feat: improve metadata processing, don't download the entire track

deno.lock

···
         "npm:@jsr/bradenmacdonald__s3-lite-client@0.9",
         "npm:@jsr/std__media-types@^1.1.0",
         "npm:@picocss/pico@^2.1.1",
+        "npm:@tokenizer/http@~0.9.2",
+        "npm:@tokenizer/range@0.13",
         "npm:@types/throttle-debounce@^5.0.2",
         "npm:astro-purgecss@^5.2.2",
         "npm:astro-scope@^3.0.1",

package-lock.json

···
         "@bradenmacdonald/s3-lite-client": "npm:@jsr/bradenmacdonald__s3-lite-client@^0.9.0",
         "@picocss/pico": "^2.1.1",
         "@std/media-types": "npm:@jsr/std__media-types@^1.1.0",
+        "@tokenizer/http": "^0.9.2",
+        "@tokenizer/range": "^0.13.0",
         "@web-applets/sdk": "https://gitpkg.vercel.app/unternet-co/web-applets/sdk?tokono.ma/experiment&scripts.postinstall=npm%20i%20%40types%2Fnode%20%26%26%20npx%20tsc",
         "iconoir": "^7.11.0",
         "idb-keyval": "^6.2.1",
         "music-metadata": "^11.2.3",
         "native-file-system-adapter": "^3.0.1",
-        "node-s3-url-encode": "^0.0.4",
         "query-string": "^9.1.2",
         "spellcaster": "^6.0.0",
         "throttle-debounce": "^5.0.2",
···
         "tslib": "^2.8.0"
       }
     },
+    "node_modules/@tokenizer/http": {
+      "version": "0.9.2",
+      "resolved": "https://registry.npmjs.org/@tokenizer/http/-/http-0.9.2.tgz",
+      "integrity": "sha512-rzJwHcqDjO3FdBPr+FK2R6dYE6Qbg6QZP7S47rhCEtG+/YqEFLqZ+gFCLcL8y5D39aYQB9vDssiwbsJlRLePPg==",
+      "license": "MIT",
+      "dependencies": {
+        "@tokenizer/range": "^0.12.0",
+        "debug": "^4.3.7",
+        "strtok3": "^10.0.0"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/Borewit"
+      }
+    },
+    "node_modules/@tokenizer/http/node_modules/@tokenizer/range": {
+      "version": "0.12.0",
+      "resolved": "https://registry.npmjs.org/@tokenizer/range/-/range-0.12.0.tgz",
+      "integrity": "sha512-xvJ1OflWjopkC5EgLge+9HrwsWStgVewQkmusoF2BxgCuGdm1KuhZAMVMNzC7h1WNei9JA6xKQlkbPNJtjZ6aw==",
+      "license": "MIT",
+      "dependencies": {
+        "debug": "^4.3.7",
+        "strtok3": "^9.1.1"
+      },
+      "engines": {
+        "node": ">=16"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/Borewit"
+      }
+    },
+    "node_modules/@tokenizer/http/node_modules/@tokenizer/range/node_modules/strtok3": {
+      "version": "9.1.1",
+      "resolved": "https://registry.npmjs.org/strtok3/-/strtok3-9.1.1.tgz",
+      "integrity": "sha512-FhwotcEqjr241ZbjFzjlIYg6c5/L/s4yBGWSMvJ9UoExiSqL+FnFA/CaeZx17WGaZMS/4SOZp8wH18jSS4R4lw==",
+      "license": "MIT",
+      "dependencies": {
+        "@tokenizer/token": "^0.3.0",
+        "peek-readable": "^5.3.1"
+      },
+      "engines": {
+        "node": ">=16"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/Borewit"
+      }
+    },
+    "node_modules/@tokenizer/http/node_modules/peek-readable": {
+      "version": "5.4.2",
+      "resolved": "https://registry.npmjs.org/peek-readable/-/peek-readable-5.4.2.tgz",
+      "integrity": "sha512-peBp3qZyuS6cNIJ2akRNG1uo1WJ1d0wTxg/fxMdZ0BqCVhx242bSFHM9eNqflfJVS9SsgkzgT/1UgnsurBOTMg==",
+      "license": "MIT",
+      "engines": {
+        "node": ">=14.16"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/Borewit"
+      }
+    },
     "node_modules/@tokenizer/inflate": {
       "version": "0.2.7",
       "resolved": "https://registry.npmjs.org/@tokenizer/inflate/-/inflate-0.2.7.tgz",
···
       },
       "engines": {
         "node": ">=18"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/Borewit"
+      }
+    },
+    "node_modules/@tokenizer/range": {
+      "version": "0.13.0",
+      "resolved": "https://registry.npmjs.org/@tokenizer/range/-/range-0.13.0.tgz",
+      "integrity": "sha512-ibLGQRU8an1g/y952+OxeZDGIj+W1HW8AQPtk26VIFWzy3tvQImmGBwYbpHJXMMAz1nhCPAAepCRptGKB8YrKg==",
+      "license": "MIT",
+      "dependencies": {
+        "debug": "^4.4.0",
+        "strtok3": "^10.2.0"
+      },
+      "engines": {
+        "node": ">=16"
       },
       "funding": {
         "type": "github",
···
       "resolved": "https://registry.npmjs.org/node-mock-http/-/node-mock-http-1.0.0.tgz",
       "integrity": "sha512-0uGYQ1WQL1M5kKvGRXWQ3uZCHtLTO8hln3oBjIusM75WoesZ909uQJs/Hb946i2SS+Gsrhkaa6iAO17jRIv6DQ==",
       "dev": true
-    },
-    "node_modules/node-s3-url-encode": {
-      "version": "0.0.4",
-      "resolved": "https://registry.npmjs.org/node-s3-url-encode/-/node-s3-url-encode-0.0.4.tgz",
-      "integrity": "sha512-l0IizfnxE1hb9dadzYBpA27syfL9LFkPzCKH6YWrssv2sPLjVuCent67A8GPe4isdj4bEsbgdPWLTcV4gxEg9w==",
-      "license": "MIT"
     },
     "node_modules/normalize-path": {
       "version": "3.0.0",

