blogger_markdown.py

#!/usr/bin/python
'''
Convert Blogger xml-exported file to markdown
Primarily from https://gist.github.com/larsks/4022537 with some modifications
'''
import os
import sys
import argparse
import iso8601
import re
import subprocess
import logging

from lxml import etree

reload(sys)
sys.setdefaultencoding("utf-8")

namespaces = {
    'atom': 'http://www.w3.org/2005/Atom',
    'app': 'http://purl.org/atom/app#',
}
kind_post = 'http://schemas.google.com/blogger/2008/kind#post'

if not os.path.exists('posts'):
    os.makedirs('posts')


def parse_args():
    p = argparse.ArgumentParser()
    p.add_argument('--html2text',
                   action='store_const', const='html2text', dest='converter')
    p.add_argument('--output-dir', '-d', default='posts')
    p.add_argument('input')
    p.set_defaults(converter='html2text')
    return p.parse_args()


def markdownify_html2text(html):
    p = subprocess.Popen(['html2text', '-d', '-b', '0', ],
                         stdin=subprocess.PIPE,
                         stdout=subprocess.PIPE)
    stdout, stderr = p.communicate(input=html.decode('utf-8'))

    return stdout


def process_entry(entry):
    kind = entry.xpath(
        'atom:category[@scheme="http://schemas.google.com/g/2005#kind"]',
        namespaces=namespaces)[0]
    if kind.get('term') != kind_post:
        return

    eid = entry.xpath('atom:id',
                      namespaces=namespaces)[0].text

    title = entry.xpath('atom:title[@type="text"]',
                        namespaces=namespaces)[0]
    title = title.text.strip().replace('\n', ' ')
    title = re.sub(' +', ' ', title)

    published = entry.xpath('atom:published', namespaces=namespaces)[0].text
    published = iso8601.parse_date(published)
    published = '%s-%s-%s' % (
        published.year,
        published.month,
        published.day)

    tags = entry.xpath(
        'atom:category[@scheme="http://www.blogger.com/atom/ns#"]',
        namespaces=namespaces)
    tags = [x.get('term') for x in tags]

    try:
        href = entry.xpath(
            'atom:link[@rel="alternate" and @type="text/html"]',
            namespaces=namespaces)[0].get('href')
    except IndexError:
        logging.error('no link for id %s' % eid)
        return

    slug = href.split('/')[-1].replace('.html', '')

    content = entry.xpath('atom:content',
                          namespaces=namespaces)[0].text

    return dict(
        id=eid,
        title=title,
        date=published,
        tags=tags,
        href=href,
        content=content,
        slug=slug,
    )


def write_entry(entry, data, opts):
    if opts.converter == 'html2text':
        mdfunc = markdownify_html2text
    else:
        raise ValueError('Unknown converter (%s)' % opts.converter)

    # Write Markdown to posts/slug.md
    md = mdfunc(data['content'].encode('utf-8'))
    with open(os.path.join(opts.output_dir, '%s.md' % data['slug']), 'w') as fd:
        fd.write('title: %(title)s\n' % data)
        fd.write('date: %(date)s\n' % data)
        fd.write('published: true\n')
        fd.write('tags: [%s]\n' % ', '.join(data['tags']))
        fd.write('\n')
        fd.write(md)


def main():
    opts = parse_args()

    logging.basicConfig(
        level=logging.INFO,
        format='%(asctime)s [%(levelname)s] %(name)s: %(message)s',
        datefmt='%Y-%m-%d %H:%M:%S')

    with open(opts.input) as fd:
        logging.info('parsing feed')
        doc = etree.parse(fd)

    for entry in doc.xpath('//atom:entry', namespaces=namespaces):
        data = process_entry(entry)
        if data is None:
            continue
        write_entry(entry, data, opts)

if __name__ == '__main__':
    main()

Can you please add comments

@saviour123, What parts do you need help with?