Merge branch 'main' into main

README.md

@@ -14,9 +14,13 @@
         <a href="https://discord.gg/pQPt8HBUPu">
             💬 community discord server
         </a>
+        <br/>
         <a href="https://x.com/justusecobalt">
             🐦 twitter
         </a>
+        <a href="https://bsky.app/profile/cobalt.tools">
+            🦋 bluesky
+        </a>
     </p>
     <br/>
 </div>
@@ -34,11 +38,10 @@ this monorepo includes source code for api, frontend, and related packages:
 it also includes documentation in the [docs tree](/docs/):
 - [cobalt api documentation](/docs/api.md)
 - [how to run a cobalt instance](/docs/run-an-instance.md)
-- [how to protect a cobalt instance](/docs/protect-an-instance.md)
-- [how to configure a cobalt instance for youtube](/docs/configure-for-youtube.md)
+- [how to protect a cobalt instance](/docs/protect-an-instance.md) (recommended if you host a public instance)
 
 ### thank you
-cobalt is sponsored by [royalehosting.net](https://royalehosting.net/?partner=cobalt) and the main processing servers are hosted on their network. we really appreciate their kindness and support!
+cobalt is sponsored by [royalehosting.net](https://royalehosting.net/?partner=cobalt). a part of our infrastructure is hosted on their network. we really appreciate their kindness and support!
 
 ### ethics
 cobalt is a tool that makes downloading public content easier. it takes **zero liability**.
@@ -50,7 +53,7 @@ it can only download free & publicly accessible content.
 same content can be downloaded via dev tools of any modern web browser.
 
 ### contributing
-thank you for considering making a contribution to cobalt! please check the [contributing guidelines here](/CONTRIBUTING.md) before making a pull request.
+if you're considering contributing to cobalt, first of all, thank you! check the [contribution guidelines here](/CONTRIBUTING.md) before getting started, they'll help you do your best right away.
 
 ### licenses
 for relevant licensing information, see the [api](api/README.md) and [web](web/README.md) READMEs.

api/README.md

@@ -11,12 +11,9 @@ we recommend [deploying your own instance](/docs/run-an-instance.md) if you wish
 
 you can read [the api documentation here](/docs/api.md).
 
-> [!WARNING]
-> the v7 public api (/api/json) will be shut down on **november 11th, 2024**.
-> you can access documentation for it [here](https://github.com/imputnet/cobalt/blob/7/docs/api.md).
-
 ## supported services
-this list is not final and keeps expanding over time. if support for a service you want is missing, create an issue (or a pull request 👀).
+this list is not final and keeps expanding over time!
+if the desired service isn't supported yet, feel free to create an appropriate issue (or a pull request 👀).
 
 | service           | video + audio | only audio | only video | metadata | rich file names |
 | :--------         | :-----------: | :--------: | :--------: | :------: | :-------------: |
@@ -39,12 +36,13 @@ this list is not final and keeps expanding over time. if support for a service y
 | twitter/x         | ✅            | ✅         | ✅         | ➖         | ➖              |
 | vimeo             | ✅            | ✅         | ✅         | ✅         | ✅              |
 | vk videos & clips | ✅            | ❌         | ✅         | ✅         | ✅              |
+| xiaohongshu       | ✅            | ✅         | ✅         | ➖         | ➖              |
 | youtube           | ✅            | ✅         | ✅         | ✅         | ✅              |
 
 | emoji   | meaning                 |
 | :-----: | :---------------------- |
 | ✅      | supported               |
-| ➖      | impossible/unreasonable |
+| ➖      | unreasonable/impossible |
 | ❌      | not supported           |
 
 ### additional notes or features (per service)
@@ -71,36 +69,35 @@ as long as you:
 - provide a link to the license and indicate if changes to the code were made, and
 - release the code under the **same license**
 
-## acknowledgements
+## open source acknowledgements
 ### ffmpeg
-cobalt heavily relies on ffmpeg for converting and merging media files. it's an absolutely amazing piece of software offered for anyone for free, yet doesn't receive as much credit as it should.
+cobalt relies on ffmpeg for muxing and encoding media files. ffmpeg is absolutely spectacular and we're privileged to have an ability to use it for free, just like anyone else. we believe it should be way more recognized.
 
 you can [support ffmpeg here](https://ffmpeg.org/donations.html)!
 
