rewrite: Vite+React → Astro 5 + Content Collections
Some checks failed
deploy / deploy (push) Failing after 12s
Some checks failed
deploy / deploy (push) Failing after 12s
- Бэкап старой версии на ветке vite-react-backup - Stack: Astro 5 + nginx:alpine runtime, образ ~50 МБ (был ~600 МБ) - @astrojs/rss заменён ручным buildRss() — гарантия CDATA в content:encoded для IPB Importer - @astrojs/sitemap → sitemap-index.xml + sitemap.txt - 152-ФЗ cookie consent + privacy.astro + Analytics с gating - AI-файлы: robots.txt с явным allow для AI-краулеров, ai.txt, llms.txt - Гибридный визуал: фото-фон шапки (аэрофото Пушкино) + PT Serif + IBM Plex Sans - Иерархия: hero "Главная история" с рамкой + "Ещё из истории" + "Хроника" - Категория "main" (псевдо) скрыта из плашек и из Рубрик в сайдбаре - hideFromList: true для технических постов - featuredImage в frontmatter для постов без хорошей первой <img> - WP resized-URL (-WxH.ext) автоматически → оригинал - CI/CD: .gitea/workflows/deploy.yml (push → SSH-build) - Внешние RSS: scripts/pull-external-rss.mjs пишет news.json в bind-mount, фронт фетчит client-side
This commit is contained in:
striker
@@ -34,9 +34,13 @@ def slugify_ru(s: str) -> str:
|
||||
return res or 'untitled'
|
||||
|
||||
UPLOAD_RE = re.compile(r'https?://(?:www\.)?pushkinohistory\.ru/wp-content/uploads/[^/]+/[^/]+/([^"\'\s)]+)')
|
||||
# WP-resized варианты: file-1024x768.png → file.png. У нас в /uploads/ лежит только оригинал.
|
||||
RESIZED_RE = re.compile(r'(/uploads/[^"\'\s)]+?)-\d+x\d+(\.\w+)')
|
||||
|
||||
def rewrite_uploads(html: str) -> str:
|
||||
return UPLOAD_RE.sub(r'/uploads/\1', html)
|
||||
html = UPLOAD_RE.sub(r'/uploads/\1', html)
|
||||
html = RESIZED_RE.sub(r'\1\2', html)
|
||||
return html
|
||||
|
||||
CATEGORIES = {
|
||||
20: [], 23: [], 73: [], 94: [], # pages — no category
|
||||
|
||||
90
scripts/convert_to_markdown.py
Normal file
90
scripts/convert_to_markdown.py
Normal file
@@ -0,0 +1,90 @@
|
||||
#!/usr/bin/env python3
|
||||
"""Convert src/content/{posts,pages}.json → src/content/{posts,pages}/<slug>.md
|
||||
for Astro Content Collections."""
|
||||
import json
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
ROOT = Path(__file__).resolve().parent.parent
|
||||
CONTENT = ROOT / "src" / "content"
|
||||
POSTS_JSON = CONTENT / "posts.json"
|
||||
PAGES_JSON = CONTENT / "pages.json"
|
||||
POSTS_DIR = CONTENT / "posts"
|
||||
PAGES_DIR = CONTENT / "pages"
|
||||
|
||||
POSTS_DIR.mkdir(parents=True, exist_ok=True)
|
||||
PAGES_DIR.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# Ручные флаги по slug'у для главной ленты.
