patch(js): utf-8 encode multipart parts (#1487)

js/ls.vitest.config.ts

@@ -5,5 +5,6 @@ export default defineConfig({
     include: ["**/*.vitesteval.?(c|m)[jt]s"],
     reporters: ["./src/vitest/reporter.ts"],
     setupFiles: ["dotenv/config"],
+    hookTimeout: 30000,
   },
 });

js/package.json

@@ -1,6 +1,6 @@
 {
   "name": "langsmith",
-  "version": "0.3.4",
+  "version": "0.3.5",
   "description": "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform.",
   "packageManager": "[email protected]",
   "files": [

js/src/client.ts

@@ -67,7 +67,7 @@ import { parsePromptIdentifier } from "./utils/prompts.js";
 import { raiseForStatus } from "./utils/error.js";
 import { _getFetchImplementation } from "./singletons/fetch.js";
 
-import { stringify as stringifyForTracing } from "./utils/fast-safe-stringify/index.js";
+import { serialize as serializePayloadForTracing } from "./utils/fast-safe-stringify/index.js";
 
 export interface ClientConfig {
   apiUrl?: string;
@@ -405,7 +405,7 @@ export class AutoBatchQueue {
       // https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise/Promise
       itemPromiseResolve = resolve;
     });
-    const size = stringifyForTracing(item.item).length;
+    const size = serializePayloadForTracing(item.item).length;
     this.items.push({
       action: item.action,
       payload: item.item,
@@ -943,7 +943,7 @@ export class Client implements LangSmithTracingClientInterface {
       {
         method: "POST",
         headers,
-        body: stringifyForTracing(mergedRunCreateParam),
+        body: serializePayloadForTracing(mergedRunCreateParam),
         signal: AbortSignal.timeout(this.timeout_ms),
         ...this.fetchOptions,
       }
@@ -1020,11 +1020,11 @@ export class Client implements LangSmithTracingClientInterface {
       }
     }
     if (batchChunks.post.length > 0 || batchChunks.patch.length > 0) {
-      await this._postBatchIngestRuns(stringifyForTracing(batchChunks));
+      await this._postBatchIngestRuns(serializePayloadForTracing(batchChunks));
     }
   }
 
-  private async _postBatchIngestRuns(body: string) {
+  private async _postBatchIngestRuns(body: Uint8Array) {
     const headers = {
       ...this.headers,
       "Content-Type": "application/json",
@@ -1143,7 +1143,7 @@ export class Client implements LangSmithTracingClientInterface {
           originalPayload;
         const fields = { inputs, outputs, events };
         // encode the main run payload
-        const stringifiedPayload = stringifyForTracing(payload);
+        const stringifiedPayload = serializePayloadForTracing(payload);
         accumulatedParts.push({
           name: `${method}.${payload.id}`,
           payload: new Blob([stringifiedPayload], {
@@ -1155,7 +1155,7 @@ export class Client implements LangSmithTracingClientInterface {
           if (value === undefined) {
             continue;
           }
-          const stringifiedValue = stringifyForTracing(value);
+          const stringifiedValue = serializePayloadForTracing(value);
           accumulatedParts.push({
             name: `${method}.${payload.id}.${key}`,
             payload: new Blob([stringifiedValue], {
@@ -1301,7 +1301,7 @@ export class Client implements LangSmithTracingClientInterface {
       {
         method: "PATCH",
         headers,
-        body: stringifyForTracing(run),
+        body: serializePayloadForTracing(run),
         signal: AbortSignal.timeout(this.timeout_ms),
         ...this.fetchOptions,
       }
@@ -4124,15 +4124,15 @@ export class Client implements LangSmithTracingClientInterface {
       };
 
       // Add main example data
-      const stringifiedExample = stringifyForTracing(exampleBody);
+      const stringifiedExample = serializePayloadForTracing(exampleBody);
       const exampleBlob = new Blob([stringifiedExample], {
         type: "application/json",
       });
       formData.append(exampleId, exampleBlob);
 
