i/Aether
1
0
Fork 0
mirror of https://github.com/fawney19/Aether.git synced 2026-09-03 01:40:21 +08:00
Code Issues Packages Projects Releases Wiki Activity

refactor: 抽离 AI pipeline 与调度共享能力逻辑

Browse Source
This commit is contained in:
fawney19
2026-04-10 01:46:14 +08:00
parent b901a6ffc7
commit 5014e2f5fd
255 changed files with 15057 additions and 3115 deletions
Download Patch File Download Diff File Expand all files Collapse all files

933
apps/aether-gateway/src/tests/ai_execute/finalize_local_cli/compact.rs
View File

@@ -25,461 +25,6 @@ use aether_data_contracts::repository::provider_catalog::{
};
use sha2::{Digest, Sha256};
#[tokio::test]
async fn gateway_executes_openai_compact_cross_format_upstream_stream_via_local_finalize_response()
{
use base64::Engine as _;
#[derive(Debug, Clone)]
struct SeenRemoteExecutionRuntimeRequest {
trace_id: String,
request_id: String,
url: String,
authorization: String,
endpoint_tag: String,
provider_model: String,
has_model_field: bool,
}
fn hash_api_key(value: &str) -> String {
let mut hasher = Sha256::new();
hasher.update(value.as_bytes());
format!("{:x}", hasher.finalize())
}
fn sample_auth_snapshot(api_key_id: &str, user_id: &str) -> StoredAuthApiKeySnapshot {
StoredAuthApiKeySnapshot::new(
user_id.to_string(),
"alice".to_string(),
Some("alice@example.com".to_string()),
"user".to_string(),
"local".to_string(),
true,
false,
Some(serde_json::json!(["openai", "gemini"])),
Some(serde_json::json!(["openai:compact"])),
Some(serde_json::json!(["gpt-5"])),
api_key_id.to_string(),
Some("default".to_string()),
true,
false,
false,
Some(60),
Some(5),
Some(4_102_444_800_i64),
Some(serde_json::json!(["openai", "gemini"])),
Some(serde_json::json!(["openai:compact"])),
Some(serde_json::json!(["gpt-5"])),
)
.expect("auth snapshot should build")
}
fn sample_candidate_row() -> StoredMinimalCandidateSelectionRow {
StoredMinimalCandidateSelectionRow {
provider_id: "provider-openai-compact-gemini-finalize-local-1".to_string(),
provider_name: "gemini".to_string(),
provider_type: "custom".to_string(),
provider_priority: 10,
provider_is_active: true,
endpoint_id: "endpoint-openai-compact-gemini-finalize-local-1".to_string(),
endpoint_api_format: "gemini:cli".to_string(),
endpoint_api_family: Some("gemini".to_string()),
endpoint_kind: Some("cli".to_string()),
endpoint_is_active: true,
key_id: "key-openai-compact-gemini-finalize-local-1".to_string(),
key_name: "prod".to_string(),
key_auth_type: "bearer".to_string(),
key_is_active: true,
key_api_formats: Some(vec!["gemini:cli".to_string()]),
key_allowed_models: None,
key_capabilities: None,
key_internal_priority: 5,
key_global_priority_by_format: Some(serde_json::json!({"gemini:cli": 1})),
model_id: "model-openai-compact-gemini-finalize-local-1".to_string(),
global_model_id: "global-model-openai-compact-gemini-finalize-local-1".to_string(),
global_model_name: "gpt-5".to_string(),
global_model_mappings: None,
global_model_supports_streaming: Some(true),
model_provider_model_name: "gemini-2.5-pro-upstream".to_string(),
model_provider_model_mappings: Some(vec![StoredProviderModelMapping {
name: "gemini-2.5-pro-upstream".to_string(),
priority: 1,
api_formats: Some(vec!["gemini:cli".to_string()]),
}]),
model_supports_streaming: Some(true),
model_is_active: true,
model_is_available: true,
}
}
fn sample_provider_catalog_provider() -> StoredProviderCatalogProvider {
StoredProviderCatalogProvider::new(
"provider-openai-compact-gemini-finalize-local-1".to_string(),
"gemini".to_string(),
Some("https://example.com".to_string()),
"custom".to_string(),
)
.expect("provider should build")
.with_transport_fields(
true,
false,
false,
None,
Some(2),
None,
Some(20.0),
None,
None,
)
}