|
||||
# featured=True — пинится наверх как hero (отдельная карточка с рамкой).
|
||||
# hideFromList=True — не показывается в общей ленте (виден только в рубрике/архиве).
|
||||
# featuredImage — переопределяет первую <img> из тела для миниатюры/hero.
|
||||
SLUG_FLAGS: dict[str, dict] = {
|
||||
'voronino': {'featured': True, 'featuredImage': '/uploads/IMG_2156.jpg'},
|
||||
'staroe-staroe-selo': {'featured': True, 'featuredImage': '/uploads/IMG_2754.jpg'},
|
||||
'vnimanie-texnicheskie-raboty': {'hideFromList': True},
|
||||
'vnimanie-texnicheskie-raboty-2': {'hideFromList': True},
|
||||
'c-nastupayushhim-novym-2014-godom': {'hideFromList': True},
|
||||
}
|
||||
|
||||
def yaml_escape(s: str) -> str:
|
||||
return s.replace('"', '\\"')
|
||||
|
||||
def make_md(item: dict, kind: str) -> str:
|
||||
fm = [
|
||||
'---',
|
||||
f'title: "{yaml_escape(item["title"])}"',
|
||||
f'slug: {item["slug"]}',
|
||||
f'legacyId: {item["id"]}',
|
||||
]
|
||||
if item.get('date'):
|
||||
date = item['date'].replace(' ', 'T') + '+03:00'
|
||||
if kind == 'post':
|
||||
fm.append(f'pubDate: {date}')
|
||||
else:
|
||||
fm.append(f'pubDate: {date}')
|
||||
if item.get('excerpt'):
|
||||
fm.append(f'description: "{yaml_escape(item["excerpt"])}"')
|
||||
else:
|
||||
fm.append('description: ""')
|
||||
if kind == 'post':
|
||||
if item.get('categories'):
|
||||
fm.append('categories:')
|
||||
for c in item['categories']:
|
||||
fm.append(f' - "{yaml_escape(c)}"')
|
||||
else:
|
||||
fm.append('categories: []')
|
||||
if item.get('categorySlugs'):
|
||||
fm.append('categorySlugs:')
|
||||
for s in item['categorySlugs']:
|
||||
fm.append(f' - "{s}"')
|
||||
else:
|
||||
fm.append('categorySlugs: []')
|
||||
fm.append('author: "История города Пушкино"')
|
||||
if item.get('oldSlug') and item['oldSlug'] != item['slug']:
|
||||
fm.append(f'oldSlug: "{item["oldSlug"]}"')
|
||||
if kind == 'post':
|
||||
flags = SLUG_FLAGS.get(item['slug'], {})
|
||||
if flags.get('featured'):
|
||||
fm.append('featured: true')
|
||||
if flags.get('hideFromList'):
|
||||
fm.append('hideFromList: true')
|
||||
if flags.get('featuredImage'):
|
||||
fm.append(f'featuredImage: "{flags["featuredImage"]}"')
|
||||
fm.append('---')
|
||||
fm.append('')
|
||||
fm.append(item['html'])
|
||||
return '\n'.join(fm) + '\n'
|
||||
|
||||
def convert(json_path: Path, out_dir: Path, kind: str) -> int:
|
||||
items = json.loads(json_path.read_text(encoding='utf-8'))
|
||||
for item in items:
|
||||
md = make_md(item, kind)
|
||||
(out_dir / f'{item["slug"]}.md').write_text(md, encoding='utf-8')
|
||||
return len(items)
|
||||
|
||||
if __name__ == '__main__':
|
||||
n_posts = convert(POSTS_JSON, POSTS_DIR, 'post')
|
||||
n_pages = convert(PAGES_JSON, PAGES_DIR, 'page')
|
||||
print(f'posts: {n_posts} → src/content/posts/')
|
||||
print(f'pages: {n_pages} → src/content/pages/')
|
||||
124
scripts/pull-external-rss.mjs
Normal file
124
scripts/pull-external-rss.mjs
Normal file
@@ -0,0 +1,124 @@
|
||||
#!/usr/bin/env node
|
||||
/**
|
||||
* Тянет внешние RSS-фиды из src/data/feeds.json и записывает агрегированный
|
||||
* news.json в DATA_DIR (по умолчанию ./data). Запускается по cron на хосте.
|
||||
*
|
||||
* Использование:
|
||||
* node scripts/pull-external-rss.mjs # пишет в ./data/news.json
|
||||
* DATA_DIR=/abs/path node scripts/pull-external-rss.mjs
|
||||
*/
|
||||
import fs from 'node:fs';
|
||||
import path from 'node:path';
|
||||
import { fileURLToPath } from 'node:url';
|
||||
import { XMLParser } from 'fast-xml-parser';
|
||||
|
||||
const __dirname = path.dirname(fileURLToPath(import.meta.url));
|
||||
const ROOT = path.resolve(__dirname, '..');
|
||||
const FEEDS_FILE = path.join(ROOT, 'src', 'data', 'feeds.json');
|
||||
const DATA_DIR = process.env.DATA_DIR || path.join(ROOT, 'data');
|
||||
const OUT_FILE = path.join(DATA_DIR, 'news.json');
|
||||
