tangle-md2tex

Extract LaTeX fenced code blocks from literate Markdown files.

中文文档见 README.zh-CN.md.

What It Does

tangle-md2tex reads Markdown files ending in .lit.md, extracts fenced code blocks whose info string is exactly latex, and writes the extracted LaTeX to a sibling .tex file.

For example:

```latex
First paragraph.

Second paragraph.
```

This prose explains why the LaTeX above is written this way.

becomes:

First paragraph.

Second paragraph.

Literate Markdown Format

Use .lit.md as the source file extension:

paper.lit.md -> paper.tex
sections/intro.lit.md -> sections/intro.tex

Only latex fenced code blocks are extracted:

```latex
\section{Introduction}

This is the manuscript text.
```

Explanation, notes, outlines, TODOs, or review comments can live here. They are
not copied into the generated `.tex` file.

Fence contents are copied as-is from markdown-it-py's token content. Leading blank lines, paragraph-separating blank lines, and trailing blank lines inside a latex fence are preserved because LaTeX uses blank lines to separate paragraphs.

Install

pip install tangle-md2tex

For local development:

python -m pip install -e .

Usage

Process one file:

tanglemd2tex paper.lit.md

Process every .lit.md file under a directory:

tanglemd2tex sections/

If no path is given, the current directory is scanned recursively:

tanglemd2tex

You can also run it as a module:

python -m tanglemd2tex paper.lit.md

Python API

from pathlib import Path

from tanglemd2tex import tangle_file, tangle_text

tex_text = tangle_text(markdown_text)
tex_path = tangle_file(Path("paper.lit.md"))

Rules

• Source files must end with .lit.md.
• Output files are written next to the source file with the .tex extension.
• Only fenced code blocks with info string latex are extracted.
• Non-LaTeX Markdown prose is treated as explanation and ignored by the output.
• The tool does not parse or validate LaTeX.

License

MIT