fix: preserve markdown fences when chunking

src/auto-reply/chunk.test.ts

@@ -1,6 +1,29 @@
 import { describe, expect, it } from "vitest";
 
-import { chunkText, resolveTextChunkLimit } from "./chunk.js";
+import {
+  chunkMarkdownText,
+  chunkText,
+  resolveTextChunkLimit,
+} from "./chunk.js";
+
+function expectFencesBalanced(chunks: string[]) {
+  for (const chunk of chunks) {
+    let open: { markerChar: string; markerLen: number } | null = null;
+    for (const line of chunk.split("\n")) {
+      const match = line.match(/^( {0,3})(`{3,}|~{3,})(.*)$/);
+      if (!match) continue;
+      const marker = match[2];
+      if (!open) {
+        open = { markerChar: marker[0], markerLen: marker.length };
+        continue;
+      }
+      if (open.markerChar === marker[0] && marker.length >= open.markerLen) {
+        open = null;
+      }
+    }
+    expect(open).toBe(null);
+  }
+}
 
 describe("chunkText", () => {
   it("keeps multi-line text in one chunk when under limit", () => {
@@ -72,3 +95,79 @@ describe("resolveTextChunkLimit", () => {
     expect(resolveTextChunkLimit(cfg, "telegram")).toBe(4000);
   });
 });
+
+describe("chunkMarkdownText", () => {
+  it("keeps fenced blocks intact when a safe break exists", () => {
+    const prefix = "p".repeat(60);
+    const fence = "```bash\nline1\nline2\n```";
+    const suffix = "s".repeat(60);
+    const text = `${prefix}\n\n${fence}\n\n${suffix}`;
+
+    const chunks = chunkMarkdownText(text, 40);
+    expect(chunks.some((chunk) => chunk.trimEnd() === fence)).toBe(true);
+    expectFencesBalanced(chunks);
+  });
+
+  it("reopens fenced blocks when forced to split inside them", () => {
+    const text = `\`\`\`txt\n${"a".repeat(500)}\n\`\`\``;
+    const limit = 120;
+    const chunks = chunkMarkdownText(text, limit);
+    expect(chunks.length).toBeGreaterThan(1);
+    for (const chunk of chunks) {
+      expect(chunk.length).toBeLessThanOrEqual(limit);
+      expect(chunk.startsWith("```txt\n")).toBe(true);
+      expect(chunk.trimEnd().endsWith("```")).toBe(true);
+    }
+    expectFencesBalanced(chunks);
+  });
+
+  it("supports tilde fences", () => {
+    const text = `~~~sh\n${"x".repeat(600)}\n~~~`;
+    const limit = 140;
+    const chunks = chunkMarkdownText(text, limit);
+    expect(chunks.length).toBeGreaterThan(1);
+    for (const chunk of chunks) {
+      expect(chunk.length).toBeLessThanOrEqual(limit);
+      expect(chunk.startsWith("~~~sh\n")).toBe(true);
+      expect(chunk.trimEnd().endsWith("~~~")).toBe(true);
+    }
+    expectFencesBalanced(chunks);
+  });
+
+  it("supports longer fence markers for close", () => {
+    const text = `\`\`\`\`md\n${"y".repeat(600)}\n\`\`\`\``;
+    const limit = 140;
+    const chunks = chunkMarkdownText(text, limit);
+    expect(chunks.length).toBeGreaterThan(1);
+    for (const chunk of chunks) {
+      expect(chunk.length).toBeLessThanOrEqual(limit);
+      expect(chunk.startsWith("````md\n")).toBe(true);
+      expect(chunk.trimEnd().endsWith("````")).toBe(true);
+    }
+    expectFencesBalanced(chunks);
+  });
+
+  it("preserves indentation for indented fences", () => {
+    const text = `  \`\`\`js\n  ${"z".repeat(600)}\n  \`\`\``;
+    const limit = 160;
+    const chunks = chunkMarkdownText(text, limit);
+    expect(chunks.length).toBeGreaterThan(1);
+    for (const chunk of chunks) {
+      expect(chunk.length).toBeLessThanOrEqual(limit);
+      expect(chunk.startsWith("  ```js\n")).toBe(true);
+      expect(chunk.trimEnd().endsWith("  ```")).toBe(true);
+    }
+    expectFencesBalanced(chunks);
+  });
+
+  it("never produces an empty fenced chunk when splitting", () => {
+    const text = `\`\`\`txt\n${"a".repeat(300)}\n\`\`\``;
+    const chunks = chunkMarkdownText(text, 60);
+    for (const chunk of chunks) {
+      const nonFenceLines = chunk
+        .split("\n")
+        .filter((line) => !/^( {0,3})(`{3,}|~{3,})(.*)$/.test(line));
+      expect(nonFenceLines.join("\n").trim()).not.toBe("");
+    }
+  });
+});