fix(memory): snippets split emoji when truncated (#102478)

qingminglong · steipete · web-flow · commit c3d567db95d0 · 2026-07-09T08:27:42.000+01:00
* fix(memory): snippets split emoji when truncated

* refactor(memory): reuse shared UTF-16 truncation

---------

Co-authored-by: Peter Steinberger &lt;steipete@gmail.com&gt;
diff --git a/packages/memory-host-sdk/src/host/read-file-shared.test.ts b/packages/memory-host-sdk/src/host/read-file-shared.test.ts
@@ -34,4 +34,40 @@ describe("memory read result slicing", () => {
       lines: 2,
     });
   });
+
+  it("does not split surrogate pairs when truncating a single line", () => {
+    expect(
+      buildMemoryReadResultFromSlice({
+        selectedLines: ["abc🤖tail"],
+        relPath: "memory/test.md",
+        startLine: 1,
+        maxChars: 4,
+        suggestReadFallback: true,
+      }),
+    ).toEqual({
+      text: "abc\n\n[More content available. Requested excerpt exceeded the default maxChars budget. If you need the full raw line, use read on the source file.]",
+      path: "memory/test.md",
+      from: 1,
+      lines: 1,
+      truncated: true,
+    });
+  });
+
+  it("keeps the continuation notice when a leading surrogate pair is dropped", () => {
+    expect(
+      buildMemoryReadResultFromSlice({
+        selectedLines: ["🤖tail"],
+        relPath: "memory/test.md",
+        startLine: 1,
+        maxChars: 1,
+        suggestReadFallback: true,
+      }),
+    ).toEqual({
+      text: "\n\n[More content available. Requested excerpt exceeded the default maxChars budget. If you need the full raw line, use read on the source file.]",
+      path: "memory/test.md",
+      from: 1,
+      lines: 1,
+      truncated: true,
+    });
+  });
 });
diff --git a/packages/memory-host-sdk/src/host/read-file-shared.ts b/packages/memory-host-sdk/src/host/read-file-shared.ts
@@ -1,4 +1,5 @@
 // Memory Host SDK module implements read file shared behavior.
+import { truncateUtf16Safe } from "@openclaw/normalization-core/utf16-slice";
 import type { MemoryReadResult } from "./types.js";
 
 // Shared memory-file read result shaping and truncation notices.
@@ -48,7 +49,7 @@ function fitLinesToCharBudget(params: { lines: string[]; maxChars: number }): {
   }
 
   return {
-    text: text.slice(0, maxChars),
+    text: truncateUtf16Safe(text, maxChars),
     includedLines: 1,
     hardTruncatedSingleLine: true,
   };
@@ -85,7 +86,7 @@ export function buildMemoryReadResultFromSlice(params: {
       : undefined;
   const truncated = charCapTruncated || moreSourceLinesRemain;
   const text =
-    truncated && fitted.text
+    truncated && (fitted.text || fitted.hardTruncatedSingleLine)
       ? `${fitted.text}${buildContinuationNotice({
           nextFrom,
           suggestReadFallback: fitted.hardTruncatedSingleLine && params.suggestReadFallback,
diff --git a/packages/memory-host-sdk/src/host/response-snippet.test.ts b/packages/memory-host-sdk/src/host/response-snippet.test.ts
@@ -35,6 +35,12 @@ describe("readResponseTextSnippet", () => {
     expect(canceled).toBe(true);
   });
 
+  it("does not split surrogate pairs when truncating text snippets", async () => {
+    await expect(readResponseTextSnippet(new Response("abc🤖tail"), { maxChars: 4 })).resolves.toBe(
+      "abc... [truncated]",
+    );
+  });
+
   it("cancels snippet body reads when the caller signal aborts", async () => {
     let canceled = false;
     const response = stallingResponse(() => {
diff --git a/packages/memory-host-sdk/src/host/response-snippet.ts b/packages/memory-host-sdk/src/host/response-snippet.ts
@@ -1,4 +1,6 @@
 // Memory Host SDK module implements response snippet behavior.
+import { truncateUtf16Safe } from "@openclaw/normalization-core/utf16-slice";
+
 const DEFAULT_ERROR_BODY_MAX_BYTES = 8 * 1024;
 const DEFAULT_ERROR_BODY_MAX_CHARS = 1_000;
 const DEFAULT_JSON_BODY_MAX_BYTES = 64 * 1024 * 1024;
@@ -41,7 +43,7 @@ export async function readResponseTextSnippet(
     return "";
   }
   if (prefix.truncated || collapsed.length > maxChars) {
-    return `${collapsed.slice(0, maxChars)}${TRUNCATED_SUFFIX}`;
+    return `${truncateUtf16Safe(collapsed, maxChars)}${TRUNCATED_SUFFIX}`;
   }
   return collapsed;
 }

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,6 @@`
`1`	`1`	`// Memory Host SDK module implements response snippet behavior.`
	`2`	`+import { truncateUtf16Safe } from "@openclaw/normalization-core/utf16-slice";`
	`3`	`+`
`2`	`4`	`const DEFAULT_ERROR_BODY_MAX_BYTES = 8 * 1024;`
`3`	`5`	`const DEFAULT_ERROR_BODY_MAX_CHARS = 1_000;`
`4`	`6`	`const DEFAULT_JSON_BODY_MAX_BYTES = 64 * 1024 * 1024;`
`@@ -41,7 +43,7 @@ export async function readResponseTextSnippet(`
`41`	`43`	`return "";`
`42`	`44`	`}`
`43`	`45`	`if (prefix.truncated \|\| collapsed.length > maxChars) {`
`44`		- return `${collapsed.slice(0, maxChars)}${TRUNCATED_SUFFIX}`;
	`46`	+ return `${truncateUtf16Safe(collapsed, maxChars)}${TRUNCATED_SUFFIX}`;
`45`	`47`	`}`
`46`	`48`	`return collapsed;`
`47`	`49`	`}`