apr", "May"],
    "Sales":  [85, 92, 78, 96, 110],
    "Target": [90, 90, 90, 90, 90]
  },
  "series": [
    { "label": "Sales",  "type": "bar",  "colour": "#3498db" },
    { "label": "Target", "type": "line", "colour": "#e74c3c", "marker": false }
  ]
}
```

Mermaid diagrams

Use standard fenced code blocks in your markdown:

```mermaid
graph TD
    A[Start] --> B[End]
```

mddoco detects mermaid blocks automatically and loads the Mermaid.js CDN only when needed.