genhtml-markdown 1.0.8

bring python to markdown to generate HTML inline

# genhtml-markdown
[Python-Markdown](http://pythonhosted.org/Markdown/) plugin allowing to build HTML from inline python source.

Direct applications includes [charts and other plots](https://plot.ly/python/) in markdown documents, and tooling for [blogging](https://blog.getpelican.com/).

pip install genhtml-markdown

See the [compiled examples and their sources](examples/) for an introductory tour, the [Makefile](Makefile) for the process, or look at next section:


## Basic example with plotly
Let's take an [offline scatter plot example](https://plot.ly/python/getting-started/#initialization-for-offline-plotting), modified to print the generated HTML :

```python
import plotly.offline
import plotly.graph_objs as go

data = go.Scatter(x=[1, 2, 3, 4], y=[4, 3, 2, 1])
layout = go.Layout(title="hello world")

figure = go.Figure(data=[data], layout=layout)
print(offline.plot(figure, output_type='div'))
```

You could want to include it in your article. With genhtml-markdown extension, it's easy:

```genhtml
import plotly.offline
import plotly.graph_objs as go

data = go.Scatter(x=[1, 2, 3, 4], y=[4, 3, 2, 1])
layout = go.Layout(title="hello world")

figure = go.Figure(data=[data], layout=layout)
print(offline.plot(figure, output_type='div'))
```

You think it's verbose ? I do too. That's the reason we have *headers* and *footers*, patches of python that will be put respectively *before* and *after* our specific code. By default, the header is full of imports, including the plotly ones. And the two last lines are provided by the *plot* footer. So, here is our final code:

```genhtml footer=plot
data = go.Scatter(x=[1, 2, 3, 4], y=[4, 3, 2, 1])
layout = go.Layout(title="hello world")
```

Imports and offline plotting boilerplate code will be added by headers/footers, the total code will be ran, and finally the output will be included in place as an interactive plot/chart.


## Features

### Ready-to-use headers and footers
You can see the full list of [headers](headers/) and [footers](footers/) in their respective directories. You can also pass `headers_dir` and `footers_dir` [parameters for the extensions](https://python-markdown.github.io/cli/#using-extensions) in order to provide your own !

For instance, with the parameter `-c config.json` added to `python -m markdown` call, you can feed markdown with the following parameters:

```json
{
"genhtml": {
"headers_dir": "alt-headers"
}
}
```

Indicating that genhtml will also look in the `alt-headers/` directory for headers.
Options can also be set when calling [markdown module programatically](https://python-markdown.github.io/extensions/api/#configsettings) with something like `markdown.Markdown(extensions=[GenHTMLMarkdownExtension(headers_dir='~/my-headers-dir'])`.

Finally, note that the [default header](headers/default.py) provides a lot of imports.

### Generate and show raw images
As shown in [images example](examples/images.mkd),
it is quite easy to build and show an image inline :

```genhtml format=png footer=png-image
from PIL import Image
image = Image.new('RGB', (60, 30), color='green')
```

The `format=png` options tells that the printed data is (base64-encoded) raw png data,
and the [*png-image* footer](genhtml/footers/png-image.py) reads something like:

```python
import io
import base64
with io.BytesIO() as output:
image.save(output, format='png')
print(base64.b64encode(output.getvalue()).decode(), end='')
```

Other formats are `jpg` and `svg`, allowing you to [bring gizeh to your markdown](https://github.com/Zulko/gizeh).


### Graphs
Using the previous feature, it becomes possible to draw graphs from their networkx definition, as shown in [the related example](examples/networkx_and_dot.mkd).

Using the *dot-png* footer, you can just build your graph and print it in no time:

```genhtml format=png footer=dot-png
graph = nx.fast_gnp_random_graph(10, 0.5)
# draw the first one in beige
graph.nodes[1]['style'] = 'filled'
graph.nodes[1]['fillcolor'] = 'beige'
```

### Show sources, and other code manipulations
See the related [example](examples/arbitrary-python.mkd).

### Nested code generation
Is fully supported. See [pyception example](examples/pyception.mkd).


### Use static data
You have in the code access to active directory.
For instance, if using [pelican](https://blog.getpelican.com/), you can read all your blog articles in few lines:

```genhtml
import glob
print('Follows a list of published articles:')
for article in glob.glob('content/articles/*.mkd'):
if any(line.strip() == 'status: published' for line in open(article)):
print(f'- {article}')
```

Following this logic, if you are using recurrent CSV data, you could put it in a dedicated directory,
and access it easily from all your integrated python code.

### Global environment
An environment is shared among codes in the same document.
As shown in [related example](examples/env-management.mkd), you can access it with the flag `global-env=true`.


## Incoming features
- access to all markdown article from python (allowing to access the article itself)
- headers/footers for [biseau](https://gitlab.inria.fr/lbourneu/biseau), in order to allow ASP in markdown (!)
- [graphs](http://js.cytoscape.org/) and [powergraphs with cytoscape.js](https://github.com/Aluriak/bubble-tools/#conversion-to-cytoscapejs)