fn sample_provider_catalog_endpoint() -> StoredProviderCatalogEndpoint {
StoredProviderCatalogEndpoint::new(
"endpoint-openai-compact-gemini-finalize-local-1".to_string(),
"provider-openai-compact-gemini-finalize-local-1".to_string(),
"gemini:cli".to_string(),
Some("gemini".to_string()),
Some("cli".to_string()),
true,
)
.expect("endpoint should build")
.with_transport_fields(
"https://generativelanguage.googleapis.com".to_string(),
Some(serde_json::json!([
{"action":"set","key":"x-endpoint-tag","value":"openai-compact-gemini-finalize-cross-format"}
])),
None,
Some(2),
None,
None,
None,
None,
)
.expect("endpoint transport should build")
}
fn sample_provider_catalog_key() -> StoredProviderCatalogKey {
StoredProviderCatalogKey::new(
"key-openai-compact-gemini-finalize-local-1".to_string(),
"provider-openai-compact-gemini-finalize-local-1".to_string(),
"prod".to_string(),
"bearer".to_string(),
None,
true,
)
.expect("key should build")
.with_transport_fields(
Some(serde_json::json!(["gemini:cli"])),
encrypt_python_fernet_plaintext(
DEVELOPMENT_ENCRYPTION_KEY,
"sk-upstream-openai-compact-gemini-finalize",
)
.expect("api key should encrypt"),
None,
None,
Some(serde_json::json!({"gemini:cli": 1})),
None,
None,
None,
None,
)
.expect("key transport should build")
}
let seen_remote_execution_runtime =
Arc::new(Mutex::new(None::<SeenRemoteExecutionRuntimeRequest>));
let seen_remote_execution_runtime_clone = Arc::clone(&seen_remote_execution_runtime);
let report_hits = Arc::new(Mutex::new(0usize));
let report_hits_clone = Arc::clone(&report_hits);
let finalize_hits = Arc::new(Mutex::new(0usize));
let finalize_hits_clone = Arc::clone(&finalize_hits);
let decision_hits = Arc::new(Mutex::new(0usize));
let decision_hits_clone = Arc::clone(&decision_hits);
let plan_hits = Arc::new(Mutex::new(0usize));
let plan_hits_clone = Arc::clone(&plan_hits);
let public_hits = Arc::new(Mutex::new(0usize));
let public_hits_clone = Arc::clone(&public_hits);
let request_candidate_repository = Arc::new(InMemoryRequestCandidateRepository::default());
let usage_repository = Arc::new(InMemoryUsageReadRepository::default());
let upstream = Router::new()
.route(
"/api/internal/gateway/decision-sync",
any(move |_request: Request| {
let decision_hits_inner = Arc::clone(&decision_hits_clone);
async move {
*decision_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"action": "proxy_public"}))
}
}),
)
.route(
"/api/internal/gateway/plan-sync",
any(move |_request: Request| {
let plan_hits_inner = Arc::clone(&plan_hits_clone);
async move {
*plan_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"action": "proxy_public"}))
}
}),
)
.route(
"/api/internal/gateway/finalize-sync",
any(move |_request: Request| {
let finalize_hits_inner = Arc::clone(&finalize_hits_clone);
async move {
*finalize_hits_inner.lock().expect("mutex should lock") += 1;
(
StatusCode::IM_A_TEAPOT,
Body::from("finalize-sync-should-not-be-hit"),
)
}
}),
)
.route(
"/api/internal/gateway/report-sync",
any(move |request: Request| {
let report_hits_inner = Arc::clone(&report_hits_clone);
async move {
let (_parts, body) = request.into_parts();
let _raw_body = to_bytes(body, usize::MAX).await.expect("body should read");
*report_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"ok": true}))
}
}),
)
.route(
"/v1/responses/compact",
any(move |_request: Request| {
let public_hits_inner = Arc::clone(&public_hits_clone);
async move {
*public_hits_inner.lock().expect("mutex should lock") += 1;
(StatusCode::IM_A_TEAPOT, Body::from("public-route-hit"))
}
}),
);
let execution_runtime = Router::new().route(
"/v1/execute/sync",
any(move |request: Request| {
let seen_remote_execution_runtime_inner = Arc::clone(&seen_remote_execution_runtime_clone);
async move {