-#### ffmpeg-static
-we use [ffmpeg-static](https://github.com/eugeneware/ffmpeg-static) to get binaries for ffmpeg depending on the platform.
-
-you can support the developer via various methods listed on their github page! (linked above)
-
 ### youtube.js
-cobalt relies on [youtube.js](https://github.com/LuanRT/YouTube.js) for interacting with the innertube api, it wouldn't have been possible without it.
+cobalt relies on **[youtube.js](https://github.com/LuanRT/YouTube.js)** for interacting with youtube's innertube api, it wouldn't have been possible without this package.
 
-you can support the developer via various methods listed on their github page! (linked above)
+you can support the developer via various methods listed on their github page!
+(linked above)
 
 ### many others
-cobalt also depends on:
-
-- [content-disposition-header](https://www.npmjs.com/package/content-disposition-header) to simplify the provision of `content-disposition` headers.
-- [cors](https://www.npmjs.com/package/cors) to manage cross-origin resource sharing within expressjs.
-- [dotenv](https://www.npmjs.com/package/dotenv) to load environment variables from the `.env` file.
-- [express](https://www.npmjs.com/package/express) as the backbone of cobalt servers.
-- [express-rate-limit](https://www.npmjs.com/package/express-rate-limit) to rate limit api endpoints.
-- [hls-parser](https://www.npmjs.com/package/hls-parser) to parse `m3u8` playlists for certain services.
-- [ipaddr.js](https://www.npmjs.com/package/ipaddr.js) to parse ip addresses (for rate limiting).
-- [nanoid](https://www.npmjs.com/package/nanoid) to generate unique (temporary) identifiers for each requested stream.
-- [psl](https://www.npmjs.com/package/psl) as the domain name parser.
-- [set-cookie-parser](https://www.npmjs.com/package/set-cookie-parser) to parse cookies that cobalt receives from certain services.
-- [undici](https://www.npmjs.com/package/undici) for making http requests.
-- [url-pattern](https://www.npmjs.com/package/url-pattern) to match provided links with supported patterns.
+cobalt-api also depends on:
+
+- **[content-disposition-header](https://www.npmjs.com/package/content-disposition-header)** to simplify the provision of `content-disposition` headers.
+- **[cors](https://www.npmjs.com/package/cors)** to manage cross-origin resource sharing within expressjs.
+- **[dotenv](https://www.npmjs.com/package/dotenv)** to load environment variables from the `.env` file.
+- **[express](https://www.npmjs.com/package/express)** as the backbone of cobalt servers.
+- **[express-rate-limit](https://www.npmjs.com/package/express-rate-limit)** to rate limit api endpoints.
+- **[ffmpeg-static](https://www.npmjs.com/package/ffmpeg-static)** to get binaries for ffmpeg depending on the platform.
+- **[hls-parser](https://www.npmjs.com/package/hls-parser)** to parse HLS playlists according to spec (very impressive stuff).
+- **[ipaddr.js](https://www.npmjs.com/package/ipaddr.js)** to parse ip addresses (used for rate limiting).
+- **[nanoid](https://www.npmjs.com/package/nanoid)** to generate unique identifiers for each requested tunnel.
+- **[set-cookie-parser](https://www.npmjs.com/package/set-cookie-parser)** to parse cookies that cobalt receives from certain services.
+- **[undici](https://www.npmjs.com/package/undici)** for making http requests.
+- **[url-pattern](https://www.npmjs.com/package/url-pattern)** to match provided links with supported patterns.
+- **[zod](https://www.npmjs.com/package/zod)** to lock down the api request schema.
+- **[@datastructures-js/priority-queue](https://www.npmjs.com/package/@datastructures-js/priority-queue)** for sorting stream caches for future clean up (without redis).
+- **[@imput/psl](https://www.npmjs.com/package/@imput/psl)** as the domain name parser, our fork of [psl](https://www.npmjs.com/package/psl).
 
