feat(route): add Cool Papers

lib/v2/papers/index.js

@@ -0,0 +1,103 @@
+const got = require('@/utils/got');
+const cheerio = require('cheerio');
+const { parseDate } = require('@/utils/parse-date');
+const { art } = require('@/utils/render');
+const path = require('path');
+
+module.exports = async (ctx) => {
+    const { category = 'arxiv/cs.CL' } = ctx.params;
+    const limit = ctx.query.limit ? parseInt(ctx.query.limit, 10) : 30;
+
+    const rootUrl = 'https://papers.cool';
+    const currentUrl = new URL(category, rootUrl).href;
+    const apiKimiUrl = new URL(`${category.split(/\//)[0]}/kimi/`, rootUrl).href;
+
+    const { data: response } = await got(currentUrl);
+
+    const $ = cheerio.load(response);
+
+    const pubDate = parseDate(
+        $('p.info')
+            .first()
+            .text()
+            .match(/(\d+\s\w+\s\d{4})/)[1],
+        ['DD MMM YYYY', 'D MMM YYYY']
+    );
+
+    let items = $('div.panel')
+        .slice(0, limit)
+        .toArray()
+        .map((item) => {
+            item = $(item);
+
+            const doi = item.prop('id');
+            const enclosureUrl =
+                item
+                    .find('a.pdf-preview')
+                    .prop('onclick')
+                    .match(/'(http.*?)'/)?.[1] ?? undefined;
+
+            return {
+                title: item.find('span[id]').first().text(),
+                link: item.find('a').first().prop('href'),
+                description: art(path.join(__dirname, 'templates/description.art'), {
+                    summary: item.find('p.summary').text(),
+                }),
+                author: item
+                    .find('p.authors a')
+                    .toArray()
+                    .map((a) => $(a).text())
+                    .join('; '),
+                guid: `${currentUrl}#${doi}`,
+                pubDate,
+                enclosure_url: enclosureUrl,
+                enclosure_type: enclosureUrl ? 'application/pdf' : undefined,
+                doi,
+            };
+        });
+
+    items = await Promise.all(
+        items.map(async (item) => {
+            const kimiURL = new URL(item.doi, apiKimiUrl).href;
+
+            const cache = await ctx.cache.get(kimiURL);
+            if (cache) {
+                return Promise.resolve(JSON.parse(cache));
+            }
+
+            try {
+                const { data: detailResponse } = await got(kimiURL, {
+                    timeout: {
+                        request: 5000,
+                    },
+                });
+
+                // Another cache with content by Kiwi Chat.
+
+                item.guid = `${item.guid}?kiwi`;
+                item.description += art(path.join(__dirname, 'templates/description.art'), {
+                    kimi: detailResponse,
+                });
+
+                ctx.cache.set(kimiURL, JSON.stringify(item));
+            } catch (e) {
+                // no-empty
+            }
+
+            return Promise.resolve(item);
+        })
+    );
+
+    const title = $('title').text();
+    const icon = new URL('favicon.ico', rootUrl).href;
+
+    ctx.state.data = {
+        item: items,
+        title: title.split(/-/)[0].trim(),
+        link: currentUrl,
+        description: title,
+        icon,
+        logo: icon,
+        subtitle: $('h1').first().text(),
+    };
+};

lib/v2/papers/maintainer.js

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

lib/v2/papers/radar.js

@@ -0,0 +1,17 @@
+module.exports = {
+    'papers.cool': {
+        _name: 'Cool Papers',
+        '.': [
+            {
+                title: 'Category',
+                docs: 'https://docs.rsshub.app/routes/journal#cool-papers-category',
+                source: ['/:category*'],
+                target: (params) => {
+                    const category = params.category;
+
+                    return `/papers${category ? `/${category}` : ''}`;
+                },
+            },
+        ],
+    },
+};

lib/v2/papers/router.js

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

lib/v2/papers/templates/description.art

@@ -0,0 +1,7 @@
+{{ if summary }}
+  <p>{{ summary }}</p>
+{{ /if }}
+
+{{ if kimi }}
+  {{@ kimi }}
+{{ /if }}

website/docs/routes/journal.mdx

@@ -114,6 +114,21 @@
   Including 'cell', 'cancer-cell', 'cell-chemical-biology', 'cell-host-microbe', 'cell-metabolism', 'cell-reports', 'cell-reports-physical-science', 'cell-stem-cell', 'cell-systems', 'chem', 'current-biology', 'developmental-cell', 'immunity', 'joule', 'matter', 'molecular-cell', 'neuron', 'one-earth' and 'structure'.
 </Route>
 
+## Cool Papers {#cool-papers}
+
+### Category {#cool-papers-category}
+
+<Route author="nczitzk" example="/papers" path="/papers/:category?" paramsDesc={['Category, see below, `arxiv/cs.CL` by default']} radar="1" supportBT="1" supportScihub="1">
+  | Category                                              | id          |
+  | ----------------------------------------------------- | ----------- |
+  | Arxiv Computation and Language (cs.CL)                | arxiv/cs.CL |
+  | Arxiv Machine Learning (cs.LG)                        | arxiv/cs.CL |
+  | Arxiv Artificial Intelligence (cs.AI)                 | arxiv/cs.CL |
+  | Arxiv Information Retrieval (cs.IR)                   | arxiv/cs.CL |
+  | Arxiv Computer Vision and Pattern Recognition (cs.CV) | arxiv/cs.CL |
+  | Arxiv Machine Learning (stat.ML)                      | arxiv/cs.CL |
+</Route>
+
 ## Deloitte {#deloitte}
 
 ### Articles {#deloitte-articles}