let (parts, body) = request.into_parts();
let raw_body = to_bytes(body, usize::MAX).await.expect("body should read");
let payload: serde_json::Value =
serde_json::from_slice(&raw_body).expect("execution runtime payload should parse");
*seen_remote_execution_runtime_inner
.lock()
.expect("mutex should lock") = Some(SeenRemoteExecutionRuntimeRequest {
trace_id: parts
.headers
.get(TRACE_ID_HEADER)
.and_then(|value| value.to_str().ok())
.unwrap_or_default()
.to_string(),
request_id: payload
.get("request_id")
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
url: payload
.get("url")
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
authorization: payload
.get("headers")
.and_then(|value| value.get("authorization"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
endpoint_tag: payload
.get("headers")
.and_then(|value| value.get("x-endpoint-tag"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
provider_model: payload
.get("body")
.and_then(|value| value.get("json_body"))
.and_then(|value| value.get("model"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
has_model_field: payload
.get("body")
.and_then(|value| value.get("json_body"))
.and_then(|value| value.get("model"))
.is_some(),
});
Json(json!({
"request_id": "trace-openai-compact-xfmt-stream-123",
"status_code": 200,
"headers": {
"content-type": "text/event-stream"
},
"body": {
"body_bytes_b64": base64::engine::general_purpose::STANDARD.encode(
concat!(
"data: {\"responseId\":\"upstream-compact-stream-123\",\"candidates\":[{\"content\":{\"parts\":[{\"text\":\"Hello \"}],\"role\":\"model\"},\"index\":0}],\"modelVersion\":\"gemini-2.5-pro-upstream\"}\n\n",
"data: {\"responseId\":\"upstream-compact-stream-123\",\"candidates\":[{\"content\":{\"parts\":[{\"text\":\"Hello Gemini Compact\"}],\"role\":\"model\"},\"finishReason\":\"STOP\",\"index\":0}],\"modelVersion\":\"gemini-2.5-pro-upstream\",\"usageMetadata\":{\"promptTokenCount\":2,\"candidatesTokenCount\":3,\"totalTokenCount\":5}}\n\n"
)
)
},
"telemetry": {
"elapsed_ms": 31
}
}))
}
}),
);
let auth_repository = Arc::new(InMemoryAuthApiKeySnapshotRepository::seed(vec![(
Some(hash_api_key("sk-client-openai-compact-xfmt-stream")),
sample_auth_snapshot(
"api-key-openai-compact-xfmt-stream-1",
"user-openai-compact-xfmt-stream-1",
),
)]));
let candidate_selection_repository =
Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![
sample_candidate_row(),
]));
let provider_catalog_repository = Arc::new(InMemoryProviderCatalogReadRepository::seed(
vec![sample_provider_catalog_provider()],
vec![sample_provider_catalog_endpoint()],
vec![sample_provider_catalog_key()],
));
let (upstream_url, upstream_handle) = start_server(upstream).await;
let (execution_runtime_url, execution_runtime_handle) = start_server(execution_runtime).await;
let gateway_state =
build_state_with_execution_runtime_override(execution_runtime_url.clone())
.with_data_state_for_tests(
GatewayDataState::with_auth_candidate_selection_provider_catalog_request_candidates_and_usage_for_tests(
auth_repository,
candidate_selection_repository,
provider_catalog_repository,
Arc::clone(&request_candidate_repository),
Arc::clone(&usage_repository),
DEVELOPMENT_ENCRYPTION_KEY,
),
);
let gateway = build_router_with_state(gateway_state);
let (gateway_url, gateway_handle) = start_server(gateway).await;
let started_at = std::time::Instant::now();
let response = reqwest::Client::new()
.post(format!("{gateway_url}/v1/responses/compact"))
.header(http::header::CONTENT_TYPE, "application/json")
.header(
http::header::AUTHORIZATION,
"Bearer sk-client-openai-compact-xfmt-stream",
)
.header(TRACE_ID_HEADER, "trace-openai-compact-xfmt-stream-123")
.body("{\"model\":\"gpt-5\",\"input\":\"hello\"}")
.send()
