Images in Mmark

Images in Mmark are somewhat complicated, not in the least, because XML2RFC needs to output both HTML and text. To make that work you can specify multiple images in an artset and the renderer will pick the correct one, depending on the output.

To include artwork/source/images, you can:

• Use a code block. If this has a language specified, it will become a sourcecode otherwise a will be an artwork. The contents of both must be in plain text.
• Use subfigures in a figure block (!--) to group figure and potentially make them have a caption. We also use this syntax to support an artset, but only under special conditions (see below).

Code Blocks

A code block with a language will be turned into a sourcecode:

``` go
println("hello!")
```

If no language (the go above) is given it will be an artwork.

Figures in an Artset

To support artset we do the following. If multiple images are present as subfigures, we check if the name without the extension of the image destination (the file to be shown) is equal for all subfigures. If so, we assume an artset needs to be outputted and do so.

For example the following will result in a artset where an svg and an ascii-art version of the (hopefully) same image exists. Note the extension must be ascii-art because we use that to set the type and XML2RFC checks for that string.

!---
![Array vs Slice](array-vs-slice.svg "Title of the svg image")
![Array vs Slice](array-vs-slice.ascii-art "Title of the ascii-art image")
!---

Note this syntax is also supported for the manual page output and it does the same thing by only using the ascii-art version. This is true for all included imagery; only ascii-art ones are included in the output.

By some happy co-incidence a browser will not show the ascii-art version of the image when generating HTML. It remains to be seen if we need some code to actually filter these out.