fix(providers): tolerate missing created_at on OpenAI Responses stream events (#12241)

Co-authored-by: Oscar-Williams <Oscar-Williams@users.noreply.github.com>
This commit is contained in:
Oscar
2026-10-14 18:55:43 +00:00
committed by GitHub
parent 9686d67400
commit 046086967c
@@ -25,6 +25,9 @@ use std::collections::HashSet;
pub struct ResponsesApiResponse {
pub id: String,
pub object: String,
// Not every OpenAI-compatible provider echoes created_at back; the value
// is carried but never used, so tolerate its absence.
#[serde(default)]
pub created_at: i64,
pub status: String,
pub model: String,
@@ -337,6 +340,7 @@ fn parse_responses_stream_event(data_line: &str) -> anyhow::Result<Option<Respon
pub struct ResponseMetadata {
pub id: String,
pub object: String,
#[serde(default)]
pub created_at: i64,
pub status: String,
pub model: String,
@@ -1477,6 +1481,41 @@ mod tests {
Ok(())
}
// OpenAI-compatible providers do not all echo `created_at` back on
// `response.*` events; a strict i64 there terminates the whole session on
// the first event. The timestamp is carried but never used, so a missing
// field must degrade to the default instead of killing the stream.
#[tokio::test]
async fn test_responses_stream_tolerates_missing_created_at() -> anyhow::Result<()> {
let lines = vec![
r#"data: {"type":"response.created","sequence_number":1,"response":{"id":"resp_1","object":"response","status":"in_progress","model":"gpt-5.2-pro","output":[]}}"#.to_string(),
r#"data: {"type":"response.output_text.delta","sequence_number":2,"item_id":"msg_1","output_index":0,"content_index":0,"delta":"Hello"}"#.to_string(),
r#"data: {"type":"response.output_text.delta","sequence_number":3,"item_id":"msg_1","output_index":0,"content_index":0,"delta":" world"}"#.to_string(),
r#"data: {"type":"response.completed","sequence_number":4,"response":{"id":"resp_1","object":"response","status":"completed","model":"gpt-5.2-pro","output":[],"usage":{"input_tokens":10,"output_tokens":4,"total_tokens":14}}}"#.to_string(),
"data: [DONE]".to_string(),
];
let response_stream = tokio_stream::iter(lines.into_iter().map(Ok));
let messages = responses_api_to_streaming_message(response_stream);
futures::pin_mut!(messages);
let mut text_parts = Vec::new();
while let Some(item) = messages.next().await {
let (message, _) = item?;
if let Some(msg) = message {
for content in msg.content {
if let MessageContentBlock::Text(text) = content {
text_parts.push(text.text.clone());
}
}
}
}
assert_eq!(text_parts.concat(), "Hello world");
Ok(())
}
#[tokio::test]
async fn test_responses_stream_marks_output_token_limit_when_incomplete() -> anyhow::Result<()>
{