mirror of
https://github.com/imsyy/DailyHotApi.git
synced 2026-01-12 13:14:55 +08:00
修复虎嗅接口
改为抓取虎嗅移动端页面的 window.__NUXT__ 内联数据,使用移动 UA/Referer 获取有效列表。 解析 NUxT 数据时处理 moment_id/origin_publish_time,生成正确的链接与时间戳。 重新拆分内容:以首行作标题(去掉末尾句号),其余行合并为描述,避免标题夹带全文、描述为空。 增加结构缺失和解析失败的错误提示,便于排查。
This commit is contained in:
65
src/routes/huxiu.ts
Normal file → Executable file
65
src/routes/huxiu.ts
Normal file → Executable file
@@ -16,36 +16,57 @@ export const handleRoute = async (_: undefined, noCache: boolean) => {
|
||||
return routeData;
|
||||
};
|
||||
|
||||
// 标题处理
|
||||
const titleProcessing = (text: string) => {
|
||||
const paragraphs = text.split("<br><br>");
|
||||
const title = paragraphs.shift()?.replace(/。$/, "");
|
||||
const intro = paragraphs.join("<br><br>");
|
||||
return { title, intro };
|
||||
};
|
||||
|
||||
const getList = async (noCache: boolean) => {
|
||||
const url = `https://www.huxiu.com/moment/`;
|
||||
// 使用移动端页面,数据通过 window.__NUXT__ 内联
|
||||
const url = `https://m.huxiu.com/moment/`;
|
||||
const result = await get({
|
||||
url,
|
||||
noCache,
|
||||
headers: {
|
||||
"User-Agent":
|
||||
"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://m.huxiu.com/moment/",
|
||||
},
|
||||
responseType: "text",
|
||||
});
|
||||
// 正则查找
|
||||
const pattern =
|
||||
/<script>[\s\S]*?window\.__INITIAL_STATE__\s*=\s*(\{[\s\S]*?\});[\s\S]*?<\/script>/;
|
||||
// 正则查找内联的 NUxt 数据
|
||||
const pattern = /window.__NUXT__=(.*?);<\/script>/s;
|
||||
const matchResult = result.data.match(pattern);
|
||||
const jsonObject = JSON.parse(matchResult[1]).moment.momentList.moment_list.datalist;
|
||||
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: ${msg}`);
|
||||
}
|
||||
const list = nuxtData?.data?.[0]?.momentList || [];
|
||||
return {
|
||||
...result,
|
||||
data: jsonObject.map((v: RouterType["huxiu"]) => ({
|
||||
id: v.object_id,
|
||||
title: titleProcessing(v.content).title,
|
||||
desc: titleProcessing(v.content).intro,
|
||||
author: v.user_info.username,
|
||||
timestamp: getTime(v.publish_time),
|
||||
data: list.map((v: RouterType["huxiu"]) => {
|
||||
const content = (v.content || "").replace(/<br\s*\/?>/gi, "\n");
|
||||
const [titleLine, ...rest] = content
|
||||
.split("\n")
|
||||
.map((s) => s.trim())
|
||||
.filter(Boolean);
|
||||
const title = titleLine?.replace(/。$/, "") || "";
|
||||
const intro = rest.join("\n");
|
||||
const momentId = v.moment_id || v.object_id;
|
||||
return {
|
||||
id: momentId,
|
||||
title,
|
||||
desc: intro,
|
||||
author: v.user_info?.username || "",
|
||||
timestamp: getTime(v.origin_publish_time || v.publish_time),
|
||||
hot: undefined,
|
||||
url: v.url || `https://www.huxiu.com/moment/${v.object_id}.html`,
|
||||
mobileUrl: v.url || `https://m.huxiu.com/moment/${v.object_id}.html`,
|
||||
})),
|
||||
url: v.url || `https://www.huxiu.com/moment/${momentId}.html`,
|
||||
mobileUrl: v.url || `https://m.huxiu.com/moment/${momentId}.html`,
|
||||
};
|
||||
}),
|
||||
};
|
||||
};
|
||||
|
||||
Reference in New Issue
Block a user