Files
codex/codex-rs/core/src/compact_tests.rs
T
Ahmed Ibrahim ba4925b3c2 [codex] Compact when comp_hash changes (#27520)
## Summary
- snapshot `comp_hash` into `TurnContext` when the turn is created and
use that snapshot as the downstream source of truth
- persist the turn hash in rollout context and recover it into
previous-turn settings during resume and fork replay
- compact existing history with the previous model only when both
adjacent turns provide hashes and the values differ
- record `comp_hash_changed` as the compaction reason
- cover ordinary transitions, resume, and missing-hash compatibility
with end-to-end tests

## Why
History produced under one compaction-compatible model configuration may
not be safe to carry directly into another. Compacting at the turn
boundary converts that history before context updates and the new user
message are added. Persisting the turn snapshot in `TurnContextItem`
makes the same protection work after resuming a rollout.

A missing hash is not treated as evidence of incompatibility. `None →
Some`, `Some → None`, and `None → None` do not trigger compaction; only
`Some(previous) → Some(current)` with unequal values does.

## Stack
- depends on #27532
- #27532 is based directly on `main`

## Testing
- `just test -p codex-core pre_sampling_compact_` — 6 passed
- `just test -p codex-core
turn_context_item_uses_turn_context_comp_hash_snapshot` — passed
- `just fix -p codex-core -p codex-protocol -p codex-analytics -p
codex-models-manager`
2026-06-11 04:11:26 +00:00

