從 WordPress 遷移到 Astro 靜態部落格

1
#!/usr/bin/env python3
2
"""
3
Parse WordPress WXR (XML) export and convert posts to Astro Markdown files.
4
"""
5
import re
6
import os
7
import json
8
from datetime import datetime
9
from lxml import etree as ET
10
import html2text
11

12
XML_FILE = "linziyoudevblog.WordPress.2026-03-01.xml"
13
OUT_DIR = "posts"
14

15
os.makedirs(OUT_DIR, exist_ok=True)
16

17
# ── Namespaces ──
18
NS = {
19
    'content': 'http://purl.org/rss/1.0/modules/content/',
20
    'wp':      'http://wordpress.org/export/1.2/',
21
    'dc':      'http://purl.org/dc/elements/1.1/',
22
    'excerpt': 'http://wordpress.org/export/1.2/excerpt/',
23
}
24

25
# ── HTML → Markdown converter ──
26
h = html2text.HTML2Text()
27
h.ignore_links = False
28
h.body_width = 0
29
h.protect_links = True
30
h.wrap_links = False
31

32
h_plain = html2text.HTML2Text()
33
h_plain.ignore_links = True
34
h_plain.ignore_images = True
35
h_plain.body_width = 0
36

37
def to_markdown(html_content):
38
    if not html_content:
39
        return ''
40
    # Rewrite wp-content/uploads URLs to relative /uploads/
41
    content = re.sub(
42
        r'https?://[^/]+/wp-content/uploads/',
43
        '/uploads/',
44
        html_content
45
    )
46
    # Remove WordPress block comments
47
    content = re.sub(r'<!-- /?wp:[^>]* -->', '', content)
48
    return h.handle(content).strip()
49

50
def to_plain(html_content):
51
    if not html_content:
52
        return ''
53
    return h_plain.handle(html_content).strip()
54

55
# ── Parse XML (with recovery for invalid chars) ──
56
parser = ET.XMLParser(recover=True, encoding='utf-8')
57
with open(XML_FILE, 'rb') as f:
58
    raw = f.read()
59
# Strip invalid XML control characters
60
raw = re.sub(rb'[\x00-\x08\x0b\x0c\x0e-\x1f]', b'', raw)
61
root = ET.fromstring(raw, parser)
62
channel = root.find('channel')
63

64
posts = []
65
for item in channel.findall('item'):
66
    post_type = item.find('wp:post_type', NS)
67
    status     = item.find('wp:status', NS)
68

69
    if post_type is None or status is None:
70
        continue
71
    if post_type.text != 'post' or status.text != 'publish':
72
        continue
73

74
    title    = item.findtext('title') or ''
75
    slug     = item.findtext('wp:post_name', namespaces=NS) or ''
76
    date_str = item.findtext('wp:post_date', namespaces=NS) or ''
77
    content  = item.findtext('content:encoded', namespaces=NS) or ''
78
    excerpt  = item.findtext('excerpt:encoded', namespaces=NS) or ''
79

80
    # Categories and tags
81
    categories, tags = [], []
82
    for cat in item.findall('category'):
83
        domain = cat.get('domain', '')
84
        name   = cat.text or ''
85
        if domain == 'category':
86
            categories.append(name)
87
        elif domain == 'post_tag':
88
            tags.append(name)
89

90
    # Parse date
91
    try:
92
        dt = datetime.strptime(date_str, '%Y-%m-%d %H:%M:%S')
93
        date_iso = dt.strftime('%Y-%m-%d')
94
    except:
95
        date_iso = date_str[:10] if date_str else '2020-01-01'
96

97
    posts.append({
98
        'title': title, 'slug': slug, 'date': date_iso,
99
        'content': content, 'excerpt': excerpt,
100
        'categories': categories, 'tags': tags,
101
    })
102

103
print(f"Found {len(posts)} published posts")
104

105
# ── Write Markdown files ──
106
for post in posts:
107
    body    = to_markdown(post['content'])
108
    excerpt = to_plain(post['excerpt']).replace('\n', ' ').strip()[:200]
109

110
    from urllib.parse import unquote
111
    slug_decoded = unquote(post['slug'])
112
    filename = f"{post['date']}-{slug_decoded}.md"
113

114
    lines = ['---']
115
    lines.append(f'title: {json.dumps(post["title"], ensure_ascii=False)}')
116
    lines.append(f'published: "{post["date"]}"')
117
    if post['categories']:
118
        lines.append(f'category: {json.dumps(post["categories"][0], ensure_ascii=False)}')
119
    if post['tags']:
120
        lines.append(f'tags: {json.dumps(post["tags"], ensure_ascii=False)}')
121
    if excerpt:
122
        lines.append(f'description: {json.dumps(excerpt, ensure_ascii=False)}')
123
    lines.append('draft: false')
124
    lines.append('---')
125
    lines.append('')
126
    lines.append(body)
127

128
    out_path = os.path.join(OUT_DIR, filename)
129
    with open(out_path, 'w', encoding='utf-8') as f:
130
        f.write('\n'.join(lines))
131

132
    print(f"  [{post['date']}] {post['title'][:60]}")
133

134
print(f"\nDone! {len(posts)} files → ./{OUT_DIR}/")