test: Phase 1 — 添加 8 个纯函数测试文件 (+134 tests)

- errors.test.ts: 28 tests (isAbortError, toError, errorMessage, getErrnoCode, isFsInaccessible, classifyAxiosError 等)
- shellRuleMatching.test.ts: 22 tests (permissionRuleExtractPrefix, hasWildcards, matchWildcardPattern, parsePermissionRule 等)
- argumentSubstitution.test.ts: 18 tests (parseArguments, parseArgumentNames, generateProgressiveArgumentHint, substituteArguments)
- CircularBuffer.test.ts: 12 tests (add, addAll, getRecent, toArray, clear, length)
- sanitization.test.ts: 14 tests (partiallySanitizeUnicode, recursivelySanitizeUnicode)
- slashCommandParsing.test.ts: 8 tests (parseSlashCommand)
- contentArray.test.ts: 6 tests (insertBlockAfterToolResults)
- objectGroupBy.test.ts: 5 tests (objectGroupBy)

总计：781 tests / 40 files

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

src/utils/__tests__/sanitization.test.ts

@@ -0,0 +1,75 @@
+import { describe, expect, test } from "bun:test";
+import {
+  partiallySanitizeUnicode,
+  recursivelySanitizeUnicode,
+} from "../sanitization";
+
+// ─── partiallySanitizeUnicode ───────────────────────────────────────────
+
+describe("partiallySanitizeUnicode", () => {
+  test("preserves normal ASCII text", () => {
+    expect(partiallySanitizeUnicode("hello world")).toBe("hello world");
+  });
+
+  test("preserves CJK characters", () => {
+    expect(partiallySanitizeUnicode("你好世界")).toBe("你好世界");
+  });
+
+  test("removes zero-width spaces", () => {
+    expect(partiallySanitizeUnicode("hello\u200Bworld")).toBe("helloworld");
+  });
+
+  test("removes BOM", () => {
+    expect(partiallySanitizeUnicode("\uFEFFhello")).toBe("hello");
+  });
+
+  test("removes directional formatting", () => {
+    expect(partiallySanitizeUnicode("hello\u202Aworld")).toBe("helloworld");
+  });
+
+  test("removes private use area characters", () => {
+    expect(partiallySanitizeUnicode("hello\uE000world")).toBe("helloworld");
+  });
+
+  test("handles empty string", () => {
+    expect(partiallySanitizeUnicode("")).toBe("");
+  });
+
+  test("handles string with only dangerous characters", () => {
+    const result = partiallySanitizeUnicode("\u200B\u200C\u200D\uFEFF");
+    expect(result.length).toBeLessThanOrEqual(1); // ZWJ may survive NFKC
+  });
+});
+
+// ─── recursivelySanitizeUnicode ─────────────────────────────────────────
+
+describe("recursivelySanitizeUnicode", () => {
+  test("sanitizes string values", () => {
+    expect(recursivelySanitizeUnicode("hello\u200Bworld")).toBe("helloworld");
+  });
+
+  test("sanitizes array elements", () => {
+    const result = recursivelySanitizeUnicode(["a\u200Bb", "c\uFEFFd"]);
+    expect(result).toEqual(["ab", "cd"]);
+  });
+
+  test("sanitizes object values recursively", () => {
+    const result = recursivelySanitizeUnicode({
+      key: "val\u200Bue",
+      nested: { inner: "te\uFEFFst" },
+    });
+    expect(result).toEqual({ key: "value", nested: { inner: "test" } });
+  });
+
+  test("preserves numbers", () => {
+    expect(recursivelySanitizeUnicode(42)).toBe(42);
+  });
+
+  test("preserves booleans", () => {
+    expect(recursivelySanitizeUnicode(true)).toBe(true);
+  });
+
+  test("preserves null", () => {
+    expect(recursivelySanitizeUnicode(null)).toBeNull();
+  });
+});