mirror of
https://github.com/DIYgod/RSSHub.git
synced 2026-08-31 01:01:11 +08:00
fix(route/huggingface): restore article list and author parsing (#20599)
* fix(route/huggingface): restore article list and author parsing - lib/routes/huggingface/blog-zh.ts: Read articles from data.props.allBlogs and construct /blog/zh/:slug links; use authorsData (fullname or name) for author names. - lib/routes/huggingface/blog.ts: Read articles from data.props.allBlogs and switch author parsing to authorsData (fullname or name); preserve upvotes and thumbnail handling. - lib/routes/huggingface/daily-papers.ts: Relax selector to read data.props from div[data-target="DailyPapers"] so DailyPapers props can be retrieved. * fix(route/huggingface): use api instead of http; title is at right property.
This commit is contained in:
Jinghao Tu
@@ -1,8 +1,8 @@
|
||||
import { load } from 'cheerio';
|
||||
|
||||
import type { Route } from '@/types';
|
||||
import type { DataItem, Route } from '@/types';
|
||||
import cache from '@/utils/cache';
|
||||
import got from '@/utils/got';
|
||||
import ofetch from '@/utils/ofetch';
|
||||
import { parseDate } from '@/utils/parse-date';
|
||||
|
||||
export const route: Route = {
|
||||
@@ -38,18 +38,18 @@ export const route: Route = {
|
||||
url: 'huggingface.co/blog/community',
|
||||
};
|
||||
|
||||
interface CommunityBlog {
|
||||
classNames: string;
|
||||
posts: Post[];
|
||||
pagination: Pagination;
|
||||
sort: string;
|
||||
showBackLink: boolean;
|
||||
}
|
||||
|
||||
interface Pagination {
|
||||
numItemsPerPage: number;
|
||||
numTotalItems: number;
|
||||
pageIndex: number;
|
||||
interface AuthorData {
|
||||
_id?: string;
|
||||
avatarUrl: string;
|
||||
fullname: string;
|
||||
name: string;
|
||||
type: 'org' | 'user';
|
||||
isPro?: boolean;
|
||||
isHf: boolean;
|
||||
isHfAdmin: boolean;
|
||||
isMod: boolean;
|
||||
followerCount: number;
|
||||
isEnterprise?: boolean;
|
||||
}
|
||||
|
||||
interface Post {
|
||||
@@ -60,60 +60,45 @@ interface Post {
|
||||
numCoauthors: number;
|
||||
publishedAt: string;
|
||||
slug: string;
|
||||
status: Status;
|
||||
status: string;
|
||||
title: string;
|
||||
upvotes: number;
|
||||
thumbnail?: string;
|
||||
}
|
||||
|
||||
interface AuthorData {
|
||||
_id?: string;
|
||||
avatarUrl: string;
|
||||
fullname: string;
|
||||
name: string;
|
||||
type: Type;
|
||||
isPro?: boolean;
|
||||
isHf: boolean;
|
||||
isHfAdmin: boolean;
|
||||
isMod: boolean;
|
||||
followerCount: number;
|
||||
isEnterprise?: boolean;
|
||||
}
|
||||
|
||||
export enum Type {
|
||||
Org = 'org',
|
||||
User = 'user',
|
||||
}
|
||||
|
||||
export enum Status {
|
||||
Published = 'published',
|
||||
interface CommunityBlogApiResponse {
|
||||
posts: Post[];
|
||||
pagination: {
|
||||
numItemsPerPage: number;
|
||||
numTotalItems: number;
|
||||
pageIndex: number;
|
||||
};
|
||||
}
|
||||
|
||||
async function handler(ctx) {
|
||||
const { sort = 'trending' } = ctx.req.param();
|
||||
const { body: response } = await got(`https://huggingface.co/blog/community?sort=${sort}`);
|
||||
const $ = load(response);
|
||||
const response = await ofetch<CommunityBlogApiResponse>(`https://huggingface.co/api/blog/community?sort=${sort}`);
|
||||
|
||||
const data = $('div[data-target="CommunityBlogsContainer"]').attr('data-props');
|
||||
if (!data) {
|
||||
throw new Error('Failed to fetch data from Huggingface Community Articles');
|
||||
}
|
||||
const articles: CommunityBlog = JSON.parse(data);
|
||||
const lists = articles.posts.map((item) => ({
|
||||
const { posts } = response;
|
||||
|
||||
const lists = posts.map((item) => ({
|
||||
title: item.title,
|
||||
link: `https://huggingface.co/blog/${item.authorData.name}/${item.slug}`,
|
||||
pubDate: parseDate(item.publishedAt),
|
||||
author: item.authorData.name,
