string2path

The string2path R package converts a text to paths of the outlines of each glyph, based on a font data. Under the hood, this package is powered by the savvy framework to use these two Rust crates:

• ttf-parser for parsing font data. TrueType font (.ttf) and OpenType font (.otf) are supported.
• lyon for tessellation of polygons and flattening the curves.

Installation

install.packages("string2path")

Development version

The development version is available on R-universe:

install.packages("string2path",
  repos = c(
    yutannihilation = "https://yutannihilation.r-universe.dev",
    CRAN = "https://cloud.r-project.org"
  )
)

If you want to install from source, you need to have Rust toolchain installed before trying to install this package. See https://www.rust-lang.org/tools/install for the installation instructions.

Example

string2path()

library(string2path)
library(ggplot2)

d <- string2path("カラテが\n高まる。", "Noto Sans JP", font_weight = "bold")

d <- tibble::rowid_to_column(d)

ggplot(d) +
  geom_path(aes(x, y, group = path_id, colour = factor(glyph_id)),
            linewidth = 1.5) +
  theme_minimal() +
  coord_equal() +
  theme(legend.position = "top") +
  scale_colour_viridis_d(option = "H")

library(gganimate)
d <- string2path("蹴", "Noto Sans JP")
d <- tibble::rowid_to_column(d)

ggplot(d) +
  geom_path(aes(x, y, group = path_id),
            linewidth = 2, colour = "purple2", lineend = "round") +
  theme_minimal() +
  coord_equal() +
  transition_reveal(rowid)

dump_fontdb()

Note that "Noto Sans JP" above (and "Iosevka SS08" below) is the font installed on my local machine, so the same code might not run on your environment. You can use dump_fontdb() to see the available combination of font family (e.g. "Arial"), weight (e.g. "bold"), and style (e.g. "italic").

dump_fontdb()
#> # A tibble: 449 × 5
#>    source                                  index family        weight style 
#>    <chr>                                   <int> <chr>         <chr>  <chr> 
#>  1 "C:\\WINDOWS\\Fonts\\arial.ttf"             0 Arial         normal normal
#>  2 "C:\\WINDOWS\\Fonts\\arialbd.ttf"           0 Arial         bold   normal
#>  3 "C:\\WINDOWS\\Fonts\\arialbi.ttf"           0 Arial         bold   italic
#>  4 "C:\\WINDOWS\\Fonts\\ariali.ttf"            0 Arial         normal italic
#>  5 "C:\\WINDOWS\\Fonts\\ariblk.ttf"            0 Arial         black  normal
#>  6 "C:\\WINDOWS\\Fonts\\bahnschrift.ttf"       0 Bahnschrift   normal normal
#>  7 "C:\\WINDOWS\\Fonts\\BIZ-UDGothicB.ttc"     0 BIZ UDGothic  bold   normal
#>  8 "C:\\WINDOWS\\Fonts\\BIZ-UDGothicB.ttc"     1 BIZ UDPGothic bold   normal
#>  9 "C:\\WINDOWS\\Fonts\\BIZ-UDGothicR.ttc"     0 BIZ UDGothic  normal normal
#> 10 "C:\\WINDOWS\\Fonts\\BIZ-UDGothicR.ttc"     1 BIZ UDPGothic normal normal
#> # ℹ 439 more rows

You can also specify the font file directly. Pomicons is a font available on gabrielelana/pomicons, licensed under SIL OFL 1.1.

pomicons_file <- here::here("fonts", "Pomicons.ttf")
if (!file.exists(pomicons_file)) {
  dir.create(dirname(pomicons_file))
  curl::curl_download(
    "https://github.com/gabrielelana/pomicons/blob/master/fonts/Pomicons.ttf?raw=true",
    destfile = pomicons_file
  )
}

d_tmp <- string2path("\uE007", pomicons_file)

ggplot(d_tmp) +
  geom_path(aes(x, y, group = path_id), linewidth = 5, colour = "#26d1a9") +
  theme_minimal() +
  coord_equal()

string2fill()

d <- string2fill("abc", "Iosevka SS08", font_weight = "bold", font_style = "italic")

ggplot(d) +
  geom_polygon(aes(x, y, group = triangle_id, fill = factor(triangle_id %% 7)),
               colour = "grey", linewidth = 0.1) +
  theme_minimal() +
  coord_equal() +
  theme(legend.position = "none") +
  scale_fill_viridis_d(option = "H")

string2stroke()

for (w in 1:9 * 0.01) {
  d <- string2stroke("abc","Iosevka SS08", font_weight = "bold", font_style = "italic", line_width = w)
  
  p <- ggplot(d) +
    geom_polygon(aes(x, y, group = triangle_id, fill = factor(triangle_id %% 2)),
                 colour = "grey", linewidth = 0.1) +
    theme_minimal() +
    coord_equal() +
    theme(legend.position = "none") +
    scale_fill_manual(values = c("purple", "pink"))
  plot(p)
}

tolerance

tolerance controls resolution of the tessellation. You can reduce tolerance to get higher resolutions. In most of the cases, 1e-5 ~ 1e-6 should be enough. For more details, please refer to lyon’s official document.

for (tolerance in c(1e-1, 1e-2, 1e-3, 1e-4, 1e-5, 1e-6, 1e-7)) {
  d <- string2fill("abc", "Iosevka SS08", font_weight = "bold", font_style = "italic", tolerance = tolerance)
  
  p <- ggplot(d) +
    geom_polygon(aes(x, y, group = triangle_id),
                 fill = "transparent", colour = "black", linewidth = 0.5) +
    theme_minimal() +
    coord_equal() +
    ggtitle(paste0("tolerance: ", tolerance))
  plot(p)
}