package.json

···
     "@bradenmacdonald/s3-lite-client": "npm:@jsr/bradenmacdonald__s3-lite-client@^0.9.0",
     "@picocss/pico": "^2.1.1",
     "@std/media-types": "npm:@jsr/std__media-types@^1.1.0",
+    "@tokenizer/http": "^0.9.2",
+    "@tokenizer/range": "^0.13.0",
     "@web-applets/sdk": "https://gitpkg.vercel.app/unternet-co/web-applets/sdk?tokono.ma/experiment&scripts.postinstall=npm%20i%20%40types%2Fnode%20%26%26%20npx%20tsc",
     "iconoir": "^7.11.0",
     "idb-keyval": "^6.2.1",

src/pages/configurator/input/_applet.astro

···
     </p>
   </div>
   <p>
-    <small
-      ><em><strong>More options coming soon!</strong><br />S3-compatible APIs, Dropbox, etc.</em
-      ></small
-    >
+    <small><em><strong>More options coming soon!</strong></em></small>
   </p>
 </main>
 

src/pages/core/types.d.ts

···
 export interface TrackTags {
   album?: string;
   artist?: string;
+  disc: { no: number; of?: number };
+  genre?: string;
   title: string;
+  track: { no: number; of?: number };
+  year?: number;
 }

src/pages/orchestrator/input-cache/_applet.astro

···
   // SETUP
   ////////////////////////////////////////////
   const context = applets.register<{ ready: boolean }>();
+  const topContext = self.top || self.parent;
 
   // Initial data
   context.data = {
···
 
   // Applet connections
   const configurator = {
-    input: await applet("../../configurator/input", { context: self.top || self.parent }),
+    input: await applet("../../configurator/input", { context: topContext }),
   };
 
   const orchestrator = {
     output: await applet<Output>("../../orchestrator/output-management", {
-      context: self.parent,
+      context: topContext,
     }),
   };
 
   const processor = {
-    metadataFetcher: await applet("../../processor/metadata-fetcher", { context: self.parent }),
+    metadataFetcher: await applet("../../processor/metadata-fetcher", {
+      context: topContext,
+    }),
   };
 
   // 🚀
···
       async (promise: Promise<Track[]>, track: Track) => {
         const acc = await promise;
 
-        if (track.tags) return [...acc, track];
+        if (track.tags && track.stats) return [...acc, track];
 
         const getURL = await configurator.input.sendAction<string | undefined>(
           "resolve",
···
           },
         );
 
-        if (!getURL) return acc;
+        const headURL = await configurator.input.sendAction<string | undefined>(
+          "resolve",
+          { method: "HEAD", uri: track.uri },
+          {
+            timeoutDuration: 60000,
+          },
+        );
 