       // Add inputs
       if (example.inputs) {
-        const stringifiedInputs = stringifyForTracing(example.inputs);
+        const stringifiedInputs = serializePayloadForTracing(example.inputs);
         const inputsBlob = new Blob([stringifiedInputs], {
           type: "application/json",
         });
@@ -4141,7 +4141,7 @@ export class Client implements LangSmithTracingClientInterface {
 
       // Add outputs if present
       if (example.outputs) {
-        const stringifiedOutputs = stringifyForTracing(example.outputs);
+        const stringifiedOutputs = serializePayloadForTracing(example.outputs);
         const outputsBlob = new Blob([stringifiedOutputs], {
           type: "application/json",
         });
@@ -4168,7 +4168,7 @@ export class Client implements LangSmithTracingClientInterface {
       }
 
       if (example.attachments_operations) {
-        const stringifiedAttachmentsOperations = stringifyForTracing(
+        const stringifiedAttachmentsOperations = serializePayloadForTracing(
           example.attachments_operations
         );
         const attachmentsOperationsBlob = new Blob(
@@ -4224,22 +4224,22 @@ export class Client implements LangSmithTracingClientInterface {
       };
 
       // Add main example data
-      const stringifiedExample = stringifyForTracing(exampleBody);
+      const stringifiedExample = serializePayloadForTracing(exampleBody);
       const exampleBlob = new Blob([stringifiedExample], {
         type: "application/json",
       });
       formData.append(exampleId, exampleBlob);
 
       // Add inputs
-      const stringifiedInputs = stringifyForTracing(example.inputs);
+      const stringifiedInputs = serializePayloadForTracing(example.inputs);
       const inputsBlob = new Blob([stringifiedInputs], {
         type: "application/json",
       });
       formData.append(`${exampleId}.inputs`, inputsBlob);
 
       // Add outputs if present
       if (example.outputs) {
-        const stringifiedOutputs = stringifyForTracing(example.outputs);
+        const stringifiedOutputs = serializePayloadForTracing(example.outputs);
         const outputsBlob = new Blob([stringifiedOutputs], {
           type: "application/json",
         });

js/src/evaluation/_runner.ts

@@ -877,7 +877,6 @@ export class _ExperimentManager {
     }
 
     await this.client.updateProject(experiment.id, {
-      endTime: new Date().toISOString(),
       metadata: projectMetadata,
     });
   }

js/src/index.ts

@@ -18,4 +18,4 @@ export { RunTree, type RunTreeConfig } from "./run_trees.js";
 export { overrideFetchImplementation } from "./singletons/fetch.js";
 
 // Update using yarn bump-version
-export const __version__ = "0.3.4";
+export const __version__ = "0.3.5";

js/src/tests/batch_client.test.ts

@@ -7,12 +7,19 @@ import { convertToDottedOrderFormat } from "../run_trees.js";
 import { _getFetchImplementation } from "../singletons/fetch.js";
 import { RunCreate } from "../schemas.js";
 