.await
.expect("request should succeed");
let elapsed = started_at.elapsed();
assert_eq!(response.status(), StatusCode::OK);
let response_json: serde_json::Value = response.json().await.expect("body should parse");
assert_eq!(
response_json,
json!({
"id": "upstream-compact-stream-123",
"object": "response",
"status": "completed",
"model": "gemini-2.5-pro-upstream",
"output": [{
"type": "message",
"id": "upstream-compact-stream-123_msg",
"role": "assistant",
"status": "completed",
"content": [{
"type": "output_text",
"text": "Hello Gemini Compact",
"annotations": []
}]
}],
"usage": {
"input_tokens": 2,
"output_tokens": 3,
"total_tokens": 5
}
})
);
assert!(
elapsed < std::time::Duration::from_millis(10_000),
"response took unexpectedly long for local finalize path: elapsed={elapsed:?} finalize_hits={} report_hits={}",
*finalize_hits.lock().expect("mutex should lock"),
*report_hits.lock().expect("mutex should lock"),
);
let seen_remote_execution_runtime_request = seen_remote_execution_runtime
.lock()
.expect("mutex should lock")
.clone()
.expect("remote execution runtime plan should be captured");
assert_eq!(
seen_remote_execution_runtime_request.trace_id,
"trace-openai-compact-xfmt-stream-123"
);
assert_eq!(
seen_remote_execution_runtime_request.request_id,
"trace-openai-compact-xfmt-stream-123"
);
assert_eq!(
seen_remote_execution_runtime_request.url,
"https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-pro-upstream:generateContent"
);
assert_eq!(
seen_remote_execution_runtime_request.authorization,
"Bearer sk-upstream-openai-compact-gemini-finalize"
);
assert_eq!(
seen_remote_execution_runtime_request.endpoint_tag,
"openai-compact-gemini-finalize-cross-format"
);
assert_eq!(
seen_remote_execution_runtime_request.provider_model,
"gemini-2.5-pro-upstream"
);
assert!(seen_remote_execution_runtime_request.has_model_field);
let mut stored_candidates = Vec::new();
for _ in 0..50 {
stored_candidates = request_candidate_repository
.list_by_request_id("trace-openai-compact-xfmt-stream-123")
.await
.expect("request candidate trace should read");
if stored_candidates.len() == 1
&& stored_candidates[0].status == RequestCandidateStatus::Success
{
break;
}
tokio::time::sleep(std::time::Duration::from_millis(10)).await;
}
assert_eq!(stored_candidates.len(), 1);
assert_eq!(stored_candidates[0].status, RequestCandidateStatus::Success);
tokio::time::sleep(std::time::Duration::from_millis(100)).await;
assert_eq!(
*report_hits.lock().expect("mutex should lock"),
0,
"report-sync should stay local when request candidate persistence is available"
);
assert_eq!(*finalize_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*decision_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*plan_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*public_hits.lock().expect("mutex should lock"), 0);
gateway_handle.abort();
execution_runtime_handle.abort();
upstream_handle.abort();
}
#[tokio::test]
async fn gateway_executes_openai_compact_openai_family_upstream_stream_via_local_finalize_response()
{
@@ -716,12 +261,13 @@ async fn gateway_executes_openai_compact_openai_family_upstream_stream_via_local
let execution_runtime = Router::new().route(
"/v1/execute/sync",
any(move |request: Request| {
let seen_remote_execution_runtime_inner = Arc::clone(&seen_remote_execution_runtime_clone);
let seen_remote_execution_runtime_inner =
Arc::clone(&seen_remote_execution_runtime_clone);
async move {
let (parts, body) = request.into_parts();
let raw_body = to_bytes(body, usize::MAX).await.expect("body should read");
let payload: serde_json::Value =
serde_json::from_slice(&raw_body).expect("execution runtime payload should parse");
let payload: serde_json::Value = serde_json::from_slice(&raw_body)