|
||||
author: item.authorData.fullname || item.authorData.name,
|
||||
upvotes: item.upvotes,
|
||||
image: item.thumbnail ? new URL(item.thumbnail, 'https://huggingface.co').toString() : undefined,
|
||||
}));
|
||||
|
||||
const items = await Promise.all(
|
||||
const items: DataItem[] = await Promise.all(
|
||||
lists.map((item) =>
|
||||
cache.tryGet(item.link, async () => {
|
||||
const { body: response } = await got(item.link);
|
||||
const response = await ofetch(item.link);
|
||||
const $ = load(response);
|
||||
$('.mb-4, h1, .mb-6, .not-prose').remove();
|
||||
$('.mb-4, .mb-6, .not-prose, h1').remove();
|
||||
return {
|
||||
...item,
|
||||
description: $('.blog-content').html(),
|
||||
description: $('.blog-content').html() ?? undefined,
|
||||
};
|
||||
})
|
||||
)
|
||||
|
||||
@@ -1,9 +1,29 @@
|
||||
import { load } from 'cheerio';
|
||||
|
||||
import type { Route } from '@/types';
|
||||
import got from '@/utils/got';
|
||||
import type { DataItem, Route } from '@/types';
|
||||
import cache from '@/utils/cache';
|
||||
import ofetch from '@/utils/ofetch';
|
||||
import { parseDate } from '@/utils/parse-date';
|
||||
|
||||
interface AuthorData {
|
||||
fullname?: string;
|
||||
name: string;
|
||||
}
|
||||
|
||||
interface BlogItem {
|
||||
slug: string;
|
||||
title: string;
|
||||
publishedAt: string;
|
||||
authorsData: AuthorData[];
|
||||
upvotes: number;
|
||||
thumbnail?: string;
|
||||
tags: string[];
|
||||
}
|
||||
|
||||
interface BlogApiResponse {
|
||||
allBlogs: BlogItem[];
|
||||
}
|
||||
|
||||
export const route: Route = {
|
||||
path: '/blog-zh',
|
||||
categories: ['programming'],
|
||||
@@ -29,29 +49,36 @@ export const route: Route = {
|
||||
};
|
||||
|
||||
async function handler() {
|
||||
const { body: response } = await got('https://huggingface.co/blog/zh');
|
||||
const $ = load(response);
|
||||
const response = await ofetch<BlogApiResponse>('https://huggingface.co/api/blog/zh');
|
||||
|
||||
/** @type {Array<{blog: {local: string, title: string, author: string, thumbnail: string, date: string, tags: Array<string>}, blogUrl: string, lang: 'zh', link: string}>} */
|
||||
const papers = $('div[data-target="BlogThumbnail"]')
|
||||
.toArray()
|
||||
.map((item) => {
|
||||
const props = $(item).data('props');
|
||||
const link = $(item).find('a').attr('href');
|
||||
return {
|
||||
...props,
|
||||
link,
|
||||
};
|
||||
});
|
||||
const { allBlogs } = response;
|
||||
|
||||
const items = papers.map((item) => ({
|
||||
title: item.blog.title,
|
||||
link: `https://huggingface.co${item.link}`,
|
||||
category: item.blog.tags,
|
||||
pubDate: parseDate(item.blog.publishedAt),
|
||||
author: item.blog.author,
|
||||
const lists = allBlogs.map((blog) => ({
|
||||
title: blog.title,
|
||||
link: `https://huggingface.co/blog/zh/${blog.slug}`,
|
||||
pubDate: parseDate(blog.publishedAt),
|
||||
author: blog.authorsData.map((author) => ({
|
||||
name: author.fullname || author.name,
|
||||
})),
|
||||
upvotes: blog.upvotes,
|
||||
image: blog.thumbnail ? new URL(blog.thumbnail, 'https://huggingface.co').toString() : undefined,
|
||||
category: blog.tags,
|
||||
}));
|
||||
|
||||
const items: DataItem[] = await Promise.all(
|
||||
lists.map((item) =>
|
||||
cache.tryGet(item.link, async () => {
|
||||
const response = await ofetch(item.link);
|
||||
const $ = load(response);
|
||||
$('.mb-4, .mb-6, .not-prose, h1').remove();
|
||||
return {
|
||||
...item,
|
||||
description: $('.blog-content').html() ?? undefined,
|
||||
};
|
||||
})
|
||||
)
|
||||
);
|
||||
|
||||
return {
|
||||
title: 'Huggingface 中文博客',
|
||||
link: 'https://huggingface.co/blog/zh',
|
||||
|
||||
@@ -1,9 +1,29 @@
|
||||
import { load } from 'cheerio';
|
||||
|
||||
import type { DataItem, Route } from '@/types';
|
||||
import got from '@/utils/got';
