Who needs pandoc when you have Sphinx?

Using Sphinx doesn't necessarily mean using reStructuredText for input and HTML for output. We explore Sphinx's newfound support for Markdown as well as it's broad range output formats available, before moving onto an overview of how you can develop your parser and builder extensions.

Sphinx is the documentation tool of choice for an increasing amount of projects, both inside and outside Python. Thanks to the success of platforms like Read the Docs, building documentation with this toolkit has never been easier. Most people associate Sphinx with the reStructuredText syntax provided by docutils, and typically output documents to HTML and, occasionally, PDF. However, Sphinx is capable of so much more. Since Sphinx 1.8, it is possible to use source documents written in the CommonMark syntax, while the amount of builders provided both in-tree and out-of-tree continue to grow.

Through this talk, we explore how one can build their documentation using Sphinx and Markdown source documents. We detail the variety of builders available as part of the standard Sphinx installation and as third-party extensions, including some basic configuration tips for the more commonly useful ones like man pages and LaTeX/PDF. Finally, we provide a high-level overview on how you can go about writing your own extensions to parse other plain-text documentation formats and output in additional documentation format. The latter of these will touch on docutils, a foundational component of Sphinx, so one can understand the intermediate documentation model it provides.

Transcript

1. Who needs Pandoc when you have Sphinx? An exploration of the parsers and builders of the Sphinx documentation tool FOSDEM 2019 @stephenfin
2. reStructuredText, Docutils & Sphinx 1
3. A little reStructuredText ========================= This document demonstrates some basic features of |rst|. You can use **bold** and *italics*, along with ``literals``. It’s quite similar to `Markdown`_ but much more extensible. CommonMark may one day approach this [1]_, but today is not that day. `Docutils`__ does all this for us. .. |rst| replace:: **reStructuredText** .. _Markdown: https://daringfireball.net/projects/markdown/ .. [1] https://talk.commonmark.org/t/444 __ http://docutils.sourceforge.net/ intro.rst
4. A little reStructuredText ========================= This document demonstrates some basic features of |rst|. You can use **bold** and *italics*, along with ``literals``. It’s quite similar to `Markdown`_ but much more extensible. CommonMark may one day approach this [1]_, but today is not that day. `Docutils`__ does all this for us. .. |rst| replace:: **reStructuredText** .. _Markdown: https://daringfireball.net/projects/markdown/ .. [1] https://talk.commonmark.org/t/444 __ http://docutils.sourceforge.net/ intro.rst
5. A little reStructuredText This document demonstrates some basic features of reStructuredText. You can use bold and italics, along with literals. It’s quite similar to Markdown but much more extensible. CommonMark may one day approach this [1], but today is not that day. Docutils does all this for us. [1] https://talk.commonmark.org/t/444/ intro.html
6. A little more reStructuredText ============================== The extensibility really comes into play with directives and roles. We can do things like link to RFCs (:RFC:`2324`, anyone?) or generate some more advanced formatting (I do love me some H\ :sub:`2`\ O). .. warning:: The power can be intoxicating. Of course, all the stuff we showed previously *still works!* The only limit is your imagination/interest. more.rst
7. A little more reStructuredText ============================== The extensibility really comes into play with directives and roles. We can do things like link to RFCs (:RFC:`2324`, anyone?) or generate some more advanced formatting (I do love me some H\ :sub:`2`\ O). .. warning:: The power can be intoxicating. Of course, all the stuff we showed previously *still works!* The only limit is your imagination/interest. more.rst
8. A little more reStructuredText The extensibility really comes into play with directives and roles. We can do things like link to RFCs (RFC 2324, anyone?) or generate some more advanced formatting (I do love me some H 2 O). Warning The power can be intoxicating. Of course, all the stuff we showed previously still works! The only limit is your imagination/interest. more.html
