import sanitizeHtml from 'sanitize-html';
import { SITE_URL } from '../consts';
/** Sanitize и нормализация HTML тела поста для RSS ``. */
export function sanitizeForRss(html: string): string {
const absolutized = html.replace(/(src|href)=["']\/uploads\//g, `$1="${SITE_URL}/uploads/`);
return sanitizeHtml(absolutized, {
allowedTags: [
'p', 'br', 'strong', 'em', 'b', 'i', 'u', 's',
'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
'a', 'img',
'ul', 'ol', 'li',
'blockquote', 'pre', 'code',
'table', 'thead', 'tbody', 'tr', 'th', 'td',
'div', 'span', 'hr',
],
allowedAttributes: {
a: ['href', 'title', 'rel', 'target'],
img: ['src', 'alt', 'title', 'width', 'height'],
div: ['align'],
span: ['align'],
p: ['align'],
td: ['colspan', 'rowspan'],
th: ['colspan', 'rowspan'],
},
transformTags: {
a: (tagName, attribs) => ({
tagName,
attribs: {
...attribs,
target: '_blank',
rel: 'noopener noreferrer',
},
}),
},
allowedSchemes: ['http', 'https', 'mailto'],
});
}
/** Экранировать строку для XML-атрибутов и текстовых нод. */
export function xmlEscape(s: string): string {
return String(s)
.replace(/&/g, '&')
.replace(//g, '>')
.replace(/"/g, '"')
.replace(/'/g, ''');
}
/** Завернуть в CDATA, нейтрализовав `]]>` внутри. */
export function cdata(s: string): string {
return `/g, ']]]]>')}]]>`;
}
/** Текстовое описание из HTML (для ``). */
export function plainTextExcerpt(html: string, max = 500): string {
const txt = html
.replace(//gi, ' ')
.replace(//gi, ' ')
.replace(/<[^>]+>/g, ' ')
.replace(/ /g, ' ')
.replace(/&/g, '&')
.replace(/</g, '<')
.replace(/>/g, '>')
.replace(/"/g, '"')
.replace(/\s+/g, ' ')
.trim();
if (txt.length <= max) return txt;
return txt.slice(0, max).replace(/\s+\S*$/, '') + '…';
}
/** Построить полный RSS 2.0 XML с CDATA-завёрнутым content:encoded. */
export function buildRss(opts: {
title: string;
description: string;
selfUrl: string;
language: string;
items: Array<{
title: string;
link: string;
pubDate: Date;
description: string;
contentHtml: string;
author: string;
categories: string[];
}>;
}): string {
const { title, description, selfUrl, language, items } = opts;
const lastBuild = new Date().toUTCString();
const itemsXml = items
.map((it) => ` -
${cdata(it.title)}
${xmlEscape(it.link)}
${xmlEscape(it.link)}
${it.pubDate.toUTCString()}
${cdata(it.author)}
${it.categories.map((c) => ` ${cdata(c)}`).join('\n')}
${cdata(it.description)}
${cdata(it.contentHtml)}
`)
.join('\n');
return `
${cdata(title)}
${xmlEscape(SITE_URL + '/')}
${cdata(description)}
${language}
${lastBuild}
60
${itemsXml}
`;
}