-const parseMockRequestBody = async (body: string | ArrayBuffer) => {
+const parseMockRequestBody = async (
+  body: string | ArrayBuffer | Uint8Array
+) => {
   if (typeof body === "string") {
     return JSON.parse(body);
   }
+
   // Typing is missing
   const rawMultipart = new TextDecoder().decode(body);
+
+  if (rawMultipart.trim().startsWith("{")) {
+    return JSON.parse(rawMultipart);
+  }
   // Parse the multipart form data boundary from the raw text
   const boundary = rawMultipart.split("\r\n")[0].trim();
   // Split the multipart body into individual parts
@@ -143,7 +150,7 @@ describe.each(ENDPOINT_TYPES)(
         _getFetchImplementation(),
         expectedTraceURL,
         expect.objectContaining({
-          body: expect.any(endpointType === "batch" ? String : ArrayBuffer),
+          body: expect.any(endpointType === "batch" ? Uint8Array : ArrayBuffer),
         })
       );
     });
@@ -257,7 +264,7 @@ describe.each(ENDPOINT_TYPES)(
         _getFetchImplementation(),
         expectedTraceURL,
         expect.objectContaining({
-          body: expect.any(endpointType === "batch" ? String : ArrayBuffer),
+          body: expect.any(endpointType === "batch" ? Uint8Array : ArrayBuffer),
         })
       );
     });
@@ -338,7 +345,7 @@ describe.each(ENDPOINT_TYPES)(
         _getFetchImplementation(),
         expectedTraceURL,
         expect.objectContaining({
-          body: expect.any(endpointType === "batch" ? String : ArrayBuffer),
+          body: expect.any(endpointType === "batch" ? Uint8Array : ArrayBuffer),
         })
       );
     });
@@ -935,7 +942,7 @@ describe.each(ENDPOINT_TYPES)(
         _getFetchImplementation(),
         "https://api.smith.langchain.com/runs/batch",
         expect.objectContaining({
-          body: expect.any(String),
+          body: expect.any(Uint8Array),
         })
       );
     });
@@ -1027,7 +1034,7 @@ describe.each(ENDPOINT_TYPES)(
         _getFetchImplementation(),
         expectedTraceURL,
         expect.objectContaining({
-          body: expect.any(endpointType === "batch" ? String : ArrayBuffer),
+          body: expect.any(endpointType === "batch" ? Uint8Array : ArrayBuffer),
         })
       );
     });

js/src/tests/utils/tree.ts

@@ -30,20 +30,29 @@ export function getAssumedTreeFromCalls(calls: unknown[][]) {
       { method: string; body: string }
     ];
     const req = `${fetchArgs.method} ${new URL(url as string).pathname}`;
-    const body: Run = JSON.parse(fetchArgs.body);
+    let body: Run;
+    if (typeof fetchArgs.body === "string") {
+      body = JSON.parse(fetchArgs.body);
+    } else {
+      const decoded = new TextDecoder().decode(fetchArgs.body);
+
+      if (decoded.trim().startsWith("{")) {
+        body = JSON.parse(decoded);
+      }
+    }
 
     if (req === "POST /runs") {
-      const id = body.id;
+      const id = body!.id;
       upsertId(id);
-      nodeMap[id] = { ...nodeMap[id], ...body };
+      nodeMap[id] = { ...nodeMap[id], ...body! };
       if (nodeMap[id].parent_run_id) {
         // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
         edges.push([nodeMap[id].parent_run_id!, nodeMap[id].id]);
       }
     } else if (req.startsWith("PATCH /runs/")) {
       const id = req.substring("PATCH /runs/".length);
       upsertId(id);
-      nodeMap[id] = { ...nodeMap[id], ...body };
+      nodeMap[id] = { ...nodeMap[id], ...body! };
     }
   }
 

js/src/tests/wrapped_openai.int.test.ts

@@ -11,6 +11,13 @@ import { z } from "zod";
 import { UsageMetadata } from "../schemas.js";
 import fs from "fs";
 
+function parseRequestBody(body: any) {
+  // eslint-disable-next-line no-instanceof/no-instanceof
+  return body instanceof Uint8Array
+    ? JSON.parse(new TextDecoder().decode(body))
+    : JSON.parse(body);
+}
+
 test("wrapOpenAI should return type compatible with OpenAI", async () => {
   let originalClient = new OpenAI();
   // eslint-disable-next-line @typescript-eslint/no-unused-vars
@@ -323,8 +330,8 @@ test.concurrent("chat completions with tool calling", async () => {
   for (const call of callSpy.mock.calls) {
     // eslint-disable-next-line @typescript-eslint/no-explicit-any
     expect(["POST", "PATCH"]).toContain((call[2] as any)["method"]);
-    // eslint-disable-next-line @typescript-eslint/no-explicit-any
-    expect(JSON.parse((call[2] as any).body).extra.metadata).toEqual({
+    const body = parseRequestBody((call[2] as any).body);
+    expect(body.extra.metadata).toEqual({
       thing1: "thing2",
       ls_model_name: "gpt-3.5-turbo",
       ls_model_type: "chat",
@@ -566,8 +573,8 @@ test.concurrent("beta.chat.completions.parse", async () => {
   for (const call of callSpy.mock.calls) {
     // eslint-disable-next-line @typescript-eslint/no-explicit-any
     expect(["POST", "PATCH"]).toContain((call[2] as any)["method"]);
-    // eslint-disable-next-line @typescript-eslint/no-explicit-any
-    expect(JSON.parse((call[2] as any).body).extra.metadata).toEqual({
+    const body = parseRequestBody((call[2] as any).body);
+    expect(body.extra.metadata).toEqual({
       ls_model_name: "gpt-4o-mini",
       ls_model_type: "chat",
       ls_provider: "openai",
@@ -655,7 +662,7 @@ describe("Usage Metadata Tests", () => {
         const requestBodies: any = {};
         for (const call of callSpy.mock.calls) {
           const request = call[2] as any;
-          const requestBody = JSON.parse(request.body);
+          const requestBody = parseRequestBody(request.body);
           if (request.method === "POST") {
             requestBodies["post"] = [requestBody];
           }