.expect("execution runtime payload should parse");
*seen_remote_execution_runtime_inner
.lock()
.expect("mutex should lock") = Some(SeenRemoteExecutionRuntimeRequest {
@@ -808,18 +354,17 @@ async fn gateway_executes_openai_compact_openai_family_upstream_stream_via_local
let (upstream_url, upstream_handle) = start_server(upstream).await;
let (execution_runtime_url, execution_runtime_handle) = start_server(execution_runtime).await;
let gateway_state =
build_state_with_execution_runtime_override(execution_runtime_url.clone())
.with_data_state_for_tests(
GatewayDataState::with_auth_candidate_selection_provider_catalog_request_candidates_and_usage_for_tests(
auth_repository,
candidate_selection_repository,
provider_catalog_repository,
Arc::clone(&request_candidate_repository),
Arc::clone(&usage_repository),
DEVELOPMENT_ENCRYPTION_KEY,
),
);
let gateway_state = build_state_with_execution_runtime_override(execution_runtime_url.clone())
.with_data_state_for_tests(
GatewayDataState::with_auth_candidate_selection_provider_catalog_request_candidates_and_usage_for_tests(
auth_repository,
candidate_selection_repository,
provider_catalog_repository,
Arc::clone(&request_candidate_repository),
Arc::clone(&usage_repository),
DEVELOPMENT_ENCRYPTION_KEY,
),
);
let gateway = build_router_with_state(gateway_state);
let (gateway_url, gateway_handle) = start_server(gateway).await;
@@ -926,451 +471,3 @@ async fn gateway_executes_openai_compact_openai_family_upstream_stream_via_local
execution_runtime_handle.abort();
upstream_handle.abort();
}
#[tokio::test]
async fn gateway_executes_openai_compact_openai_family_upstream_stream_via_local_finalize_response_even_when_conversion_flagged(
) {
use base64::Engine as _;
#[derive(Debug, Clone)]
struct SeenRemoteExecutionRuntimeRequest {
trace_id: String,
request_id: String,
url: String,
model: String,
authorization: String,
endpoint_tag: String,
}
fn hash_api_key(value: &str) -> String {
let mut hasher = Sha256::new();
hasher.update(value.as_bytes());
format!("{:x}", hasher.finalize())
}
fn sample_auth_snapshot(api_key_id: &str, user_id: &str) -> StoredAuthApiKeySnapshot {
StoredAuthApiKeySnapshot::new(
user_id.to_string(),
"alice".to_string(),
Some("alice@example.com".to_string()),
"user".to_string(),
"local".to_string(),
true,
false,
Some(serde_json::json!(["openai"])),
Some(serde_json::json!(["openai:compact"])),
Some(serde_json::json!(["gpt-5"])),
api_key_id.to_string(),
Some("default".to_string()),
true,
false,
false,
Some(60),
Some(5),
Some(4_102_444_800_i64),
Some(serde_json::json!(["openai"])),
Some(serde_json::json!(["openai:compact"])),
Some(serde_json::json!(["gpt-5"])),
)
.expect("auth snapshot should build")
}
fn sample_candidate_row() -> StoredMinimalCandidateSelectionRow {
StoredMinimalCandidateSelectionRow {
provider_id: "provider-openai-compact-openai-family-conversion-local-1".to_string(),
provider_name: "openai".to_string(),
provider_type: "custom".to_string(),
provider_priority: 10,
provider_is_active: true,
endpoint_id: "endpoint-openai-compact-openai-family-conversion-local-1".to_string(),
endpoint_api_format: "openai:cli".to_string(),
endpoint_api_family: Some("openai".to_string()),
endpoint_kind: Some("cli".to_string()),
endpoint_is_active: true,
key_id: "key-openai-compact-openai-family-conversion-local-1".to_string(),
key_name: "prod".to_string(),
key_auth_type: "bearer".to_string(),
key_is_active: true,
key_api_formats: Some(vec!["openai:cli".to_string()]),
key_allowed_models: None,
key_capabilities: None,
key_internal_priority: 5,
key_global_priority_by_format: Some(serde_json::json!({"openai:cli": 1})),
model_id: "model-openai-compact-openai-family-conversion-local-1".to_string(),
