Converting blog to .md (.txt)

#!/usr/bin/env python3
"""
wxr_to_doc.py — Convert WordPress WXR export XML file(s) into one readable
Markdown document. Works entirely offline; no WordPress or database needed.

USAGE
    python3 wxr_to_doc.py export1.xml export2.xml -o my_writing.md

OPTIONS
    -o, --output       output filename (default: writing_compiled.md)
    --post-types       comma-separated post types to include (default: post)
                        (use e.g. "post,poem" if you export a custom post type)
    --status           comma-separated statuses to include (default: publish)
                        (use "publish,draft,private" to include everything)
    --sort             date (default) or title
    --categories       comma-separated category slugs/names to include
                        (omit to include all categories)

EXAMPLES
    # Basic: both export files, only published posts, chronological
    python3 wxr_to_doc.py posts1.xml posts2.xml -o compiled.md

    # Include drafts too, sorted alphabetically
    python3 wxr_to_doc.py posts1.xml posts2.xml --status publish,draft --sort title

    # Only a specific category (e.g. "poetry")
    python3 wxr_to_doc.py posts1.xml posts2.xml --categories poetry
"""

import argparse
import html
import re
from datetime import datetime
from xml.etree import ElementTree as ET

NS = {
    'content': 'http://purl.org/rss/1.0/modules/content/',
    'wp': 'http://wordpress.org/export/1.2/',
    'dc': 'http://purl.org/dc/elements/1.1/',
    'excerpt': 'http://wordpress.org/export/1.2/excerpt/',
}


def html_to_readable(raw_html):
    text = raw_html or ''
    # WXR often double-encodes entities
    text = html.unescape(html.unescape(text))
    # strip Gutenberg block comments
    text = re.sub(r'<!--\s*/?wp:[^>]*-->', '', text)
    # line breaks / paragraphs
    text = re.sub(r'<br\s*/?>', '\n', text, flags=re.I)
    text = re.sub(r'</p>\s*<p[^>]*>', '\n\n', text, flags=re.I)
    text = re.sub(r'</?p[^>]*>', '', text, flags=re.I)
    # basic emphasis, preserved as markdown
    text = re.sub(r'<(strong|b)[^>]*>(.*?)</\1>', r'**\2**', text, flags=re.I | re.S)
    text = re.sub(r'<(em|i)[^>]*>(.*?)</\1>', r'*\2*', text, flags=re.I | re.S)
    # blockquotes -> simple indent marker
    text = re.sub(r'<blockquote[^>]*>(.*?)</blockquote>', r'> \1', text, flags=re.I | re.S)
    # strip anything else
    text = re.sub(r'<[^>]+>', '', text)
    # collapse excess blank lines
    text = re.sub(r'\n{3,}', '\n\n', text)
    return text.strip()


def parse_wxr(path, post_types, statuses, categories_filter):
    tree = ET.parse(path)
    root = tree.getroot()
    channel = root.find('channel')
    items = []
    for item in channel.findall('item'):
        post_type = item.findtext('wp:post_type', default='', namespaces=NS)
        status = item.findtext('wp:status', default='', namespaces=NS)
        if post_type not in post_types or status not in statuses:
            continue

        cats = [c.text for c in item.findall('category') if c.get('domain') == 'category']
        if categories_filter and not (set(c.lower() for c in cats) & categories_filter):
            continue

        title = item.findtext('title', default='(untitled)') or '(untitled)'
        pub_date_raw = item.findtext('wp:post_date', default='', namespaces=NS)
        try:
            pub_date = datetime.strptime(pub_date_raw, '%Y-%m-%d %H:%M:%S')
        except ValueError:
            pub_date = None

        content_el = item.find('content:encoded', NS)
        content_raw = content_el.text if content_el is not None and content_el.text else ''

        items.append({
            'title': title.strip(),
            'date': pub_date,
            'content': html_to_readable(content_raw),
            'categories': cats,
        })
    return items


def main():
    parser = argparse.ArgumentParser(
        description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter
    )
    parser.add_argument('files', nargs='+', help='WXR .xml export files')
    parser.add_argument('-o', '--output', default='writing_compiled.md')
    parser.add_argument('--post-types', default='post')
    parser.add_argument('--status', default='publish')
    parser.add_argument('--sort', choices=['date', 'title'], default='date')
    parser.add_argument('--categories', default='')
    args = parser.parse_args()

    post_types = set(t.strip() for t in args.post_types.split(','))
    statuses = set(s.strip() for s in args.status.split(','))
    categories_filter = set(c.strip().lower() for c in args.categories.split(',') if c.strip())

    all_items = []
    for f in args.files:
        all_items.extend(parse_wxr(f, post_types, statuses, categories_filter))

    if args.sort == 'date':
        all_items.sort(key=lambda x: x['date'] or datetime.min)
    else:
        all_items.sort(key=lambda x: x['title'].lower())

    with open(args.output, 'w', encoding='utf-8') as out:
        for it in all_items:
            out.write(f"# {it['title']}\n\n")
            meta_bits = []
            if it['date']:
                meta_bits.append(it['date'].strftime('%B %d, %Y'))
            if it['categories']:
                meta_bits.append(', '.join(it['categories']))
            if meta_bits:
                out.write(f"*{' — '.join(meta_bits)}*\n\n")
            out.write(it['content'])
            out.write('\n\n---\n\n')

    print(f"Wrote {len(all_items)} entries to {args.output}")


if __name__ == '__main__':
    main()