ghost2zola.py

from collections import defaultdict
import gzip
import json
import os
import os.path
import urllib.request


EXPORT_FILE = 'crepererum-net.ghost.2019-02-23.json'
OUT_PATH = 'out'
SITE_URL = 'https://crepererum.net'

TEMPLATE = '''+++
title = "{title}"
date = {date}

[taxonomies]
categories = ["blog"]
tags = [{tags}]
+++
{markdown}
'''


def get_markdown(post):
    mobiledoc = json.loads(post['mobiledoc'])
    cards = mobiledoc['cards']
    if len(cards) != 1:
        return None
    assert len(cards) == 1

    card0 = cards[0]
    assert len(card0) == 2

    ctype, cdata = card0
    assert ctype == 'card-markdown'
    return cdata['markdown']


def download_uncompressed(url, target):
    response = urllib.request.urlopen(url)
    data = response.read()
    if data.startswith(b'\x1f\x8b'):
        data = gzip.decompress(data)

    with open(target, 'wb') as fp:
        fp.write(data)


def fetch_images(md, basedir):
    while True:
        start = md.find('/content/images')
        if start == -1:
            break
        ends = (
            md.find(marker, start)
            for marker in (')', '"')
        )
        end = min((
            p
            for p in ends
            if p != -1
        ))

        path = md[start:end]
        basename = os.path.basename(path)
        outpath = os.path.join(basedir, basename)
        url = SITE_URL + path
        print(f'  fetch {url}')
        download_uncompressed(url, outpath)

        md = md[:start] + basename + md[end:]


    return md


def get_date(post):
    return post['published_at'].split(' ')[0]


def process_post(post, tag_dict):
    slug = post['slug']
    print(f'converting {slug}:')

    status = post['status']
    if status != 'published':
        print('  skipping (not published)')
        return

    pid = post['id']
    tags = tag_dict.get(pid, [])

    title = post['title']
    md = get_markdown(post)
    if md is None:
        print('skipped (not markdown)')
        return
    date = get_date(post)

    basedir = os.path.join(
        OUT_PATH,
        slug,
    )
    try:
        os.makedirs(basedir)
    except FileExistsError:
        pass

    md = fetch_images(md, basedir)

    index_md = TEMPLATE.format(
        markdown=md,
        date=date,
        title=title,
        tags=', '.join(
            f'"{t}"'
            for t in sorted(tags)
        )
    )
    with open(os.path.join(basedir, 'index.md'), 'w') as fp:
        fp.write(index_md)

    print('  done')


def process_tags(db0_data):
    tag_ids = {
        tag['id']: tag['name']
        for tag in db0_data['tags']
    }
    tag_dict = defaultdict(list)
    for relation in db0_data['posts_tags']:
        tag_dict[relation['post_id']].append(tag_ids[relation['tag_id']])
    return dict(tag_dict)


with open(EXPORT_FILE) as fp:
    data = json.load(fp)

db0 = data['db'][0]
db0_data = db0['data']
posts = db0_data['posts']
tag_dict = process_tags(db0_data)

for post in posts:
    process_post(post, tag_dict)