Merge branch 'master' into enable_multi_star_for_private_vercel

readme.md

@@ -94,6 +94,7 @@ Visit <https://indiafightscorona.giveindia.org> and make a small donation to hel
     -   [Wakatime Card Exclusive Option](#wakatime-card-exclusive-options)
 -   [Deploy Yourself](#deploy-on-your-own-vercel-instance)
     -   [Disable rate limit protections](#disable-rate-limit-protections)
+    - [Keep your fork up to date](#keep-your-fork-up-to-date)
 
 # Important Notice
 
@@ -454,6 +455,9 @@ See [the Vercel documentation](https://vercel.com/docs/concepts/projects/environ
 
 > **Warning**
 > Please be aware that disabling these rate limit protections increases the execution time of the serverless function as now more data needs to be fetched. As a result, GitHub users with many repositories might hit Vercel's [Serverless Function Execution Timeout](https://vercel.com/docs/concepts/limits/overview).
+### Keep your fork up to date
+
+You can keep your fork, and thus your private Vercel instance up to date with the upstream using GitHubs' [Sync Fork button](https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork). You can also use the [pull](https://github.com/wei/pull) package created by [@wei](https://github.com/wei) to automate this process.
 
 ## :sparkling_heart: Support the project
 

tests/api.test.js

@@ -161,6 +161,16 @@ describe("Test /api/", () => {
     ]);
   });
 
+  it("should not store cache when error", async () => {
+    const { req, res } = faker({}, error);
+    await api(req, res);
+
+    expect(res.setHeader.mock.calls).toEqual([
+      ["Content-Type", "image/svg+xml"],
+      ["Cache-Control", `no-store`],
+    ]);
+  });
+
   it("should set proper cache with clamped values", async () => {
     {
       let { req, res } = faker({ cache_seconds: 200000 }, data_stats);