Move the building of the <figure /> and <figcaption /> elements into
subfunctions to improve readability.
Support images nested inside of links
Because Python's ElementTree's XPath syntax is not really complete, we
need a strange matching function in order to recognize the images.
Preserve attributes set by attr_list
In order to preserve attributes of the image set by the built-in
attr_list extension, yafg has to have a different priority. 8 seems to
be the sweet spot, 9 doesn't work. Note that this is just a hacky
heuristic and may break other plugins.
Additionally, another processor seems to run too early or late now,
because we have to manually add a newline after each figure tag in
order to preserve the current behaviour.
This will haunt me one day.
Provide IDs for figures
Given IDs, the figures can be link targets.
Make the figure number title user-changeable
Allow custom classes for figure numbers
Add the possibility for figure numbers
requirements.txt: Allow for newer Markdown versions
Add support for CSS classes
Support both Python 2.7 and Python 3.6
Also test both versions in the CI pipeline
Instead of installing with
pip install -e .
during development and handling dependencies that way, just use
virtualenv as everyone else does.
Add a .build.yml
That file will run all unit tests and if the latest git tag matches the
version noted in setup.py, it will upload the package to PyPI.
This includes a first working version of the yafg Markdown plugin. It
can handle multi-line captions and has an option to strip the original
title from the image.
It also includes a setup.py which should enable an upload to PyPI, but
no .build.yaml, yet.