Reproduced split unicode error

Author: matt-aitken

internal-packages/clickhouse/src/fixtures/bad-clickhouse-output.json

@@ -0,0 +1,3 @@
+{
+  "title": "❜ 𝐒 𝐏𝗈𝗌𝗍 . . . 𝐍𝖾𝗐 𝐂𝗈𝗇𝗍𝖾𝗇𝗍 ꒰ ⚔️ ꒱ 𝐒𝐋 ❜ 𝐔𝐋\n\n꒰ ❤️ ꒱ 𓃊 𝐋𝗲𝗮𝘃𝗲 𝖺 𝗹𝗶𝗸𝗲 𝖺𝗇\ud835"
+}

internal-packages/clickhouse/src/taskRuns.test.ts

@@ -2,6 +2,7 @@ import { clickhouseTest } from "@internal/testcontainers";
 import { z } from "zod";
 import { ClickhouseClient } from "./client/client.js";
 import { getTaskRunsQueryBuilder, insertRawTaskRunPayloads, insertTaskRuns } from "./taskRuns.js";
+import { readFile } from "node:fs/promises";
 
 describe("Task Runs V2", () => {
   clickhouseTest("should be able to insert task runs", async ({ clickhouseContainer }) => {
@@ -395,4 +396,93 @@ describe("Task Runs V2", () => {
       );
     }
   );
+
+  clickhouseTest(
+    "should be able to insert task runs with invalid output",
+    async ({ clickhouseContainer }) => {
+      const client = new ClickhouseClient({
+        name: "test",
+        url: clickhouseContainer.getConnectionUrl(),
+        logLevel: "debug",
+      });
+
+      const insert = insertTaskRuns(client, {
+        async_insert: 0, // turn off async insert for this test
+      });
+
+      const output = await readFile(`${__dirname}/fixtures/bad-clickhouse-output.json`, "utf-8");
+
+      const [insertError, insertResult] = await insert([
+        {
+          environment_id: "env_1234",
+          environment_type: "DEVELOPMENT",
+          organization_id: "org_1234",
+          project_id: "project_1234",
+          run_id: "run_1234",
+          friendly_id: "friendly_1234",
+          attempt: 1,
+          engine: "V2",
+          status: "PENDING",
+          task_identifier: "my-task",
+          queue: "my-queue",
+          schedule_id: "schedule_1234",
+          batch_id: "batch_1234",
+          created_at: Date.now(),
+          updated_at: Date.now(),
+          completed_at: undefined,
+          tags: ["tag1", "tag2"],
+          output: JSON.parse(output),
+          error: {
+            type: "BUILT_IN_ERROR",
+            name: "Error",
+            message: "error",
+            stackTrace: "stack trace",
+          },
+          usage_duration_ms: 1000,
+          cost_in_cents: 100,
+          task_version: "1.0.0",
+          sdk_version: "1.0.0",
+          cli_version: "1.0.0",
+          machine_preset: "small-1x",
+          is_test: true,
+          span_id: "span_1234",
+          trace_id: "trace_1234",
+          idempotency_key: "idempotency_key_1234",
+          expiration_ttl: "1h",
+          root_run_id: "root_run_1234",
+          parent_run_id: "parent_run_1234",
+          depth: 1,
+          _version: "1",
+        },
+      ]);
+
+      expect(insertError).toBeNull();
+      expect(insertResult).toEqual(expect.objectContaining({ executed: true }));
+      expect(insertResult?.summary?.written_rows).toEqual("1");
+
+      const query = client.query({
+        name: "query-task-runs",
+        query: "SELECT * FROM trigger_dev.task_runs_v2",
+        schema: z.object({
+          environment_id: z.string(),
+          run_id: z.string(),
+        }),
+        params: z.object({
+          run_id: z.string(),
+        }),
+      });
+
+      const [queryError, result] = await query({ run_id: "run_1234" });
+
+      expect(queryError).toBeNull();
+      expect(result).toEqual(
+        expect.arrayContaining([
+          expect.objectContaining({
+            environment_id: "env_1234",
+            run_id: "run_1234",
+          }),
+        ])
+      );
+    }
+  );
 });