Update worker to avoid downloading entire wheel (#168)

The previous solution downloaded the entire wheel in-memory. This
solution reads lazily.

Closes https://github.com/astral-sh/puffin/issues/153.

workers/pypi-metadata/package-lock.json

@@ -8,6 +8,7 @@
       "name": "api",
       "version": "0.0.0",
       "dependencies": {
+        "@zip.js/zip.js": "^2.7.30",
         "install": "^0.13.0",
         "jszip": "^3.10.1",
         "npm": "^10.2.1",
@@ -339,6 +340,16 @@
       "integrity": "sha512-TgfOX+mGY/NyNxJLIbDWrO9DjGoVSW9+aB8H2yy1fy32jsvxijhmyJI9fDFgvz3YP4lvJaq9DzdR/M1bOgVc9g==",
       "dev": true
     },
+    "node_modules/@zip.js/zip.js": {
+      "version": "2.7.30",
+      "resolved": "https://registry.npmjs.org/@zip.js/zip.js/-/zip.js-2.7.30.tgz",
+      "integrity": "sha512-nhMvQCj+TF1ATBqYzFds7v+yxPBhdDYHh8J341KtC1D2UrVBUIYcYK4Jy1/GiTsxOXEiKOXSUxvPG/XR+7jMqw==",
+      "engines": {
+        "bun": ">=0.7.0",
+        "deno": ">=1.0.0",
+        "node": ">=16.5.0"
+      }
+    },
     "node_modules/anymatch": {
       "version": "3.1.3",
       "resolved": "https://registry.npmjs.org/anymatch/-/anymatch-3.1.3.tgz",

workers/pypi-metadata/package.json

@@ -13,8 +13,8 @@
     "fmt": "prettier --cache -w ."
   },
   "dependencies": {
+    "@zip.js/zip.js": "^2.7.30",
     "install": "^0.13.0",
-    "jszip": "^3.10.1",
     "npm": "^10.2.1",
     "prettier": "^3.0.3"
   }

workers/pypi-metadata/src/index.ts

@@ -1,4 +1,4 @@
-import JSZip from "jszip";
+import * as zip from "@zip.js/zip.js";
 
 export interface Env {}
 
@@ -27,7 +27,6 @@ export default {
       //   /packages/d2/3d/fa76db83bf75c4f8d338c2fd15c8d33fdd7ad23a9b5e57eb6c5de26b430e/click-7.1.2-py2.py3-none-any.whl
       const url = new URL(request.url);
       const path = url.pathname;
-      const query = url.search;
 
       if (path.startsWith("/packages/")) {
         // Given the path, extract `click-7.1.2`.
@@ -35,22 +34,11 @@ export default {
         const name = parts[parts.length - 1].split("-")[0];
         const version = parts[parts.length - 1].split("-")[1];
 
-        // Extract the zip contents.
+        // Read the metadata.
-        // Now, fetch "https://files.pythonhosted.org/packages/d2/3d/fa76db83bf75c4f8d338c2fd15c8d33fdd7ad23a9b5e57eb6c5de26b430e/click-7.1.2-py2.py3-none-any.whl"
+        const reader = new zip.ZipReader(
-        response = await fetch(`https://files.pythonhosted.org${path}`, {
+          new zip.HttpRangeReader(`https://files.pythonhosted.org${path}`),
-          cf: {
+        );
-            // Always cache this fetch regardless of content type
+        const file = await readMetadata(reader, name, version);
-            // for a max of 5 seconds before revalidating the resource
-            cacheTtl: 5,
-            cacheEverything: true,
-          },
-        });
-        const buffer = await response.arrayBuffer();
-        const archive = await JSZip.loadAsync(buffer);
-        const file = await archive
-          .folder(`${name}-${version}.dist-info`)
-          ?.file("METADATA")
-          ?.async("string");
         if (!file) {
           return new Response("Not found", { status: 404 });
         }
@@ -64,19 +52,6 @@ export default {
           },
         });
 
-        ctx.waitUntil(cache.put(cacheKey, response.clone()));
-      } else if (path.startsWith("/simple/")) {
-        // Pass the request on to `https://pypi.org/`. Include query string.
-        // Propagate headers.
-        response = await fetch(`https://pypi.org${path}${query}`, {
-          cf: {
-            // Always cache this fetch regardless of content type
-            // for a max of 5 seconds before revalidating the resource
-            cacheTtl: 5,
-            cacheEverything: true,
-          },
-        });
-
         ctx.waitUntil(cache.put(cacheKey, response.clone()));
       } else {
         return new Response("Not found", { status: 404 });
@@ -88,3 +63,20 @@ export default {
     return response;
   },
 };
+
+/**
+ * Read the `METADATA` file from the given wheel.
+ */
+async function readMetadata(
+  reader: zip.ZipReader<any>,
+  name: string,
+  version: string,
+) {
+  const entries = await reader.getEntriesGenerator();
+  for await (const entry of entries) {
+    if (entry.filename == `${name}-${version}.dist-info/METADATA`) {
+      return await entry.getData!(new zip.TextWriter());
+    }
+  }
+  return null;
+}