NbConvert is both a library and command line tool that allows you to convert notebooks to other formats. It ships with many common formats: html, latex, markdown, python, rst, and slides
NbConvert relys on the Jinja templating engine, so implementing a new format or tweeking an existing one is easy.
You can invoke nbconvert by running
$ ipython nbconvert <options and arguments>
Call ipython nbconvert with the --help flag or without any aruments to display the basic help. For detailed configuration help, use the --help-all flag.
As a test, the Index.ipynb notebook in the directory will be convert.
If you're converting a notebook with code in it, make sure to run the code cells that you're interested in before attempting to convert the notebook. Unless explicitly requested, nbconvert does not execute the code cells of the notebooks that it converts.
In [ ]:
%%bash
ipython nbconvert 'Index.ipynb'
Html is the (configurable) default value. The verbose form of the same command as above is
In [ ]:
%%bash
ipython nbconvert --to=html 'Index.ipynb'
You can also convert to latex, which will extract the embeded images. If the embeded images are SVGs, inkscape is used to convert them to pdf:
In [ ]:
%%bash
ipython nbconvert --to=latex 'Index.ipynb'
Note that the latex conversion creates latex, not a PDF. To create a PDF you need the required third party packages to compile the latex.
A --post flag is provided for convinience which allows you to have nbconvert automatically compile a PDF for you from your output.
In [ ]:
%%bash
ipython nbconvert --to=latex 'Index.ipynb' --post=pdf
Look at the first 20 lines of the python exporter
In [ ]:
pyfile = !ipython nbconvert --to python 'Index.ipynb' --stdout
for l in pyfile[20:40]:
print l
From the code, you can see that non-code cells are also exported. If you want to change this behavior, you can use a custom template. The custom template inherits from the Python template and overwrites the markdown blocks so that they are empty.
In [ ]:
%%writefile simplepython.tpl
{% extends 'python.tpl'%}
{% block markdowncell -%}
{% endblock markdowncell %}
## we also want to get rig of header cell
{% block headingcell -%}
{% endblock headingcell %}
## and let's change the appearance of input prompt
{% block in_prompt %}
# This was input cell with prompt number : {{ cell.prompt_number if cell.prompt_number else ' ' }}
{%- endblock in_prompt %}
In [ ]:
pyfile = !ipython nbconvert --to python 'Index.ipynb' --stdout --template=simplepython.tpl
for l in pyfile[4:40]:
print l
print '...'
For details about the template syntax, refer to Jinja's manual.
The notebook file format supports attaching arbitrary JSON metadata to each cell. Here, as an exercise, you will use the metadata to tags cells.
First you need to choose another notebook you want to convert to html, and tag some of the cells with metadata. You can refere to the file soln/celldiff.js as an example or follow the Javascript tutorial to figure out how do change cell metadata. Assuming you have a notebook with some of the cells tagged as Easy|Medium|Hard|<None>, the notebook can be converted specially using a custom template. Design your template in the cells provided below.
The following, unorganized lines of code, may be of help:
{% extends 'html_full.tpl'%}
{% block any_cell %}
{{ super() }}
<div style="background-color:red">
<div style='background-color:orange'>If your key name under cell.metadata.example.difficulty, the following code would get the value of it:
cell['metadata'].get('example',{}).get('difficulty','')
Tip: Use %%writefile to edit the template in the notebook.
In [ ]:
%%bash
# ipython nbconvert --to html <your chosen notebook.ipynb> --template=<your template file>
In [ ]:
%loadpy soln/coloreddiff.tpl
In [ ]:
# ipython nbconvert --to html '04 - Custom Display Logic.ipynb' --template=soln/coloreddiff.tpl
IPython nbconvert can be configured using the default profile or a profile specified via the --profile flag. Additionally, if a config.py file exist in current working directory, nbconvert will use that as config.