link patterns

link-patterns extra

The link-patterns extra enables mapping of given regex patterns in text to links. The link_patterns list on the Markdown class provides the mapping.

For example, the following with map "recipe NNNN" and "komodo bug NNNN" to appropriate links:

>>> import markdown2, re
>>> link_patterns = [
...    (re.compile("recipe\s+(\d+)", re.I), r"http://code.activestate.com/recipes/\1/"),
...    (re.compile("(?:komodo\s+)?bug\s+(\d+)", re.I), r"http://bugs.activestate.com/show_bug.cgi?id=\1"),
... ]
>>> markdown2.markdown('Recipe 123 and Komodo bug 234 are related.'
...     extras=["link-patterns"], link_patterns=link_patterns)
'<p><a href="http://code.activestate.com/recipes/123/">Recipe 123</a> and \
<a href="http://bugs.activestate.com/show_bug.cgi?id=234">Komodo bug 234</a> are related.</p>'

Here is a script that uses link-patterns to auto-link WikiWords (for Wiki syntaxes that do that):

https://github.com/trentm/python-markdown2/blob/master/sandbox/wiki.py

link patterns file

For the command-line interface, the --link-patterns-file option has been added. A "link patterns file" has one link pattern per line of the form:

<regex-pattern> <href-pattern>

For example:

/recipe\s+(\d+)/i               http://code.activestate.com/recipes/\1/
/(?:komodo\s+)?bug\s+(\d+)/i    http://bugs.activestate.com/show_bug.cgi?id=\1

Spaces are not allowed in the <href-pattern> (to simplify parsing). Lines beginning with a hash (#) are comments.

$ python markdown2.py -x link-patterns --link-patterns-file patterns foo.text

Converting links into links automatically

from markdown2 import Markdown
import re
pattern = re.compile(
    r"""
        \b
        (
            (?:https?://|(?<!//)www\.)    # prefix - https:// or www.
            \w[\w_\-]*(?:\.\w[\w_\-]*)*   # host
            [^<>\s"']*                    # rest of url
            (?<![?!.,:*_~);])             # exclude trailing punctuation
            (?=[?!.,:*_~);]?(?:[<\s]|$))  # make sure that we're not followed by " or ', i.e. we're outside of href="...".
        )
    """,
    re.X
)
markdown = Markdown(
    extras=["link-patterns"], 
    link_patterns=[(pattern, r'\1')]
)
markdown.convert('http://www.google.com')
markdown.convert('<a href="http://www.google.com">link</a>')
markdown.convert('[link](http://www.google.com)')

Output

u'<p><a href="http://www.google.com">http://www.google.com</a></p>\n'
u'<p><a href="http://www.google.com">link</a></p>\n'
u'<p><a href="http://www.google.com">link</a></p>\n'

Default link behaviour

Without specifying the link-patterns extra, the module will use standard markdown link format [text](link). Please note that it only allows the URL schemas http(s) and ftp; otherwise it outputs a bookmark link #.

(Return to Extras page.)