Extract posts from WordPress to Jekyll posts

extract_posts.py

import xml.etree.ElementTree as etree

ns = {'wp': 'http://wordpress.org/export/1.2/', 'content': 'http://purl.org/rss/1.0/modules/content/'}
blog = etree.parse('data.xml').getroot().find('channel')

for item in blog.findall('item'):
    header = []
    title = item.find('title').text
    slug = item.find('wp:post_name', ns).text
    content = item.find('content:encoded', ns).text
    date = item.find('wp:post_date', ns).text.split(' ')[0]

    tags = []; cate = []
    for cate in item.findall('category'):
        if cate.get('domain')=='post_tag':
            tags.append(cate.get('nicename'))
        else:
            cate.append(cate.get('nicename'))
    tags = ' '.join(tags); cate = ' '.join(cate)

    fname = '_posts/%s-%s.md'%(date, slug)
    header.append('---')
    header.append('title: %s'%title)
    header.append('category: %s'%cate)
    header.append('tags: %s'%tags)
    header.append('---')
    
    with open(fname, 'w') as f:
        f.write('\n'.join(header))
        f.write('\n\n')
        f.write(unicode(content).encode('utf-8'))

    print fname