const TIMEOUT_MS = 15000;
|
||||
const HARD_CAP = 200;
|
||||
|
||||
const parser = new XMLParser({
|
||||
ignoreAttributes: false,
|
||||
attributeNamePrefix: '@_',
|
||||
textNodeName: '#text',
|
||||
});
|
||||
|
||||
async function fetchFeed(url, timeoutMs) {
|
||||
const ctl = new AbortController();
|
||||
const t = setTimeout(() => ctl.abort(), timeoutMs);
|
||||
try {
|
||||
const r = await fetch(url, {
|
||||
signal: ctl.signal,
|
||||
headers: { 'User-Agent': 'pushkinohistory-ru-v2 RSS aggregator' },
|
||||
});
|
||||
if (!r.ok) throw new Error(`HTTP ${r.status}`);
|
||||
return await r.text();
|
||||
} finally {
|
||||
clearTimeout(t);
|
||||
}
|
||||
}
|
||||
|
||||
function stripHtml(s) {
|
||||
if (!s) return '';
|
||||
return String(s).replace(/<[^>]+>/g, ' ').replace(/\s+/g, ' ').trim().slice(0, 400);
|
||||
}
|
||||
|
||||
function extractItems(xml, feed) {
|
||||
const parsed = parser.parse(xml);
|
||||
const rssItems = parsed?.rss?.channel?.item;
|
||||
if (rssItems) {
|
||||
const arr = Array.isArray(rssItems) ? rssItems : [rssItems];
|
||||
return arr.map((it) => ({
|
||||
title: typeof it.title === 'string' ? it.title : it.title?.['#text'] || '',
|
||||
link: typeof it.link === 'string' ? it.link : it.link?.['#text'] || '',
|
||||
guid: typeof it.guid === 'string' ? it.guid : it.guid?.['#text'] || it.link || '',
|
||||
pubDate: it.pubDate ? new Date(it.pubDate).toISOString() : null,
|
||||
description: stripHtml(it.description || it['content:encoded'] || ''),
|
||||
source: feed.name,
|
||||
}));
|
||||
}
|
||||
const atomEntries = parsed?.feed?.entry;
|
||||
if (atomEntries) {
|
||||
const arr = Array.isArray(atomEntries) ? atomEntries : [atomEntries];
|
||||
return arr.map((e) => {
|
||||
const link = Array.isArray(e.link)
|
||||
? e.link[0]?.['@_href']
|
||||
: e.link?.['@_href'] || e.link;
|
||||
return {
|
||||
title: typeof e.title === 'string' ? e.title : e.title?.['#text'] || '',
|
||||
link: link || '',
|
||||
guid: e.id || link || '',
|
||||
pubDate: e.updated || e.published ? new Date(e.updated || e.published).toISOString() : null,
|
||||
description: stripHtml(e.summary?.['#text'] || e.summary || e.content?.['#text'] || ''),
|
||||
source: feed.name,
|
||||
};
|
||||
});
|
||||
}
|
||||
return [];
|
||||
}
|
||||
|
||||
async function main() {
|
||||
const feeds = JSON.parse(fs.readFileSync(FEEDS_FILE, 'utf8')).filter((f) => f.enabled);
|
||||
if (feeds.length === 0) {
|
||||
console.log('no enabled feeds — writing empty news.json');
|
||||
fs.mkdirSync(DATA_DIR, { recursive: true });
|
||||
fs.writeFileSync(OUT_FILE, JSON.stringify({ updatedAt: new Date().toISOString(), items: [] }, null, 2));
|
||||
return;
|
||||
}
|
||||
|
||||
const all = [];
|
||||
for (const feed of feeds) {
|
||||
try {
|
||||
const xml = await fetchFeed(feed.url, TIMEOUT_MS);
|
||||
const items = extractItems(xml, feed);
|
||||
const max = feed.max || 20;
|
||||
all.push(...items.slice(0, max));
|
||||
console.log(`OK ${feed.name}: ${items.length} (kept ${Math.min(items.length, max)})`);
|
||||
} catch (e) {
|
||||
console.warn(`FAIL ${feed.name}: ${e.message}`);
|
||||
}
|
||||
}
|
||||
|
||||
const seen = new Set();
|
||||
const deduped = [];
|
||||
for (const it of all) {
|
||||
const key = it.guid || it.link;
|
||||
if (!key || seen.has(key)) continue;
|
||||
seen.add(key);
|
||||
deduped.push(it);
|
||||
}
|
||||
deduped.sort((a, b) => (b.pubDate || '').localeCompare(a.pubDate || ''));
|
||||
|
||||
const out = { updatedAt: new Date().toISOString(), items: deduped.slice(0, HARD_CAP) };
|
||||
fs.mkdirSync(DATA_DIR, { recursive: true });
|
||||
fs.writeFileSync(OUT_FILE, JSON.stringify(out, null, 2));
|
||||
console.log(`-> ${OUT_FILE}: ${out.items.length} items`);
|
||||
}
|
||||
|
||||
main().catch((e) => {
|
||||
console.error(e);
|
||||
process.exit(1);
|
||||
});
|
||||
Reference in New Issue
Block a user