elmgates/DailyHotApi
1
0
Fork 0
mirror of https://github.com/imsyy/DailyHotApi.git synced 2026-01-12 13:14:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

虎嗅接口：从移动端页面爬取改为原生feed API

Browse Source 
移动端页面内容爬取不稳定，标题和描述拆分易出错；原生feed接口获取的内容更为稳定。
This commit is contained in:
xuan
2025-12-19 05:58:16 +08:00
parent 82f983793c
commit 17d1658818
1 changed files with 17 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
src/routes/huxiu.ts
View File

@@ -1,7 +1,7 @@
import type { RouterData } from "../types.js"; import type { RouterData } from "../types.js";
import type { RouterType } from "../router.types.js"; import type { RouterType } from "../router.types.js";
import { get } from "../utils/getData.js";
import { getTime } from "../utils/getTime.js"; import { getTime } from "../utils/getTime.js";
import axios from "axios";
export const handleRoute = async (_: undefined, noCache: boolean) => { export const handleRoute = async (_: undefined, noCache: boolean) => {
const listData = await getList(noCache); const listData = await getList(noCache);
@@ -17,37 +17,21 @@ export const handleRoute = async (_: undefined, noCache: boolean) => {
}; };
const getList = async (noCache: boolean) => { const getList = async (noCache: boolean) => {
// 使用移动端页面，数据通过 window.__NUXT__ 内联 // PC 端接口
const url = `https://m.huxiu.com/moment/`; const url = `https://moment-api.huxiu.com/web-v3/moment/feed?platform=www`;
const result = await get({ const res = await axios.get(url, {
url,
noCache,
headers: { headers: {
"User-Agent": "User-Agent": "Mozilla/5.0",
"Mozilla/5.0 (iPhone; CPU iPhone OS 15_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/15.0 Mobile/15E148 Safari/604.1", Referer: "https://www.huxiu.com/moment/",
Referer: "https://m.huxiu.com/moment/",
}, },
responseType: "text", timeout: 10000,
}); });
// 正则查找内联的 NUxt 数据 const list: RouterType["huxiu"][] = res.data?.data?.moment_list?.datalist || [];
const pattern = /window.__NUXT__=(.*?);<\/script>/s;
const matchResult = result.data.match(pattern);
if (!matchResult || !matchResult[1]) {
throw new Error("虎嗅页面结构变更，未找到内联数据");
}
const expr = matchResult[1].trim();
let nuxtData: any;
try {
// 直接求值表达式 (function(...){return {...}})(...)
// eslint-disable-next-line no-eval
nuxtData = eval(expr);
} catch (err) {
const msg = err instanceof Error ? err.message : "未知错误";
throw new Error(`虎嗅数据解析失败: ${msg}`);
}
const list = nuxtData?.data?.[0]?.momentList || [];
return { return {
...result, fromCache: false,
updateTime: res.data?.data?.moment_list?.last_id
? getTime(res.data.data.moment_list.last_id)
: undefined,
data: list.map((v: RouterType["huxiu"]) => { data: list.map((v: RouterType["huxiu"]) => {
const content = (v.content || "").replace(/<br\s*\/?>/gi, "\n"); const content = (v.content || "").replace(/<br\s*\/?>/gi, "\n");
const [titleLine, ...rest] = content const [titleLine, ...rest] = content
@@ -56,16 +40,16 @@ const getList = async (noCache: boolean) => {
.filter(Boolean); .filter(Boolean);
const title = titleLine?.replace(/。$/, "") || ""; const title = titleLine?.replace(/。$/, "") || "";
const intro = rest.join("\n"); const intro = rest.join("\n");
const momentId = v.moment_id || v.object_id; const momentId = v.object_id;
return { return {
id: momentId, id: momentId,
title, title,
desc: intro, desc: intro,
author: v.user_info?.username || "", author: v.user_info?.username || "",
timestamp: getTime(v.origin_publish_time || v.publish_time), timestamp: getTime(v.publish_time),
hot: undefined, hot: v.count_info?.agree_num,
url: v.url || `https://www.huxiu.com/moment/${momentId}.html`, url: `https://www.huxiu.com/moment/${momentId}.html`,
mobileUrl: v.url || `https://m.huxiu.com/moment/${momentId}.html`, mobileUrl: `https://m.huxiu.com/moment/${momentId}.html`,
}; };
}), }),
}; };