global_model_id: "global-model-openai-compact-openai-family-conversion-local-1"
.to_string(),
global_model_name: "gpt-5".to_string(),
global_model_mappings: None,
global_model_supports_streaming: Some(true),
model_provider_model_name: "gpt-5-upstream".to_string(),
model_provider_model_mappings: Some(vec![StoredProviderModelMapping {
name: "gpt-5-upstream".to_string(),
priority: 1,
api_formats: Some(vec!["openai:cli".to_string()]),
}]),
model_supports_streaming: Some(true),
model_is_active: true,
model_is_available: true,
}
}
fn sample_provider_catalog_provider() -> StoredProviderCatalogProvider {
StoredProviderCatalogProvider::new(
"provider-openai-compact-openai-family-conversion-local-1".to_string(),
"openai".to_string(),
Some("https://example.com".to_string()),
"custom".to_string(),
)
.expect("provider should build")
.with_transport_fields(
true,
false,
false,
None,
Some(2),
None,
Some(20.0),
None,
None,
)
}
fn sample_provider_catalog_endpoint() -> StoredProviderCatalogEndpoint {
StoredProviderCatalogEndpoint::new(
"endpoint-openai-compact-openai-family-conversion-local-1".to_string(),
"provider-openai-compact-openai-family-conversion-local-1".to_string(),
"openai:cli".to_string(),
Some("openai".to_string()),
Some("cli".to_string()),
true,
)
.expect("endpoint should build")
.with_transport_fields(
"https://api.openai.example".to_string(),
Some(serde_json::json!([
{"action":"set","key":"x-endpoint-tag","value":"openai-cli-compact-family-finalize-local"}
])),
None,
Some(2),
Some("/custom/v1/responses".to_string()),
None,
None,
None,
)
.expect("endpoint transport should build")
}
fn sample_provider_catalog_key() -> StoredProviderCatalogKey {
StoredProviderCatalogKey::new(
"key-openai-compact-openai-family-conversion-local-1".to_string(),
"provider-openai-compact-openai-family-conversion-local-1".to_string(),
"prod".to_string(),
"bearer".to_string(),
None,
true,
)
.expect("key should build")
.with_transport_fields(
Some(serde_json::json!(["openai:cli"])),
encrypt_python_fernet_plaintext(
DEVELOPMENT_ENCRYPTION_KEY,
"sk-upstream-openai-cli-family-finalize",
)
.expect("api key should encrypt"),
None,
None,
Some(serde_json::json!({"openai:cli": 1})),
None,
None,
None,
None,
)
.expect("key transport should build")
}
let seen_remote_execution_runtime =
Arc::new(Mutex::new(None::<SeenRemoteExecutionRuntimeRequest>));
let seen_remote_execution_runtime_clone = Arc::clone(&seen_remote_execution_runtime);
let report_hits = Arc::new(Mutex::new(0usize));
let report_hits_clone = Arc::clone(&report_hits);
let finalize_hits = Arc::new(Mutex::new(0usize));
let finalize_hits_clone = Arc::clone(&finalize_hits);
let decision_hits = Arc::new(Mutex::new(0usize));
let decision_hits_clone = Arc::clone(&decision_hits);
let plan_hits = Arc::new(Mutex::new(0usize));
let plan_hits_clone = Arc::clone(&plan_hits);
let public_hits = Arc::new(Mutex::new(0usize));
let public_hits_clone = Arc::clone(&public_hits);
let request_candidate_repository = Arc::new(InMemoryRequestCandidateRepository::default());
let usage_repository = Arc::new(InMemoryUsageReadRepository::default());
let upstream = Router::new()
.route(
"/api/internal/gateway/decision-sync",
any(move |_request: Request| {
let decision_hits_inner = Arc::clone(&decision_hits_clone);
async move {
*decision_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"action": "proxy_public"}))
}
}),
)
.route(
"/api/internal/gateway/plan-sync",
any(move |_request: Request| {
let plan_hits_inner = Arc::clone(&plan_hits_clone);
async move {
*plan_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"action": "proxy_public"}))
}
}),
)
.route(
"/api/internal/gateway/finalize-sync",
any(move |_request: Request| {
let finalize_hits_inner = Arc::clone(&finalize_hits_clone);