9. reStructuredText provides the syntax Docutils provides the parsing and file generation
10. reStructuredText provides the syntax Docutils provides the parsing and file generation Sphinx provides the cross-referencing
11. Docutils use readers, parsers, transforms, and writers Docutils works with individual files
12. Docutils use readers, parsers, transforms, and writers Docutils works with individual files Sphinx uses readers, parsers, transforms, writers and builders Sphinx works with multiple, cross-referenced files
13. How Does Docutils Work? 2
14. About me ======== Hello, world. I am **bold** and *maybe* I am brave. index.rst
15. $ rst2html index.rst
16. About me Hello, world. I am bold and maybe I am brave. index.html
17. index.rst index.html
18. $ rst2pseudoxml index.rst
19. <document ids="about-me" names="about\ me" source="index.rst" title="About me"> <title> About me <paragraph> Hello, world. I am <strong> bold and <emphasis> maybe I am brave. index.xml
20. $ ./docutils/tools/quicktest.py index.rst
21. <document source="index.rst"> <section ids="about-me" names="about\ me"> <title> About me <paragraph> Hello, world. I am <strong> bold and <emphasis> maybe I am brave. index.xml
22. Readers (reads from source and passes to the parser) Parsers (creates a doctree model from the read file) Transforms (add to, prune, or otherwise change the doctree model) Writers (converts the doctree model to a file)
23. Readers (reads from source and passes to the parser) Parsers (creates a doctree model from the read file) Transforms (add to, prune, or otherwise change the doctree model) Writers (converts the doctree model to a file)
24. What About Sphinx? 3
25. About me ======== Hello, world. I am **bold** and *maybe* I am brave. index.rst
26. master_doc = 'index' conf.py
27. $ sphinx-build -b html . _build
28. About me Hello, world. I am bold and maybe I am brave. index.html
29. Readers (reads from source and passes to the parser) Parsers (creates a doctree model from the read file) Transforms (add to, prune, or otherwise change the doctree model) Writers (converts the doctree model to a file)
30. Builders (call the readers, parsers, transformers, writers) Application (calls the builder(s)) Environment (store information for future builds)
31. Builders (call the readers, parsers, transformers, writers) Application (calls the builder(s)) Environment (store information for future builds)
32. ... updating environment: 1 added, 0 changed, 0 removed reading sources... [100%] index looking for now-outdated files... none found pickling environment... done checking consistency... done preparing documents... done generating indices... done writing additional pages... done copying static files... done copying extra files... done dumping search index in English (code: en) ... done dumping object inventory... done build succeeded.
33. Docutils provides almost 100 node types document section title subtitle paragraph block_quote bullet_list note ... (the root element of the document tree) (the main unit of hierarchy for documents) (stores the title of a document, section, ...) (stores the subtitle of a document) (contains the text and inline elements of a single paragraph) (used for quotations set off from the main text) (contains list_item elements marked with bullets) (an admonition, a distinctive and self-contained notice) ...
34. Sphinx provides its own custom node types translatable not_smartquotable toctree versionmodified seealso productionlist manpage pending_xref ... (indicates content which supports translation) (indicates content which does not support smart-quotes) (node for inserting a "TOC tree") (version change entry) (custom "see also" admonition) (grammar production lists) (reference to a man page) (cross-reference that cannot be resolved yet) ...
35. Docutils provides dozens of transforms DocTitle DocInfo SectNum Contents Footnotes Messages SmartQuotes Admonitions ... (promote title elements to the document level) (transform initial field lists to docinfo elements) (assign numbers to the titles of document sections) (generate a table of contents from a document or sub-node) (resolve links to footnotes, citations and their references) (place system messages into the document) (replace ASCII quotation marks with typographic form) (transform specific admonitions to generic ones) ...
