Open jgm opened 2 years ago
Do you mean supporting a feature where you can write something like, "see figure 4", and that will render as a link to an image (as in ![figure 4](figure-4.png)
?
No, this is about creating the figure itself: an environment that may contain one or more images, or perhaps other things, plus a caption and a label that can be the target of references.
Would the current way to get something similar to this be to add the figure as an image, like this:
![figure 4](figure-4.png){#fig-4}
and then link to it with "[see figure 4](#fig-4)
"?
In pandoc's markdown that will produce a figure (implicit_figures
extension), but we haven't implemented that here. I think I'd prefer to have something more explicit for generating figures.
What about
!!!label
![alt text](path/to/image.png)
![other alt text](path/to/other.png)
:: Short caption on one line
Long caption which may have multiple lines/paragraphs.
!!!
And if you need a grid you simply use table structure (pipe or list).
Ok, I understand. Thanks.
I wonder if there's a way to re-use the div syntax for this, like
::: figure
foo ...
:::
Similar to the list tables idea https://github.com/jgm/djot/issues/27 .
Actually, if I were "delimiter-shopping" for a punctuation character to use for figures, it would be hard to beat &
--- I mean, it looks just like a figure that a figure skater might make. Stealing @bpj 's example:
&&& label
![alt text](path/to/image.png)
![other alt text](path/to/other.png)
Short caption on one line
Long caption which may have multiple
lines/paragraphs.
&&&
The point of using the exclamation point is that it is already used to differentiate inline images from links.
Yeah, I see that, though I don't think the !!!
looks good.
If ---
is free, this looks good:
--- figure 4
![alt text](path/to/image.png)
![other alt text](path/to/other.png)
Short caption on one line
Long caption which may have multiple
lines/paragraphs.
---
I don't know how you could enhance that to support figures containing subfigures, as in "figure 4(a), 4(b), etc." Maybe using a definition list inside the figure block, e.g.:
--- figure 4
: figure 4(a)
![alt text](path/to/figure-4-a.png)
Short caption for fig 4(a), on one line
Long caption which may have multiple
lines/paragraphs.
: figure 4(b)
![alt text](path/to/figure-4-b.png)
Short caption for fig 4(b), on one line
Long caption which may have multiple
lines/paragraphs.
---
Whatever solution is considered for figures and subfigures, captions should probably be generalized with a consistent syntax across blocks where they are appropriate:
^
caption syntax currently...)See also #28
FWIW: Given how flexible Djot’s syntax already is, it may not be necessary to introduce new syntax – e.g.:
[Figure](#fig:diagram){!float} visualizes how everything is connected.
{!float for="Figure" #fig:diagram}
:::
![](img/venn-diagram.svg){width="277.164" height="176.5"}
---
Visual explanation.
:::
[Table](#tbl:summary){!float} contains a summary.
{!float for="Table" #tbl:summary}
:::
| Output | Vector | Bitmap |
|--------|--------|--------|
| HTML | `.svg` | `.jpg` |
| LaTeX | `.pdf` | `.jpg` |
---
Brief summary.
:::
Notes:
!float
is a command (which I use in my Markdown dialect), it could also be the class .float
. Commands work much like classes but have a separate namespace and don’t produce output – unless there is a plugin/filter.
We need a general way of producing a figure with a caption and label.
Pandoc's "implicit figures" are too limiting. Figures can include multiple images, and also non-image content like code.