 ...and many other packages that these packages rely on.

api/package.json

@@ -1,7 +1,7 @@
 {
     "name": "@imput/cobalt-api",
     "description": "save what you love",
-    "version": "10.5.4",
+    "version": "10.6",
     "author": "imput",
     "exports": "./src/cobalt.js",
     "type": "module",
@@ -39,7 +39,7 @@
         "set-cookie-parser": "2.6.0",
         "undici": "^5.19.1",
         "url-pattern": "1.0.3",
-        "youtubei.js": "^12.2.0",
+        "youtubei.js": "^13.0.0",
         "zod": "^3.23.8"
     },
     "optionalDependencies": {

api/src/core/api.js

@@ -313,7 +313,7 @@ export const runAPI = async (express, app, __dirname, isPrimary = true) => {
             ...Object.entries(req.headers)
         ]);
 
-        return stream(res, { type: 'internal', ...streamInfo });
+        return stream(res, { type: 'internal', data: streamInfo });
     };
 
     app.get('/itunnel', itunnelHandler);

api/src/misc/utils.js

@@ -1,8 +1,16 @@
-export function getRedirectingURL(url) {
-    return fetch(url, { redirect: 'manual' }).then((r) => {
-        if ([301, 302, 303].includes(r.status) && r.headers.has('location'))
+const redirectStatuses = new Set([301, 302, 303, 307, 308]);
+
+export async function getRedirectingURL(url, dispatcher) {
+    const location = await fetch(url, {
+        redirect: 'manual',
+        dispatcher,
+    }).then((r) => {
+        if (redirectStatuses.has(r.status) && r.headers.has('location')) {
             return r.headers.get('location');
+        }
     }).catch(() => null);
+
+    return location;
 }
 
 export function merge(a, b) {
@@ -29,3 +37,7 @@ export function splitFilenameExtension(filename) {
         return [ parts.join('.'), ext ]
     }
 }
+
+export function zip(a, b) {
+    return a.map((value, i) => [ value, b[i] ]);
+}

api/src/processing/match-action.js

@@ -15,7 +15,8 @@ export default function({ r, host, audioFormat, isAudioOnly, isAudioMuted, disab
             filename: r.filenameAttributes ?
                     createFilename(r.filenameAttributes, filenameStyle, isAudioOnly, isAudioMuted) : r.filename,
             fileMetadata: !disableMetadata ? r.fileMetadata : false,
-            requestIP
+            requestIP,
+            originalRequest: r.originalRequest
         },
         params = {};
 
@@ -47,7 +48,7 @@ export default function({ r, host, audioFormat, isAudioOnly, isAudioMuted, disab
             });
 
         case "photo":
-            responseType = "redirect";
+            params = { type: "proxy" };
             break;
 
         case "gif":
@@ -83,6 +84,7 @@ export default function({ r, host, audioFormat, isAudioOnly, isAudioMuted, disab
                 case "twitter":
                 case "snapchat":
                 case "bsky":
+                case "xiaohongshu":
                     params = { picker: r.picker };
                     break;
 
@@ -143,6 +145,7 @@ export default function({ r, host, audioFormat, isAudioOnly, isAudioMuted, disab
                 case "ok":
                 case "vk":
                 case "tiktok":
+                case "xiaohongshu":
                     params = { type: "proxy" };
                     break;
 

api/src/processing/match.js

@@ -28,6 +28,7 @@ import snapchat from "./services/snapchat.js";
 import loom from "./services/loom.js";
 import facebook from "./services/facebook.js";
 import bluesky from "./services/bluesky.js";
+import xiaohongshu from "./services/xiaohongshu.js";
 
 let freebind;
 