36. Sphinx also provides additional transforms MoveModuleTargets AutoNumbering CitationReferences SphinxSmartQuotes DoctreeReadEvent ManpageLink SphinxDomains Locale ... (promote initial module targets to the section title) (register IDs of tables, figures and literal blocks to assign numbers) (replace citation references with pending_xref nodes) (custom SmartQuotes to avoid transform for some extra node types) (emit doctree-read event) (find manpage section numbers and names) (collect objects to Sphinx domains for cross referencing) (replace translatable nodes with their translated doctree) ...
37. Using Additional Parsers 4
38. There are a number of parsers available reStructuredText (part of docutils) Markdown (part of recommonmark) Jupyter Notebooks (part of nbsphinx)
39. # About me Hello, world. I am **bold** and *maybe* I am brave. index.md
40. $ cm2html index.md
41. About me Hello, world. I am bold and maybe I am brave. index.html
42. $ cm2pseudoxml index.md
43. <document ids="about-me" names="about\ me" source="index.md" title="About me"> <title> About me <paragraph> Hello, world. I am <strong> bold and <emphasis> maybe I am brave. index.xml
44. # About me Hello, world. I am **bold** and *maybe* I am brave. index.md
45. from recommonmark.parser import CommonMarkParser master_doc = 'index' source_parsers = {'.md': CommonMarkParser} source_suffix = '.md' conf.py
46. from recommonmark.parser import CommonMarkParser master_doc = 'index' source_parsers = {'.md': CommonMarkParser} source_suffix = '.md' conf.py
47. $ sphinx-build -b html . _build
48. About me Hello, world. I am bold and maybe I am brave. index.html
49. Using Additional Writers, Builders 5
50. Docutils provides a number of in-tree writers docutils_xml html4css1 latex2e manpage null odf_odt pep_html pseudoxml ... (simple XML document tree Writer) (simple HTML document tree Writer) (LaTeX2e document tree Writer) (simple man page Writer) (a do-nothing Writer) (ODF Writer) (PEP HTML Writer) (simple internal document tree Writer) ...
51. $ rst2html5 index.rst
52. from docutils.core import publish_file from docutils.writers import html5_polyglot with open('README.rst', 'r') as source: publish_file(source=source, writer=html5_polyglot.Writer())
53. $ pip install rst2txt
54. $ rst2txt index.rst
55. from docutils.core import publish_file from rst2txt with open('README.rst', 'r') as source: publish_file(source=source, writer=rst2txt.Writer())
56. html qthelp epub latex text man texinfo xml ... (generates output in HTML format) (like html but also generates Qt help collection support files) (like html but also generates an epub file for eBook readers) (generates output in LaTeX format) (generates text files with most rST markup removed) (generates manual pages in the groff format) (generates textinfo files for use with makeinfo) (generates Docutils-native XML files) ... Sphinx provides its own in-tree builders
57. $ sphinx-build -b html . _build
58. $ pip install sphinx-asciidoc
59. $ sphinx-build -b asciidoc . _build
60. Writing Your Own Parsers, Writers 6
61. Reading (reads from source and passes to the parser) Parsing (creates a doctree model from the read file) Transforming (applies transforms to the doctree model) Writing (converts the doctree model to a file)
62. from docutils import parsers class Parser(parsers.Parser): supported = ('null',) config_section = 'null parser' config_section_dependencies = ('parsers',) def parse(self, inputstring, document): pass docutils/parsers/null.py
63. We’re not covering Compilers 101
64. We’re not covering Compilers 101 We’re going to cheat
65. <?xml version="1.0" encoding="utf-8"?> <document source="index.rst"> <section ids="about-me" names="about\ me"> <title>About me</title> <paragraph>Hello, world. I am <strong>bold</strong> and <emphasis>maybe</emphasis> I am brave.</paragraph> </section> </document> index.xml
