mirror of
https://github.com/openai/codex.git
synced 2026-11-15 21:50:46 +00:00
Enable server-side output truncation for Code Mode sessions (#57870)
## Why Code Mode sessions always used client-side output handling, leaving server truncation settings and observation budgets unused. ## What changed - Honor each session's `output_truncation` setting in the runtime and both host transports, and advertise the bounded-output capability. - Apply `max_output_tokens` independently to execute, wait, and terminate responses in server mode. Keep client mode as the default when the setting is omitted. - Use bounded output collection and rendering in server mode, preserving head and tail output with omission notices. Include script errors in the output budget while retaining failure status and a separate diagnostic capped at 64 KiB. - Estimate audio tokens without the Tokio-dependent cache when rendering bounded output. ## Testing Add runtime and host regression tests for session mode selection, per-request budgets, output overflow, oversized images, zero-token audio output, diagnostic preservation, and termination under output backpressure. GitOrigin-RevId: 3d9661edcdd2dde31526bb92eb83fa0614f0fa88
This commit is contained in:
@@ -1,4 +1,4 @@
|
||||
//! Covers per-item short-WAV omissions without disrupting other Code Mode outputs.
|
||||
//! Covers short-WAV handling and Server audio output alongside other Code Mode outputs.
|
||||
|
||||
use super::cell_id;
|
||||
use super::execute;
|
||||
@@ -11,6 +11,9 @@ use crate::WaitOutcome;
|
||||
use crate::WaitRequest;
|
||||
use base64::Engine;
|
||||
use base64::engine::general_purpose::STANDARD as BASE64_STANDARD;
|
||||
use codex_code_mode_protocol::CodeModeOutputTruncation;
|
||||
use codex_code_mode_protocol::CodeModeSessionOptions;
|
||||
use codex_code_mode_protocol::host::BOUNDED_OUTPUT_MAX_BYTES;
|
||||
use pretty_assertions::assert_eq;
|
||||
|
||||
const SHORT_AUDIO_OMISSION_TEXT: &str =
|
||||
@@ -168,3 +171,45 @@ async fn audio_helper_bounds_wav_data_and_preserves_outputs_after_yield() {
|
||||
})
|
||||
);
|
||||
}
|
||||
|
||||
#[tokio::test]
|
||||
async fn server_omits_audio_above_the_text_allowance_at_zero_tokens() {
|
||||
let session = InProcessCodeModeSession::with_options(
|
||||
CodeModeSessionOptions {
|
||||
output_truncation: CodeModeOutputTruncation::Server,
|
||||
..CodeModeSessionOptions::default()
|
||||
},
|
||||
/*task_failure_handler*/ None,
|
||||
);
|
||||
// Each clip exceeds the text allowance, but the clips fit the separate
|
||||
// audio allowance and reach the zero-token formatter.
|
||||
let wav = pcm16_wav(/*sample_rate*/ 24_000, BOUNDED_OUTPUT_MAX_BYTES / 2);
|
||||
let payload = BASE64_STANDARD.encode(wav);
|
||||
let source = format!(
|
||||
"const clip = 'data:audio/wav;base64,{payload}'; for (let i = 0; i < 6; i++) {{ audio(clip); text(''); }}"
|
||||
);
|
||||
let response = execute(
|
||||
&session,
|
||||
ExecuteRequest {
|
||||
yield_time_ms: Some(60_000),
|
||||
max_output_tokens: Some(0),
|
||||
..execute_request(&source)
|
||||
},
|
||||
)
|
||||
.await;
|
||||
let RuntimeResponse::Result {
|
||||
content_items,
|
||||
error_text,
|
||||
..
|
||||
} = response
|
||||
else {
|
||||
panic!("expected completed cell");
|
||||
};
|
||||
assert_eq!(error_text, None);
|
||||
assert_eq!(
|
||||
content_items,
|
||||
[FunctionCallOutputContentItem::InputText {
|
||||
text: "[omitted 6 audio items ...]".to_owned(),
|
||||
}]
|
||||
);
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user