|
||||
import cache from '@/utils/cache';
|
||||
import ofetch from '@/utils/ofetch';
|
||||
import { parseDate } from '@/utils/parse-date';
|
||||
|
||||
interface AuthorData {
|
||||
fullname?: string;
|
||||
name: string;
|
||||
}
|
||||
|
||||
interface BlogItem {
|
||||
slug: string;
|
||||
title: string;
|
||||
publishedAt: string;
|
||||
authorsData: AuthorData[];
|
||||
upvotes: number;
|
||||
thumbnail: string;
|
||||
tags: string[];
|
||||
}
|
||||
|
||||
interface BlogApiResponse {
|
||||
allBlogs: BlogItem[];
|
||||
}
|
||||
|
||||
export const route: Route = {
|
||||
path: '/blog',
|
||||
categories: ['programming'],
|
||||
@@ -28,58 +48,37 @@ export const route: Route = {
|
||||
url: 'huggingface.co/blog',
|
||||
};
|
||||
|
||||
interface Author {
|
||||
user: string;
|
||||
guest: boolean;
|
||||
org?: string;
|
||||
}
|
||||
|
||||
interface Blog {
|
||||
authors: Author[];
|
||||
canonical: boolean;
|
||||
isUpvotedByUser: boolean;
|
||||
publishedAt: string;
|
||||
slug: string;
|
||||
title: string;
|
||||
upvotes: number;
|
||||
thumbnail: string;
|
||||
guest: boolean;
|
||||
}
|
||||
|
||||
interface BlogData {
|
||||
blog: Blog;
|
||||
blogUrl: string;
|
||||
lang: string;
|
||||
loggedInUser: string;
|
||||
}
|
||||
|
||||
async function handler() {
|
||||
const { body: response } = await got('https://huggingface.co/blog');
|
||||
const $ = load(response);
|
||||
const response = await ofetch<BlogApiResponse>('https://huggingface.co/api/blog');
|
||||
|
||||
/** @type {Array<{blog: {local: string, title: string, author: string, thumbnail: string, date: string, tags: Array<string>}, blogUrl: string, lang: 'zh', link: string}>} */
|
||||
const papers = $('div[data-target="BlogThumbnail"]')
|
||||
.toArray()
|
||||
.map((item) => {
|
||||
const props = $(item).data('props') as BlogData;
|
||||
const link = $(item).find('a').attr('href');
|
||||
return {
|
||||
...props,
|
||||
link,
|
||||
};
|
||||
});
|
||||
const { allBlogs } = response;
|
||||
|
||||
const items: DataItem[] = papers.map((item) => ({
|
||||
title: item.blog.title,
|
||||
link: `https://huggingface.co${item.link}`,
|
||||
pubDate: parseDate(item.blog.publishedAt),
|
||||
author: item.blog.authors.map((author) => ({
|
||||
name: author.user,
|
||||
const lists = allBlogs.map((blog) => ({
|
||||
title: blog.title,
|
||||
link: `https://huggingface.co/blog/${blog.slug}`,
|
||||
pubDate: parseDate(blog.publishedAt),
|
||||
author: blog.authorsData.map((author) => ({
|
||||
name: author.fullname || author.name,
|
||||
})),
|
||||
upvotes: item.blog.upvotes,
|
||||
image: new URL(item.blog.thumbnail, 'https://huggingface.co').toString(),
|
||||
upvotes: blog.upvotes,
|
||||
image: blog.thumbnail ? new URL(blog.thumbnail, 'https://huggingface.co').toString() : undefined,
|
||||
category: blog.tags,
|
||||
}));
|
||||
|
||||
const items: DataItem[] = await Promise.all(
|
||||
lists.map((item) =>
|
||||
cache.tryGet(item.link, async () => {
|
||||
const response = await ofetch(item.link);
|
||||
const $ = load(response);
|
||||
$('.mb-4, .mb-6, .not-prose, h1').remove();
|
||||
return {
|
||||
...item,
|
||||
description: $('.blog-content').html() ?? undefined,
|
||||
};
|
||||
})
|
||||
)
|
||||
);
|
||||
|
||||
return {
|
||||
title: 'Huggingface 英文博客',
|
||||
link: 'https://huggingface.co/blog',
|
||||
|
||||
@@ -66,7 +66,7 @@ async function handler(ctx) {
|
||||
|
||||
const { body: response } = await got(url);
|
||||
const $ = load(response);
|
||||
const papers = $('main > div[data-target="DailyPapers"]').data('props') as PapersData;
|
||||
const papers = $('div[data-target="DailyPapers"]').data('props') as PapersData;
|
||||
|
||||
const items = papers.dailyPapers
|
||||
.filter((item) => item.paper.upvotes >= voteFliter)
|
||||
|
||||
Reference in New Issue
Block a user