fix: Leading whitespace chunks break partial parser

openai · Dec 17, 2024 · 62254df · 62254df
1 parent f361a0c
commit 62254df
Show file tree

Hide file tree

Showing 3 changed files with 78 additions and 1 deletion.
diff --git a/src/lib/ChatCompletionStream.ts b/src/lib/ChatCompletionStream.ts
@@ -509,7 +509,9 @@ export class ChatCompletionStream<ParsedT = null>
         choice.message.content = (choice.message.content || '') + content;
 
         if (!choice.message.refusal && this.#getAutoParseableResponseFormat()) {
-          choice.message.parsed = partialParse(choice.message.content);
+          // Even a partial parser does not accept empty string
+          const trimmed = choice.message.content.trimStart();
+          choice.message.parsed = trimmed ? partialParse(trimmed) : null;
         }
       }
 

diff --git a/tests/lib/ChatCompletionStream.test.ts b/tests/lib/ChatCompletionStream.test.ts
@@ -45,6 +45,47 @@ describe('.stream()', () => {
     `);
   });
 
+  it('is robust against leading newline chunks', async () => {
+    const stream = await makeStreamSnapshotRequest((openai) =>
+      openai.beta.chat.completions.stream({
+        model: 'gpt-4o-2024-08-06',
+        messages: [
+          {
+            role: 'user',
+            content: "What's the weather like in SF?",
+          },
+        ],
+        response_format: zodResponseFormat(
+          z.object({
+            city: z.string(),
+            units: z.enum(['c', 'f']).default('f'),
+          }),
+          'location',
+        ),
+      }),
+    );
+
+    expect((await stream.finalChatCompletion()).choices[0]).toMatchInlineSnapshot(`
+      {
+        "finish_reason": "stop",
+        "index": 0,
+        "logprobs": null,
+        "message": {
+          "content": "
+
+      {"city":"San Francisco","units":"c"}",
+          "parsed": {
+            "city": "San Francisco",
+            "units": "c",
+          },
+          "refusal": null,
+          "role": "assistant",
+          "tool_calls": [],
+        },
+      }
+    `);
+  });
+
   it('emits content logprobs events', async () => {
     var capturedLogProbs: ChatCompletionTokenLogprob[] | undefined;
 

diff --git a/tests/lib/__snapshots__/ChatCompletionStream.test.ts.snap b/tests/lib/__snapshots__/ChatCompletionStream.test.ts.snap
@@ -99,3 +99,37 @@ data: [DONE]
 
 "
 `;
+
+exports[`.stream() is robust against leading newline chunks 1`] = `
+"data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"\\n\\n"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"{\\""},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"city"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"\\":\\""},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"San"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":" Francisco"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"\\",\\""},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"units"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"\\":\\""},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"c"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{"content":"\\"}"},"logprobs":null,"finish_reason":null}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[{"index":0,"delta":{},"logprobs":null,"finish_reason":"stop"}]}
+
+data: {"id":"chatcmpl-9tZXEmwtoDf6vqCqEWSvDP8jx9OXe","object":"chat.completion.chunk","created":1723031664,"model":"gpt-4o-2024-08-06","system_fingerprint":"fp_845eaabc1f","choices":[],"usage":{"prompt_tokens":17,"completion_tokens":10,"total_tokens":27}}
+
+data: [DONE]
+
+"
+`;