Skip to main content

Sphinx configuration and libraries for Safari Books Online documentation

Project description

Overview

Safari technical documentation is now being written and collected in a form that can be processed by Sphinx, a utility for generating documentation in HTML, PDF, Epub, and other formats from text files using reST (reStructuredText) wiki markup. In addition to writing docs directly, we can have Sphinx grab API documentation from our core programming languages:

  • Python docstrings are collected using the sphinx-apidoc command.

  • JSDoc-formatted comments in JavaScript are collected using the JsDoc Toolkit RST-Template library, which in turn uses jsdoc-toolkit.

Web service APIs can be documented using httpdomain from sphinx-contrib.

Installation

pip install sbo-sphinx

To build JavaScript API documentation, you’ll also need java and ant.

Settings

sbo-sphinx uses the standard Sphinx conf.py file, but offloads the vast majority of the configuration to an sbo_sphinx.conf module which should be appropriate for most SBO projects. Hence a minimal docs/conf.py file can be as simple as:

from sbo_sphinx.conf import *

project = 'my_project_name'

There should also be a docs/index.rst file to serve as the documentation home page; see the one in this project for an example.

There are additional settings for the extensions which auto-generate Python and JavaScript API documentation. See sbo_sphinx.apidoc and sbo_sphinx.jsdoc for details.

Usage

Use the standard sphinx-build syntax. For the usual case of wanting to generate the documentation in HTML format:

sphinx-build -b html . _build

External Files

reStructuredText not inside the docs directory hierarchy can’t be directly included in a table of contents. To include a README.rst file from the repository’s root directory in the generated documentation, create a placeholder inside the docs directory which uses an include directive to pull in its content:

.. include:: ../README.rst

For an example, see docs/readme.rst in this project.

PyPI Description Validation

If PyPI encounters something it doesn’t know how to handle in a reStructuredText package description, it just silently shows it as plain text instead of formatting it as expected. To get some warning of this before uploading your package, install the readme package and run:

python setup.py check --restructuredtext --strict

Note that for this to work, long_description must be set in setup.py (usually by loading it from a README.rst file or such). If you instead count on setting description-file in setup.cfg, the reStructuredText that will be used for the PyPI page isn’t available to the check command.

Markdown

Sphinx currently has no real support for Markdown-style wiki markup. If a project has a README.md which you want to include in the documentation, there are a few options:

  • Convert it to README.rst instead, changing the markup accordingly. pandoc may do a reasonably good job of automating this conversion.

  • Add a reStructuredText-formatted copy of the file to the docs directory and include that in the documentation instead. This does run the risk of the copy getting out of sync with the original, however.

  • Implement a Sphinx extension which uses pandoc to automatically convert and copy the Markdown files specified in a configured list. The drawback with this approach is that it requires pandoc to be installed on each system on which the documentation will be generated.

Read the Docs

sbo-sphinx was written to be mostly compatible with the Read the Docs service, but keep in mind that private source code repositories cannot be used on the public Read the Docs service (but can be on a suitably configured private installation).

Notes

  • The table of contents page for Python modules is generated at docs/python/index. The equivalent file for JavaScript (if generated) is at docs/javascript/index, and there is also a list of processed JS files at docs/javascript/files. These should be added to a toctree directive in the documentation. Again, see this project’s docs/index.rst for an example.

  • The RST-Template library for creating reST files from JSDoc comments currently uses jsdoc-toolkit, which is no longer in active development. If we decide that its successor JSDoc 3 has enough useful improvements, we can look into updating the library to use that instead.

Troubleshooting

  • error: unrecognized arguments - If this pops up and breaks the build while parsing the code being documented, odds are that file has code at the module level which uses argparse or optparse, and it’s unsuccessfully trying to parse the command line parameters which were given to sphinx-build. Put such code inside a function which is only called inside an if __name__ == '__main__' condition (i.e., if that script was the one called).

References

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sbo-sphinx-2.2.0.tar.gz (965.1 kB view details)

Uploaded Source

Built Distribution

sbo_sphinx-2.2.0-py2.py3-none-any.whl (1.0 MB view details)

Uploaded Python 2 Python 3

File details

Details for the file sbo-sphinx-2.2.0.tar.gz.

File metadata

  • Download URL: sbo-sphinx-2.2.0.tar.gz
  • Upload date:
  • Size: 965.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No

File hashes

Hashes for sbo-sphinx-2.2.0.tar.gz
Algorithm Hash digest
SHA256 4fd5072da6096eb746773eac9ee4db862f87b245d1e52cc352b07445af5f6216
MD5 40b5f34856756ede523e386486fd1217
BLAKE2b-256 ca3c8568bfcef9abaac70dc2be4b4e46190950550ac56f56249c49572024652f

See more details on using hashes here.

File details

Details for the file sbo_sphinx-2.2.0-py2.py3-none-any.whl.

File metadata

File hashes

Hashes for sbo_sphinx-2.2.0-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 87539b4909000f2b285f8660243d0ca02918936a04ecedf191abe7fbb0fb7ccf
MD5 e5c803bce16754ec651b3bee6f3ff833
BLAKE2b-256 ef8cf440db98d04af72e6f91deff3c5199fb3da5c52ca4656cd07c243d2d4ad7

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page