This is a pre-production deployment of Warehouse. Changes made here affect the production instance of PyPI (
Help us improve Python packaging - Donate today!

A Python interface for Discount, the C Markdown parser

Project Description

This Python package is a ctypes binding of David Loren’s Discount, a C implementation of John Gruber’s Markdown.


Markdown is a text-to-HTML conversion tool for web writers. Markdown allows you to write using an easy-to-read, easy-to-write plain text format, then convert it to structurally valid XHTML (or HTML).

The discount Python module contains two things of interest:

  • libmarkdown, a submodule that provides access to the public C functions defined by Discount.
  • Markdown, a helper class built on top of libmarkdown, providing a more familiar Pythonic interface

Using the Markdown class

The Markdown class wraps the C functions exposed in the libmarkdown submodule and handles the ctypes leg work for you. If you want to use the Discount functions directly, skip to the next section about libmarkdown.

Let’s take a look at a simple example:

import sys
import discount

mkd = discount.Markdown(sys.stdin)

Markdown takes one required argument, input_file_or_string, which is either a file object or a string-like object.

Note: There are limitations to what kind of file-like objects can be passed to Markdown. File-like objects like StringIO can’t be handled at the C level in the same way as OS file objects like sys.stdin and sys.stdout, or file objects returned by the builtin open() method.

Markdown also has methods for getting the output as a string, instead of writing to a file-like object. Let’s look at a modified version of the first example, this time using strings:

import discount

mkd = discount.Markdown('`test`')
print mkd.get_html_content()

Currently, Markdown does not manage character encoding, since the Markdown wraps C functions that support any character encoding that is a superset of ASCII. However, when working with unicode objects in Python, you will need to pass them as bytestrings to Markdown, and then convert them back to unicode afterwards. Here is an example of how you could do this:

import discount

txt = u'\xeb' # a unicode character, an e with an umlaut
mkd = discount.Markdown(txt.encode('utf-8'))
out = mkd.get_html_content()
val = out.decode('utf-8')

The Markdown class constructor also takes optional boolean keyword arguments that map to Discount flags compilation flags.

Generate table-of-contents headers (each generated <h1>, <h2>, etc will include a id=”name” argument.) Use get_html_toc() or write_html_toc() to generate the table-of-contents itself.
Disable relaxed emphasis and superscripts.
Greedily expand links; if a url is encountered, convert it to a hyperlink even if it isn’t surrounded with <>s.
Be paranoid about how [][] is expanded into a link - if the url isn’t a local reference, http://, https://, ftp://, or news://, it will not be converted into a hyperlink.
Do not process the pandoc document header, but treat it like regular text.
Do not allow <a or expand [][] into a link.
Do not allow <img or expand ![][] into a image.
Don’t process PHP Markdown Extra tables.
Disable SmartyPants processing.
Disable all embedded HTML by replacing all <’s with &lt;.
Do not process pseudo-protocols.

Pandoc header elements can be retrieved with the methods get_pandoc_title(), get_pandoc_author() and get_pandoc_date().

The converted HTML document parts can be retrieved as a string with the get_html_css(), get_html_toc() and get_html_content() methods, or written to a file with the write_html_css(fp), write_html_toc(fp) and write_html_content(fp) methods, where fp is the output file descriptor.

Discount provides two hooks for manipulating links while processing markdown. The first lets you rewrite urls specified by []() markup or <link/> tags, and the second lets you add additional HTML attributes on <a/> tags generated by Discount. You can pass callback functions to Markdown’s rewrite_links_func and link_attrs_func keyword arguments respectively.

Here is an example of a callback function that adds the domain name to internal pages:

def add_basepath(url):
    if url.startswith('/'):
        return '' % url

md = Markdown(

Here is an example that opens external pages in another window/tab:

def add_target_blank(url):
    if url.startswith('http://'):
        return 'target="_blank"'

md = Markdown(

Alternatively, you can attach these callbacks using decorators:

md = Markdown('`[a](/a.html)`')

def add_basepath(url):
    # same as above

md = Markdown('`[a](`')

def add_target_blank(url):
    # same as above

Under some conditions, the functions in libmarkdown may return integer error codes. These errors are raised as a MarkdownError exceptions when using the Markdown class.

Using libmarkdown

If you are familiar with using the C library and would rather use Discount library directly, libmarkdown is what you are looking for; it’s simply a thin wrapper around the original C implementation. libmarkdown exposes the public functions and flags documented on the Discount homepage.

In Python you’ll need to do some extra work preparing Python objects you want to pass to libmarkdown’s functions.

Most of these functions accept FILE* and char** types as their arguments, which require some additional ctypes boilerplate.

To get a FILE* from a Python file descriptor for use with libmarkdown, use the following pattern:

i = ctypes.pythonapi.PyFile_AsFile(sys.stdin)
o = ctypes.pythonapi.PyFile_AsFile(sys.stdout)
doc = libmarkdown.mkd_in(i, 0)
libmarkdown.markdown(doc, o, 0))

For libmarkdown functions to which you pass a char**, use the following pattern:

cp = ctypes.c_char_p('')
ln = libmarkdown.mkd_document(doc, ctypes.byref(cp))
html_text = cp.value[:ln]

It is important to initialize c_char_p with an empty string.

Running the test suite

Tests are available with the source distibution of discount in the file. The C shared object should be compiled first:

python build_ext

Then you can run the tests:


Source code and reporting bugs

You can obtain the source code and report bugs on GitHub project page.


See the LICENSE file in the source distribution for details.


discount is maintained by Tamas Kemenczy, and is funded by Trapeze. The Discount C library is written and maintained by David Loren and contributors. See the AUTHORS file for details.

Release History

Release History

This version
History Node


History Node


History Node


History Node


History Node


Download Files

Download Files

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

File Name & Checksum SHA256 Checksum Help Version File Type Upload Date
discount-0.2.1STABLE.tar.gz (14.9 kB) Copy SHA256 Checksum SHA256 Source Mar 1, 2010

Supported By

WebFaction WebFaction Technical Writing Elastic Elastic Search Pingdom Pingdom Monitoring Dyn Dyn DNS Sentry Sentry Error Logging CloudAMQP CloudAMQP RabbitMQ Heroku Heroku PaaS Kabu Creative Kabu Creative UX & Design Fastly Fastly CDN DigiCert DigiCert EV Certificate Rackspace Rackspace Cloud Servers DreamHost DreamHost Log Hosting