import { z } from 'zod';
import { ArticleRepository } from '../database/repository.js';
/**
* Input schema for rss_media tool
*/
export const RssMediaInputSchema = z.object({
article_id: z.string().optional().describe('Specific article ID'),
feed_id: z.string().optional().describe('Get media from specific feed'),
media_type: z.enum(['image', 'video', 'all']).default('all').describe('Type of media to extract'),
limit: z.number().default(20).describe('Number of articles to process')
});
/**
* Output schema for rss_media tool
*/
export const RssMediaOutputSchema = z.object({
media_items: z.array(z.object({
id: z.string(),
article_id: z.string(),
article_title: z.string(),
media_type: z.string(),
url: z.string(),
width: z.number().nullable(),
height: z.number().nullable(),
caption: z.string().nullable()
})),
total: z.number()
});
export type RssMediaInput = z.infer<typeof RssMediaInputSchema>;
export type RssMediaOutput = z.infer<typeof RssMediaOutputSchema>;
/**
* Extract media (images/videos) from articles
*/
export async function handleRssMedia(
input: RssMediaInput,
articleRepository: ArticleRepository
): Promise<RssMediaOutput> {
const { article_id, feed_id, media_type, limit } = input;
const db = articleRepository['db'];
let articles: any[];
if (article_id) {
articles = [db.prepare('SELECT * FROM articles WHERE id = ?').get(article_id)];
} else if (feed_id) {
articles = db.prepare(`
SELECT * FROM articles
WHERE feed_id = ?
ORDER BY pub_date DESC
LIMIT ?
`).all(feed_id, limit) as any[];
} else {
articles = db.prepare(`
SELECT * FROM articles
ORDER BY pub_date DESC
LIMIT ?
`).all(limit) as any[];
}
const mediaItems: any[] = [];
for (const article of articles) {
if (!article) continue;
const content = article.content || article.description || '';
// Extract images
if (media_type === 'image' || media_type === 'all') {
const imgRegex = /<img[^>]+src="([^">]+)"/g;
let match;
while ((match = imgRegex.exec(content)) !== null) {
const url = match[1];
// Try to extract dimensions
const widthMatch = content.match(/width[=:"]\s*(\d+)/i);
const heightMatch = content.match(/height[=:"]\s*(\d+)/i);
mediaItems.push({
id: `media_${Date.now()}_${Math.random().toString(36).substr(2, 9)}`,
article_id: article.id,
article_title: article.title,
media_type: 'image',
url,
width: widthMatch ? parseInt(widthMatch[1]) : null,
height: heightMatch ? parseInt(heightMatch[1]) : null,
caption: null
});
}
}
// Extract videos
if (media_type === 'video' || media_type === 'all') {
const videoRegex = /<video[^>]+src="([^">]+)"|<source[^>]+src="([^">]+)"/g;
let match;
while ((match = videoRegex.exec(content)) !== null) {
const url = match[1] || match[2];
mediaItems.push({
id: `media_${Date.now()}_${Math.random().toString(36).substr(2, 9)}`,
article_id: article.id,
article_title: article.title,
media_type: 'video',
url,
width: null,
height: null,
caption: null
});
}
// Check for YouTube/Vimeo embeds
const youtubeRegex = /youtube\.com\/embed\/([a-zA-Z0-9_-]+)|youtu\.be\/([a-zA-Z0-9_-]+)/g;
while ((match = youtubeRegex.exec(content)) !== null) {
const videoId = match[1] || match[2];
mediaItems.push({
id: `media_${Date.now()}_${Math.random().toString(36).substr(2, 9)}`,
article_id: article.id,
article_title: article.title,
media_type: 'video',
url: `https://www.youtube.com/watch?v=${videoId}`,
width: null,
height: null,
caption: null
});
}
}
}
return {
media_items: mediaItems,
total: mediaItems.length
};
}