async move {
*finalize_hits_inner.lock().expect("mutex should lock") += 1;
(
StatusCode::IM_A_TEAPOT,
Body::from("finalize-sync-should-not-be-hit"),
)
}
}),
)
.route(
"/api/internal/gateway/report-sync",
any(move |request: Request| {
let report_hits_inner = Arc::clone(&report_hits_clone);
async move {
let (_parts, body) = request.into_parts();
let _raw_body = to_bytes(body, usize::MAX).await.expect("body should read");
*report_hits_inner.lock().expect("mutex should lock") += 1;
Json(json!({"ok": true}))
}
}),
)
.route(
"/v1/responses/compact",
any(move |_request: Request| {
let public_hits_inner = Arc::clone(&public_hits_clone);
async move {
*public_hits_inner.lock().expect("mutex should lock") += 1;
(StatusCode::IM_A_TEAPOT, Body::from("public-route-hit"))
}
}),
);
let execution_runtime = Router::new().route(
"/v1/execute/sync",
any(move |request: Request| {
let seen_remote_execution_runtime_inner = Arc::clone(&seen_remote_execution_runtime_clone);
async move {
let (parts, body) = request.into_parts();
let raw_body = to_bytes(body, usize::MAX).await.expect("body should read");
let payload: serde_json::Value =
serde_json::from_slice(&raw_body).expect("execution runtime payload should parse");
*seen_remote_execution_runtime_inner
.lock()
.expect("mutex should lock") = Some(SeenRemoteExecutionRuntimeRequest {
trace_id: parts
.headers
.get(TRACE_ID_HEADER)
.and_then(|value| value.to_str().ok())
.unwrap_or_default()
.to_string(),
request_id: payload
.get("request_id")
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
url: payload
.get("url")
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
model: payload
.get("body")
.and_then(|value| value.get("json_body"))
.and_then(|value| value.get("model"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
authorization: payload
.get("headers")
.and_then(|value| value.get("authorization"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
endpoint_tag: payload
.get("headers")
.and_then(|value| value.get("x-endpoint-tag"))
.and_then(|value| value.as_str())
.unwrap_or_default()
.to_string(),
});
Json(json!({
"request_id": "trace-openai-compact-openai-family-conversion-123",
"status_code": 200,
"headers": {
"content-type": "text/event-stream"
},
"body": {
"body_bytes_b64": base64::engine::general_purpose::STANDARD.encode(
concat!(
"event: response.created\n",
"data: {\"type\":\"response.created\",\"response\":{\"id\":\"resp_compact_openai_family_conversion_123\",\"object\":\"response\",\"model\":\"gpt-5\",\"status\":\"in_progress\",\"output\":[]}}\n\n",
"event: response.output_text.delta\n",
"data: {\"type\":\"response.output_text.delta\",\"output_index\":0,\"content_index\":0,\"delta\":\"Hello Compact\"}\n\n",
"event: response.completed\n",
"data: {\"type\":\"response.completed\",\"response\":{\"id\":\"resp_compact_openai_family_conversion_123\",\"object\":\"response\",\"model\":\"gpt-5\",\"status\":\"completed\",\"output\":[],\"usage\":{\"input_tokens\":2,\"output_tokens\":3,\"total_tokens\":5}}}\n\n"
)
)
},
"telemetry": {
"elapsed_ms": 31
}
}))
}
}),
);
let auth_repository = Arc::new(InMemoryAuthApiKeySnapshotRepository::seed(vec![(
Some(hash_api_key(
"sk-client-openai-compact-openai-family-conversion",
)),
sample_auth_snapshot(
"api-key-openai-compact-openai-family-conversion-1",
"user-openai-compact-openai-family-conversion-1",
),
)]));
let candidate_selection_repository =
Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![
sample_candidate_row(),
]));
let provider_catalog_repository = Arc::new(InMemoryProviderCatalogReadRepository::seed(
vec![sample_provider_catalog_provider()],
vec![sample_provider_catalog_endpoint()],
vec![sample_provider_catalog_key()],
));
