xiaoxue_admin
/
RSSHub
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix(route): researchgate (#11429)

This commit is contained in:
Tony 2022-12-09 22:41:37 +05:00 committed by GitHub
parent cd0b88f1c2
commit 8ed999f1b5
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
10 changed files with 102 additions and 73 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
docs/en/new-media.md
View File

@ -581,12 +581,6 @@ Compared to the official one, this feed:
</RouteEn>
## Research Gate
### Publications
<RouteEn author="nczitzk" example="/researchgate/publications/Somsak-Panha" path="/researchgate/publications/:username" :paramsDesc="['Username, can be found in URL']"/>
## RSS3
### Blog

6
docs/en/study.md
View File

@ -108,6 +108,12 @@ pageClass: routes
<RouteEn author="OrangeEd1t" example="/orcid/0000-0002-4731-9700" path="/orcid/:id" :paramsDesc="['Open Researcher and Contributor ID']"/>
## ResearchGate
### Publications
<RouteEn author="nczitzk" example="/researchgate/publications/Somsak-Panha" path="/researchgate/publications/:username" :paramsDesc="['Username, can be found in URL']" puppeteer="1" anticrawler="1"/>
## X-MOL
### News

6
docs/new-media.md
View File

@ -1187,12 +1187,6 @@ IPFS 网关有可能失效，那时候换成其他网关。
</Route>
## Research Gate
### Publications
<Route author="nczitzk" example="/researchgate/publications/Somsak-Panha" path="/researchgate/publications/:username" :paramsDesc="['用户名，可在用户页地址栏中找到']"/>
## RSS3
### Blog

6
docs/study.md
View File

@ -234,6 +234,12 @@ path="/ctfhub/upcoming/:limit?"
</Route>
## ResearchGate
### Publications
<Route author="nczitzk" example="/researchgate/publications/Somsak-Panha" path="/researchgate/publications/:username" :paramsDesc="['用户名，可在用户页地址栏中找到']" puppeteer="1" anticrawler="1"/>
## X-MOL 平台
### 新闻

2
lib/router.js
View File

@ -4028,7 +4028,7 @@ router.get('/sbs/chinese/:category?/:id?/:dialect?/:language?', lazyloadRouteHan
// router.get('/asiantolick/:category?/:keyword?', lazyloadRouteHandler('./routes/asiantolick'));
// Research Gate
router.get('/researchgate/publications/:id', lazyloadRouteHandler('./routes/researchgate/publications'));
// router.get('/researchgate/publications/:id', lazyloadRouteHandler('./routes/researchgate/publications'));
// QuestMobile
router.get('/questmobile/report/:category?/:label?', lazyloadRouteHandler('./routes/questmobile/report'));

60
lib/routes/researchgate/publications.js
View File

@ -1,60 +0,0 @@
const got = require('@/utils/got');
const cheerio = require('cheerio');
module.exports = async (ctx) => {
const id = ctx.params.id;
const rootUrl = 'https://www.researchgate.net';
const currentUrl = `${rootUrl}/profile/${id}`;
const response = await got({
method: 'get',
url: currentUrl,
});
const $ = cheerio.load(response.data);
const list = $('div[itemprop="headline"] a')
.slice(0, 15)
.map((_, item) => {
item = $(item);
return {
title: item.text(),
link: item.attr('href'),
};
})
.get();
const items = await Promise.all(
list.map(
async (item) =>
await ctx.cache.tryGet(item.link, async () => {
const detailResponse = await got({
method: 'get',
url: item.link,
});
const content = cheerio.load(detailResponse.data);
item.doi = content('meta[property="citation_doi"]').attr('content');
item.pubDate = Date.parse(content('meta[property="citation_publication_date"]').attr('content'));
const authors = [];
content('meta[property="citation_author"]').each(function () {
authors.push(content(this).attr('content'));
});
item.author = authors.join(', ');
item.description = content('div[itemprop="description"]').html();
return item;
})
)
);
ctx.state.data = {
title: `${$('meta[property="profile:username"]').attr('content')}'s Publications - ResearchGate`,
link: currentUrl,
item: items,
};
};

3
lib/v2/researchgate/maintainer.js Normal file
View File

@ -0,0 +1,3 @@
module.exports = {
'/publications/:username': ['nczitzk'],
};

70
lib/v2/researchgate/publications.js Normal file
View File

@ -0,0 +1,70 @@
const cheerio = require('cheerio');
const { parseDate } = require('@/utils/parse-date');
module.exports = async (ctx) => {
const id = ctx.params.id;
const rootUrl = 'https://www.researchgate.net';
const currentUrl = `${rootUrl}/profile/${id}`;
const browser = await require('@/utils/puppeteer')();
const page = await browser.newPage();
await page.setRequestInterception(true);
page.on('request', (request) => {
request.resourceType() === 'document' || request.resourceType() === 'script' ? request.continue() : request.abort();
});
await page.goto(currentUrl);
const response = await page.evaluate(() => document.documentElement.innerHTML);
await page.close();
const $ = cheerio.load(response);
const list = $('div[itemprop="headline"] a')
.toArray()
.slice(0, ctx.query.limit ? Number(ctx.query.limit) : 15)
.map((item) => {
item = $(item);
return {
title: item.text(),
link: item.attr('href'),
};
});
const items = await Promise.all(
list.map((item) =>
ctx.cache.tryGet(item.link, async () => {
const page = await browser.newPage();
await page.setRequestInterception(true);
page.on('request', (request) => {
request.resourceType() === 'document' || request.resourceType() === 'script' ? request.continue() : request.abort();
});
await page.goto(item.link);
const detailResponse = await page.evaluate(() => document.documentElement.innerHTML);
await page.close();
const content = cheerio.load(detailResponse);
item.doi = content('meta[property="citation_doi"]').attr('content');
item.pubDate = parseDate(content('meta[property="citation_publication_date"]').attr('content'));
const authors = [];
content('meta[property="citation_author"]').each(function () {
authors.push(content(this).attr('content'));
});
item.author = authors.join(', ');
item.description = content('div[itemprop="description"]').html();
return item;
})
)
);
await browser.close();
ctx.state.data = {
title: `${$('meta[property="profile:username"]').attr('content')}'s Publications - ResearchGate`,
link: currentUrl,
item: items,
};
};

13
lib/v2/researchgate/radar.js Normal file
View File

@ -0,0 +1,13 @@
module.exports = {
'researchgate.net': {
_name: 'ResearchGate',
'.': [
{
title: 'Publications',
docs: 'https://docs.rsshub.app/study.html#researchgate',
source: ['/profile/:username'],
target: '/researchgate/publications/:username',
},
],
},
};

3
lib/v2/researchgate/router.js Normal file
View File

@ -0,0 +1,3 @@
module.exports = (router) => {
router.get('/publications/:id', require('./publications'));
};