Closed
Description
Explain the problem.
When converting a reStructuredText figure
directive to HTML, options like figclass
and align
are only applied to the image and not the entire figure in the output. Based on the reference documentation for figure
, I think these options are intended to apply to the entire figure.
For example, if I run the following command:
$ pandoc --from=rst --to=html5 <<EOF
.. figure:: example.img
:figclass: foobar
This is a caption.
EOF
I would expect the output to be:
<figure class="foobar">
<img src="example.img" alt="This is a caption." /><figcaption aria-hidden="true">This is a caption.</figcaption>
</figure>
Instead, I get the following, where the class=foobar
attribute is part of the img
tag.
<figure>
<img src="example.img" class="foobar" alt="This is a caption." /><figcaption aria-hidden="true">This is a caption.</figcaption>
</figure>
This issue causes the image and caption to get separated if the CSS stylesheet tries using something like float
to align the figure.
Pandoc version?
I can reproduce this issue with the current development version (git) and pandoc 2.14.0.2 (pacman version: 2.14.0.2-2).
My OS is Arch Linux.