let (upstream_url, upstream_handle) = start_server(upstream).await;
let (execution_runtime_url, execution_runtime_handle) = start_server(execution_runtime).await;
let gateway_state =
build_state_with_execution_runtime_override(execution_runtime_url.clone())
.with_data_state_for_tests(
GatewayDataState::with_auth_candidate_selection_provider_catalog_request_candidates_and_usage_for_tests(
auth_repository,
candidate_selection_repository,
provider_catalog_repository,
Arc::clone(&request_candidate_repository),
Arc::clone(&usage_repository),
DEVELOPMENT_ENCRYPTION_KEY,
),
);
let gateway = build_router_with_state(gateway_state);
let (gateway_url, gateway_handle) = start_server(gateway).await;
let started_at = std::time::Instant::now();
let response = reqwest::Client::new()
.post(format!("{gateway_url}/v1/responses/compact"))
.header(http::header::CONTENT_TYPE, "application/json")
.header(
http::header::AUTHORIZATION,
"Bearer sk-client-openai-compact-openai-family-conversion",
)
.header(
TRACE_ID_HEADER,
"trace-openai-compact-openai-family-conversion-123",
)
.body("{\"model\":\"gpt-5\",\"input\":\"hello\"}")
.send()
.await
.expect("request should succeed");
let elapsed = started_at.elapsed();
assert_eq!(response.status(), StatusCode::OK);
let response_json: serde_json::Value = response.json().await.expect("body should parse");
assert_eq!(
response_json,
json!({
"id": "resp_compact_openai_family_conversion_123",
"object": "response",
"model": "gpt-5",
"status": "completed",
"output": [],
"usage": {
"input_tokens": 2,
"output_tokens": 3,
"total_tokens": 5
}
})
);
assert!(
elapsed < std::time::Duration::from_millis(10_000),
"response took unexpectedly long for local finalize path: elapsed={elapsed:?} finalize_hits={} report_hits={}",
*finalize_hits.lock().expect("mutex should lock"),
*report_hits.lock().expect("mutex should lock"),
);
let seen_remote_execution_runtime_request = seen_remote_execution_runtime
.lock()
.expect("mutex should lock")
.clone()
.expect("remote execution runtime plan should be captured");
assert_eq!(
seen_remote_execution_runtime_request.trace_id,
"trace-openai-compact-openai-family-conversion-123"
);
assert_eq!(
seen_remote_execution_runtime_request.request_id,
"trace-openai-compact-openai-family-conversion-123"
);
assert_eq!(
seen_remote_execution_runtime_request.url,
"https://api.openai.example/custom/v1/responses"
);
assert_eq!(
seen_remote_execution_runtime_request.model,
"gpt-5-upstream"
);
assert_eq!(
seen_remote_execution_runtime_request.authorization,
"Bearer sk-upstream-openai-cli-family-finalize"
);
assert_eq!(
seen_remote_execution_runtime_request.endpoint_tag,
"openai-cli-compact-family-finalize-local"
);
let mut stored_candidates = Vec::new();
for _ in 0..50 {
stored_candidates = request_candidate_repository
.list_by_request_id("trace-openai-compact-openai-family-conversion-123")
.await
.expect("request candidate trace should read");
if stored_candidates.len() == 1
&& stored_candidates[0].status == RequestCandidateStatus::Success
{
break;
}
tokio::time::sleep(std::time::Duration::from_millis(10)).await;
}
assert_eq!(stored_candidates.len(), 1);
assert_eq!(stored_candidates[0].status, RequestCandidateStatus::Success);
tokio::time::sleep(std::time::Duration::from_millis(100)).await;
assert_eq!(
*report_hits.lock().expect("mutex should lock"),
0,
"report-sync should stay local when request candidate persistence is available"
);
assert_eq!(*finalize_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*decision_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*plan_hits.lock().expect("mutex should lock"), 0);
assert_eq!(*public_hits.lock().expect("mutex should lock"), 0);
gateway_handle.abort();
execution_runtime_handle.abort();
upstream_handle.abort();
}