js/src/utils/fast-safe-stringify/index.ts

@@ -6,22 +6,29 @@ var CIRCULAR_REPLACE_NODE = { result: "[Circular]" };
 var arr = [];
 var replacerStack = [];
 
+const encoder = new TextEncoder();
+
 function defaultOptions() {
   return {
     depthLimit: Number.MAX_SAFE_INTEGER,
     edgesLimit: Number.MAX_SAFE_INTEGER,
   };
 }
 
+function encodeString(str: string): Uint8Array {
+  return encoder.encode(str);
+}
+
 // Regular stringify
-export function stringify(obj, replacer?, spacer?, options?) {
+export function serialize(obj, replacer?, spacer?, options?) {
   try {
-    return JSON.stringify(obj, replacer, spacer);
+    const str = JSON.stringify(obj, replacer, spacer);
+    return encodeString(str);
   } catch (e: any) {
     // Fall back to more complex stringify if circular reference
     if (!e.message?.includes("Converting circular structure to JSON")) {
       console.warn("[WARNING]: LangSmith received unserializable value.");
-      return "[Unserializable]";
+      return encodeString("[Unserializable]");
     }
     console.warn(
       "[WARNING]: LangSmith received circular JSON. This will decrease tracer performance."
@@ -31,28 +38,28 @@ export function stringify(obj, replacer?, spacer?, options?) {
     }
 
     decirc(obj, "", 0, [], undefined, 0, options);
-    var res;
+    let res: string;
     try {
       if (replacerStack.length === 0) {
         res = JSON.stringify(obj, replacer, spacer);
       } else {
         res = JSON.stringify(obj, replaceGetterValues(replacer), spacer);
       }
     } catch (_) {
-      return JSON.stringify(
+      return encodeString(
         "[unable to serialize, circular reference is too complex to analyze]"
       );
     } finally {
       while (arr.length !== 0) {
-        var part = arr.pop();
+        const part = arr.pop();
         if (part.length === 4) {
           Object.defineProperty(part[0], part[1], part[3]);
         } else {
           part[0][part[1]] = part[2];
         }
       }
     }
-    return res;
+    return encodeString(res);
   }
 }
 

python/langsmith/beta/_evals.py

@@ -169,7 +169,7 @@ def convert_runs_to_test(
         client.create_run(**new_run, project_name=test_project_name)
 
     _ = client.update_project(
-        project.id, end_time=datetime.datetime.now(tz=datetime.timezone.utc)
+        project.id,
     )
     return project
 

python/langsmith/evaluation/_arunner.py

@@ -4,7 +4,6 @@
 
 import asyncio
 import concurrent.futures as cf
-import datetime
 import io
 import logging
 import pathlib
@@ -1137,8 +1136,6 @@ async def _aend(self) -> None:
         project_metadata["dataset_splits"] = await self._get_dataset_splits()
         self.client.update_project(
             experiment.id,
-            end_time=experiment.end_time
-            or datetime.datetime.now(datetime.timezone.utc),
             metadata={
                 **experiment.metadata,
                 **project_metadata,