624 lines
19 KiB
Rust

use super::*;
use codex_model_provider_info::ModelProviderInfo;
use codex_model_provider_info::WireApi;
use codex_protocol::models::DEFAULT_IMAGE_DETAIL;
use pretty_assertions::assert_eq;
async fn process_compacted_history_with_test_session(
compacted_history: Vec<ResponseItem>,
previous_turn_settings: Option<&PreviousTurnSettings>,
) -> (Vec<ResponseItem>, Vec<ResponseItem>) {
let (session, turn_context) = crate::session::tests::make_session_and_context().await;
session
.set_previous_turn_settings(previous_turn_settings.cloned())
.await;
let initial_context = session.build_initial_context(&turn_context).await;
let refreshed = crate::compact_remote::process_compacted_history(
&session,
&turn_context,
compacted_history,
InitialContextInjection::BeforeLastUserMessage,
)
.await;
(refreshed, initial_context)
}
fn user_message(text: &str) -> ResponseItem {
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: text.to_string(),
}],
phase: None,
}
}
#[test]
fn content_items_to_text_joins_non_empty_segments() {
let items = vec![
ContentItem::InputText {
text: "hello".to_string(),
},
ContentItem::OutputText {
text: String::new(),
},
ContentItem::OutputText {
text: "world".to_string(),
},
];
let joined = content_items_to_text(&items);
assert_eq!(Some("hello\nworld".to_string()), joined);
}
#[test]
fn content_items_to_text_ignores_image_only_content() {
let items = vec![ContentItem::InputImage {
image_url: "file://image.png".to_string(),
detail: Some(DEFAULT_IMAGE_DETAIL),
}];
let joined = content_items_to_text(&items);
assert_eq!(None, joined);
}
#[test]
fn collect_user_messages_extracts_user_text_only() {
let items = vec![
ResponseItem::Message {
id: Some("assistant".to_string()),
role: "assistant".to_string(),
content: vec![ContentItem::OutputText {
text: "ignored".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: Some("user".to_string()),
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "first".to_string(),
}],
phase: None,
},
ResponseItem::Other,
];
let collected = collect_user_messages(&items);
assert_eq!(vec!["first".to_string()], collected);
}
#[test]
fn collect_user_messages_filters_session_prefix_entries() {
let items = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: r#"# AGENTS.md instructions for project
<INSTRUCTIONS>
do things
</INSTRUCTIONS>"#
.to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "<ENVIRONMENT_CONTEXT>cwd=/tmp</ENVIRONMENT_CONTEXT>".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "real user message".to_string(),
}],
phase: None,
},
];
let collected = collect_user_messages(&items);
assert_eq!(vec!["real user message".to_string()], collected);
}
#[test]
fn collect_user_messages_filters_legacy_warnings() {
let items = vec![
user_message(
"Warning: The maximum number of unified exec processes you can keep open is 60 and you currently have 61 processes open. Reuse older processes or close them to prevent automatic pruning of old processes",
),
user_message(
"Warning: apply_patch was requested via exec_command. Use the apply_patch tool instead of exec_command.",
),
user_message(
"Warning: Your account was flagged for potentially high-risk cyber activity and this request was routed to gpt-5.2 as a fallback. To regain access to gpt-5.3-codex, apply for trusted access: https://chatgpt.com/cyber or learn more: https://developers.openai.com/codex/concepts/cyber-safety",
),
user_message("real user message"),
];
let collected = collect_user_messages(&items);
assert_eq!(vec!["real user message".to_string()], collected);
}
#[test]
fn build_token_limited_compacted_history_truncates_overlong_user_messages() {
// Use a small truncation limit so the test remains fast while still validating
// that oversized user content is truncated.
let max_tokens = 16;
let big = "word ".repeat(200);
let history = super::build_compacted_history_with_limit(
Vec::new(),
std::slice::from_ref(&big),
"SUMMARY",
max_tokens,
);
assert_eq!(history.len(), 2);
let truncated_message = &history[0];
let summary_message = &history[1];
let truncated_text = match truncated_message {
ResponseItem::Message { role, content, .. } if role == "user" => {
content_items_to_text(content).unwrap_or_default()
}
other => panic!("unexpected item in history: {other:?}"),
};
assert!(
truncated_text.contains("tokens truncated"),
"expected truncation marker in truncated user message"
);
assert!(
!truncated_text.contains(&big),
"truncated user message should not include the full oversized user text"
);
let summary_text = match summary_message {
ResponseItem::Message { role, content, .. } if role == "user" => {
content_items_to_text(content).unwrap_or_default()
}
other => panic!("unexpected item in history: {other:?}"),
};
assert_eq!(summary_text, "SUMMARY");
}
#[test]
fn build_token_limited_compacted_history_appends_summary_message() {
let initial_context: Vec<ResponseItem> = Vec::new();
let user_messages = vec!["first user message".to_string()];
let summary_text = "summary text";
let history = build_compacted_history(initial_context, &user_messages, summary_text);
assert!(
!history.is_empty(),
"expected compacted history to include summary"
);
let last = history.last().expect("history should have a summary entry");
let summary = match last {
ResponseItem::Message { role, content, .. } if role == "user" => {
content_items_to_text(content).unwrap_or_default()
}
other => panic!("expected summary message, found {other:?}"),
};
assert_eq!(summary, summary_text);
}
#[test]
fn should_use_remote_compact_task_for_azure_provider() {
let provider = ModelProviderInfo {
name: "Azure".into(),
base_url: Some("https://example.com/openai".into()),
env_key: Some("AZURE_OPENAI_API_KEY".into()),
env_key_instructions: None,
experimental_bearer_token: None,
auth: None,
aws: None,
wire_api: WireApi::Responses,
query_params: None,
http_headers: None,
env_http_headers: None,
request_max_retries: None,
stream_max_retries: None,
stream_idle_timeout_ms: None,
websocket_connect_timeout_ms: None,
requires_openai_auth: false,
supports_websockets: false,
};
assert!(should_use_remote_compact_task(&provider));
}
#[tokio::test]
async fn process_compacted_history_replaces_developer_messages() {
let compacted_history = vec![
ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "stale permissions".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "stale personality".to_string(),