-        // TODO: Do we need to pass the HEAD URL too?
-        const meta = await processor.metadataFetcher.sendAction("extract", getURL, {
-          timeoutDuration: 60000,
-        });
+        if (!getURL) return acc;
 
-        const stats: TrackStats = {
-          duration: meta.format.duration,
-        };
+        const { stats, tags } = await processor.metadataFetcher.sendAction(
+          "extract",
+          { urls: { get: getURL, head: headURL || getURL } },
+          {
+            timeoutDuration: 60000,
+          },
+        );
 
-        const tags: TrackTags = {
-          album: meta.common.album,
-          artist: meta.common.artist,
-          title: meta.common.title,
-        };
+        console.log(stats, tags);
 
         return [...acc, { ...track, stats, tags }];
       },

src/pages/orchestrator/single-queue/_applet.astro

···
   };
 
   const orchestrator = {
-    output: await applet<Output>("../../orchestrator/output-management", { context: self.parent }),
+    output: await applet<Output>("../../orchestrator/output-management", {
+      context: self.top || self.parent,
+    }),
   };
 
   ////////////////////////////////////////////

src/pages/processor/metadata-fetcher/_applet.astro

···
 <script>
   import { applets } from "@web-applets/sdk";
-  import { parseWebStream } from "music-metadata";
+  import { parseFromTokenizer } from "music-metadata";
   import { contentType } from "@std/media-types";
   import * as URI from "uri-js";
+  import * as HTTP_TOKENIZER from "@tokenizer/http";
+  import * as RANGE_TOKENIZER from "@tokenizer/range";
+
+  import { TrackStats, TrackTags } from "@applets/core/types";
 
   ////////////////////////////////////////////
   // SETUP
   ////////////////////////////////////////////
   const context = applets.register();
 
+  type Extraction = { stats: TrackStats; tags: TrackTags };
+  type Urls = { get: string; head: string };
+
   ////////////////////////////////////////////
   // ACTIONS
   ////////////////////////////////////////////
   context.setActionHandler("extract", extract);
 
-  async function extract(url: string) {
-    const uri = URI.parse(url);
+  async function extract({ urls }: { urls: Urls }) {
+    // Construct records
+    // TODO: Use other metadata lib as fallback: https://github.com/buzz/mediainfo.js
+    const { stats, tags } = await musicMetadataTags(urls, false);
+
+    // Fin
+    return { stats, tags };
+  }
+
+  // 🛠️
+  async function musicMetadataTags(urls: Urls, covers: boolean = false): Promise<Extraction> {
+    const httpClient = new HTTP_TOKENIZER.HttpClient(urls.head, { resolveUrl: false });
+    httpClient.resolvedUrl = urls.get;
+
+    const tokenizer = await RANGE_TOKENIZER.tokenizer(httpClient);
+    const meta = await parseFromTokenizer(tokenizer, { skipCovers: covers });
+
+    const uri = URI.parse(urls.get);
     const pathParts = uri.path?.split("/");
-    const mimeType = pathParts?.[pathParts.length - 1]?.includes(".")
-      ? contentType(pathParts[pathParts.length - 1].split(".").reverse()[0])
-      : undefined;
-    const resp = await fetch(url);
-    const stream = resp.body;
-    const metadata = await parseWebStream(stream, { mimeType });
+    const filename = pathParts?.[pathParts.length - 1];
+
+    const stats: TrackStats = {
+      duration: meta.format.duration,
+    };
+
+    const tags: TrackTags = {
+      album: meta.common.album,
+      artist: meta.common.artist,
+      disc: { no: meta.common.disk.no || 1, of: meta.common.disk.of ?? undefined },
+      genre: Array.isArray(meta.common.genre) ? meta.common.genre[0] : meta.common.genre,
+      title: meta.common.title || filename || urls.head,
+      track: { no: meta.common.track.no || 1, of: meta.common.track.of ?? undefined },
+      year: meta.common.year,
+    };
 
-    return metadata;
+    return {
+      stats,
+      tags,
+    };
   }
 </script>

src/pages/processor/metadata-fetcher/_manifest.json

···
       "title": "Extract",
       "description": "Get the metadata for a given URL.",
       "params_schema": {
-        "type": "string",
-        "description": "URL"
+        "type": "object",
+        "properties": {
+          "urls": {
+            "type": "object",
+            "properties": {
+              "get": { "type": "string" },
+              "head": { "type": "string" }
+            },
+            "required": ["get", "head"]
+          }
+        },
+        "required": ["urls"]
       }
     }
   }