webstreamr-github/src/extractor/VidSrc.ts
WebStreamr 9ee91dff31
feat(extractor): add back VidSrc
This reverts commit f5f47403cc.
2026-01-14 13:38:09 +00:00

87 lines
3.3 KiB
TypeScript

import * as cheerio from 'cheerio';
import { BlockedError, TooManyRequestsError } from '../error';
import { Context, Format, Meta, NonEmptyArray, UrlResult } from '../types';
import { Fetcher, guessHeightFromPlaylist } from '../utils';
import { Extractor } from './Extractor';
export class VidSrc extends Extractor {
public readonly id = 'vidsrc';
public readonly label = 'VidSrc';
public override readonly ttl: number = 10800000; // 3h
private readonly domains: NonEmptyArray<string>;
public constructor(fetcher: Fetcher, domains: NonEmptyArray<string>) {
super(fetcher);
this.domains = domains;
}
public supports(_ctx: Context, url: URL): boolean {
return null !== url.host.match(/vidsrc|vsrc/);
}
protected async extractInternal(ctx: Context, url: URL, meta: Meta): Promise<UrlResult[]> {
// While this is a crappy thing to do, they seem to be blocking overly strict IMO
const randomIp = `${Math.floor(Math.random() * 223) + 1}.${Math.floor(Math.random() * 256)}.${Math.floor(Math.random() * 256)}.${Math.floor(Math.random() * 256)}`;
const newCtx = { ...ctx, ip: randomIp };
return this.extractUsingRandomDomain(newCtx, url, meta, [...this.domains]);
};
private async extractUsingRandomDomain(ctx: Context, url: URL, meta: Meta, domains: string[]): Promise<UrlResult[]> {
const domainIndex = Math.floor(Math.random() * domains.length);
const [domain] = domains.splice(domainIndex, 1) as [string];
const newUrl = new URL(url);
newUrl.hostname = domain;
let html: string;
try {
html = await this.fetcher.text(ctx, newUrl, { queueLimit: 1 });
} catch (error) {
/* istanbul ignore next */
if (domains.length && (error instanceof TooManyRequestsError || error instanceof BlockedError)) {
return this.extractUsingRandomDomain(ctx, url, meta, domains);
}
/* istanbul ignore next */
throw error;
}
const $ = cheerio.load(html.replace('<!--', '').replace('-->', '')); // server HTML is commented-out
const iframeUrl = new URL(($('#player_iframe').attr('src') as string).replace(/^\/\//, 'https://'));
const title = $('title').text().trim();
return Promise.all(
$('.server')
.map((_i, el) => ({ serverName: $(el).text(), dataHash: $(el).data('hash') }))
.toArray()
.filter(({ serverName }) => serverName === 'CloudStream Pro')
.map(async ({ serverName, dataHash }) => {
const rcpUrl = new URL(`/rcp/${dataHash}`, iframeUrl.origin);
const iframeHtml = await this.fetcher.text(ctx, rcpUrl, { headers: { Referer: newUrl.origin } });
const srcMatch = iframeHtml.match(`src:\\s?'(.*)'`) as string[];
const playerHtml = await this.fetcher.text(ctx, new URL(srcMatch[1] as string, iframeUrl.origin), { headers: { Referer: rcpUrl.href } });
const fileMatch = playerHtml.match(`(https:\\/\\/.*?{v\\d}.*?) or`) as string[];
const m3u8Url = new URL((fileMatch[1] as string).replace(/{v\d}/, iframeUrl.host));
return {
url: m3u8Url,
format: Format.hls,
label: serverName,
ttl: this.ttl,
meta: {
...meta,
height: await guessHeightFromPlaylist(ctx, this.fetcher, m3u8Url, { headers: { Referer: iframeUrl.href } }),
title,
},
};
}),
);
}
}