Generate Typst from R • r2typ

r2typ lets you generate Typst markup using R functions. It supports all of the following:

✅ almost all Typst functions
✅ conversions from R to Typst (NULL -> none, TRUE -> true, etc.)
✅ Typst colors, alignment, units and direction natively
✅ set and (simple) show rules
✅ works well with Quarto
✅ extremely simple syntax
✅ very lightweight

[!WARNING]
The project is still early stage and contains a few bugs.

Installation

# install.packages("pak")
pak::pkg_install("y-sunflower/r2typ")

Quick start

Basic usage

library(r2typ)

heading(level = 2, numbering = "1.1", "Hello world")
#> #heading(level: 2, numbering: "1.1")[Hello world]

text(size = pt(12), baseline = em(1.2), overhang = FALSE, "hey there")
#> #text(size: 12pt, baseline: 1.2em, overhang: false)[hey there]

image(width = percent(80), height = auto, "link.svg")
#> #image(width: 80%, height: auto, "link.svg")

circle(fill = blue, "hey")
#> #circle(fill: blue)[hey]

circle(radius = pt(100), "hey", linebreak(), "there")
#> #circle(radius: 100pt)[hey #linebreak() there]

place(top + left, dy = pt(15), square(size = pt(35), fill = red))
#> #place(top + left, dy: 15pt)[#square(size: 35pt, fill: red)]

Validating Typst

Functions in r2typ accept all positional and named arguments! This means that you’re responsible of making sure the arguments you’re using are valid!

But to help you in that process, there is a is_valid_typst() function that will return either TRUE or FALSE depending on whether your Typst can be compiled successfully.

place(
  top + left,
  dy = pt(15),
  square(size = pt(35), fill = red)
) |>
  is_valid_typst()
#> TRUE

Also note that all examples in the r2typ documentation are valid Typst examples.

Types conversion

r2typ converts some R types into Typst types:

NULL becomes none

image("image.png", width = percent(80), alt = NULL)
#> #image(width: 80%, alt: none, \"image.png\")

TRUE/FALSE become true/false

list_(tight = FALSE, "hey", "you")
#> #list(tight: false, [hey], [you])

c() vectors and unnamed list() (such as list("a", "b") become arrays:

text(`stylistic-set` = c(1, 2, 3), "10 years ago")
#> #text(stylistic-set: (1, 2, 3))[10 years ago]

text(`stylistic-set` = list(1, 2, 3), "10 years ago") # equivalent
#> #text(stylistic-set: (1, 2, 3))[10 years ago]

Named list() (such as list(a = "hello", b = "world")) become dictionnaries:

text(costs = list(hyphenation = percent(100), runt = percent(100)))
#> #text(costs: (hyphenation: 100%, runt: 100%))

Set and show rules

set_text(red, size = pt(20))
#> #set text(size: 20pt, red)

show_("heading", set_text(red, size = pt(20)))
#> #show heading: set text(size: 20pt, red)

Complete example

A complete example that generates a PDF using R only:

c(
  set_text(purple),
  set_circle(width = percent(50)),
  align(
    center + horizon,
    circle(
      fill = aqua,
      stroke = pt(5) + red,
      align(
        right,
        text(
          font = "Roboto",
          size = em(1.2),
          "My favorite food is cookies!"
        )
      )
    )
  )
) |>
  typst_write() |>
  typst_compile(output = "here.pdf")

Learn more in the get started vignette.

Markup VS Code mode in Typst

r2typ generates Typst markup, not Typst code. Most people, when writing native Typst, rely primarily on markup mode. Code mode is mainly used to add logic or create functions.

This is an important distinction to keep in mind, but the core difference is that function calls start with a # (e.g., #text("hey") VS text("hey")).

You can learn more about it here.

Missing functions

✅ Text
✅ Foundations
✅ Model, everything except: cite, link, numbering, ref, terms
✅ Layout, everything except: columns, layout, measure, repeat, rotate
✅ Visualize, everything except: curve, gradient, path, stroke, polygon, tiling

something’s missing? Please open an issue!

r2typ (R to Typst): An R package for generating Typst markup