Exporting rackerhacker.com from Wordpress to Markdown

wordpress-to-markdown.py

#!/usr/bin/env python
import codecs
import feedparser
import html2text
from pprint import pprint
import sys
import time


def slugmaker(linktext):
    """Turn a post link into a slug"""
    linkpieces = linktext.split('/')
    slugpieces = linkpieces[3:7]
    return "/%s" % '/'.join(slugpieces)


# Forget line wrapping
html2text.BODY_WIDTH=0

# Pick up the XML export
d = feedparser.parse('rackerhacker.wordpress.2012-05-13.xml')

for entry in d.entries:
    # Skip if it's an unpublished entry
    if entry.wp_status != 'publish':
        continue

    # Skip if it's not a post
    if entry.wp_post_type != 'post':
        continue

    # Empty dict to hold this posts' data
    elements = {}

    # The easy stuff
    elements['title'] = entry.title
    elements['slug'] = slugmaker(entry.link)
    elements['date'] = time.strftime('%Y-%m-%d %H:%M', entry.published_parsed)
    elements['author'] = "Major Hayden"

    # Get the tags (if they exist)
    taglist = []
    if hasattr(entry, 'tags'):
        for tag in entry.tags:
            if tag.scheme == 'post_tag':
                taglist.append(tag.term)
    elements['tags'] = ', '.join(taglist)

    # Snag the actual content of the post
    elements['content'] = html2text.html2text(entry.content[0].value)

    # Generate a filename to hold the markdown text
    elements['filename'] = "%s-%s.md" % (
        time.strftime('%Y-%m-%d', entry.published_parsed),
        entry.link.split('/')[6])

    # Basic markdown template
    file_contents = """Title: %(title)s
Date: %(date)s
Slug: %(slug)s
Tags: %(tags)s
Author: %(author)s


%(content)s
""" % elements

    # Write the data (I hate unicode)
    with codecs.open("src/%s" % elements['filename'], 'wb', "utf-8") as f:
        f.write(file_contents)