}],
phase: None,
},
];
let (refreshed, mut expected) = process_compacted_history_with_test_session(
compacted_history,
/*previous_turn_settings*/ None,
)
.await;
expected.push(ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
});
assert_eq!(refreshed, expected);
}
#[tokio::test]
async fn process_compacted_history_reinjects_full_initial_context() {
let compacted_history = vec![ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
}];
let (refreshed, mut expected) = process_compacted_history_with_test_session(
compacted_history,
/*previous_turn_settings*/ None,
)
.await;
expected.push(ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
});
assert_eq!(refreshed, expected);
}
#[tokio::test]
async fn process_compacted_history_drops_non_user_content_messages() {
let compacted_history = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: r#"# AGENTS.md instructions for /repo
<INSTRUCTIONS>
keep me updated
</INSTRUCTIONS>"#
.to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: r#"<environment_context>
<cwd>/repo</cwd>
<shell>zsh</shell>
</environment_context>"#
.to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: r#"<turn_aborted>
<turn_id>turn-1</turn_id>
<reason>interrupted</reason>
</turn_aborted>"#
.to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "stale developer instructions".to_string(),
}],
phase: None,
},
];
let (refreshed, mut expected) = process_compacted_history_with_test_session(
compacted_history,
/*previous_turn_settings*/ None,
)
.await;
expected.push(ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
});
assert_eq!(refreshed, expected);
}
#[tokio::test]
async fn process_compacted_history_drops_legacy_warnings() {
let latest_user = user_message("latest user");
let compacted_history = vec![
user_message(
"Warning: The maximum number of unified exec processes you can keep open is 60 and you currently have 61 processes open. Reuse older processes or close them to prevent automatic pruning of old processes",
),
user_message(
"Warning: apply_patch was requested via exec_command. Use the apply_patch tool instead of exec_command.",
),
user_message(
"Warning: Your account was flagged for potentially high-risk cyber activity and this request was routed to gpt-5.2 as a fallback. To regain access to gpt-5.3-codex, apply for trusted access: https://chatgpt.com/cyber or learn more: https://developers.openai.com/codex/concepts/cyber-safety",
),
latest_user.clone(),
];
let (refreshed, initial_context) = process_compacted_history_with_test_session(
compacted_history,
/*previous_turn_settings*/ None,
)
.await;
let mut expected = initial_context;
expected.push(latest_user);
assert_eq!(refreshed, expected);
}
#[tokio::test]
async fn process_compacted_history_inserts_context_before_last_real_user_message_only() {
let compacted_history = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "older user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: format!("{SUMMARY_PREFIX}\nsummary text"),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "latest user".to_string(),
}],
phase: None,
},
];
let (refreshed, initial_context) = process_compacted_history_with_test_session(
compacted_history,
/*previous_turn_settings*/ None,
)
.await;
let mut expected = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "older user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: format!("{SUMMARY_PREFIX}\nsummary text"),
}],
phase: None,
},
];
expected.extend(initial_context);
expected.push(ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "latest user".to_string(),
}],
phase: None,
});
assert_eq!(refreshed, expected);
}
#[tokio::test]
async fn process_compacted_history_reinjects_model_switch_message() {
let compacted_history = vec![ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
}];
let previous_turn_settings = PreviousTurnSettings {
model: "previous-regular-model".to_string(),
comp_hash: None,
realtime_active: None,
};
let (refreshed, initial_context) = process_compacted_history_with_test_session(
compacted_history,
Some(&previous_turn_settings),
)
.await;
let ResponseItem::Message { role, content, .. } = &initial_context[0] else {
panic!("expected developer message");
};
assert_eq!(role, "developer");
let [ContentItem::InputText { text }, ..] = content.as_slice() else {
panic!("expected developer text");
};
assert!(text.contains("<model_switch>"));
let mut expected = initial_context;
expected.push(ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "summary".to_string(),
}],
phase: None,
});
assert_eq!(refreshed, expected);
}
#[test]
fn insert_initial_context_before_last_real_user_or_summary_keeps_summary_last() {
let compacted_history = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "older user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "latest user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: format!("{SUMMARY_PREFIX}\nsummary text"),
}],
phase: None,
},
];
let initial_context = vec![ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "fresh permissions".to_string(),
}],
phase: None,
}];
let refreshed =
insert_initial_context_before_last_real_user_or_summary(compacted_history, initial_context);
let expected = vec![
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "older user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "fresh permissions".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: "latest user".to_string(),
}],
phase: None,
},
ResponseItem::Message {
id: None,
role: "user".to_string(),
content: vec![ContentItem::InputText {
text: format!("{SUMMARY_PREFIX}\nsummary text"),
}],
phase: None,
},
];
assert_eq!(refreshed, expected);
}
#[test]
fn insert_initial_context_before_last_real_user_or_summary_keeps_compaction_last() {
let compacted_history = vec![ResponseItem::Compaction {
encrypted_content: "encrypted".to_string(),
}];
let initial_context = vec![ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "fresh permissions".to_string(),
}],
phase: None,
}];
let refreshed =
insert_initial_context_before_last_real_user_or_summary(compacted_history, initial_context);
let expected = vec![
ResponseItem::Message {
id: None,
role: "developer".to_string(),
content: vec![ContentItem::InputText {
text: "fresh permissions".to_string(),
}],
phase: None,
},
ResponseItem::Compaction {
encrypted_content: "encrypted".to_string(),
},
];
assert_eq!(refreshed, expected);
}