perf: url fetch

This commit is contained in:
archer
2023-08-17 23:26:44 +08:00
parent 59ccc8565b
commit c5b5c440ca
5 changed files with 388 additions and 53 deletions

View File

@@ -1,51 +0,0 @@
// pages/api/fetchContent.ts
import { NextApiRequest, NextApiResponse } from 'next';
import axios from 'axios';
import { JSDOM } from 'jsdom';
import { Readability } from '@mozilla/readability';
import { jsonRes } from '@/service/response';
const fetchContent = async (req: NextApiRequest, res: NextApiResponse) => {
const { url } = req.body;
if (!url) {
return res.status(400).json({ error: 'URL is required' });
}
try {
const response = await axios.get(url, {
httpsAgent: new (require('https').Agent)({
rejectUnauthorized: false,
}),
});
const dom = new JSDOM(response.data, {
url,
contentType: 'text/html',
});
const reader = new Readability(dom.window.document);
const article = reader.parse();
if (!article) {
jsonRes(res, {
code: 500,
error: '页面获取失败或页面为空'
});
return;
}
jsonRes(res, {
code: 200,
data: article.content
});
} catch (error:any) {
jsonRes(res, {
code: 500,
error: error
});
}
};
export default fetchContent;

View File

@@ -0,0 +1,64 @@
// pages/api/fetchContent.ts
import { NextApiRequest, NextApiResponse } from 'next';
import axios from 'axios';
import { JSDOM } from 'jsdom';
import { Readability } from '@mozilla/readability';
import { jsonRes } from '@/service/response';
import { authUser } from '@/service/utils/auth';
type FetchResultItem = {
url: string;
title: string;
content: string;
};
export type UrlFetchResponse = FetchResultItem[];
const fetchContent = async (req: NextApiRequest, res: NextApiResponse) => {
try {
const { urlList = [] } = req.body as { urlList: string[] };
if (!urlList || urlList.length === 0) {
throw new Error('urlList is empty');
}
await authUser({ req });
const response = (
await Promise.allSettled(
urlList.map(async (url) => {
const fetchRes = await axios.get(url, {
timeout: 30000
});
const dom = new JSDOM(fetchRes.data, {
url,
contentType: 'text/html'
});
const reader = new Readability(dom.window.document);
const article = reader.parse();
return {
url,
title: article?.title || '',
content: article?.textContent || ''
};
})
)
)
.filter((item) => item.status === 'fulfilled')
.map((item: any) => item.value)
.filter((item) => item.content);
jsonRes<UrlFetchResponse>(res, {
data: response
});
} catch (error: any) {
jsonRes(res, {
code: 500,
error: error
});
}
};
export default fetchContent;