66. from docutils import parsers import xml.etree.ElementTree as ET class Parser(parsers.Parser): supported = ('xml',) config_section = 'XML parser' config_section_dependencies = ('parsers',) def parse(self, inputstring, document): xml = ET.fromstring(inputstring) self._parse(document, xml) ... xml_parser.py
67. ... def _parse(self, node, xml): for attrib, value in xml.attrib.items(): # NOTE(stephenfin): this isn't complete! setattr(node, attrib, value) for child in xml: child_node = getattr(nodes, child.tag)(text=child.text) node += self._parse(child_node, child) if xml.tail: return node, nodes.Text(xml.tail) return node xml_parser.py
68. Reading (reads from source and passes to the parser) Parsing (creates a doctree model from the read file) Transforming (applies transforms to the doctree model) Writing (converts the doctree model to a file)
69. from docutils import writers class Writer(writers.Writer): supported = ('pprint', 'pformat', 'pseudoxml') config_section = 'pseudoxml writer' config_section_dependencies = ('writers',) output = None def translate(self): self.output = self.document.pformat() docutils/writers/pseudoxml.py
70. from docutils import writers class Writer(writers.Writer): supported = ('pprint', 'pformat', 'pseudoxml') config_section = 'pseudoxml writer' config_section_dependencies = ('writers',) output = None def translate(self): self.output = self.document.pformat() docutils/writers/pseudoxml.py
71. from docutils import nodes, writers class TextWriter(writers.Writer): supported = ('text',) config_section = 'text writer' config_section_dependencies = ('writers',) output = None def translate(self): visitor = TextTranslator(self.document) self.document.walkabout(visitor) self.output = visitor.body rst2txt/writer.py
72. from docutils import nodes, writers class TextWriter(writers.Writer): supported = ('text',) config_section = 'text writer' config_section_dependencies = ('writers',) output = None def translate(self): visitor = TextTranslator(self.document) self.document.walkabout(visitor) self.output = visitor.body rst2txt/writer.py
73. ... class TextTranslator(nodes.NodeVisitor): ... def visit_document(self, node): pass def depart_document(self, node): pass def visit_section(self, node): pass rst2txt/writer.py
74. from sphinx.builders import Builder class TextBuilder(Builder): name = 'text' def __init__(self): pass def get_outdated_docs(self): pass def get_target_uri(self): pass sphinx/builders/text.py
75. ... def prepare_writing(self, docnames): pass def write_doc(self, docnames, doctree): pass def finish(self): pass sphinx/builders/text.py
76. Wrap Up 6
77. Sphinx and Docutils share most of the same architecture… Readers Parsers Transforms Writers
78. …but Sphinx builds upon and extends Docutils’ core functionality Builders Application Environment
79. There are multiple writers/builders provided by both… HTML Manpage LaTeX XML texinfo (Sphinx only) ODF (Docutils only) ...
80. ...and many more writers/builders available along with readers Markdown (reader and builder) Text (writer) ODF (builder) AsciiDoc (builder) EPUB2 (builder) reStructuredText (builder) ...
81. It’s possible to write your own
82. It’s possible to write your own
83. Fin
84. Who needs Pandoc when you have Sphinx? An exploration of the parsers and builders of the Sphinx documentation tool FOSDEM 2019 @stephenfin
85. Useful Packages and Tools • recommonmark (provides a Markdown reader) • sphinx-markdown-builder (provides a Markdown builder) • sphinx-asciidoc (provides an AsciiDoc builder) • rst2txt (provides a plain text writer) • asciidoclive.com (online AsciiDoc Editor) • rst.ninjs.org (online rST Editor)
86. References • Quick reStructuredText • Docutils Reference Guide ◦ reStructuredText Markup Specification ◦ reStructuredText Directives ◦ reStructuredText Interpreted Text Roles • Docutils Hacker’s Guide • PEP-258: Docutils Design Specification
87. References • A brief tutorial on parsing reStructuredText (reST) -- Eli Bendersky • A lion, a head, and a dash of YAML -- Stephen Finucane () • OpenStack + Sphinx In A Tree -- Stephen Finucane () • Read the Docs & Sphinx now support Commonmark -- Read the Docs Blog