chore: sweep em dashes from our own source and README (#13)

src/__tests__/voice.test.ts

@@ -0,0 +1,38 @@
+import { readFileSync, readdirSync } from "node:fs";
+import path from "node:path";
+import { describe, expect, it } from "vitest";
+
+const EM_DASH = "\u2014";
+const SRC = path.resolve(__dirname, "..");
+const README = path.resolve(__dirname, "../../README.md");
+
+function sourceFiles(directory: string): string[] {
+  return readdirSync(directory, { withFileTypes: true }).flatMap((entry) => {
+    if (entry.isDirectory()) {
+      // Tests are skipped: unicode test data legitimately contains an em dash,
+      // and the rule is about what we ship rather than about our fixtures.
+      return entry.name === "__tests__" ? [] : sourceFiles(path.join(directory, entry.name));
+    }
+
+    return /\.tsx?$/.test(entry.name) ? [path.join(directory, entry.name)] : [];
+  });
+}
+
+describe("voice", () => {
+  it("ships no em dashes in the source", () => {
+    // The default prompt tells the model never to write one, so the extension
+    // should not ship one either. Guarding the whole tree rather than picking
+    // out UI strings by hand keeps the rule simple enough to actually hold, and
+    // comment rewrites are cheap next to a documented standard that quietly
+    // drifts.
+    const offenders = sourceFiles(SRC)
+      .filter((file) => readFileSync(file, "utf8").includes(EM_DASH))
+      .map((file) => path.relative(SRC, file));
+
+    expect(offenders).toEqual([]);
+  });
+
+  it("ships no em dashes in the README", () => {
+    expect(readFileSync(README, "utf8")).not.toContain(EM_DASH);
+  });
+});