@@ -239,6 +240,15 @@ export default async function({ host, patternMatch, params }) {
                 });
                 break;
 
+            case "xiaohongshu":
+                r = await xiaohongshu({
+                    ...patternMatch,
+                    h265: params.tiktokH265,
+                    isAudioOnly,
+                    dispatcher,
+                });
+                break;
+
             default:
                 return createResponse("error", {
                     code: "error.api.service.unsupported"

api/src/processing/service-config.js

@@ -166,6 +166,14 @@ export const services = {
         subdomains: ["m"],
         altDomains: ["vkvideo.ru", "vk.ru"],
     },
+    xiaohongshu: {
+        patterns: [
+            "explore/:id?xsec_token=:token",
+            "discovery/item/:id?xsec_token=:token",
+            "a/:shareId"
+        ],
+        altDomains: ["xhslink.com"],
+    },
     youtube: {
         patterns: [
             "watch?v=:id",

api/src/processing/service-patterns.js

@@ -71,4 +71,8 @@ export const testers = {
 
     "bsky": pattern =>
         pattern.user?.length <= 128 && pattern.post?.length <= 128,
+
+    "xiaohongshu": pattern =>
+        pattern.id?.length <= 24 && pattern.token?.length <= 64
+        || pattern.shareId?.length <= 12,
 }

api/src/processing/services/bluesky.js

@@ -71,6 +71,24 @@ const extractImages = ({ getPost, filename, alwaysProxy }) => {
     return { picker };
 }
 
+const extractGif = ({ url, filename }) => {
+    const gifUrl = new URL(url);
+
+    if (!gifUrl || gifUrl.hostname !== "media.tenor.com") {
+        return { error: "fetch.empty" };
+    }
+
+    // remove downscaling params from gif url
+    // such as "?hh=498&ww=498"
+    gifUrl.search = "";
+
+    return {
+        urls: gifUrl,
+        isPhoto: true,
+        filename: `${filename}.gif`,
+    }
+}
+
 export default async function ({ user, post, alwaysProxy, dispatcher }) {
     const apiEndpoint = new URL("https://public.api.bsky.app/xrpc/app.bsky.feed.getPostThread?depth=0&parentHeight=0");
     apiEndpoint.searchParams.set(
@@ -102,22 +120,37 @@ export default async function ({ user, post, alwaysProxy, dispatcher }) {
     const embedType = getPost?.thread?.post?.embed?.$type;
     const filename = `bluesky_${user}_${post}`;
 
-    if (embedType === "app.bsky.embed.video#view") {
-        return extractVideo({
-            media: getPost.thread?.post?.embed,
-            filename,
-        })
-    }
-
-    if (embedType === "app.bsky.embed.recordWithMedia#view") {
-        return extractVideo({
-            media: getPost.thread?.post?.embed?.media,
-            filename,
-        })
-    }
-
-    if (embedType === "app.bsky.embed.images#view") {
-        return extractImages({ getPost, filename, alwaysProxy });
+    switch (embedType) {
+        case "app.bsky.embed.video#view":
+            return extractVideo({
+                media: getPost.thread?.post?.embed,
+                filename,
+            });
+
+        case "app.bsky.embed.images#view":
+            return extractImages({
+                getPost,
+                filename,
+                alwaysProxy
+            });
+
+        case "app.bsky.embed.external#view":
+            return extractGif({
+                url: getPost?.thread?.post?.embed?.external?.uri,
+                filename,
+            });
+
+        case "app.bsky.embed.recordWithMedia#view":
+            if (getPost?.thread?.post?.embed?.media?.$type === "app.bsky.embed.external#view") {
+                return extractGif({
+                    url: getPost?.thread?.post?.embed?.media?.external?.uri,
+                    filename,
+                });
+            }
+            return extractVideo({
+                media: getPost.thread?.post?.embed?.media,
+                filename,
+            });
     }
 
     return { error: "fetch.empty" };